Click for full size

(Sort of) in the tradition of Mike Fagan's Fightlines, I bring you UFC 142's Dirty Dozen, the twelve fighters surrounding this Saturday's featherweight title bout. Under ideal circumstances, this bracket would of course feature a Sweet Sixteen so to speak, but the unpredictability of MMA makes it difficult for its practitioners to consistently perform well against others who do the same.

Blue fighters were seeded into the first round of the bracket, with no special consideration. Green fighters received one level of privilege, either a bye to the second round, or an extra seed elsewhere in the first. Yellow fighters were granted some combination of the two (for example, a two-round bye, or a ridiculous three total seeds).

In each individual match, the then-betting favorite is positioned above the underdog. Only two upset losses occurred en route to Aldo vs. Mendes, and they were both suffered by Mike Brown, who was initially afforded the most favorable conditions to reach the final round (two seeds and a single bye).

Seeding, as extrapolated from betting odds and pairings:

1. Aldo
2. Mendes
3. MTB
4. Florian
5. MTB (again)
   
6. Yahya
7. Omigawa
8. Hominick
9. Roop
10. Escovedo
11. Nunes
12. Vazquez
13. Gamburyan

Not much to analyze here, it's just an alternative way to look at the title picture at 145.

What-ifs are a popular part of our sport. Maybe it's a holdover from the not so long ago days when we would have to wait a month (or months!) between major events. Maybe it's a symptom of an internet-age hardcore fanbase. Maybe it's just the natural tendency of any sports fans (see draft predictions, fantasy leagues, March madness). Whatever the reason, speculation in MMA most definitely runs the gamut from harmless imaginings to overblown ruminative monstrosities. Let it be known that I expect this fanpost to number among the latter.

In times of either extreme instability or stagnation within a weight class, a tournament, no matter how impractical or unlikely, is often submitted as a solution. With the increasing number of fighters defeated by Georges St. Pierre, but whom otherwise continue to win, I've heard several longtime fans suggest that a tournament involving both those former challengers and new rising contenders is the fairest way to determine the recipient of the next title shot. That sentiment was reiterated when the 170-pound champion was shelved with consecutive knee injuries, necessitating an interim belt. The Strikeforce heavyweight division, in turmoil following a drought in activity from their champion, as well as the first in #1-ranked Fedor Emelianenko's recent collection of losses, played host to an ongoing grand prix in an attempt to bring its title picture back into focus. Bellator was predicated upon the concept of perpetual tournaments across all weight classes. Even the UFC itself seeded a four-man bracket to inaugurate their flyweight division and crown its first titleholder. In short, there are numerous examples of both envisioned and executed bracket-style matchmaking to be found.

But the tournament format can be just as interesting when considered in reverse. Backfilling a bracket from a title fight can give us a unique sense of the circumstances that allowed (or forced) the matchup to happen. It can also lead us to some bizarre what-if scenarios, as well as some possible answers. This dual perspective, "How did this happen?" and "What could have happened instead?" will be the focus of this fanpost, and any future installments.

For this freshman effort, I'll actually be looking back to last Saturday's championship bout in the 205-pound division. We all know the result, a second round chokeout by retaining champ Jon Jones, but what road did Lyoto Machida travel, all to find himself with his neck in the crook of Jones' elbow? Before that analysis, allow me to explain how I invented a tournament backwards from the final match, and what I expected to find.

Since Machida was selected to challenge for the title, his otherwise most recent fight could be retroactively considered a de facto #1 contendership bout, even if it wasn't regarded as such at the time (more on this later). This matchup will serve as the final round of the "tournament," with one side of the bracket belonging to Machida himself, and the other side ostensibly dominated by his last opponent before Jones, in this case Randy Couture. Backfilling through two further rounds, using each fighter's victories and those of their previously successful opponents should, under ideal circumstances, yield an opening quarterfinal round with eight unique competitors.

In this instance, pretty much nothing goes as expected, which results in just about the most broken would-be tournament conceivable, a bracket that would be mercilessly ridiculed if uttered aloud. Here's a look:

• Machida actually lost his fight with Rampage immediately before defeating Couture, which means that the Machida/Rampage fight has no place in the bracket. Consequently, were this a true tournament, a two-round bye would be the only explanation for Machida's spot in the finals.
• Couture actually did win two fights before being booked opposite Machida. However, his opponent in the "semifinal" round was MMA newcomer James Toney, who, in this hypothetical tournament scenario, must have been awarded a single-round bye and seeded directly into the semifinals, where he awaited the winner of:
• Couture vs. Coleman, the lone contest in what I projected would be a four-fight quarterfinal round.

The end product looks more like an easy ladder on Mortal Kombat than a classic Pride grand prix. Nonetheless, these were really and truly the circumstances surrounding and leading to Machida's latest title "run". Why does the bracket look so bad? Earlier, I mentioned that backfilling matchups in this manner ought to produce a pretty good-looking eight-man tournament. The reason this should happen is because it would mean that the latest title challenger beat guys whom beat other guys. Even though UFC matchmaking is officially linear, fighters don't typically get title shots on the strength of a single non-consecutive win.

One possible reason for a murky title picture (and crappy bracket) is a shallow weight division. Heavyweight fighters in all the major promotions (except Bellator!) have historically been given title shots after a lone victory because there often just weren't enough compelling matchups to do otherwise. The UFC once allowed a heavyweight fighter to challenge for the belt coming off a loss! Fortunately, depth is a non-issue at 205 pounds. While it has in recent years been surpassed in terms of competitive insanity, light heavyweight is still a very robust division with no shortage of contenders.

If not divisional poverty, then questionable matchmaking could be the culprit. While it's very difficult to find faults in Joe Silva's plying of his trade, there are some arguable missteps in this small collection of bouts. The booking of James Toney in any capacity is the top offender, though that may have been more the machination of Dana White. I find it hard to imagine a world in which Toney's contribution to the UFC outweighed the sum he was paid as a participant. Given that, it would have made more sense to pair Toney with lowly-ranked Coleman, and let the winner square off against Couture. Machida vs. Couture wasn't a crown jewel of matchmaking either, but it struck a certain balance in having a former champion coming off a top-ten loss stand opposite a (distant) former champion riding a flimsy win streak.

The real cause of this weak showing, at which most readers have no doubt already arrived, was injuries. Machida didn't receive a title shot because Joe Silva went insane, or earn one by beating Couture; he inherited one from a sidelined Rashad Evans, the intended title challenger. If not for Evans' misfortune, we would be analyzing a very different fantasy bracket today:

While not perfect, this setup is certainly not as plagued by seeding errors and mismatches as the one above. And beyond directly changing Jon Jones' dance card, training injuries were also the cause of Machida and Evans being in separate brackets at all, as they were originally meant to rematch for #1 contendership, until Machida was injured and replaced by Tito Ortiz.

What about the other guys?

Had Mark Coleman somehow defeated Couture in the quarterfinals, it would be easy to imagine the UFC booking him to welcome loudmouth James Toney to the octagon. A likely win over Toney would theoretically pit him against Machida. While this would be a criminally irresponsible mismatch, Coleman is a veteran of the sport, and an athletic commission recently cleared him to fight Shogun.

For his part, a win over either Coleman or Couture would probably not earn Toney an appointment opposite Machida. Joe Silva wouldn't dream of it, and none but the sleaziest of ACs would give it the green light.

Now for the crazy part: what if Couture had bested Machida in the finals? What if Coleman won the whole tournament? We've established that Machida, having not even two consecutive wins at the time, was awarded a title shot primarily due to training injuries sustained by official challenger Rashad Evans. Under these extenuating circumstances, one would have to imagine that a Couture or Coleman victory over Machida would heavily qualify either man as a replacement for Evans.

Had Machida been defeated by Couture or Coleman, it's also possible that a different fighter altogether would have received the call. Likely candidates could have been Dan Henderson (which would have very sadly removed him from his brawl with Shogun) or Phil Davis, who in that case would have been fast-tracked to a title fight in a manner not unlike Jones' own.

Alright assdicks, the last scheduled event of the MMA Salary Cap Challenge beta is bearing down, and given the game's shamefully undeserved lack of popularity, it's anyone's guess whether or not it will resurface after this Saturday. That being said, I'm gonna try to sink this ship in style, won't you join me?

The first main card fight of UFC 139 pits Ultimate Fighter darling Bonnar against Ultimate Fighter lab experiment Kingsbury. Though the younger, more athletic Kingsbury should take this fight handily, it's interesting to note Bonnar's apparent defensive success across three contests. He has generously outgrappled his past two opponents (Pokrajac and Soszynski twice... kinda racist), going so far as to deny every single grappling maneuver attempted against him (assuming there were any at all, of course). While Kingsbury is a gifted wrestler, he may be better off avoiding the mat and risking a Bonnar submission; instead using his striking power to batter the skilled but physically weaker veteran.

Story vs. Kampmann is as even a contest as I've seen in the past couple months of playing MSCC. Both fighters are evenly matched not only in total expected points, but over four of the five point categories. Kampmann's most marked advantage is his striking power, but Story has thus far displayed a tough beard. On a card stuffed with flashy strikers, this is my dark horse pick for Fight of the Night, ending in a hard-earned decision in Story's favor or a third round TKO win for Kampmann.

It's a little strange to glance at the Knockdowns row for this matchup and see nothing but zeroes, but it's actually true that neither Faber nor Bowles have hit the mark with any bombs in quite awhile. As a matter of fact, points-wise, this is the least compelling matchup on the main card. Fortunately, it's my expectation that these two can bring out the absolute best in each other. Bowles' higher Takedowns stat is a bit of a red herring, as he completed those attempts against wrestlers highly inferior to Faber. With that supposed edge erased, Faber should be able to pick up the win wherever he so chooses.

Bring out your dead, bring out your dead. I have a sad and overwhelming suspicion that Wanderlei Silva, in whom we've seen both bloodthirsty demon god and mild-mannered advice columnist, will not get the better of this fight. Silva's best chance appears to hang off the possibility that Cung Le's busy acting itinerary will render him rusty and uncomfortable in the cage. While that's certainly a possibility (as Rampage once demonstrated), my understanding is that Le is not only a prizefighter, but a lifelong martial artist with matching pride and dedication. It would surprise me if he used his film career as a "vacation" from his craft, the way some fighters might. Much as it pains me to say, I expect Wanderlei's eroding chin to prove a frequent and unfortunate target for Le's speedy combinations.

Like the co-main event before it, Saturday's main attraction will spell doom for a decorated Pride veteran. The difference in this bout is that both contestants carry that distinction. Two of Rua's past three performances have been vintage beatdowns of top-tier opponents. Even so, Henderson leads him in expected fantasy points by a fairly comfortable margin. Both men have well-documented knockout power, but Rua was outwrestled badly in his title loss to Jon Jones. Henderson will be no more merciful if given the chance, and I don't see him falling for any kneebar attempts from Rua. Hendo just has more ways to win, and I think he'll be adding another impressive win to an already sterling record.

• 21 on Silva ($147)
• 10 on Henderson ($150)
• 11 on Bonnar ($143)
• 5 on Story ($60)
• 47 plays total ($500)

The MSCC staff posted some interesting prices for this event. Wanderlei Silva, while very likely to get trounced, should be able to perform well enough to be a steal at $7. Not only that, but if dreams really do come true and he manages to axe murder Le, those 21 buys will rake in an absurd amount of points. Henderson and Bonnar's purchase costs are a little more defensible, and still great deals.

With $60 left, I would normally make as many buys as I could on the #4 ranked fighter (Martin Kampmann) on the value leaderboard. This time, however, it crossed my mind to work out the expected payoff from laying that $60 on each of the remaining fighters, since the prices don't always divide into the budget evenly (for example, $60 on Kampmann ($11) would secure 5 buys, but leave $5 unspent). Sure enough, the best return came from #7 Rick Story, whose $12 cost divides evenly into my leftover funds, preventing any wasted dollars.

Continue reading this post »

Howdy howdy howdy! No analysis this week, because I spent my writing time coming up with an explanatory section on the MMA Salary Cap Game and how to use my stat tables. That whole rigamarole is after the break. Let me know if anything is still unclear, and happy pickin!

My Buys

• 15 on Pablo Garza ($150)
• 15 on Clay Guida ($150)
• 5 on Cain Velasquez ($100)
• 35 plays total ($400)

Continue reading this post »

This weekend marks the beginning of the first multi-event fantasy game at the MMA Salary Cap Challenge. This should prove to be similar to the previous single-event games, with the exception of more opportunities for players to earn points and rattle the leaderboards, as well as the possibility that fantasy points scored after an event (along with any unspent cash) become your budget for the next one.

The MSCC roster for UFC 138 was just posted yesterday, so for this event I'll just be sharing the usual data (expected fantasy points, points-per-dollar, and my buys), with very little analysis. Next week I'll try to be more in-depth.

The real mystery of fantasy play for UFC 138 is what the hell to do about Alves vs. Abedi. Abedi has spent his career on the European regional circuit, and thus Fightmetric has no pertinent data to contribute on his behalf.

