SBNation.com User Blog: Hermus

Pie vs. Cake Survey Results

Hermus — Mon, 04 Feb 2013 19:51:58 -0000

Thanks everyone for taking the Pie vs. Cake Survey! We got a total of 49 responses. Before I get to the results, I'd like to say a few words about the purpose of this survey. This started in Friday's links thread where RMN member FireItUp posted the results of another Pie vs. Cake tournament bracket that started a lot of discussion. Several of us took issue with the results of that contest, and there was interest in doing our own RMN Pie vs. Cake bracket challenge. In order to set up the brackets though, we would need to seed the entries, and I thought the best way to do that would be to survey RMN members ourselves.

So, think of these not so much as the final results of RMN pie/cake preferences, but as a launching point for further discussion. And hopefully, if I or anyone else can find an online bracket template that would allow us to vote on our favorites, we'll ultimately be able to do our very own RMN pie vs. cake tournament challenge (I'm looking into it but no promises).

Team Pie vs. Team Cake

First off, how many of us are Team Pie vs. Team Cake? It was not as close as I imagined. Of the 49 respondents, 25 were Team Pie (51%), 13 were Team Cake (26.5%), and 11 were Neutral or Other (22.4%). Looks like we'll need to start referring to Team Cake as a minority special interest group around here.

Seeding the Cakes

So let's start with the cakes. Here are the responses given to Single Favorite Cake:

I was a bit surprised that carrot cake was so far ahead of the rest, especially given the beating carrot cake took in the comments on Friday. We'll come back to cheesecake in a minute. Also, I got some entertaining trollers. For the most part I removed these from results but I left a few of the entertaining ones.

Now for Top Five Cakes:

I don't have room to post all the cakes that got only a single vote, but rest assured they factored into the seeding if they also got votes elsewhere (there were an additional 21 types of cake that got only a single vote here). Coffee cake got some love here, apparently although it may not be considered a favorite, it is still liked by many. Pound cake too.

Next up are the Least Favorite Cakes:

German Chocolate is quite the polarizing cake, being among the most as well as least favorites. Fruitcake rears its ugly head here.

In order to seed the cakes, I gave each cake 3 points for each Single Favorite vote, 1 point for each Top Five vote, and -3 points for each Least Favorite vote. I took the liberty of combining a few entries, such as "Yellow" and "Yellow with Chocolate Icing". I also left out any cake that was only suggested by one person. I also left out the two generic categories of Birthday and Wedding cake. I know these are loved by some, but they can cover so many actual varieties of cake that it doesn't seem fair to include them here. Finally, I left out Cheesecake. That may anger some of you Cheesecake devotees, but as you'll see it also got votes in Pies, and I think the judicious thing is to leave cheesecake out of the debate altogether. That resulted in 24 cakes, which is convenient for seeding:

I'm open to suggestions if anyone has an issue with my methods. But overall I think this would be a good starting seeding for the cake bracket.

Seeding the Pies

Now for the pies, starting again with the Single Favorites:

Again I was a bit surprised at the love for Key Lime, considering that it's not as much in the standard repertoire as Apple, Pecan, Pumpkin, etc. But you all do love you some key lime pie. Note the cheesecake votes again.

Onto the Top Five Pies:

Wow, Apple was quiet among the favorites, but really took command here. Truly America's pie? Banana Cream also came out of the shadows. Strawberry Rhubarb does better than either Strawberry or Rhubarb, proving that some mergers are truly synergistic. Again I don't have room to list all the single-vote-getters, but there were 17 of them.

Finally the Least Favorites:

The unloved Mincemeat gets thrust upon the plank. Blackbird may have been a joke here, but why not.

I seeded the pies using the same 3,1,-3 point system described above. I combined a few types such as Chocolate, Chocolate Meringue, and Chocolate Cream, although I recognize these aren't necessarily one and the same. As with cakes I left out those that got votes from only one person. Through some judicious combinations I was conveniently able to get the list to exactly 24 pies:

Summary

So there you have it. Our favorite pies and cakes. These would work great as seeds for a 48-entry bracket with pies on one side and cakes on the other. The top 8 seeds on each side would get a first round bye to get down to 32 entries.

