Baseball Prospectus' Jason Parks wrote a short piece on Wednesday about a pitch thrown by Yu Darvish on Tuesday April 24. Parks, for those of you who are don't know (and you should know him), is one of BPro's scouts of minor league talent and does good work alongside Kevin Goldstein.

Parks and "several other scouts he knows", were befuddled by one pitch in particular on the night Yu struck out ten Yankees:

I hadn’t noticed that he was throwing what I’ve seen described as a shuuto, or a reverse slider.

I’m not certain of the classification—Gameday and Brooks Baseball classified the pitch as a two-seam fastball.

I was confused until the seventh inning, when catcher Mike Napoli appeared to be using a different sign for the pitch than the standard fastball, placing an L between his legs when he wanted the pitch with the extreme arm-side run. The announcers referred to the offering as a two-seamer, but as I stated, the movement was way more extreme, as was evident by the catcher’s pre-pitch setup and anticipation of the pitch (he expected run). The best example can be seen in the sixth and final pitch of Darvish’s seventh-inning matchup against Raul Ibanez, as the standard fastball morphs into a reverse slider that runs away from the hitter like the hitter has cooties. The image below shows the trajectory from Gameday, but you can also view the pitch at the 44-second mark here.

I’m not overly familiar with PITCHf/x, which is to say I’m aware of its existence and I respect those that have a mastery of such data, but I’m not fluent in its language. That said, I’d love to see a breakdown of Darvish’s pitch movements to see if my eye was just playing tricks on me or if he was manipulating this particular ball for effect, as was suggested by the sign from the catcher and the outcome of the offering.

--Jason Parks, "Does Darvish Throw a Shuuto"

Well, let's oblige Mr. Parks' request and see what this pitch is and how it moves, and whether it is indeed some new Japanese pitch or not.

Continue reading this post »

Hockey Prospectus (formerly "Puck Prospectus") has created a projection system, like their more well-known owner Baseball Prospectus' PECOTA, called VUKOTA (yes, after the ex-Isle) to predict how well players in the NHL would do in the future. Before the 2011-2012 season, in the Hockey Prospectus 2011-2012 Annual, HP used the system to project how every player in the NHL (no prospects) would perform and how well each team would do in the standings.

In this post, I'm going to go through the individual projections of VUKOTA at the beginning of the year and see how certain players are exceeding/failing-to-meet expectations and to see really where the Isles would be if not for injuries.

Continue reading this post »

A little more than a month ago, I wrote the following:

1. Nabokov's high SV% is not REAL - it's a facade caused by luck on special teams that is likely to crumble. Nabokov has essentially been the 22nd best goalie in the league this year - NOT the 7th best as his ordinary SV% suggests. He is NOT an elite goalie (though he is certainly average) and the longer he plays for the Isles, the more likely his true talent will cause his results to drop, and lower his trade value.
......

As a result, Hockey statisticians rely instead upon Even Strength Save % (ESSV%) to judge goalies, as that way we eliminate the problems caused by including Special teams Data. And by this measure, Nabokov is actually 22nd in the league among NHL goalies with 20 games or more played (.925 SV%). The Reason why Nabokov is 7th in overall SV% is simple: He's gotten insanely lucky on the Islanders PK, with a .951 PKSV% (2nd in the league)! Before anyone asks, this is nowhere near anything Nabokov has done in his career, where he usually has PK SV%s of .880 or so.

None of this should be surprising to any of you - Nabokov's current SV% would be his career high at age 36 and far above his .913 carer average.

Long story short, Nabokov's #s are a good deal smoke and mirrors - and are far far more likely to get worse than go up

-Garik16, February 12, 2012

And here were a few of the responses (Names withheld to protect the non-believers):

Continue reading this post »

Editor's Note: FanPost front-paged for this hot topic of the next two weeks.

Evgeni Nabakov has been amazing for the Isles over the past few weeks, and he's certainly been responsible for the wins against LA and Philly in the last week. If he could continue to play like this, the Isles MIGHT have a shot at the playoffs.

And yet the Islanders need to trade Nabokov for a prospect/draft pick. Preferably soon.

There are two reasons for this. The first is less important, but is easier to understand so I'll start with that:

Continue reading this post »

One of the more interesting developments in the study of baseball over the last few years — whether you're interested in sabermetrics or not — has been the public release of pitch-tracking data, or PITCHf/x. This data, available for free on the MLB website (hidden within files at gd2.mlb.com), allows anyone to look at every individual pitch thrown by their favorite pitchers (or not-so-favorite pitchers) to discover what pitches these pitchers throw, how often they throw them, how these pitches move, and of course, how effective those pitches are.