To estimate very, very tentative expected fantasy points figures, I combined four provided values: The number of fantasy points scored both for (8.16) and against (9.69) Alves in his last three bouts, his cost ($15), and Abedi's ($7). As the play cost of each fighter tends to represent MSCC's performance expectations, I summed Alves' fantasy point stats (17.85, a reflection of his recent strengths and weaknesses) and awarded 15/22 of the total to Alves himself, and the remaining 7/22 to Abedi. This levels their points/dollar scores, making both fighters equally "worth it" or "not worth it" from a fantasy perspective. If Alves has impressive numbers compared to the other main card combatants, then he (and Abedi) will rank highly on the points/dollar leaderboard. Otherwise, both fighters will probably avoid being recommended as valuable plays.

Making Plays

My Plays:

• 15 on Perosh ($150)
• 11 on Leben ($143)
• 21 on Faaloloto ($147)
• 4 on Barao ($52)
• 1 on Abedi ($7)
• 52 plays total ($499)

The value leaderboard is much more shuffled this time than last. The best and worst plays are still opposing fighters, but the other pairings don't appear to follow any discernable pattern. Perosh is the underdog to Diabate, but I think he'll find opportunities to score on the mat. Same goes for Barao against Pickett. Munoz is poised to outscore Leben overall, but I think he's too pricey to count on. Abedi gets a play because I had some pocket change leftover after buying the four big earners.

I really encourage everyone here at BE to sign up and play at MSCC. I was surprised that player count dropped between the first and second games- this is really fun and well-executed. Any fans of fantasy football or baseball need to do yourselves a favor and give this a spin!

UFC 137 is limping to the starting line, which means it's time for everyone to make their picks for the second MMA Salary Cap game. Below are my stat tables for each main card fight. I'll always calculate Expected Fantasy Points the same way (the average of Fighter A's points and Fighter B's opponents' points), but this time around I'll be sharing my plays, which are now based on much more sensible criteria.

As Japan's last great hope for the foreseeable future, Hioki enters UFC 137 under considerable pressure, against a dangerous and underrated opponent in Roop. From a statistical standpoint, this looks to be a striker vs. grappler matchup, but Hioki has more striking talent than it seems here. He held his own on the feet opposite Marlon Sandro, a much more threatening striker than Roop. For his part, Roop is currently on the right side of a knock out or be knocked out campaign, but I'm not confident that he can put Hioki away before getting dragged to the ground and submitted.

While I'm a fan of both, Jorgensen vs. Curran is the frontrunner for slow burn of the night. Over their past six fights combined, neither man has made an especially dramatic hammer or nail. Jorgensen tends to get hit, but there isn't really any version of Curran that can effectively exploit that fact. I believe the numbers will spoil the ending for us this time: Jorgensen is good at takedowns; Curran is good at getting taken down. Curran will inspire moments of hope with spirited sub attempts, but for the most part Jorgensen will keep him under wraps and beat him up with enough ground and pound to take a clear decision.

The statistics here would have you believe that Mirko Cro Cop is a more talented grappler than Roy Nelson. While the opposite is almost certainly true, this inconsistency actually highlights a legitimate flaw in Nelson's approach: poor gameplanning. His shortage of points earned on the mat is a result of refusing to force or even oblige his opponent in going there. Instead, he's elected to absorb punishment on the feet from top 10 fighters, while trying to land the one hitter quitter that paid off against lesser opponents. Fortunately for Nelson, the general opinion seems to be that a slow, aging Cro Cop is the perfect victim for such an approach, and no better off if the fight hits the floor.

Interesting to note is Kongo's significant grappling advantage in his bout with Mitrione. The Frenchman- not at all known for his wrestling acumen- is the most prolific takedown artist (as of late) on UFC 137's main card. If memory serves, a large portion of this statistic is thanks to the defensive shortcomings of Paul Buentello, but Mitrione hasn't exactly faced a murderers' row himself. Either way, Mitrione should be capable of fending off Kongo's takedowns and racking up points via the combination of his punching power and Kongo's suspect chin.

High expectations for Penn's performance are built largely upon his blistering knockout of Matt Hughes last year, not to mention Diaz's stubborn (and awesome) willingness to wade into high-volume exchanges. The two headlining fighters are poised to have the highest point outputs of anyone on the main card, and that's hardly taking into account the ground war they're very liable to get into. Next, I'll discuss whether or not their tempting point values are worth any fantasy buys.

Making Plays

From this point on, I will be employing and sharing a new method of making plays, based entirely on the MMA Salary Cap game's system of calculating an event's "Ideal Points."

Once an event is complete, each fighter's point output is divided by their dollar cost, yielding a points-per-dollar value. Ideally, the fighter with the highest score in this category would have contributed the maximum amount of plays to a fantasy player (remember that $150 is the most that may be spent on a fighter). This would continue down the list until the $500 budget was depleted. A fantasy player who mimicked these exact buys, whether by luck or true prescience, would receive the ideal number of points for the event.

My new system of making plays is nearly identical to the above process, except for replacing actual fantasy points with the expected fantasy points from the previous tables. For the last event's game, I tried spreading my buys out, making between one to five plays on all ten fighters. If I had used this new method, my earnings would have jumped from the low 500s to low 600s.

Cro Cop, whom I believe will earn 9.22 points whether in victory or defeat, has by far the highest expected points-per-dollar value, with an event-low cost of $7. His opponent, the highly-favored Nelson, is a terrible value, likely earning just about a half fantasy point per dollar spent. It's an interesting coincidence that the two opponents are on mirrored ends of the value spectrum; one that becomes downright strange when one realizes the same applies to all five pairings, all the way down to Diaz and Penn meeting in the middle. It is also without exception that the betting underdogs are the best deals for fantasy play. I'll be sure to check for this pattern in future events, as it may speak to how purchase costs are assigned to fighters.

At any rate, here are my buys for UFC 137's Salary Cap Game.

• 21 on Filipovic ($147)
• 15 on Roop ($150)
• 21 on Curran ($147)
• 5 on Kongo ($55)
• 62 plays total ($499)

Just to clarify, I fully expect all of these fighters to receive medium to heavy ass-beatings. I would not place any victory bets on them (although a savvier gambler might). However, since the salary cap game wholly disregards win/loss outcomes, I believe that the fighters I've chosen will likely kick up enough dust to justify the dollar amount spent on them.

To make your picks, sign up or log in at MMASalaryCap.com.

About fifteen minutes after I published my previous post on the new MMA Salary Cap game, I thought of a much better way to make stat-based play decisions. Calculating the spread between two fighters' stat differentials was interesting, but the results spoke more to who might win the fight than the quality of performance each contestant could produce. Since players collect points based on the latter and not at all the former, my initial approach needed some redirection to be of any particular use.

Expected Fantasy Points

The MMA Salary Cap Challenge assigns fantasy points to fighters in five categories. The number of points in each category is equal to the number of successful techniques executed per minute, inflated by a predetermined multiplier, as follows:

• Knockdowns: x25
• Significant Strikes: x2
• Takedowns: x10
• Submission Attempts: x20
• Ground Passes: x10

The goal of the game is to predict how all ten individual fighters might perform, based on how their strength and weaknesses match up across the five bouts.

A simple but effective way to estimate these performances would be to find the average of one fighter's technical success over his last three bouts and his upcoming opponent's past failure to avoid being damaged or controlled.

As an example, Frankie Edgar has averaged 6.67 fantasy points in the significant strikes category over his last three fights. This translates to 3.33 strikes landed per minute, increased by the game's x2 multiplier. Conversely, Gray Maynard has failed to avoid an average of 3.13 strikes per minute from his past three opponents, or 6.27 fantasy points after applying the multiplier. The average of these values yields the 6.47 fantasy points that Edgar could be expected to earn from significant strikes landed on Maynard. For his part, Maynard's past fantasy points in the same category amount to 4.38, while Edgar's past opponents averaged 4.75, for an expected score of 4.565. Based on the data provided, Edgar is more likely to outstrike Maynard in their approaching encounter, but that's only one piece of the puzzle. Here's my take on the provided FightMetric data:

With the expected fantasy points calculated, the most useful way to organize them is on a leaderboard of individual fighters, disregarding the match-ups.

Remember that this leaderboard doesn't reflect how "good" a fighter is. The fighters are arranged according to their rate of offensive output compared to the defensive shortcomings of their respective opponents. Further, fighters who have in the past excelled at performing techniques with high multipliers will easily climb the leaderboard, especially if they face an adversary with a known disadvantage against those maneuvers.

The last thing to mention is that making a play at the Salary Cap game has a different cost for each fighter, based on how well the game's design team expects that fighter to perform. For UFC 136, Melvin Guillard is the most expensive fighter (at $28 per play), little wonder. However, according to his and Joe Lauzon's recent fights, Lauzon is just barely more likely to outperform Guillard than vice versa. Even if Lauzon gets flattened in the third round, he has historically demonstrated the striking defense, takedown and grappling acumen, and frenetic rate of activity necessary to stay ahead on points (fantasy points, to be specific). All that being said, any amount of plays on Lauzon at $18 could very well turn out to be much wiser investments than the same amount on Guillard at $28.

(Edit: I added a cost/cost rank column. If a fighter's cost rank is higher than his fantasy points rank, I would consider him overpriced. That doesn't mean he wouldn't be worth a play, just that there may be higher gains elsewhere.)

Yesterday, the fine folks at FightMetric announced the launch of their innovative MMA fantasy game, The MMA Salary Cap Challenge, and they're not wasting any time getting started.

The game is awesomely close to fantasy games for stick and ball sports, owing to FightMetric's meticulous stat-collecting methods and extensive database. Fighters are assigned point values based on their last three performances, and the return on making a play is based on fighters' performances (by statistical category) at the event in question. No points are awarded or lost based on who actually wins the fight.

To help folks make their plays, FightMetric has provided the pertinent per-minute stats (knockdowns, significant strikes, takedowns, sub attempts, and ground passes) in four ways per match:

1. Fighter A's averages over his last three fights.
2. The averages of Fighter A's last three opponents.
3. Fighter B's averages over his last three fights.
4. The averages of Fighter B's last three opponents.

This is undoubtedly important and helpful information, but it will take a few more steps to yield any particularly illuminating data.

Differentials and Spreads

FightMetric already provides a striking differential leaderboard on their UFC records webpage, and applying this concept to the game's remaining four statistical categories should prove to be quite useful.

For example, the Salary Cap Challenge's website credits Frankie Edgar with landing an average of 3.33 significant strikes per minute in his last three contests. On the other hand, the opponents he faced during those three bouts were able to average 2.37 per minute against him. Edgar's differential is therefore +0.96, nearly a one strike per minute surplus. (This is a little more impressive when considering his having gone the distance in all three fights, not to mention their being title fights.)

His opponent, Gray Maynard, averaged 2.19 strikes/minute, while his last three adversaries outstruck him overall with a 3.13. His differential comes to -0.94, a deficit nearly the exact opposite of Edgar's.

The last stat to calculate is the spread between their differentials. When subtracting Maynard's from Edgar's, a positive result indicates Edgar's advantage, a negative one Maynard's. The spread in this case is +1.9, an ostensibly comfortable lead for Edgar.

The rest of this post will present the differentials and spreads across all categories, with some clarifying comments, for all five fights in the UFC 136 game. I'll leave it to one of our elder statesmen to make an official BE pool; in fact, I'm sure it will be up and running before this is even published!

Phan seems poised to get justice in his rematch against Garcia. The stats support the notion that both men are right at home on their feet. The difference lies in the absurd amount of strikes that land on Garcia during his exchanges. Phan also appears to enjoy a slight edge on the mat, something he may consider taking advantage of, given Garcia's recent submission loss via twister.

Guillard is the rightful favorite in this match, but that doesn't mean that Lauzon is going to get steamrolled. While Guillard's speed and power can't be denied, both men are talented strikers who give better than they get. Lauzon's best shot is clearly to get the fight to the ground, but it's hard to expect that from any Guillard opponents since the beating he put on Evan Dunham.

Stann leads Sonnen in three stats out of five, but the categories aren't necessarily of equal value. Sonnen's takedown ability (averaging one to two per round) could neutralize Stann's apparent striking advantage. Of course, Sonnen's greatest strength so frequently delivers him to the door of his most glaring weakness, a discipline (submission grappling) in which Stann has lately found some success.

I was initially surprised by Florian's pronounced underdog status leading into this event, but the FightMetric data may have exposed my bias. Unlike the combatants previously discussed, Aldo possesses not only the two striking advantages over Florian, but the edge in takedowns as well. This suggests that Aldo should be successful more often than not at keeping the fight standing, where he'll be free to pepper Florian with fast strikes and the occasional haymaker.