I'll look into setting this up. I've found a few sites that offer brackets, but as far as I can tell they are just for recording results, you can't actually vote within the brackets. Maybe we could vote through polls here on RMN (or Survey Monkey) and then I could use the brackets just to show results. I don't know. But I do know I've already spent way too much time on this today, so that will have to wait for another time.

Anyway, please discuss! Let me know if you see any flaws in the seeding system. Remember these are just for starting seeds, the actual bracket votes might not turn out this way.

--Hermus

Team Pie vs. Team Cake

Seeding the Cakes

So let's start with the cakes. Here are the responses given to Single Favorite Cake:

Now for Top Five Cakes:

Next up are the Least Favorite Cakes:

German Chocolate is quite the polarizing cake, being among the most as well as least favorites. Fruitcake rears its ugly head here.

I'm open to suggestions if anyone has an issue with my methods. But overall I think this would be a good starting seeding for the cake bracket.

Seeding the Pies

Now for the pies, starting again with the Single Favorites:

Onto the Top Five Pies:

Finally the Least Favorites:

The unloved Mincemeat gets thrust upon the plank. Blackbird may have been a joke here, but why not.

Summary

Anyway, please discuss! Let me know if you see any flaws in the seeding system. Remember these are just for starting seeds, the actual bracket votes might not turn out this way.

--Hermus

With this Brilliant Compromise, all of college football's problems will be solved.

Hermus — Thu, 29 Mar 2012 15:04:45 -0000

So we're hearing a lot of whisperings about how the impending 4-team college football playoff will go down (I'm taking it as a given at this point that there will be 4 teams involved in some form of playoff/plus-1 starting in 2014). It appears that one of the most contentious issues on the table is whether to include a conference champion requirement (that is, you must be a conference champion to qualify for the playoffs). Several conference commissioners not residing in the former Confederacy favor the requirement. Mike Slive and Stewart Mandel (among probably a host of other sportswriters) oppose it.

Being from Missouri, the home of perhaps the most famous compromise in our nation's history, I am instinctively compelled to seek a middle ground in the face of any divisive argument. I therefore propose the following Brilliant Compromise (okay I'm kidding about the Brilliant part, but I actually think this is a pretty good idea)...

First off, I actually like the idea of requiring playoff participants to be conference champions. It moves the national championship one step closer to being settled on the field, and a step further away from being a beauty contest. It moves the criteria away from some subjective idea of who we all think the best teams are, and towards a set of objective requirements (win your conference championship based on clearly defined criteria in each conference). Of course it doesn't solve everything -- you still have to decide who the "best" four conference champions are (and you have to deal with Notre Dame), but that seems much less contentious than the current situation.

But I also understand that any system that doesn't include the "best" two teams would be a deal-breaker for many, and in cases like this past season just wouldn't pass the laugh test.

So here's what I propose: take the top 4 conference champions, UNLESS the 1st or 2nd ranked team (using whatever ranking system is in place) is not a conference champion. In that case, the wildcard team would bump the 4th ranked conference champ. Note that this is different from having 3 champs and a wildcard. In this case the wildcard would ONLY be eligible if ranked #1-2.

I'd like to think this would satisfy most people on both sides of the issue. We're allowing for a wildcard, but that wildcard would have a higher bar to pass than in most sports. The conference champion requirement would be the rule and would apply in probably 9 of 10 seasons. The wildcard would be the exception and would allow for cases like Alabama in 2011. If the system were in place last year, Alabama would have bumped #10 Wisconsin.

There are other details to sort out. Would the wildcard get the seed corresponding to their ranking, or drop to the 4 seed as a penalty for not winning their conference? What would happen if both 1 and 2 were not conference champs (two wildcards)? How would you account for Notre Dame? I could also consider an argument for relaxing the wildcard ranking threshold (maybe include the top ranked non-champ if ranked in top 4, for example). But those are relatively minor in comparison.

I like it, it seems like a rational compromise, so of course it will never happen (unless Mike Slive is reading this on his way home from KC).