Originally, this information was really only useful to those who were willing to trawl through the individual data (think large large excel spreadsheets) and knew how to manipulate it. Over the last few years, however, a bunch of websites have culled the useful data (movement, effectiveness of each pitch, etc.) and presented it in human-readable form. These sites included Texasleaguers.com and JoeLefkowitz.com, which provided a series of graphs and charts using the data. Even Fangraphs had a PITCHf/x page for every pitcher in the database.

In other words, if you ever wanted to become an expert on the pitches of your favorite pitcher, your most-hated pitcher, or just a pitcher you find interesting, now you could!

But there was a problem for most people who wanted to use these sites to learn about various pitchers. The graphs/charts/data on these sites would be presented like this:

Continue reading this post »

So by now I'm pretty sure everyone here has heard of ESPN's infamous Top 25 players under 25 list by Neil Greenberg (To see the full list and reasonings if you have ESPN Insider, go HERE) which excluded John Tavares. More to the point, it excluded John Tavares in favor of Michael Neuvirth, Sergei Kostitsyn, Semyon Varlamov and obviously a bunch of others. I think all of us here disagree with Tavares' exclusion from this list. In fact, Howie Rose even mentioned it on the air Saturday night.

That said, given the arguments made by Neil Greenberg, which to a certain extent involve statistics, it bothers me to see the following argument (in some shape or form) given as a rebuttal to Greenberg:

Obviously he doesn't watch hockey at all.

As someone who tries to learn about the latest in sports statistics and how they can be used to get a better understanding of the sports I love, this type of argument drives me a little bit crazy. Why? Because in essence it's showing ignorance - it's basically saying "I don't know what exactly you mean with your explanations, but I don't care to find out and I'm just going to ignore your conclusion anyhow." It's the type of closed minded reasoning that prevents people from learning anything new, and getting a better understanding of what's going on.

Now, mind you, it's everyone's own prerogative to bother to learn or even to pay attention to advanced stats - everyone can enjoy sports in their own way. But when someone uses those stats to make an argument, if you choose to not learn those stats, you can't really argue strongly "Well you're wrong" if you don't even bother to understand the arguments being made. (You can ignore the argument if you want, but you can't try to refute the argument).

And here, this is a case where there's a strong argument that Greenberg is wrong. So let's make that argument, rather than simply attack Greenberg with an ignorant response.

Continue reading this post »

Last Month, on November 7th, I noted how the Islanders real problem wasn't finishing their opportunities, but simply that their opponents were getting more opportunities than the Islanders were. I'm really busy right now in real life (Finals+Lots of other stuff=Crazyness) so I haven't finished my look at Isles Forwards, but thought I could quickly during class update everyone on how the possession numbers look for the NHL one month later.

Once again - Possession % simply is the % of time a team has the puck in the offensive zone (neutral zone time is ignored) instead of the defensive zone. Possession #s above 50% mean that the team is generally winning the possession battle - below 50% means that they are losing.

Three quick notes:
First, I've changed the metric slightly I'm using to measure possession - we are now looking essentially at the possession #s (Corsi) ONLY when the score is CLOSE - aka within one goal. The reason for this is that when games are blowouts (or just not close), the leading team tends to stop really trying to score and drive possession, resulting in their numbers looking worse than they actually are. I didn't use this measure last time due to sample size.

Second, I've added in the chart below a measure labeled "Effective Possession %". This is an alternative estimate to possession % using Fenwick. What does this mean? Well this measure of possession does NOT COUNT BLOCKED SHOTS. If a hypothetical team is in its own zone 60% of the time (for a Possession % of 40%) but blocks every shot on net, it's not really losing the possession battle - after all, it will always be outshooting the opponent! Thus Fenwick eliminates blocked shots from its numbers to take a measure of EFFECTIVE Possession.*

*Teams have been found to have an actual measure of control over blocking shots (This makes sense), though they don't over preventing their own shots from being blocked. However, teams don't seem to have a measure of control over making opponents miss on their shots, so that is still included in Fenwick.

You should note that the Effective Possession numbers result in more or less the same rankings as regular possession #s, so you don't really need to worry about them.

Third, this data is not including the games of 12/5. It's up to date as of Sunday 12/4.

Continue reading this post »

Quick Interlude in this series - we'll have more on the forwards after Thanksgiving, but this needed to be said first.

Two related complaints that have gotten a lot of traction lately are the complaints that Marty Reasoner has been a failure for this team and that this team could really use Zenon Konopka. Some of you might recall I sung Garth's praises for acquiring Reasoner in the offseason (short bits in this comment here and this post here) and called him a "HUGE Upgrade over Zenon." So what's happened?  