UFC 136 features the third match between Edgar and Maynard, and according to the numbers, it will likely be every bit as competitive as their previous bout. Maynard has a slight edge in three stats, while Edgar enjoys a sizeable lead in the significant strikes per minute differential. It's worth mentioning that much (if not all) of Maynard's knockdown advantage comes from his last scrap with Edgar, wherein he conducted a first round clinic on the champ. Edgar of course roared back to even the score by the final bell, which is reason enough to buy equal plays on both fighters- regardless of the winner, they're likely to have a similar number of fantasy points when the dust settles.

In Closing

I have a lot more ideas on how to translate the raw FightMetric stats into helpful data for playing the Salary Cap Challenge, and I'll try to get them polished and incorporate them into future installments. In the meantime, have fun playing, and make sure to give the guys at FightMetric your feedback.

The MMA Salary Cap Challenge

Hey guys, I just wanted to leave a quick note letting any inquiring minds know what's going on with tasormetrics.

Having a reasonably large pool of results to study, I took this uncommonly long lull between UFC events to step out of my mechanical data collecting to see how the small figures were adding up on the way to generating the more weighty ones.

Overall I'm pleased with the accuracy coughed up by such a basic consideration of only a tiny part of the momentum that fighters bring to each appointment, but I'm having an increasingly harder time reconciling with the some of the wildly inaccurate smaller values (even if they result in a correct prediction).

Combined with the arrival of Spring, a distant memory of a place called "outside", and an inexplicable desire to go to it, this dissatisfaction has prompted me to put the tasormetrics project on hold, pending the acquisition of some fresh air, and later, an overhaul of its core formulas.

Thanks to all who offered encouragement and suggestions; obviously I'll still be on BE all the time. See you all in the comments sections!

The question of age's influence on fight results is not an easy one to address. While it's commonly known that a certain age range reflects most fighters' athletic peaks, and it's easy to observe the decline of attributes such as speed and reflexes past a given age, the number of times a fighter has blown out birthday candles doesn't consistently reflect his strength, skill level, experience/composure, etc.

Regardless, it's a curiosity that merits investigation. Here is a surface-layer summary of the data made available to us by Fightmetric.com:

These preliminary figures suggest that youth holds the advantage. Across 2600 fights, the younger contestant prevailed 57% of the time. Youthful fighters also outperformed their seniors in all three methods of victory, excelling in striking-based finishes, but less so on judges' scorecards.

Some of the statistics fall under the "duh" category (obviously the average winner's age will be lower than the average loser's age when considering only bouts won by younger men), but most are at least somewhat thought-provoking. It's interesting to note, for example, that despite a tremendous total age range (16 to 51!), the overall average winner is still a full year younger than the overall average loser. Also notice the wider mean age difference across fights with more youthful victors. This could reflect yet another nod for fledgling fighters: perhaps it's easier for youngsters to win in spite of dramatic age disparities, while veterans tend to find success over more comparably seasoned opponents. Finally, there appears to be a faint correlation between bout length and winner's age (relative to that of his inferior). Younger winners clock their victories about a minute quicker than older winners, sensible, given the former's inclination toward (T)KO wins and the latter's toward decisions.

Though these figures are quite harmonious, they are too broad to be definitive. Consistent trends across narrower intervals would present a more convincing argument, and that's what I will attempt to uncover.

Continue reading this post »

More tasormetrics coming at you, on the heels of some spectacular performances at UFC 128. Brendan Schaub, Jim Miller, and Jon Jones all picked up emphatic knockout wins, while Nate Marquardt and Urijah Faber persevered through some frustrating moments en route to their victories via the judges' scorecards. Marquardt was threatened by a couple of sneaky guillotine setups from Dan Miller, but otherwise beat up his foe pretty convincingly. Eddie Wineland put on a hell of a first round against Faber, deftly resisting a half-dozen takedown attempts, and retaliating with a beautiful bodyslam that nobody seems to remember. In the end, Faber kept his wits about him and turned the tables to secure the nod. These results make no significant difference to tasormetrics prediction accuracy, which holds fast at 57% (27 correct out of 47).

I'm pulling the plug on the value bet segment of this series, and just barely too late to save it from shooting 0 for 4 last weekend. In the near future, I'll continue examining the bookmaker odds and their relation to ASR scores, but the focus will be shifted towards written analysis instead of black or white gambling recommendations. I haven't yet decided exactly how to implement this change, but it should be ready in time for UFC 129 next month. Hopefully this approach will continue to address anomalies, while no longer being a horrible embarrassing failure. The final accuracy for the old betting formula goes in the books at 10/31, or 32%. Awesome.

Chan Sung Jung	W	L	D	WLD	Win%	RSR	ASR	vs.	Leonard Garcia	W	L	D	WLD	Win%	RSR	ASR
	7	3	0	10	70%	(0.9539)	1.0293			4	3	1	8	56%	(1.1164)	0.8092
Choi	0	0	0	0	0%	0.0000			Pulver	3	3	0	6	50%	0.7071
Choi	1	0	0	1	100%	0.5774			Brown	8	0	0	8	100%	1.6330
Jo	2	0	0	2	100%	0.8165			Massouh	10	5	0	15	67%	1.4907
Omigawa	4	5	0	9	44%	0.7698			Gamburyan	3	3	0	6	50%	0.7071
Son	3	2	0	5	60%	0.7746			Roop	5	5	0	10	50%	0.9129
Ishiwatari	6	0	3	9	83%	1.4434			Jung	10	1	0	11	91%	1.7408
Kanehara	11	3	2	16	75%	1.7321			Hominick	4	1	0	5	80%	1.0328
Jaggers	11	5	0	16	69%	1.5877			Phan	3	3	0	6	50%	0.7071
Garcia	4	3	1	8	56%	0.9186
Roop	4	3	1	8	56%	0.9186

One man's shot at redemption becomes... a different man's shot at redemption, as an injured Nam Phan was replaced by Jung for a chance to rewrite history against Garcia. Jung was actually the victim of the less heinous robbery, but it will be interesting to see how his gameplan has changed, as he vowed to abandon the reckless "Korean Zombie" style after being beheaded by, of all people, George Roop.

Jung enjoys a theoretical advantage, courtesy of his relatively consistent performances before the controversial and not-so-controversial losses to Garcia and Roop, respectively. It's important to note that this disparity would be even wider, had the correct scores been awarded at the conclusion of Jung/Garcia I. It wasn't too long ago that Jung was proving himself to be an effective submission grappler, and it would benefit him to revisit and apply that skill set against Garcia. Jung wins if he clinches up, gets an ugly takedown, and keeps busy with ground and pound while looking for chokes. If he pulls it off, we can call him the Korean Giant Squid. Otherwise, we might be calling him "fired from the UFC."

DaMarques Johnson	W	L	D	WLD	Win%	RSR	ASR	vs.	Amir Sadollah	W	L	D	WLD	Win%	RSR	ASR
	7	3	0	10	70%	(1.0143)	1.0506			4	2	0	6	67%	(1.1354)	0.8763
Nitsuma	4	2	0	6	67%	0.9428			Dollaway	6	0	0	6	100%	1.4142
Lucambio	6	3	0	9	67%	1.1547			Hendricks	5	0	0	5	100%	1.2910
Vincent	0	0	0	0	0%	0.0000			Baroni	3	4	0	7	43%	0.6547
Olsen	4	1	0	5	80%	1.0328			Blackburn	5	1	1	7	79%	1.2002
Williams	6	1	0	7	86%	1.3093			Kim	5	0	2	7	86%	1.3093
Wilks	2	0	0	2	100%	0.8165			Sobotta	4	2	0	6	67%	0.9428
Garcia	7	1	0	8	88%	1.4289
Blackburn	5	1	1	7	79%	1.2002
Riddle	4	1	0	5	80%	1.0328
Guymon	6	2	0	8	75%	1.2247

Poor Amir Sadollah has been dragged to both ends of the interesting match-up spectrum, and now finds himself somewhere in the middle against TUF runner-up DaMarques Johnson. Johnson replaces his former conqueror James Wilks, who himself had already stepped in for Duane Ludwig. I hope Sadollah didn't spend too much of his fight camp studying tape.

Neither fighter has managed to cover much significant ground since graduating from the reality show. They've both lost winnable fights, and their respective opposition has been similarly uninspiring. The difference here is in finishing power. Sadollah has displayed some impressive but low-power striking technique, and his defensive grappling is less polished than it needs to be on the UFC stage. Johnson, on the other hand, has been posting a wide variety of stoppage victories. I'm rooting for Amir, and I see him performing admirably, but ultimately succumbing to constant pressure from Johnson.

Dan Hardy	W	L	D	WLD	Win%	RSR	ASR	vs.	Anthony Johnson	W	L	D	WLD	Win%	RSR	ASR
	6	2	0	8	75%	(1.4589)	1.1796			4	2	0	6	67%	(1.4027)	0.9403
Reiner	17	3	0	20	85%	2.1947			Speer	11	2	0	13	85%	1.7614
Weichel	11	3	0	14	79%	1.6973			Burns	6	1	0	7	86%	1.3093
Gono	5	1	0	6	83%	1.1785			Burns	7	1	0	8	88%	1.4289
Markham	8	3	0	11	73%	1.3926			Fioravanti	8	4	0	12	67%	1.3333
Davis	11	1	0	12	92%	1.8333			Yoshida	7	1	0	8	88%	1.4289
Swick	4	1	0	5	80%	1.0328			Koscheck	6	3	0	9	67%	1.1547
St. Pierre	6	1	0	7	86%	1.3093
Condit	4	1	0	5	80%	1.0328

Much is being made of Johnson's questionable ability to make weight for this fight, which is fair play, but that issue shouldn't detract from a relevant and intriguing bit of matchmaking. Surreal as it might now seem, Hardy was once a title challenger, and while he very well may never realize his championship aspirations, he remains a valuable thresher in the welterweight division. Hardy is, to a lesser extent, a standing version of Jon Fitch-- if you can outstrike him, you're doing something right. For his part, Johnson was on the verge of entry-level contendership before finding himself on the wrong end of a slapstick sideshow with Josh Koscheck.

A win for either fighter nicely sets up a bout with the last man to have defeated his respective opponent. That is to say, Hardy could earn a highly sensible fight against Koscheck, just as a dance with Carlos Condit would be perfect for Rumble if he emerges victorious. My money is on Hardy. Johnson has treated us to several truly bone-crunching knockouts, but against no one to write home about. Meanwhile, Hardy has hung in there with some real tough customers, and even in defeat, survived multiple deep submission attempts (sloppy? Let's see you get out of them) in his title fight with Georges St. Pierre. Hardy is slicker and more battle-tested, and he'll ride those advantages to a victory on Saturday.

Phil Davis	W	L	D	WLD	Win%	RSR	ASR	vs.	Rogerio Nogueira	W	L	D	WLD	Win%	RSR	ASR
	8	0	0	8	100%	(1.0795)	1.4225			6	1	0	7	86%	(1.2246)	1.2163
Chism	7	3	0	10	70%	1.2780			Dewees	4	5	0	9	44%	0.7698
Green	0	0	0	0	0%	0.0000			Rimbon	5	2	3	10	65%	1.1867
Cohens	4	2	0	6	67%	0.9428			Matyushenko	7	0	0	7	100%	1.5275
Baggett	2	1	0	3	67%	0.6667			Staring	7	4	0	11	64%	1.2185
Stann	6	2	0	8	75%	1.2247			Cane	6	1	0	7	86%	1.3093
Gustafsson	9	0	0	9	100%	1.7321			Brilz	4	1	0	5	80%	1.0328
Wallace	9	2	0	11	82%	1.5667			Bader	7	0	0	7	100%	1.5275
Boetsch	6	2	0	8	75%	1.2247

So much for a bounce-back fight. In the third and final main card fight to be reshuffled due to injury, Little Nog goes from disposing of the leftover husk of Tito Ortiz to standing on the tracks in front of an oncoming Phil Davis. It's hard for me to imagine Rogerio winning this one, but I can admit to being somewhat blinded by his last two performances, wherein he was soundly handled by gifted wrestlers. Davis is exactly that sort of fighter, and he's brandished that weapon menacingly in all of his UFC appointments to date. The strongest argument you could make for a Nogueira win is that he's a world of experience ahead of Davis. Nogueira is undeniably the toughest assignment of Mr. Wonderful's career, while the opposite doesn't necessarily ring true. For all his ability, Davis could very well fall prey to some trick Nogueira has up his sleeve from facing such a wide variety of opponents, all over the world.

To summarize, we are being presented with a fight in which: 1.) the odds-on favorite is a young collegiate wrestling champion, a quick study, and a rapidly rising star, and 2.) the underdog is a fearsome Pride alumnus and Brazilian jiu jitsu black belt who has the inclination (and the skills) to stand and exchange strikes with any opponent. A precedent for the outcome in this type of match-up has already been set, a whole five days ago. It's not a good one for people whose brothers are also professional mixed martial artists.

Continue reading this post »

So close to a clean sweep. Kampmann was just a little too badass for his own good in Kentucky, and my numbers took a hit for it. I didn't get $160,000 for my trouble, either. That said, UFC Live 3 was the best free card since, well, the WEC farewell show. Bowles/Page went down exactly as advertised, Dollaway cried (which makes sense: he did get spanked), and there was blood everywhere. Don't give Mr. White any ideas, but shit, I would have paid for it.