But I also understand that any system that doesn't include the "best" two teams would be a deal-breaker for many, and in cases like this past season just wouldn't pass the laugh test.

I like it, it seems like a rational compromise, so of course it will never happen (unless Mike Slive is reading this on his way home from KC).

Thoughts on the MU-KU rivalry

Hermus — Thu, 23 Feb 2012 13:59:49 -0000

I don't know if I really have any truly original insights on this issue, but I've been thinking about it lately and just wanted to express some thoughts.

First, I understand Kansas' position on continuing the rivalry, I do. If Missouri were the jilted party, I'm sure we would feel a strong tendency to end the rivalry out of spite, and we would come up with similar reasons as theirs to feel justified in our position.

But let's clarify some things. Missouri did not choose to leave the Big 12 because they wanted to end this rivalry, or they don't respect the tradition of the rivalry, or because they're just tired of Kansas. On the contrary, potential loss of the rivalry was a major drawback of the move that had to be taken into consideration.

Missouri chose to leave the Big 12 because of the well-documented instability of the conference, and ANY school in the Big 12 would likely have made the same choice if they were in Missouri's position. Indeed, EVERY Big 12 school that has had even a hint of an opportunity to head to greener pastures over the past two years has made clear their preference to leave: Nebraska to the Big Ten; Colorado to the Pac-12;Texas, Oklahoma, Oklahoma St, Texas Tech, Baylor falling all over each other for the Pac-12's advances; Texas A&M and Missouri to the SEC.

As others have pointed out, it is an unfortunate coincidence of geography and demographics that Kansas, K-State, and Iowa State have been on the short end of the stick throughout all this. No question they would have been ready to leave had any of the major conferences been interested.

And that would be totally understandable. If Missouri were left behind while Kansas or any of the others left, it would be frustrating, disappointing, demoralizing even, but I would still understand the decision.

So in the end, the decision was about much more than the rivalry, and the rivalry should be about more than the decision. Granted that Missouri initiated the current impasse, but Kansas now has the power to resolve it.

Finally, I can understand the challenges of continuing the football rivalry, given limited game slots, financial considerations, and the pressures to schedule winnable non-conference games. But those issues are much less of a factor when it comes to basketball.

Hopefully saner heads will prevail in the near future.

I don't know if I really have any truly original insights on this issue, but I've been thinking about it lately and just wanted to express some thoughts.

Hopefully saner heads will prevail in the near future.

Impact of Losing/Returning All-American QBs

Hermus — Thu, 09 Feb 2012 17:31:42 -0000

After reading Bill C's post on the value of returning starters, I was curious to see what affect losing or returning an All-American quarterback would have on a team's ranking. I used consensus All-Americans (from Wikipedia - some years have two consensus) and AP Top 25 Rankings for comparison. If teams were ranked out of the top 25 the next year, I treated that as ranked 26.

The results? Over the past 10 seasons, 8 consensus all-american QBs have not returned, and their teams dropped 13 spots in the AP poll on average (likely worse if you consider non-ranked teams likely to actually be lower than 26).

Over the same time span, 5 consensus all-american QBs have returned the next season, and their teams still dropped 6.6 spots on average.

My take from this quick-&-dirty comparison is that losing an all-american QB is somewhat of a big deal, and it would make sense to include something along those lines in preseason ratings (such as weighting returning starters by position and by achievements such as all-american or all-conf). Returning an all-american QB however? Not as predictive. Perhaps because it's harder to maintain the high ranking the team achieved that year regardless of who's playing QB the next. Every team with an all-american QB ranked in the top 15 that year and most in the top 5, so there's not much room for improvement.

Over the same time span, 5 consensus all-american QBs have returned the next season, and their teams still dropped 6.6 spots on average.

A look at the 2011 conference schedule with Colorado staying

Hermus — Fri, 27 Aug 2010 16:41:41 -0000

With the recent reports that Colorado is likely to stay in the Big 12 for 2011, I started thinking about how the conference schedule would work next year. Let's assume that with 11 teams the conference will go ahead and do away with divisions and the championship game as they've been planning (keeping divisions with 11 teams would be a scheduling nightmare). A full round robin schedule doesn't seem likely, as that would require 10 conference games, leaving only two games for non-conference matchups. Possible I suppose, but I don't see it happening. What are the likely alternatives?