Well, let's look at why Reasoner was considered to be an upgrade over Konopka.  Last year, Konopka was used in an extreme defensive role due to his faceoff ability and his defensive skills. For this purpose, Konopka was used in defensive zone faceoffs as often as possible - he actually took more defensive zone faceoffs than any other player on the team at even strength, despite playing less than 9 minutes of EV time per game. (By comparison, Tavares played 15.) The team did try and give him weaker opponents when possible (it seems likely they did this by having him out there whenever the opposing first line wasn't the one taking the faceoff in our defensive zone), but for the most part, Konopka played EXTREMELY tough minutes.  

So to sum up:

Continue reading this post »

One of the main complaints about the team's management from preseason onward has been that the Islanders don't have good enough defensemen.  Islander fans were confident in their top 3 D-men...and that's about it.

Well the season is here, and the complaints about D-Men have continued, perhaps for good reason.  Interestingly enough, most of the complaints now focus on the bottom pair D-Men, rather than on the lack of a clear top 4.  

But enough about what the complaints are.  Let's talk about reality.  In the first post in this series I talked about how a major problem of the Islanders wasn't their shooting %, but their inability to drive possession.  How much of that are the D-men responsible for?  And which D-Men have been successful, and which have been poor?  

Table 1 below shows the various statistics of the Islander D-men.  The prime number to focus on in the table is the "Possession %" (Corsi %), which attempts to estimate the % of time the Isles are in the offensive zone instead of the defensive zone (neutral zone time being...well, neutral).  

Table 1 also includes two statistics to put the possession #s into context:  
Zone-Start % calculates the total percentage of faceoffs each D Man is on the ice for that are in the offensive zone (factoring out neutral zone faceoffs once again).
Quality of Competition (Relative Corsi) attempts to calculate how good the competition has been while each D Man is on the ice.  The higher the number, the better the competition.  

Without further ado, Table 1:

Table 1:  The Possession (and some other) Statistics of Islander Defensemen so far.

Continue reading this post »

The Islanders' situation this far (12 games in) to the season looks dismal. The Islanders have the second worst record in hockey - ahead of only Columbus - and look very little like they did in the more positive end of last season. Nearly every potential expected problem for the Isles has shown up this season (Goaltending, despite the three-headed monster, has not been a major problem) while very few of the expected positives have shown up (Just Tavares' improved offensive play). 

But what are the true causes of the Islanders' problems? Well let's start from a really broad team perspective. The Isles' problems appear to have come from a lack of scoring.  

This comes from two different places:  

Continue reading this post »

Hockey Prospectus released recently their annual for the upcoming NHL season. It's $9.99 for the PDF version, and I recommend you all get it. Hockey Prospectus provides some of the best statistical analyses of the game of hockey and the NHL and their work is extremely informative. If you were just interested in hockey in general, I'd recommend you buy the annual.

One thing contained in this annual is their projections for the current season, using their projection system VUKOTA.  VUKOTA is a system which is used to try and project how well each player in the NHL will play in the upcoming season based upon their past results. VUKOTA does this by taking note of each player's age and their career statistics and comparing them to players in the past who have had similar results.

Continue reading this post »

Two weeks ago, we took a look at what individual pitches in baseball were the luckiest pitches in baseball all year, or rather we looked at the pitches that owed most of their success to outside factors other than the pitcher himself.  We did this by looking at the differences between the expected run values for individual pitches and the actual run values, as often seen on fangraphs (under Pitch Type Run Values).  For more on the failures of run values and the differences between expected and normal run values, see the last article.

Today we're going to look at the opposite: the individual pitches that are the UNluckiest* pitches in baseball.  These are pitches that look mediocre to poor (generally) if you look them up on fangraphs, but really have just been affected by other factors to an extent that their run values and results are really misleading.

*I'm not sure if unluckiest is an actual word, but I like it, so there.

Obvious note:  I'm calling these pitches "unlucky" but of course it's totally possible that something about these pitches is really causing these worse results.  But it's far more likely in most cases that these pitches are suffering the worst effects of random variation.

Continue reading this post »

Bumped from FanPosts. --Chris

---

Bobby Parnell is a pitcher who relies a good deal on throwing gas.  And this year, he's done that pretty successfully - his fastball is at the highest velocity it has ever been and he's breaking 100 quite regularly.  In fact, as of 8/17, only Henry Rodriguez (37 times) and Aroldis Chapman (93 times!) have broken triple digits with a pitch more often than Bobby Parnell. 