Despite just barely losing out on the main event, tasormetrics fight prediction accuracy remains on the rise, now totaling 24/42 or 57%. The only value bet pick last time was for Sakara, who came on strong but was ultimately crushinated by dead-serious prospect Chris Weidman. Value bet accuracy hobbles along at 10/27 or 37%.

Before we get to the breakdowns, I should mention that I'm going to try something new with my analysis. Looking back, I feel that I've been too meek and apologetic when the tasormetrics formulas roll out numbers that I personally disagree with, or even just when they go against the grain. I'm officially changing this narrative. Tasormetrics is an imperfect system, but it's a system nonetheless. It analyzes the same set of data for every fighter and only produces figures solely based on that data. It's not some drunken idiot who exclusively roots for whichever fighter is the more like-minded asshole. So, from now on, I'm going to make a stronger effort to hitch my wagon to the tasormetrics favorite in each fight before I write my analysis. I'll try to mention any interesting ways in which the data influences the picking of a horse, and if an obvious underdog sneaks in, I'll focus on how he could reasonably win, not why he won't.

Brendan Schaub	W	L	D	WLD	Win%	RSR	ASR	vs.	Mirko Filipovic	W	L	D	WLD	Win%	RSR	ASR
	7	1	0	8	88%	(0.7116)	1.0833			4	2	1	7	64%	(0.9987)	0.8523
Lester	0	0	0	0	0%	0.0000			Overeem	7	5	0	12	58%	1.1667
Curtis	1	1	0	2	50%	0.4082			Choi	1	1	0	2	50%	0.4082
Rozman	0	6	0	6	0%	0.0000			Al-Turk	4	2	0	6	67%	0.9428
Spalevic	2	3	1	6	42%	0.5893			Dos Santos	7	1	0	8	88%	1.4289
Nelson	7	4	0	11	64%	1.2185			Perosh	5	2	0	7	71%	1.0911
Gormley	6	1	0	7	86%	1.3093			Barry	5	1	0	6	83%	1.1785
Tuchscherer	8	2	0	10	80%	1.4606			Mir	3	2	0	5	60%	0.7746
Gonzaga	3	3	0	6	50%	0.7071

This bout has all the makings of one of those silently sad changing of the guard matches. Our beloved Croatian Cop ostensibly already passed the torch to Junior Dos Santos (truly none more deserving), but he apparently has a spare that's up for grabs. He tried as hard as he could to give it to Pat Barry, and Mir was no young lion when they met, so now it's Schaub's turn to take a swipe at it. After being smashed to pieces by Roy Nelson, Schaub turned it around quick and had a great 2010. With victories in three very sensible match-ups, including his most recent over resident Cro Cop destroyer Gabe Gonzaga, Schaub should be ready to win this contest.

It's no mystery that Mirko has recently faced the stiffer (though not exactly sterling) competition, but much of this gap results from the inclusion of the earliest bouts from Schaub's young career. The Hybrid's current UFC win streak is the more significant factor and gives him a marked advantage in momentum over the aging killer. I don't know that his hands are crisp enough to force a stoppage, but Schaub's speed should allow him to land a strike or two and get out of the danger zone enough times to lock up a decision win. If it pans out as such, Schaub should be the last up-and-comer to further his career via Cro Cop. Give him a couple chances to retire on a senior circuit win, or let the man fish in peace.

Dan Miller	W	L	D	WLD	Win%	RSR	ASR	vs.	Nate Marquardt	W	L	D	WLD	Win%	RSR	ASR
	6	3	0	9	67%	(1.3899)	1.0730			4	3	0	7	57%	(1.3714)	0.8421
McGivern	10	5	0	15	67%	1.4907			Leites	8	1	0	9	89%	1.5396
Kimmons	17	2	0	19	89%	2.2517			Kampmann	6	0	0	6	100%	1.4142
Horwich	9	6	0	15	60%	1.3416			Gouveia	6	2	0	8	75%	1.2247
Rosholt	5	0	0	5	100%	1.2910			Maia	9	0	0	9	100%	1.7321
Sonnen	7	3	0	10	70%	1.2780			Sonnen	7	2	0	9	78%	1.3472
Maia	6	1	0	7	86%	1.3093			Palhares	6	1	0	7	86%	1.3093
Bisping	5	3	0	8	63%	1.0206			Okami	4	1	0	5	80%	1.0328
Salter	5	1	0	6	83%	1.1785
Doerksen	7	2	0	9	78%	1.3472

Man, it would have been a lot easier to sell a Marquardt win over Akiyama than for Miller over Marquardt. I'm going to take the easy way out one last time, and take this opportunity to reiterate the specific factor that ASR truly attempts to measure: momentum. The results returned by the formula are meant not to champion Miller as the better fighter of the two, but to show that, given their recent performances, Miller is rolling faster along his newly-plotted collision course with Marquardt than vice versa. This is a fair claim to make when comparing a three-year 6-3 record to 4-3 over the same time span, but there's something not quite right in this particular set of figures: no way has Miller tilled a harder patch than Nate the Great.

This discrepancy is mostly the result of a single anomalous opponent on Miller's resume: senior executive can-crusher Rob Kimmons. In the three years before he met (and was, for the record, handled by) Miller, Kimmons patched together an insane 17 wins in 19 fights. If he himself was fighting on Saturday, and we were calculating his Adjusted Strength of Record, his trophy case of jobbers would be blown wide open. Unfortunately, toward the goal of finding Miller's score, we only reference Kimmons' shallower, and overinflated, Raw Strength of Record. All told, a more accurate reflection of Kimmons' "accomplishments" still wouldn't give the overall advantage back to Marquardt, and this infrequent but ultimately unavoidable flaw is the reason that Opponents' RSR is weighted less than the other contributing variables in the ASR formula.

Jim Miller	W	L	D	WLD	Win%	RSR	ASR	vs.	Kamal Shalorus	W	L	D	WLD	Win%	RSR	ASR
	9	1	0	10	90%	(1.3798)	1.4967			7	0	2	9	89%	(1.0758)	1.3136
Palaszewski	15	4	0	19	79%	1.9868			Davis	10	4	0	14	71%	1.5430
Baron	9	2	0	11	82%	1.5667			Jones	1	2	0	3	33%	0.3333
Wiman	4	2	0	6	67%	0.9428			Evans	0	0	0	0	0%	0.0000
Maynard	6	0	1	7	93%	1.4184			Bronzoulis	7	1	0	8	88%	1.4289
Danzig	4	4	0	8	50%	0.8165			Miller	2	0	0	2	100%	0.8165
Lopez	12	1	0	13	92%	1.9215			Kerr	5	0	0	5	100%	1.2910
Ludwig	5	3	0	8	63%	1.0206			Jansen	13	0	0	13	100%	2.0817
Bocek	5	2	0	7	71%	1.0911			Varner	4	1	0	5	80%	1.0328
Tibau	4	3	0	7	57%	0.8729			Palaszewski	6	3	0	9	67%	1.1547
Oliveira	14	0	0	14	100%	2.1602

In the "Mario" role of the Super Miller Bros, Jim Miller has been hard at work, putting together a respectable two-year win streak under MMA's big top. His tapout victory over undefeated submission stylist Charles Oliveira is a big bright feather in his cap, and betrays an ever-evolving skill set. Sticking with the awesome video game metaphors, Shalorus would have to be Starman from Pro Wrestling: it's only a mild stretch to call him a space-traveling super wrestler, and his birthdate is literally unknown.

Unfortunately for Shalorus, his enigmatic shroud ends there, as ample WEC footage exists to light Miller's path to victory. One really need look no further than Shalorus's gift draw with Jamie Varner to see that, for all his base style talent, he's the perfect victim for Miller's Swiss Army knife attack. With the recent loss posted by George Sotiropoulos, a dominant win here legitimizes Miller's status as #1 lightweight contender, ahead of Anthony Pettis, despite promises made. If Clay Guida bests Pettis (entirely possible), and there's no time for a Miller/Guida eliminator (presumably the case), then Slim Jim deserves his swing at the survivor of Edgar vs. Maynard III.

Eddie Wineland	W	L	D	WLD	Win%	RSR	ASR	vs.	Urijah Faber	W	L	D	WLD	Win%	RSR	ASR
	5	1	0	6	83%	(1.1835)	1.1106			4	3	0	7	57%	(1.2564)	0.8178
Choate	7	4	0	11	64%	1.2185			Pulver	3	3	0	6	50%	0.7071
Yahya	6	3	0	9	67%	1.1547			Brown	7	1	0	8	88%	1.4289
Tapia	3	2	0	5	60%	0.7746			Pulver	3	4	0	7	43%	0.6547
Roop	6	4	0	10	60%	1.0954			Brown	9	0	0	9	100%	1.7321
Campuzano	7	1	0	8	88%	1.4289			Assuncao	6	0	0	6	100%	1.4142
Stone	7	1	0	8	88%	1.4289			Aldo	8	0	0	8	100%	1.6330
									Mizugaki	6	2	0	8	75%	1.2247

This fight is getting so short-changed on hype, it's not even funny. It appears that, other than myself, no one is even aware of it enough to lament how thoroughly it's being neglected. What we have here is presumably a #1 contendership bout between two former WEC champions with a mile-long highlight reel each. Faber, though historically the more dominant fighter, has posted a checkered set of recent performances, while Wineland has been roaring back from the bottom of the ladder after losing his WEC title to Chase Beebe in a textbook barn-burner four years ago. This fight matters, both guys are coming off of downright obscene stoppage victories, and they're going to get after each other like rabid wild dogs. Get fucking stoked.

The numbers for this bout are elementary: Faber has squared off against the stiffer competition, but the edge goes to Wineland, on the strength of his current four-fight win streak. However, when considering non-statistical factors, Wineland turns up as a clear cut underdog. So how can he win this bout? We saw Jose Aldo absolutely cripple Faber with leg kicks, but as far as I know, they're not a staple of Wineland's arsenal. Unless he's secretly mastered them in preparation for this fight, Wineland will want to stay light on his feet and keep Faber at range with fast, straight punches. Both men are going to be aggressive, but pressing forward is only going to work for Wineland if he does it with enough gusto to make Faber back up. Every time Faber locks up a clinch, to say nothing of a takedown, Eddie's shot to win will slip a little further away. On the other hand, if Wineland has the defense and discipline to stick and move for the first two rounds, and keep the power punches in his back pocket until the third, he very well may pull off the upset and catapult himself to the highest stage of his career.

Jon Jones	W	L	D	WLD	Win%	RSR	ASR	vs.	Mauricio Rua	W	L	D	WLD	Win%	RSR	ASR
	12	1	0	13	92%	(0.7035)	1.3385			3	1	0	4	75%	(1.0143)	0.8294
Bernard	0	1	0	1	0%	0.0000			Coleman	1	1	0	2	50%	0.4082
Eduardo	0	1	0	1	0%	0.0000			Liddell	3	3	0	6	50%	0.7071
Pina	0	0	0	0	0%	0.0000			Machida	7	0	0	7	100%	1.5275
Verrett	1	2	0	3	33%	0.3333			Machida	6	0	0	6	100%	1.4142
Porter	2	0	0	2	100%	0.8165
Gabin	2	1	0	3	67%	0.6667
Gusmao	5	0	0	5	100%	1.2910
Bonnar	3	2	0	5	60%	0.7746
O'Brien	4	2	0	6	67%	0.9428
Hamill	4	2	0	6	67%	0.9428
Vera	3	4	0	7	43%	0.6547
Matyushenko	6	1	0	7	86%	1.3093
Bader	6	0	0	6	100%	1.4142

Sweet baby Jesus in a tuxedo t-shirt. How often do fights like this come together? And in lieu of the intended match-up, no less. I'd like to sidestep the old guard vs. wave of the future motif for just a moment and say this: when I try to imagine this fight ending by knockout, I can see either guy crumpled on the mat, lifeless. When I picture a submission finish, I see, with identical clarity, alternating scenes of Shogun's eyes bulging as his face turns a deep and horrible purple, or Jones on both knees, dejected, nursing a severely abused and possibly broken arm, while his Brazilian assailant struts around the perimeter of the cage.  I can easily envision them awaiting a judges' decision in their post-fight gear, logo-emblazoned chests heaving, the damage or lack thereof on their faces foreshadowing any possible scorecards from 50-45 to 48-47, in favor of either contestant.

Win-loss statistics, journalists, bloggers, and oddsmakers sternly favor Jones. The prevailing soundbite from hardcore fans is: "Look, no one loves Shogun more than I do, but..." But, indeed. The guilt-stricken among us rage against the seemingly foregone conclusion, snarling in defense of the beloved champion. But reality as I see it is less deserving of such polarization. The overlooked truth is that this is one of those rare and wonderful fights wherein not only is anything possible, but nearly everything is probable. We say "I can't wait for this fight" because, without realizing it, we've been waiting for it since before it was announced; maybe since before Jones was even a blip on the radar. I'm pleased to report that the wait is nearly over, and whether a storied career marches on through another incredible chapter, or a new legacy is born, Saturday night will be a very good night to be an MMA fan.