First, a nine game schedule with each team skipping one other is mathematically impossible with 11 teams. Intuitively it seems you could pull it off using bye weeks, but it literally cannot be done. You cannot defy math! Well, unless you are Bill C. I suppose.

So, an eight game conference schedule seems most likely. That means each team would miss two other teams in the conference, and that's where the controversy would begin. At least one team ~~would~~ might get to miss both Texas and Oklahoma, while at least a few would have to play both of them. How do you choose the scheduling winners and losers? Remember, it's just for one year, so we don't really need a system to stand for the long term. We need a system that addresses the current balance of power in the conference.

A random draw would be "fair", but not very satisfying, especially if any team with a decent chance of actually winning the conference next year got to skip the Big Two. Can you imagine the backlash if Mizzou "won" the Big 12 next year by going undefeated, without having played Oklahoma or Texas?

I think a more reasonable approach to choosing which schools get to skip each other would be based on recent conference standings. This is not unlike how the NFL designs schedules. Let's use the Big 12 standings over the last three years to see how this would work.

Big Twelve Standings (excluding Nebraska), 2007-2009

Texas, 20-4
Oklahoma, 18-6
Texas Tech, 16-8
Missouri, 16-8
Oklahoma St, 15-9
Kansas, 12-12
Kansas St, 9-15
Texas A&M, 9-15
Colorado, 8-16
Iowa St, 5-19
Baylor, 3-21

I'm suggesting that the top team would skip the two bottom teams, and so on down the standings. Here's how that would work out (each team is listed with the two teams they would NOT play in the 2011 season):

Texas - Baylor, Iowa St
Oklahoma - Baylor, Colorado
Texas Tech - Iowa St, Texas A&M
Missouri - Colorado, Kansas St
Oklahoma St - Texas A&M, Kansas
Kansas - Oklahoma St, Kansas St
Kansas St - Missouri, Texas Tech
Texas A&M - Texas Tech, Oklahoma St
Colorado - Oklahoma, Missouri
Iowa St - Texas, Texas Tech
Baylor - Texas, Oklahoma

One thing that jumps out is that one of the traditional conference rivalries - Kansas vs. Kansas St - would get skipped, but we could fix that by flipping two matchups that are nearby in the standings:

Texas Tech - Iowa St, Kansas St
Kansas - Oklahoma St, Texas A&M
Kansas St - Missouri, Texas Tech
Texas A&M - Kansas, Oklahoma St

So Missouri wouldn't get to face Colorado on their farewell tour of the conference, and would miss Kansas St for one year.

Overall this approach seems pretty fair. One could argue that Kansas St and Texas A&M are likely to be better than Oklahoma St or Kansas over the next couple of years, but if you wait to include the 2010 standings then that might start to show up anyway. I think this would achieve the important goal of having every team that's likely to be a contender for the conference face every other contender. To be the best, you have to beat the best.

Big Twelve Standings (excluding Nebraska), 2007-2009

Texas, 20-4
Oklahoma, 18-6
Texas Tech, 16-8
Missouri, 16-8
Oklahoma St, 15-9
Kansas, 12-12
Kansas St, 9-15
Texas A&M, 9-15
Colorado, 8-16
Iowa St, 5-19
Baylor, 3-21

Texas - Baylor, Iowa St
Oklahoma - Baylor, Colorado
Texas Tech - Iowa St, Texas A&M
Missouri - Colorado, Kansas St
Oklahoma St - Texas A&M, Kansas
Kansas - Oklahoma St, Kansas St
Kansas St - Missouri, Texas Tech
Texas A&M - Texas Tech, Oklahoma St
Colorado - Oklahoma, Missouri
Iowa St - Texas, Texas Tech
Baylor - Texas, Oklahoma

Texas Tech - Iowa St, Kansas St
Kansas - Oklahoma St, Texas A&M
Kansas St - Missouri, Texas Tech
Texas A&M - Kansas, Oklahoma St