But Parnell had an issue with throwing his pitches (both this fastball and his slider) at their peak velocities:  for the first batter or so, Parnell often seems unable to hit his normal high velocities, as if he was still warming up.  You can see this on Figure 1 below, which shows the average velocity of Parnell's fastball through the first 15 pitches of his appearances: 

Continue reading this post »

A 53.5 MPH Dickeephus taken for a strike by poor Wes Helms in 2010.  (Thanks to Mike Fast for the image)

For most of 2010, R.A. Dickey befuddled opposing batters with 3* pitches: a fastball, a fast knuckleball, and a slow knuckleball.  But on August 18th**, R.A. Dickey unleashed his third, most devastating knuckleball - the Dickeephus*** pitch -, striking out Geoff Blum swinging with a 62 MPH knuckleball.  Dickey proceeded to throw 9 more Dickeephuses in 2010 (one on 8/24, two on 8/29, two on 9/14, one on 9/19, and 3 on 9/29) to mixed results.  The Geoff Blum pitch was actually the fastest the Dickeephus would be for the rest of 2010 - these nine last Dickeephus pitches ranged from 53.3 MPH (a called strike against Wes Helms) to 58.1 MPH. 

*Dickey actually has thrown a change-up each of the last few years, but it's really really rare and random.  It's not worth talking about.

**Dickey actually threw two Dickeephus pitches before this date, on June 17 and July 15.  But these were isolated incidents; his real use of the Dickeephus in earnest started on August 18th.  

Continue reading this post »

Pitch Type Run Values, as found on Fangraphs (using BIS classifications) and other locations, attempt to use linear weights to determine how good, or how bad, specific pitches are during each season for individual pitchers.  Fooling around with these measures on fangraphs can tell us who has the best fastball, change-up, etc and can let us know when we should (seemingly) be angry with our favorite pitchers' choice of pitches during certain individual games. Thus these values can easily lead to arguments where someone argues Pitcher A should use Pitch Type 1 more often at the expense of Pitch Type 2 to get better results.

However, these run values have numerous problems when used in this way: 
First, the results of each individual pitch type is NOT independent of the results/specifics of other pitches a pitcher can throw; in other words, a pitcher's curveball's results are to some extent affected by the results and properties of the same pitcher's fastball (and other pitches).  Thus run values can't tell whether a pitcher would be better off using certain pitches more often and others less - a pitcher reducing the use of his worst pitches could lead to that same pitchers' other "better" pitches getting worse results. 

Second, pitch type run values are heavily dependent upon the count in which pitches are thrown - strikes on 0-2 counts have a naturally higher value than on 3-0 counts, for example.  So if a pitcher throws Pitch A in early counts, or counts where he's behind (say, a fastball), and throws Pitch B (say, a curveball) only when he's ahead or on 2 strike counts, Pitch B is almost certainly likely to look better with run values, even though the pitch may in fact be worse than Pitch A without context. 

Third, and perhaps most importantly, pitch type run values do not attempt to correct for luck or defense.  If a pitcher gives up 8 line drives off of a pitch and all are right at a fielder, the run values indicate that the pitch is really a great pitch (it got all 8 balls in play out!) even though we wouldn't expect this performance to necessarily continue.  The end result of this is that the run values you see on fangraphs may tell you that a pitch is good, only for the reality to be that the pitch is just REALLY lucky (or the converse is true). 

Now it's easy for people these days to make a guesstimate as to how "lucky" or "unlucky" a pitcher overall has been during the year, whether by looking at BABIP, LOB%, or just comparing pitchers' ERAs to more advanced ERA estimators (Whether these things actually tell us exactly how much of a pitcher's ERA this year is due to luck is another question, but they do give a decent indication of one way or the other).  However, nowhere is there a metric easily listed (on fangraphs or something else) that can show the amount luck/defense has affected an individual pitch's results during a given year. 

EXPECTED RUN VALUES attempt to do just that: strip away the results of luck and defense from pitch type run values so that the only thing remains is a true evaluation of how good a pitch's results have been.*

*Feel free to ignore the next two paragraphs if you don't care how Expected Run Values work.

Expected Run Values are the same as regular run values for pitches that aren't put into play (Called Strikes, Called Balls, Foul Balls, Hit by Pitches, etc.).  However instead of using the actual results of pitches put into play, Expected Run Values simply assume that each ball hit into play has an average result based upon the batted ball type of the ball in play. 

In other words: if Expected Run Values see that a ball in play is a ground ball, it calculates the result of that individual pitch as if it was an average ground ball.  The same is true of line drives, fly balls, and pop-ups.  So if a pitcher's in play results are all pop-ups, they'll result in really good Expected Run Values, whereas if they're all line drives, his Expected Run Values will be terrible.