So... this might be the death knell for trying to identify value bets. This time around, the ramshackle system I put in place has happily encouraged wagers on the four biggest underdogs, while advising against the safest action of the night (a play on Shogun). To date, ASR values are better than a wild guess at predicting a fight winner, but the typical margin between two fighters' scores just doesn't line up ideally with the betting lines. We'll see if any Brian Ebersole-type shenanigans unfold, but otherwise it might be time to put this segment to bed.

(odds via BetUS.com)

Continue reading this post »

Promoted by Mike Fagan.

Here's your winner of the inaugural FightMetric Research Challenge. Congratulations to PistonHyundai. Additional congratulations go out to everyone who participated. Rami and I are both very happy with the turnout, though that's to be expected with the Bloody Elbow community.

What's next? We're looking for another question for the community to answer in April. What do you want to see? Leave your suggestions in the comments.

Ok, here goes nothing.

So far, there's been some great legwork done on the correlation between reach advantage and fight results. Here's a brief recap before I introduce my analysis:

When you consider MMA's dynamism as a sport, and facts like 1.) submission victories not necessarily resulting from ground-heavy fights and 2.) knockouts not being exclusive to stand-up wars, it's reasonable to say that these figures do indeed suggest an advantage for rangier fighters. At the very least, the results merit a closer look.

I feel that it would be very useful to examine the distribution of win percentages across the spectrum of reach differences. A direct relationship between win rate and the amount of reach "advantage" would lend more credibility to the idea of reach as a literal advantage. An inverse relationship would obviously suggest the opposite, although it would still highlight reach as a (weird) influence on fight results. An undecipherable cluster of data points would damage the reach advantage theory. Accordingly, I've prepared a series of histograms to explore the possibilities of such relationships.

How to Read These Graphs (Important)

It's no lie: these graphs are a little clunky. But I couldn't think of a better way to express the necessary data, and I promise that just a bit of explanation will make things significantly clearer:

1. The X axis tracks the difference in inches between the fighters' reach. "1" includes all differences greater than 0 and equal to 1, etc.
2. The last X value includes all reach differences greater than 8 inches because beyond that, there just weren't enough fights per inch to generate defensible percentages.
3. The Y axis only pertains to the orange/red longer reach win% bars. The blue bars represent the total number of fights featuring each respective reach difference, and are only meant to convey the relative amounts of data that were available for calculating the percentages.

Let's jump right in with the graph for the whole population.

DEPTH LEVEL 1

First things first: the most obvious trend is portrayed by the blue bars. To wit, as the reach difference increases, the number of fights in the data pool consistently decreases. This is a natural result of weight class stratification and probably matchmakers' preference for pairing similarly-sized fighters within those classes. 331 fights commenced with a 1 inch or shorter difference, while only 20 fights represent freakish differences of 8 or more inches. The important thing to take away is that shorter blue bars indicate less reliable orange/red win% bars.

Continue reading this post »

Quick turnaround from a bizarre UFC 127, which included the second inconclusive main event judges' decision of a very young 2011, as well as a couple of fairly shocking upsets. But I guess that's why we watch. In any case, my stat-based features return under what I hope to be a catchier banner. I've also whipped up a brief (-as-possible) primer on this particular brand of number-crunching, which I'll include after the break from now on, for new readers or those who still aren't quite sure what I'm actually trying to do here.

As far as updated tallies go, Adjusted Strength of Record-based fight predictions scored two out of four, barely reducing accuracy to 21/38 or 55%. Value bet picks fared identically, and of course, the two correct returns were the ones that I mocked as impossible. All the same, those bizarre left fielders kept value bet accuracy alive on life support, bringing the total to 10/26 or 38%.

Brian Bowles	W	L	D	WLD	Win%	RSR	ASR	vs.	Damacio Page	W	L	D	WLD	Win%	RSR	ASR
	3	1	0	4	75%	(1.4852)	0.9418			2	2	0	4	50%	(1.1969)	0.5843
Page	10	3	0	13	77%	1.6013			Bowles	5	0	0	5	100%	1.2910
Ribeiro	7	1	0	8	88%	1.4289			Galvao	3	1	1	5	70%	0.9037
Torres	9	0	0	9	100%	1.7321			Campuzano	6	0	0	6	100%	1.4142
Cruz	5	1	0	6	83%	1.1785			Johnson	5	1	0	6	83%	1.1785

This is one of the most exciting fights to come out of the WEC-UFC merger so far, and it's not even a new match-up. Two and a half years ago, Bowles and Page met each other in our favorite little blue cage, both coming off important wins. Bowles had more to gain than the longer-toothed Page, and he indeed emerged victorious, by way of a standing guillotine in the first round. Since then, both men have traveled nearly identical paths, with one submission and one hellacious knockout victory apiece, followed by losses in their respective last fights. Former champion Bowles has the clear advantage, especially with Page having lost his last contest in the same way that Bowles had previously bested him, but the true appeal of this fight lies in its very serious potential to end in a double knockout- these guys both have an extreme case of dynamite hands.

Chris Weidman	W	L	D	WLD	Win%	RSR	ASR	vs.	Alessio Sakara	W	L	D	WLD	Win%	RSR	ASR
	4	0	0	4	100%	(0.7611)	1.0049			3	0	0	3	100%	(1.0052)	1.0017
Lopes	0	0	0	0	0%	0.0000			Vedepo	7	2	0	9	78%	1.3472
Stewart	3	1	0	4	75%	0.8660			Leites	5	2	0	7	71%	1.0911
Hall	3	0	0	3	100%	1.0000			Irvin	2	2	0	4	50%	0.5774
Arraujo	5	1	0	6	83%	1.1785

An interesting short-notice pairing boils down to a dead heat according to the ASR formula. Or to be more precise, UFC debutant Chris Weidman actually leads vet Alessio Sakara by about eight hundredths of a percent. Not at all significant, and it's not hard to imagine why. Sakara is seasoned and talented, but inconsistent. He doesn't always win when he's supposed to, which isn't a great distinction to bring into a bout against a highly regarded prospect. Weidman was a D1 All-American wrestler at Hofstra University, and brought that pedigree to the Serra-Longo Fight Team, where he's been making impressive strides, including capturing the middleweight title in New Jersey's Ring of Combat promotion. To his credit, Sakara has been compiling a respectable resume at 185, but it remains to be seen if he'll blossom into any sort of worldbeater. As a still-young veteran, it's tempting to give him the edge, but I think there's truth in the ASR scores: Weidman has momentum on his side... barely.

Mark Munoz	W	L	D	WLD	Win%	RSR	ASR	vs.	CB Dollaway	W	L	D	WLD	Win%	RSR	ASR
	6	2	0	8	75%	(1.1122)	1.0776			5	2	0	7	71%	(1.0848)	0.9735
Grigsby	12	3	0	15	80%	1.7889			Sadollah	0	0	0	0	0%	0.0000
Barros	1	0	0	1	100%	0.5774			Taylor	6	2	0	8	75%	1.2247
Hamill	5	2	0	7	71%	1.0911			Massenzio	9	1	0	10	90%	1.6432
Catone	7	1	0	8	88%	1.4289			Lawlor	5	1	1	7	79%	1.2002
Jensen	5	4	0	9	56%	0.9623			Silva	5	1	0	6	83%	1.1785
Grove	3	3	0	6	50%	0.7071			Reljic	3	0	0	3	100%	1.0000
Okami	4	1	0	5	80%	1.0328			Doerksen	7	2	0	9	78%	1.3472
Simpson	6	1	0	7	86%	1.3093

This is a relevant fight, but I don't like Dollaway and I just want Munoz to smash his face.

Martin Kampmann	W	L	D	WLD	Win%	RSR	ASR	vs.	Diego Sanchez	W	L	D	WLD	Win%	RSR	ASR
	5	3	0	8	63%	(1.4029)	0.9702			4	2	0	6	67%	(1.3069)	0.9183
Rivera	6	2	0	8	75%	1.2247			Fioravanti	9	3	0	12	75%	1.5000
Marquardt	4	2	0	6	67%	0.9428			Stevenson	5	3	0	8	63%	1.0206
Barros	5	0	0	5	100%	1.2910			Guida	6	4	0	10	60%	1.0954
Condit	9	2	0	11	82%	1.5667			Penn	4	1	0	5	80%	1.0328
Daley	10	3	0	13	77%	1.6013			Hathaway	9	0	0	9	100%	1.7321
Volkmann	9	1	0	10	90%	1.6432			Thiago	8	2	0	10	80%	1.4606
Thiago	8	1	0	9	89%	1.5396
Shields	6	0	0	6	100%	1.4142

Great style match-up to top the main card- I love this fight and I think it could easily co-headline a PPV show. Kampmann and Sanchez have both demonstrated excellence over every inch of the cage, so unlike many fights with tons of potential but also the risk of devolving into an Ultimate Staring/Hugging Championship situation, I have every reason to believe that this one will deliver no matter where it goes. Sanchez is already showing off a fresh coat of paint from Greg Jackson's camp, as he can run from his old nickname for whatever silly reason he wants, but he looked every bit the Nightmare against Paulo Thiago. Kampmann will enter the cage with tight submissions and tighter kickboxing, and you have to think that he'll be equally happy to stand and trade or to try catching Sanchez with the guillotine. If I'm picturing it correctly, Kampmann should have the size advantage, and all else equal (or just weird and difficult to compare), the bigger guy should bring home the win.

Betting

The above lines are courtesy of Bookmaker, since BetUS didn't publish odds for the full main card, and both websites tend to have similar trending and overround. Most of the ASR odds run too close to the book odds to justify any action, but a play on Alessio Sakara looks pretty reasonable. As mentioned earlier, Sakara and Weidman are neck-and-neck in the ASR column, so (theoretically) there's no glaring reason to think that the oddsmaker underdog couldn't pull through.

Continue reading this post »

Recap

UFC 126 was a hit at the box office and delivered with some very fan-friendly highlights, but it was a little rough on the ASR-based predictions. Victor accuracy actually climbed a tiny bit, as Torres, Jones, and Silva all won as expected. Rocha was the bookmaker underdog but favored to win by the ASR formula, and he almost pulled it off, but Jake Ellenberger survived a hairy first round and brought home a calm and collected win from the judges (well the two that had their glasses on, anyway). Forrest Griffin looked very sharp after the longest layoff of his UFC career; I thought Franklin did a great job of defending strikes from his back, but great defense doesn't win fights. Fight predictions stand at 19/34 or 56%. (I removed the extraneous half-point originally contributed by the Edgar/Maynard draw.)

It was value bet accuracy that once again took a real beating, correctly identifying exactly zero out of three worthwhile long shots. It's getting hard to defend the collection of this data, but I never had any particularly high expectations for it. If nothing else, I'm starting to get a feel for why it fails when indeed it does. Heading into UFC 127, betting pick accuracy has sunk to 8/22, a mere 36%.

Chris Camozzi	W	L	D	WLD	Win%	RSR	ASR	vs.	Kyle Noke	W	L	D	WLD	Win%	RSR	ASR
	7	2	0	9	78%	(1.1012)	1.1584			4	0	0	4	100%	(0.9558)	1.0842
Lewis	4	3	0	7	57%	0.8729			Gouaida	0	1	0	1	0%	0.0000
Rossborough	6	8	0	14	43%	0.9258			Uscola	6	6	0	12	50%	1.0000
Duff	2	0	0	2	100%	0.8165			Bryant	9	0	0	9	100%	1.7321
Moreno	12	8	0	20	60%	1.5492			Kimmons	5	2	0	7	71%	1.0911
Taylor	9	3	0	12	75%	1.5000
Brudigan	3	3	0	6	50%	0.7071
Reiner	10	8	0	18	56%	1.3608
Hammortree	5	1	0	6	83%	1.1785
Yang	3	0	0	3	100%	1.0000

A nice little scrap between Ultimate Fighter alumni gets us started as Australia's own Kyle Noke is, for the third time, pitted against a season 11 housemate. Both men are perfect in their last four fights, though Camozzi's victories are significantly fresher. I have to admit that I'm confused by the bookmaker odds for this bout, which list Noke as a comfortable favorite. Noke does have home court advantage, and he is two for two with stoppages in the UFC (compared to Camozzi's pair of decision victories), but I have to think he's being overvalued here. Camozzi has been keeping busy and performing pretty consistently for a C-level fighter, and he's carved out his career on the more competitive North American stage. Welcome to Upset City.