Using this metric, we can look at who has the luckiest pitches in baseball so far in 2011.  After the jump we'll do exactly that.

Continue reading this post »

One of the things you sometimes hear from people talking about a hockey game is that one team is "controlling the pace of play."  What this tends to mean is that one team has had control of the puck far more often than the other and has kept the puck in the opposing zone - and out of their own zone - for most of the game.  When a team is able to do that, it seems like they are "outplaying" the other team, even if they've been yet unable to put the puck in the net.  And the feeling seems to be is that if a team manages to control the pace of play - or control possession of the puck - then eventually the goals will come and they'll lead on the scoreboard as well in the spectator's minds.

Well this thought - the team that controls possession for most of the game is outplaying the other team and will most likely win the game in the end- actually seems to be as true as one would think.  In other words, keeping possession of the puck - and more particularly, keeping it in the offensive zone - is indeed super important to winning in hockey.  

Thus a good measure of how good teams are (and how good certain players are) would be to measure the amount of time the puck is in one team's possession, or perhaps the amount of time it was in the opponent's zone, per game.  In Soccer (European Football), such a thing is tracked as an official statistic (See the "Time of Possession" measurement in this game here).  Unfortunately, the NHL doesn't track such things. 

So we're out of luck right?  Fortunately the answer is no.

Continue reading this post »

It's now pretty clear that Nassau County voted down the Islanders' referendum.  And there is a lot of doomsaying going on.  Which is just silly in my opinion: the "No" vote may be the best thing to happen to the Isles in years. It may not be: there is unquestionably far more uncertainty with the team now than with a "yes" vote - but the team had been stalled by Wang's insistence on Lighthouse project or bust - now we can see what comes after the door is officially closed.

It's admirable in some sense that Wang kept trying to keep the Isles where they were - how he was committed to Nassau County through his Lighthouse Project.  And no one is saying that Kate Murray's actions were justifiable......ever.  But that ship had sailed long ago.....and yet the team remained in limbo with no plan.

Continue reading this post »

There are two features of the game of hockey that make effective use of statistical analysis difficult. The first, as discussed in the last four parts (Part 2, linked at the bottom of this post) is the problem of Context, and as detailed in those posts, Hockey Stats have evolved so that we can take into account the effect of context on a player's numbers. 

The second is Sample Size - the issue being that the numbers we have often come from such a small sample size that we can't rely on them to make accurate judgments of a player's true talent and his worth going forward.  Now this is true in every sport (and other things as well), but it is a particular problem in hockey, far more than the average fan realizes.  Obviously a player with a career high of 33 goals/60 points who scores 4 goals in his first two games* on a new team isn't likely to break 100 goals scored, or even put up 50 goals.  But for several commonly used statistics, even a full season isn't quite enough for a user of statistics to conclude anything about a player's true talent, something that often is unrealized by the average fan.**

Continue reading this post »

So a lot of people (or maybe a few vocal people) keep commenting  on how we can't go into the season next year relying upon Travis Hamonic to keep up or improve on last year because of the "sophomore slump." 

So I took a look, using GVT (Hockey Prospectus' statistic which measures total player value, including offensive, defensive, and shootout contributions) since I lack a good measure for Defense historically, of how defensemen who played 60+ games in their first season did in their second season (assuming they had a 2nd season). We're talking 190 Defensemen on this list.  

Of the 190 D-Men on this list, 116 improved the next year, while 74 put up worse numbers. So putting up "worse" numbers is certainly possible, and happens a decent amount for such players, but it's NOT more likely to happen than an improvement in the sophomore season.  

Continue reading this post »

So far in Part 2 of this series we've talked about trying to account for the context in which a player puts up his numbers (Goals, Assists, +/-, corsi, etc.).  To this end we've talked about measuring the difficulty of player ice time due to competition and where he starts his shifts, the ease of ice time due to having stronger teammates, and what special teams ice time does the player get.  In addition, we've even talked a little bit about some measures, such as Relative Statistics discussed in Part 3, that can be used to account for differences in context and allow us to better evaluate and compare players. 

This final part of the Context series introduces the concept of the Replacement Level, a concept (seemingly pioneered in sabermetrics) that is used by statisticians in every American Sport (and probably non-American Sports as well) to evaluate players and truly do so in a way that accounts entirely for the context in which those players play in. 

A quick note before we go further: this concept is a good bit theoretical and a lot less useful than what we've talked about in parts 2.1-2.3 of this guide.  If you understand parts 2.1-2.3 of this series, you don't really need to worry if this concept if it's confusing.  But it is useful to know, and hopefully will be more useful in the future.

Continue reading this post »