Brian Ebersole	W	L	D	WLD	Win%	RSR	ASR	vs.	Chris Lytle	W	L	D	WLD	Win%	RSR	ASR
	8	1	0	9	89%	(0.8043)	1.1922			5	2	0	7	71%	(1.1765)	1.0001
Henderson	2	0	0	2	100%	0.8165			Koscheck	7	2	0	9	78%	1.3472
Lombard	15	2	1	18	86%	2.1093			Taylor	4	3	1	8	56%	0.9186
Nix	5	0	0	5	100%	1.2910			Davis	12	1	0	13	92%	1.9215
Forrester	0	0	0	0	0%	0.0000			Burns	7	2	0	9	78%	1.3472
Bradney	5	1	1	7	79%	1.2002			Foster	7	2	0	9	78%	1.3472
Schaffa	0	2	0	2	0%	0.0000			Brown	5	3	0	8	63%	1.0206
Staden	4	0	0	4	100%	1.1547			Serra	1	2	0	3	33%	0.3333
Newton	2	1	0	3	67%	0.6667
Robertson	0	0	0	0	0%	0.0000

And just as I was getting cocky, the formula drops this little surprise in my lap. It's not hard to see why the numbers favor Ebersole. Over the same time period, he's racked up more fights, more wins, and fewer losses than Lytle. Regardless, the graph hints at one telltale sign of Lytle's impending victory: all of the guys he fights are badasses. In a situation like this, the Adjusted Strength of Record formula isn't deep enough to calculate what we can plainly see: Lytle's opponents have been more successful against UFC competition than Ebersole's have been against non-UFC competition. Sure, Ebersole could maybe, maybe, lay his hands on Lytle enough to open a cut and get the fight stopped, but- and I was talking about this recently with a buddy- when you find yourself making obligatory statements like that, you've automatically indicated who you really think is going to win.

George Sotiropoulos	W	L	D	WLD	Win%	RSR	ASR	vs.	Dennis Siver	W	L	D	WLD	Win%	RSR	ASR
	6	0	0	6	100%	(1.0852)	1.2948			6	2	0	8	75%	(0.9993)	1.0398
Mitichyan	3	0	0	3	100%	1.0000			Guillard	4	3	0	7	57%	0.8729
Roop	7	3	0	10	70%	1.2780			Jacquier	3	5	0	8	38%	0.6124
Dent	7	2	0	9	78%	1.3472			Mohr	5	4	0	9	56%	0.9623
Stevenson	5	3	0	8	63%	1.0206			Hartt	4	1	0	5	80%	1.0328
Pellegrino	5	2	0	7	71%	1.0911			Kelly	6	1	0	7	86%	1.3093
Lauzon	3	2	0	5	60%	0.7746			Pearson	10	2	0	12	83%	1.6667
									Fisher	3	2	0	5	60%	0.7746
									Winner	3	3	1	7	50%	0.7638

Alright, here we go with some love for the home team. I don't harbor any ill will at all toward the Australian MMA scene, but it is young as hell and just isn't producing many interesting prospects yet. I have every belief that it will someday soon, especially with so many talented judokas nearby that might be willing to help out. Until then, George Sotiropoulos stands alone as the premier talent of his nation. The last time we saw Dennis Siver fight, I was complaining that, coming off a win, he had been given a woefully outmatched opponent. This time we're seeing a problem of the opposite kind. Siver always shows up ready to rock, but it's too early for him to be fighting a bona fide title contender. Look for Sotiropoulos to further extend his UFC win streak, tying with Georges St. Pierre (!) if he does indeed emerge victorious this weekend.

Michael Bisping	W	L	D	WLD	Win%	RSR	ASR	vs.	Jorge Rivera	W	L	D	WLD	Win%	RSR	ASR
	6	2	0	8	75%	(1.0264)	1.0491			3	1	0	4	75%	(1.3751)	0.9179
McCarthy	3	2	0	5	60%	0.7746			Kampmann	7	0	0	7	100%	1.5275
Day	15	2	0	17	88%	2.1004			Osterneck	5	1	0	6	83%	1.1785
Leben	5	3	0	8	63%	1.0206			Kimmons	11	2	0	13	85%	1.7614
Henderson	4	3	0	7	57%	0.8729			Quarry	4	1	0	5	80%	1.0328
Kang	4	3	0	7	57%	0.8729
Silva	1	4	0	5	20%	0.2582
Miller	6	2	1	9	72%	1.2509
Akiyama	4	1	1	6	75%	1.0607

Man, these guys have been kicking up some dust. I usually eat pre-fight antics right up, but this attempt kind of fell flat for me. I don't know what it is, but every time Bisping fights, I care less and less. I was much more excited when Rivera was supposed to throw down with Sakara, and Bisping is a better fighter than either of them. Win or lose, I would really like to see Bisping show some serious aggression. Not in a stupid, reckless way, but to come forward, and have any contingency plan besides backpedaling if he catches the worse of an exchange. I want him to keep his eyes open for opportunities to clinch or grab a leg. We know he has good ground and pound; he should give it more space in his gameplan. It's all easier said than done, but that's the kind of stuff that's going to make this fight interesting. Rivera's a tough dude, and he has ways to win, but they have more to do with what Bisping does than what he does himself.

Jon Fitch	W	L	D	WLD	Win%	RSR	ASR	vs.	BJ Penn	W	L	D	WLD	Win%	RSR	ASR
	6	1	0	7	86%	(1.4122)	1.2755			4	3	0	7	57%	(1.1249)	0.7882
Wilson	11	2	0	13	85%	1.7614			Sherk	4	1	0	5	80%	1.0328
St. Pierre	7	1	0	8	88%	1.4289			St. Pierre	6	1	0	7	86%	1.3093
Gono	5	2	0	7	71%	1.0911			Florian	6	1	0	7	86%	1.3093
Thiago	10	0	0	10	100%	1.8257			Sanchez	5	2	0	7	71%	1.0911
Pierce	9	2	0	11	82%	1.5667			Edgar	5	1	0	6	83%	1.1785
Saunders	4	1	0	5	80%	1.0328			Edgar	5	1	0	6	83%	1.1785
Alves	5	1	0	6	83%	1.1785			Hughes	3	2	0	5	60%	0.7746

A very interesting contrast of well, nearly everything, awaits us at the height of this weekend's fight card. Penn and Fitch each possess different styles of grappling, different physical advantages, and swathed wildly dissimilar paths to their impending face-off. Penn is a living legend, still competitive after mixed success in title fight after title fight after title fight. Fitch, after 13 wins in 14 UFC fights, has earned at least the respect of the diehard fanbase, though not quite so much from the casual audience.

From a distance, this seems like a close fight, based simply on how unbelievably skilled these guys are. A closer look, however, exposes a pretty bad hand for Penn. Even with some of the most highly lauded takedown defense in the sport, Penn was dumped to the mat several times by pint-sized Frankie Edgar. Meanwhile, Fitch's takedown acumen could probably earn him a winning record at middleweight. On top of that, Fitch's defensive grappling is a notch above Edgar's. For all his ability, I can't see Penn winning a ground war. His best shot to win lives or dies with his ability to outbox Fitch without overcommitting and getting put on his ass. I'd love for Penn to go in there and prove to me that it's not that open and shut, but we know him pretty well by now. The only real variable in a Penn fight these days is how prepared he'll be on fight night, and even at his best, he'll seriously have his work cut out for him against Jon Fitch.

WARNING: DO NOT ACTUALLY BET ON EBERSOLE OR SIVER

THAT IS ALL

(Odds from BetUS.com)

A discussion about weight class optimization is something no one wants to see. 90% of people have no idea what it means. Of the remainder, 90% don't care, and those that do tend to have conversations of this variety:

The weight classes are boned! We need a new division for every 10 pounds, so it'll be more fair to all the fighters, like in boxing.

No, that many weight classes would completely ruin the legitimacy of championship titles... like in boxing.

Well at least add cruiserweight. The gap between heavyweight and light heavyweight is nuts!

That's the last place in the world that needs a new weight class! Heavyweight is too shallow to begin with. Both divisions would be dead before they hit the ground.

But MMA is growing so fast, the empty spots will fill right up with huge collegiate wrestlers and football players. Even giant 205ers could move up to cruiserweight for the speed and conditioning advantages.

NONONONO

YESYESYESYES

The fact of the matter is that those are all valid points, and there are probably a dozen more on each side of the argument. This is good. This means that today's weight classes are getting the job done just fine, and their current standing isn't a cancer on the sport in the vein of judging incompetence. But make no mistake: things are changing. Every year there are more promotions, more events, more dollars, and more fighters, and our cobbled-together weight class system might surprise us one day by actually needing a facelift. Wouldn't it be nice to have a solution ready and waiting?

My biggest issue with the current weight divisions is the line of thinking that appears to have most influenced their definitions. I imagine it started well enough: "Let's add a few more weight classes to even things out. The differences between classes should probably start small for the smaller guys and can get bigger for the bigger guys." That makes perfect sense. But somewhere along the line, someone insisted that weight class limits be defined exclusively by multiples of 5, presumably for the "benefit" of casual observers. This is the final point of departure for my logically-inclined side.

Let me be clear: the multiple-of-5 weight limits are perfectly nice. I like them just fine. But when I was learning this sport, they didn't really make it any easier to remember the divisions. If you're good at remembering, you will, and if you're not, you won't. That's why they show the name and limit of the division before every single fight. The reason I harp on this is that the current distribution of weight limits is good, but not ideal. In a make or break situation, why should it be easier for the average 155er to move down a class than it is for a 205er? That's a 6.5% versus 9.8% difference in necessary bodyweight loss. Remember James Irvin at UFC on Versus 1? Because he sure as fuck doesn't. And from the conversation above, I'm more in line with the guy who's bewildered at the 60-pound gulf between the biggest men in the sport. No, there isn't yet enough top flight talent to stack two big boy divisions, but there absolutely will be. Every generation of prospects north of 220 pounds isn't only more talented than the last, there's also more of them, period. Defining a cruiserweight category now will seem very prescient in five years.

So, having bemoaned what has already been bemoaned, I believe I can actually propose a reasonable solution. The best way to address the inconsistent distribution of weight limits is to equalize the ratios between them. In other words, the percentage of weight difference should be the same between any adjacent divisions. That's easy enough to produce if you pick an arbitrary starting point and percentage:

• Bantamweight: 100 lbs
• Featherweight: 110 lbs (10% increase)
• Lightweight: 121 lbs (10% increase)
• Welterweight: 133 lbs (10% increase)
• etc.

However, out of respect for the current system, I'd rather not change anything that I don't actively take issue with. Accordingly, I chose to find the optimal weight classes between the modern 125- and 265-pound boundaries of the Unified Rules. Obviously I included the flyweight division, as Dana White has made clear his intentions to bring those fighters under the UFC's perpetually-expanding roof. As I mentioned previously, I also intend to identify a suitable limit for a future cruiserweight division. This totals to nine weight classes, including and between flyweight and heavyweight.

With pre-determined floor and ceiling values, finding the optimal constant ratio would be much more difficult, if not for this excellent formula:

x = ((b / a) ^ (1 / c))

a = lowest weight limit
b = highest weight limit
c = total number of weight classes, minus 1

or, "B can be calculated by C repetitions of multiplying A by X." Here it is applied to the matter at hand:

x = ((b / a) ^ (1 / c))

a = 125
b = 265
c = 8 (nine weight classes minus 1)

x = ((265 / 125) ^ (1 / 8))
x = (2.12 ^ (1 / 8))

x = 1.09848

A bit anti-climatic, but there's our magic ratio: 1.09848 to 1, or a constant decrease of 8.965% from heavyweight to flyweight. This allows for more forgiving weight cuts than the one that skeletized Irvin, but remember, the goal of this alternative is to make division-hopping equally difficult, not more or less difficult.

Division	Old Limit	New Limit
Heavyweight	265	265
Cruiserweight	n/a	241
Light Heavy	205	220
Middleweight	185	200
Welterweight	170	182
Lightweight	155	166
Featherweight	145	151
Bantamweight	135	137
Flyweight	125	125
(center of graph marked to show that "New" plot line is not a straight diagonal)

What should be immediately apparent is that the new divisions would seriously make use of the space between 125 and 265 pounds. The optimal limits are an average of 7.6 pounds higher than their original counterparts, 10.2 pounds if you exclude the static flyweight and heavyweight values. This would trigger a pretty massive shakeup, particularly in the middle third of the spectrum. Up to now, I've focused exclusively on logic, without considering logistics, but it's undeniably important to consider how these proposed changes might realistically impact the UFC roster.

At the time of this writing, the UFC employs exactly 250 fighters. Here's the breakdown across the current seven weight classes:

There's been a lot of focus lately on the unsustainable overcrowding of the lightweight division, but it's clear that welterweight and to a lesser degree middleweight are similarly bloated. Obviously this is equally contributed to by both the youth of the lighter weight classes and the relative scarcity of talented, heavier fighters. UFC management has been vocal about reducing their roster to 200 fighters, and some have interpreted this to mean that the promotion plans to house exactly 28 men per weight class. I, for one, think it's more likely that numbers in the more shallow classes will be maintained (certainly not increased), while as much pruning as possible will take place from 155 (a lot) to 205 (just a bit). All told, fifty guys will be looking for different promotions on the current course. How might weight class reorganization affect this situation?

Before proceeding, some delicate assumptions have to be made. First, I would reason that if the UFC were, in fact, to add two weight divisions, it wouldn't be necessary to shrink its roster at all. Their target average of 28 fighters per weight class would actually recalculate to a total of 252 fighters under the proposed nine-division system. However, that wouldn't necessarily save any fighters from the chopping block. As has been pointed out, adding cruiserweight would almost certainly cripple the already-thin heavyweight division, not to mention that a new 125-pound category would have to be built from scratch. Restocking these areas would once again necessitate cuts in the more saturated weight classes.

Next is the matter of how the fighters would redistribute themselves throughout the new landscape. There's a reasonable way to approximate these shifts. Notice that the old and new weight limits, for the most part, fit together like zipper teeth. For example, the old lightweight class is four pounds above and eleven pounds below the new 151- and 166-pound divisions, respectively. Those distances equate to 27% and 73% of the total fifteen-pound gap between the new categories on each end of the old one. Given these figures, I estimate that just 27% of the old lightweight fighters (14) would move up to the somewhat distant new lightweight division, while 73% (40 fighters, nearly three times as many) would make the small weight cut to the much closer new featherweight division. All of the migrations might shake down like this:

The only necessary exception to these estimations occurred in the heavyweight category. The basic argument in favor of installing a cruiserweight division is that many if not most heavyweights weigh in and fight at lighter than 265 pounds. To adjust for this, I simply averaged the weights of the 25 current UFC heavyweight fighters (as listed at UFC.com). The 250 pound average was closer to the cruiserweight limit below it, so I assigned more fighters to move down than up (without the adjustment, none of the "265ers" would have dropped to cruiserweight, which is of course unrealistic). Here are the projected populations immediately following the implementation of the nine-division system.

Truthfully, the lay of the land doesn't improve at all (keeping in mind that it wasn't expected to- it just would have been nice). The bloated 155-pound class becomes an identically bloated 151-pound class, and as predicted, the three heaviest divisions are badly understocked (although with 16 fighters, the cruiserweight division could conceivably start operating right away). Flyweight only hosts three fighters (presumably undersized current bantamweights- most fans could probably think of three off the top of their heads), but that's obviously because the division hadn't been included before the reorganization.

So how could the UFC compensate for these imbalances while maintaining a 250-fighter maximum? An acceptable solution certainly wouldn't be simple to execute, but might actually be less dramatic than one would expect. As one of many possibilities, the UFC could gradually remove 25% from the four most crowded divisions (remember that some people are currently expecting nearly 50% of the lightweight division to be released) and use the roster space to recruit prospects for the four weakest ones. For example:

I can confidently say that those numbers are not very frightening. How long would it seriously take the UFC to do that, a year? Fifteen months? Yes, signing twelve humongous top-tier heavyweights would be a challenge, but if they can't be found, then scoop up as many prospects as possible and let them sort each other out under the same roof. Clever scouting will speed up the process, but the division will need time to evolve, just like it has up to now, and just like it will anyway without these changes actually taking place. In any case, the above adjustments result in this much nicer-looking distribution:

In the end, it can't be overlooked that the current system of weight classes has served us well and, for now, continues to meet the needs of the MMA landscape. When the time comes for an updated structure, this nine-division proposal would nicely correct the few existing flaws. I absolutely agree that a change of this magnitude is not yet necessary, but I do feel that it will be, and I don't think anyone should be surprised if this or something similar becomes a reality within the decade. Change happens fast in our little sport, and for the foreseeable future, it's only going to happen faster. When a cruiserweight division is finally added, much will be made of it, it will be poked and prodded and have its effects predicted and measured... and then all of sudden it will be like it had been there the whole time. Until then, enjoy the great fights made possible by the system we have now- you might find yourself reminiscing about it sooner than you think.

Since its inception, the UFC has gone through a long series of transformations, including modifications to combat rules, various degrees of sanctioning, and even outright changes in ownership. Right alongside these has been the unfolding of weight divisions within the promotion. From the beginning to present day, the defined weight classes have evolved in a fairly logical manner, pacing our sport's rise in popularity and regulation. Most MMA community members are satisfied by the current divisions as outlined by the Unified Rules, but most of them were also inducted into their fandom with that system already in place (myself included). Are today's weight classes 100% acceptable? My thought is "absolutely yes." But are they as good as they possibly could be, from a logical viewpoint? Before addressing that question, I'd like to submit a brief history of the UFC's application of the weight class system:

UFC 1 (Autumn 1993)

Divisions:

• none

In the beginning, there was nothing. And that's no joke: the very first UFC was Mortal Kombat without lightning and fireballs, a completely openweight tournament where the winner got $50,000 and the losers got stitches. As a matter of fact, the first thirteen UFC shows (1 through 11 and both "Ultimate Ultimate" events) carried on without weight classes, and all but one of them featured the tournament format. Of course this structure was originally intended to showcase the harmless-looking Royce Gracie's ability to tie less-skilled giants into various types of knots. The Gracie family bowed out of the promotion after UFC 5, and without any competitors to replicate Royce's small size and seemingly magical techniques, it became increasingly clear that the openweight ingredient of the events was no longer essential.

UFC 12 (Winter 1997)

Divisions:

• above 200 lbs: Heavyweight
• up to 200 lbs: Lightweight

Fighters were divided into two weight classes starting at UFC 12. The event featured a four-man tournament in each division, above and below 200 pounds. It was a step in the right direction, but not a very big one, considering the "lightweight" category essentially encompassed all of today's weight classes including and below 205-pound light heavyweight. Nonetheless, the new category gave some future sub-heavyweight stars a more reasonable chance to get their careers started. Tito Ortiz and Frank Shamrock both debuted in the UFC as "little guys," but they wouldn't keep that distinction for very long.

UFC 14 (Summer 1997)

Divisions:

• above 200 lbs: Heavyweight
• up to 200 lbs: Middleweight
• up to 170 lbs: Lightweight (added at UFC 16, Winter 1997)

Less than a year later, at UFC 14, a miniature tournament for 200-pound or smaller fighters was held under the banner of the "middleweight" division. The victor, Kevin Jackson, went on to face Frank Shamrock for the championship in the UFC's first outing to Japan, Shamrock's preferred stomping ground. Jackson was vaporized in 16 seconds, and Shamrock, of course, was able to convincingly defend his title an impressive four times (a story better told by a riper banana).

While the heavyweight division carried on in its usual fashion, the real purpose of the middleweight reclassification was to welcome a new weight class for fighters tipping the scales at 170 pounds or less. Among those to seize the opportunity was Pat Miletich, who immediately won the first lightweight tournament and went on to join Frank Shamrock as a storied champion of the UFC's dark ages.

UFC 26 (Spring 2000)

Divisions:

• above 265 lbs: Super Heavyweight (added at UFC 28, Autumn 2000)
• up to 265 lbs: Heavyweight
• up to 200 lbs: Middleweight
• up to 170 lbs: Lightweight
• up to 150 lbs: Bantamweight

With three active weight classes, the UFC had created enough roster space to sit back for awhile and allow their talent pool to grow and develop naturally. Moving forward from the introduction of the 170-pound category, the promotion began to abandon the tournament format in favor of the more linear structure of boxing-style championships. It became clear that housing lighter-weight fighters was a wise decision, as it was during this incubation period that Shamrock and Miletich enjoyed their aforementioned title reigns, while the heavyweight division failed to crown a comparably dominant champion (all due respect to Randy Couture).

Perhaps in light of this trend- not to mention an ever-expanding deficit- UFC management opened the doors to the smaller still bantamweight division, for competitors at a maximum of 150 pounds. Pulverweight would have been a more apt classification, as the UFC held only three matches in that category, with Little Evil victorious in every one of them.

A final note on this evolutionary stage: UFC 28 featured the promotion's first and only super heavyweight bout, between future kind-of-a-jerk Josh Barnett and Gan McGee. This was not an attempt at christening a new weight class, but rather the result of strict sanctioning under an early draft of the Unified Rules. From that event on, the UFC eschewed super heavyweight contests by observing the heavyweight cap of 265 pounds.

UFC 31 (Spring 2001)

Divisions:

• up to 265 lbs: Heavyweight
• up to 205 lbs: Light Heavyweight
• up to 185 lbs: Middleweight
• up to 170 lbs: Welterweight
• up to 155 lbs: Lightweight
• up to 145 lbs: Featherweight (added at TUF 12, Autumn 2010)
• up to 135 lbs: Bantamweight (added at TUF 12, Autumn 2010)

At the top of 2001, Dana White and the Super Zuffa Bros finalized their purchase of the UFC from Semaphore Entertainment Group. By their second event, they had adopted the official weight classes as we know them today. Nearly a decade later, the featherweight and bantamweight classes of sister promotion WEC were absorbed into the UFC's ranks, and White has been vocal about the impending addition of a 125-pound flyweight division. This action would officially allow fighters from every weight class sanctioned by the modern Unified Rules (excluding the bizarre and vestigial super heavyweight division) to compete on MMA's largest stage.

Unfortunately, I have to ruin this pleasantly harmless summary by questioning the status quo. Today's weight categories are perfectly serviceable and there's no dire need to improve them... but the fact remains that they can be improved. In the next and final installment of this mini-series, I'll identify the (admittedly small) flaws of the current standard, and explore a couple of preferable alternatives, both rooted firmly in a sound mathematical concept.

Recap

We're back after a great night of action at UFC Fight Night 23, and the accuracy of both Adjusted Strength of Record-based predictions remains largely unchanged. Three of the five expected victors emerged in Hominick, Mitrione, and Guillard, for a new rate of 16.5/30 or 55%, while only two value bet picks came through (Wiman and Guillard again) on a card that gave them all a chance, slightly lowering betting accuracy to 8/19 or 42%.

Miguel Torres	W	L	D	WLD	Win%	RSR	ASR	vs.	Antonio Banuelos	W	L	D	WLD	Win%	RSR	ASR
	5	2	0	7	71%	(1.4249)	1.0661			4	2	0	6	67%	(1.1765)	0.8867
Beebe	12	1	0	13	92%	1.9215			Tapia	6	0	0	6	100%	1.4142
Maeda	10	3	1	14	75%	1.6202			Goldsby	5	4	0	9	56%	0.9623
Tapia	4	0	0	4	100%	1.1547			Jorgensen	6	2	0	8	75%	1.2247
Mizugaki	6	2	2	10	70%	1.2780			Osawa	3	3	1	7	50%	0.7638
Bowles	6	0	0	6	100%	1.4142			Jorgensen	7	2	0	9	78%	1.3472
Benavidez	9	1	0	10	90%	1.6432			George	7	2	0	9	78%	1.3472
Valencia	4	2	0	6	67%	0.9428

Cut and dry bout to start the main card, as former WEC 135-lb. champ Torres faces Pit Hawaiian Kempo product Antonio Banuelos. Both men were WEC staples for years, and it's initially puzzling that they never before met in the promotion's signature blue cage. A closer look reveals that Banuelos was just too inconsistent to challenge for the belt during Torres' reign. He made his way to a #1 contender's match but fell short against Manny Tapia, who was later eaten alive by Torres. This gap soon narrowed, however, as Torres lost his title and dropped his comeback effort against Joseph Benavidez, while Banuelos was once again denied a title shot, this time by the hard-nosed Scott Jorgensen. The aftermath of this shakeup is as good a time as any for these two to meet, but expect mullet to defeat mustache as Torres takes another step toward reclaiming his lost gold.

Jon Jones	W	L	D	WLD	Win%	RSR	ASR	vs.	Ryan Bader	W	L	D	WLD	Win%	RSR	ASR
	11	1	0	12	92%	(0.6443)	1.2568			6	0	0	6	100%	(0.9347)	1.2319
Bernard	0	1	0	1	0%	0.0000			Acosta	4	5	0	9	44%	0.7698
Eduardo	0	1	0	1	0%	0.0000			Magalhaes	2	2	1	5	50%	0.6455
Pina	0	0	0	0	0%	0.0000			Marrero	6	2	1	9	72%	1.2509
Verrett	1	2	0	3	33%	0.3333			Schafer	4	2	0	6	67%	0.9428
Porter	2	0	0	2	100%	0.8165			Jardine	2	4	0	6	33%	0.4714
Gabin	2	1	0	3	67%	0.6667			Nogueira	7	0	0	7	100%	1.5275
Gusmao	5	0	0	5	100%	1.2910
Bonnar	3	2	0	5	60%	0.7746
O'Brien	4	2	0	6	67%	0.9428
Hamill	4	2	0	6	67%	0.9428
Vera	3	4	0	7	43%	0.6547
Matyushenko	6	1	0	7	86%	1.3093

Interesting numbers for the most interesting fight on the card. Jones' entire pro career fits within the three-year time frame in question. This increases his ASR score for a busy rate of fight activity, but the bonus is counteracted by the weakness of his early opponents. This coincidental penalty is meant not to punish Jones for being developed properly, but rather to help remind us that three a year-old fighter is a young fighter. Ryan Bader isn't exactly a grizzled vet himself, but three years ago he was already past the "debutant vs. debutant" stage of his career. In my head, I see Jones winning this fight more often than not, but I don't buy the notion of a Bader defeat as a foregone conclusion.

Carlos Eduardo Rocha	W	L	D	WLD	Win%	RSR	ASR	vs.	Jake Ellenberger	W	L	D	WLD	Win%	RSR	ASR
	9	0	0	9	100%	(0.6484)	1.2483			7	2	0	9	78%	(1.1142)	1.1629
Kunze	0	0	0	0	0%	0.0000			Healy	12	8	0	20	60%	1.5492
Mensing	4	4	2	10	50%	0.9129			Story	4	2	0	6	67%	0.9428
Sidoni	1	0	0	1	100%	0.5774			Lakebir	2	0	0	2	100%	0.8165
Mengulluoglu	2	1	0	3	67%	0.6667			Seo	3	5	0	8	38%	0.6124
Goldberg	0	4	2	6	17%	0.2357			Seguin	6	4	0	10	60%	1.0954
Mihnjak	1	0	0	1	100%	0.5774			Alfaya	3	2	0	5	60%	0.7746
Kadlubek	5	3	0	8	63%	1.0206			Condit	7	1	0	8	88%	1.4289
Balci	2	1	0	3	67%	0.6667			Pyle	7	2	0	9	78%	1.3472
McCray	5	1	0	6	83%	1.1785			Howard	8	2	0	10	80%	1.4606

Rocha vs. Ellenberger is a good example of the difficulty in comparing fighters from mutually exclusive competitive backgrounds. Rocha is a legitimate jiu-jitsu black belt who has been reducing his opponents to neatly-wrapped little packages, but before his lone UFC contest, he'd only done so on the demonstrably bottom shelf German circuit. Ellenberger has been detoured on his march to a title shot, but only by recognizably high-level fighters. This bout reminds me of December's match-up between lightweights Charles Oliveira and Jim Miller, and despite Rocha's statistical advantage, I predict a similar outcome: the flashy submission stylist on a hot streak will get out-hustled by the no-nonsense workhorse.

Rich Franklin	W	L	D	WLD	Win%	RSR	ASR	vs.	Forrest Griffin	W	L	D	WLD	Win%	RSR	ASR
	4	2	0	6	67%	(0.7736)	0.7711			2	2	0	4	50%	(1.1581)	0.5779
Lutter	4	2	0	6	67%	0.9428			Jackson	6	0	0	6	100%	1.4142
Hamill	4	1	0	5	80%	1.0328			Evans	6	0	1	7	93%	1.4184
Henderson	4	3	0	7	57%	0.8729			Silva	8	0	0	8	100%	1.6330
Silva	2	4	0	6	33%	0.4714			Ortiz	0	2	1	3	17%	0.1667
Belfort	4	1	0	5	80%	1.0328
Liddell	1	3	0	4	25%	0.2887

Personally, I'm very excited for the contest between former champs Franklin and Griffin, but it's hard to say what a win here means for either man. This is by no means a "senior circuit" match, but these guys aren't exactly young lions either. And it's a long way to the front of the #1 contender line. Rampage has been "promised" the next title shot, but if he sneaks away with another split decision win, while other high-stakes LHW fights are resolved convincingly around him- or if he loses!- who knows what will actually happen.

Either way, the victor of Franklin vs. Griffin is not next in line to challenge for the belt, but I get a feeling from this match-up that neither fighter is being barred from the conversation. Whoever wins this fight will have beaten a highly-regarded opponent, and one more subsequent impressive victory is all the UFC would need to spin a yarn about a reborn ex-champ, a phoenix rising from the ashes, leaving a trail of devastation behind him on the road to recapture former glory. I'm a big fan of both guys, but I think the subject of all the hyperbole is going to be Ace.

Anderson Silva	W	L	D	WLD	Win%	RSR	ASR	vs.	Vitor Belfort	W	L	D	WLD	Win%	RSR	ASR
	7	0	0	7	100%	(1.2394)	1.4247			3	0	0	3	100%	(1.1101)	1.0354
Henderson	6	3	0	9	67%	1.1547			Martin	7	4	0	11	64%	1.2185
Irvin	6	3	1	10	65%	1.1867			Lindland	5	2	0	7	71%	1.0911
Cote	7	1	0	8	88%	1.4289			Franklin	5	3	0	8	63%	1.0206
Leites	6	1	0	7	86%	1.3093
Griffin	4	2	0	6	67%	0.9428
Maia	7	1	0	8	88%	1.4289
Sonnen	6	2	0	8	75%	1.2247

The last time Silva and Belfort were scheduled to fight: 10 months ago. The time before that: 12 months ago. The very first of their three cancelled engagements: 13 months ago. The last time Belfort set foot inside the cage? 17 months ago. There's a lot of rhetoric surrounding this fight that would have been much more relevant at the top of last year. That's not to say that it's completely unfounded now- Belfort is still lightning fast and has bombs for hands, but he's cooled to room temperature, and that's not good when most analysts agree that his best shot to win is in the early rounds. It's perfectly reasonable to suggest that Silva has lost a step, but by the time Belfort is fully warmed up, he's going to be behind on points, probably tired, and possibly hurt. I'm already looking forward to his next title run, but Saturday night will be Silva's night.

Betting

If there's been one recurring factor that contributes to bad ASR-based value bet picks, it's that the distance between opposing ASR probabilities is rarely as wide as it is between betting lines. To wit, Miguel Torres is the favorite in both categories, but bookmakers have him at such an advantage that, under the current criteria for a "value bet", my system will happily send you to the homeless shelter with suggestions like betting on Banuelos. After seeing some of the recommendations so far, I don't think I'm the only one who's shocked that value bet accuracy is even as high as it currently is (42%).

I'm pleased with the accordance between probabilities in the Franklin/Griffin fight; both systems of odds reasonably reflect how competitive it is. In the past I've actually attained spreads of 0, but lately that metric has been wildly erratic. In the end, I don't think these discrepancies particularly discredit one measure of likelihood more than the other; it should probably just be accepted that the suggestions of formulas and bookmakers are equally dubious.

(Odds from BetUS.com)

Recap

It's fight week once again, and just as we were winding down from a really solid offering in UFC 125. Edgar vs. Maynard was as thrilling as any fight I've ever seen, and you couldn't have scripted a better plot twist than the judges' decision of a draw. Following suit were results for Adjusted Strength of Record-based predictions, which added up to 2.5 out of 5 for fight outcome accuracy (a half-point awarded for the non-conclusive outcome of Edgar/Maynard) and 1 out of 2 for value bet accuracy. As Brandon Vera was both an infinitesimally slight ASR favorite and a more significant sportsbook underdog, a win from him would have supplied a welcome boost to both accuracy stats. Not surprisingly, he shit the bed and ruined everything. Accordingly, percentages stand at 54% for fight results (13.5 out of 25) and 43% for value bets (6 out of 14).

Cole Miller	W	L	D	WLD	Win%	RSR	ASR	vs.	Matt Wiman	W	L	D	WLD	Win%	RSR	ASR
	4	2	0	6	67%	(1.4432)	0.9492			4	2	0	6	67%	(1.3060)	0.9181
Stephens	12	2	0	14	86%	1.8516			Buchholz	7	1	0	8	88%	1.4289
Gurgel	3	2	0	5	60%	0.7746			Tavares	10	1	0	11	91%	1.7408
Browning	3	0	0	3	100%	1.0000			Miller	11	1	0	12	92%	1.8333
Escudero	9	0	0	9	100%	1.7321			Stout	4	4	0	8	50%	0.8165
Lauzon	8	1	0	9	89%	1.5396			Nelson	6	1	0	7	86%	1.3093
Pearson	11	2	0	13	85%	1.7614			Danzig	3	3	0	6	50%	0.7071

UFC's second Fight For the Troops event kicks off with a very close match-up between Ultimate Fighter classmates Cole Miller and Matt Wiman. Over three years, both men have posted identical records against similarly mid-tier competition, but the records of Miller's opponents barely outshine those of Wiman's. A worrisome line can be drawn, from Wiman's come-from-behind KTFO of Thiago Tavares, to Miller's crushing defeat at the hands of Efrain Escudero. Fortunately, Miller showed tighter stand-up in his most recent bout with Ross Pearson, and should be able to avoid a violent end of that nature. One final thing to mention is Wiman's somewhat frequent injuries. It's likely that he has just gotten hurt at inopportune times, but a legitimate lack of resiliency would be a serious disadvantage in the back nine of this fight.

Joey Beltran	W	L	D	WLD	Win%	RSR	ASR	vs.	Pat Barry	W	L	D	WLD	Win%	RSR	ASR
	9	3	0	12	75%	(0.8842)	1.1427			5	2	0	7	71%	(0.9697)	0.9377
Friedman	1	0	0	1	100%	0.5774			Delaney	2	0	0	2	100%	0.8165
Lopez	6	2	0	8	75%	1.2247			George	13	9	0	22	59%	1.6002
Fenton	0	0	0	0	0%	0.0000			Diouf	0	0	0	0	0%	0.0000
Browy	0	0	0	0	0%	0.0000			Evensen	4	3	0	7	57%	0.8729
Pendergarst	6	11	0	17	35%	0.8402			Hague	9	1	0	10	90%	1.6432
Combs	2	3	0	5	40%	0.5164			Hardonk	4	3	0	7	57%	0.8729
Willis	7	0	0	7	100%	1.5275			Filipovic	4	2	1	7	64%	0.9820
Lopez	14	1	0	15	93%	2.0870
Alexander	4	4	1	9	50%	0.8660
Gracie	3	0	0	3	100%	1.0000
Hague	6	3	0	9	67%	1.1547
Mitrione	2	0	0	2	100%	0.8165

This is a great example of how sensitive ASR-based predictions can be to tuneup fights. Barry is the clear favorite here skill-wise, and I'm very confident that he will eat Beltran alive. The reason that Beltran outscores him is something that is bound to come up when a fighter has spent a large portion of his three-year window in smaller promotions. Many little league fighters are free agents trying to build their resumes fast to attract contracts. They rack up experience because they have the freedom to fight whomever, wherever, whenever. Barry's score is additionally disadvantaged by the fact that the ASR formula time frame actually overreaches his entire pro career. Consequentially, Barry comes up short with significantly fewer fights and a comparable but still lower win percentage, despite facing the tougher opponents. Regardless, I would be shocked to see him lose this fight.

Mark Hominick	W	L	D	WLD	Win%	RSR	ASR	vs.	George Roop	W	L	D	WLD	Win%	RSR	ASR
	4	1	0	5	80%	(1.2157)	1.0123			4	3	1	8	56%	(0.8687)	0.7443
Grispi	10	1	0	11	91%	1.7408			Buschman	1	1	0	2	50%	0.4082
Young	4	5	0	9	44%	0.7698			Nelson	6	1	0	7	86%	1.3093
Caraway	9	1	0	10	90%	1.6432			Kaplan	0	2	0	2	0%	0.0000
Jabouin	4	2	0	6	67%	0.9428			Sotiropoulos	4	1	0	5	80%	1.0328
Garcia	4	2	1	7	64%	0.9820			Dell	4	3	0	7	57%	0.8729
									Wineland	4	2	0	6	67%	0.9428
									Garcia	4	4	0	8	50%	0.8165
									Jung	9	2	0	11	82%	1.5667

Keeping a busy schedule is great, but winning is pretty important too. George Roop is about as close to being a .500 fighter (over three years) as he can get, and I think he's going to hit the mark on Jan 22. Roop will definitely come forward until his limbs are broken and all his teeth are knocked out, but other than that, I don't see what he has for Hominick in this fight. Hominick is faster and more technical, and I would have to say stronger, given Roop's lanky frame. Given these serious advantages, I don't think Hominick's shorter reach will be much of an issue at all. I'll rep the ASR probability here and say Hominick finishes this fight.

Matt Mitrione	W	L	D	WLD	Win%	RSR	ASR	vs.	Tim Hague	W	L	D	WLD	Win%	RSR	ASR
	3	0	0	3	100%	(1.2441)	1.0755			7	4	0	11	64%	(1.2140)	1.0468
Jones	4	1	0	5	80%	1.0328			Petkovic	4	5	0	9	44%	0.7698
Ferguson	4	1	0	5	80%	1.0328			Kilkenny	7	3	0	10	70%	1.2780
Beltran	10	2	0	12	83%	1.6667			Lundburg	0	2	0	2	0%	0.0000
									Pendergarst	9	7	0	16	56%	1.2990
									Petkovic	5	4	0	9	56%	0.9623
									Barry	4	0	0	4	100%	1.1547
									Duffee	5	0	0	5	100%	1.2910
									Tuchscherer	8	2	0	10	80%	1.4606
									Beltran	10	2	0	12	83%	1.6667
									Jensen	9	4	0	13	69%	1.4412
									Wiuff	14	2	1	17	85%	2.0304