Saturday, January 30, 2010
Monday, January 25, 2010
Many people feel that 2009 is the year that clearly shows that a playoff is necessary for college football. Maybe an eight-team playoff would have worked pretty well this year; however, it seems likely to me that the strongest team won. While there was some controversy regarding the fact that three undefeated teams did not get a chance to play for a national championship, that seems to suggest to me that a four-team playoff this year would have been a disaster. At least one of the undefeated teams would have to be excluded and Florida with only one loss to Alabama would also have had a pretty compelling case. In fact the winners besides Alabama were the fifth and sixth teams in the BCS and very well might have been the teams excluded from a four-team playoff. It is interesting that after the season all of the playoff talk now discusses an eight-team playoff as opposed to a four-team playoff. If a four-team playoff had existed, I'm sure playoff creep would have already crept in to an eight-team or sixteen-team playoff.
Wednesday, January 20, 2010
1. The Wall Street Journal has a story on the value of being a fan of the Vikings. (H/T Marginal Revolution) See King Banaian and Phil Miller for more perspective. King in particular has a positive view of the study.
2. Yet another story on how hosting a Super Bowl is not a windfall to the local economy. The story quotes a few sports economists in arguing against large increases in economic activity due to hosting. At Division of Labor one of the economists, Craig Depken, further defends the "skeptics'" view.
Tuesday, January 19, 2010
The prospect of a new stadium for the A's in Fremont is back on the table. Given the economic and budget situation in California currently I do not see this as particularly likely. However, from a geographical point of view, Fremont is the optimal location given the rules in place. Unlike Chicago, New York and Los Angeles, the Giants and A's do not share the territorial rights for the Bay Area. The A's have Alameda County plus other areas. The Giants have San Francisco, Santa Clara County plus other areas. At some point in the past San Francisco and Oakland were probably the obvious likely locations for the two teams, but now San Jose is the largest metropolitan area in the Bay Area. It is also quite wealthy. Fremont represents the closest city to San Jose that is currently in the A's territorial rights, or more importantly outside the Giants' territorial rights. The stadium may not be built for a number of years, but if a stadium is eventually built I would not be surprised to see it in Fremont.
Sunday, January 17, 2010
With the disappointing seasons by the Cubs and Bears, I have been forced to follow my alma mater's football and basketball teams. Fortunately, William and Mary football had a pretty successful season in reaching the semifinals of the I-FCS playoffs. This result is not that surprising given that the team just missed the playoffs last year, it was the third playoff trip this decade and they reached the semifinals back in '04. What is more surprising is the performance of the basketball team, which has started off the season 14-3. William and Mary does not have a particularly renowned history in basketball, being one of the five teams that have been in the NCAA since the inception of the tournament never to have been to the tournament. When I went to school, "the game" was when the team came back from 23 down to Virginia to force overtime, which they proceeded to lose in overtime. How many teams have their most memorable games in losses?
This year, the team is winning by taking a lot of three-pointers and making a lot of them. The style of the team allows them to come back from way behind at times, as can be seen in this video of their comeback against Delaware last week, where the team came from seven-down with 32 seconds to win (H/T CAAHoops). Hopefully the success will continue all the way to the tournament.
Wednesday, January 13, 2010
There has been much consternation in Cubland over the trade of Milton Bradley for Carlos Silva. As a Cub fan, I cannot say I was happy about the trade when it happened. J.C. actually thinks it did not hurt the Cubs too much, though that is mostly based on the combination of Silva not being completely worthless and the six million dollars. But most importantly, J.C. does not value Bradley particularly highly. Interestingly, I do not think his methodology accounts for much negative value for Bradley's attitude. Baseball Analysts estimates that that could cost his team as much as 1.5 wins above replacement, though this might be hard to look at on straight value as there is certainly an asymmetric effect. Bradley's attitude costs the Cubs more than it will cost the Mariners. The analysis from the Baseball Analysts, however, generally feels the Mariners won the trade. Presumably the difference here is from a higher estimated on-the-field value for Bradley.
An example of a bad analysis of the trade comes from Fangraphs. There are two main problems with their analysis. The first problem is the assumption that Silva will replace Gorzelanny in the rotation. First, if Gorzelanny is the much better pitcher, even the Cubs would presumably start him regardless of how much they are making. Second, Lilly is hurt, so the Cubs will have to need an additional starter. Given that Lilly was the only lefty in the rotation and that Gorzelanny is a lefty and Silva is not, that one of the spots in the rotation would be Gorzelanny's to lose. Lastly, even if the Cubs have no plans for Gorzelanny, he still has trade value and the Cubs could recoup some of his value.
The second problem is completely not accounting for the difference in defensive value because Byrd will be playing CF and Bradley would be playing RF. Their assumptions of the trade being essentially Silva and Byrd for Bradley, turned out to be correct as the Cubs used most of the money from the Mariners to sign Marlon Byrd. Pretty much any analysis of their offensive contributions will conclude that the Cubs are taking a hit to their offense. Some of the commenters at the original post suggest that a defensive position update needs to be made. However, I think that is slightly incorrect. They are essentially replacing one for the other in the lineup, so the offensive contribution is correct. However, at the same time, the Cubs have dramatically improved their defense with the move. This table summarizes the defensive issue:
Fielding Bible Zone Rating
Player value in RF value in CF value in RF value in CF
Byrd -5 0 to -9.5
Fukudome 3 -5 13.1 -18.1
Bradley -7 -6.9
Even if Byrd is not much of an upgrade over Fukudome in CF, Fukudome is a big upgrade over Bradley in RF. Regardless of whether the improvement in defense outweighs the downgrade in offense, any analysis of the trade ignoring defense is going to be insufficient.
The most distressing thing about the analysis is that it comes from a sabermetric site (and a generally good one at that).
Monday, January 11, 2010
When I woke up on Sunday morning, I had attended the highest scoring NFL playoff game in history. This morning that is no longer true. For a long time that Eagles-Lions game was a trifecta for me: the only NFL game I ever attended was the highest scoring playoff game in NFL history, while being the coldest I have ever been. I have attended two Rams games in the last few years, so now only that last part is still true. I would really prefer to avoid breaking that (admittedly subjective) record.
Friday, January 1, 2010
It seems every season there is a discussion about teams resting their starters for meaningless/nearly meaningless games at the end of the NFL season. These discussions get a little confusing because there are three different questions associated with the problem. This season the questions have come up with the Colts-Jets game last week and the Bengals-Jets game this week. Since most writers at least relate to the first two issues, I will just link to a number of examples to begin with (see here,
here, here and here)
- Is it optimal for the team to rest starters? The optimal choice in this question differs between the Bengals and the Colts. When the Colts played the Jets, they still had one regular season game and were locked into a bye. A case could be made that they took their foot off the gas too soon. In the Bengals case, however, it is their last game and they will not be getting a bye, so they will have to play next week as well. Not only that, but their last playoff trip was ended when their quarterback got injured on the first series of their first playoff game. For that reason, I think they will be a little worried about injuries. I think I would be inclined to not go all out in the last game.
- Is it moral for the teams to not try/go half speed in games that don't matter? Is it ok to do so unless the other team is going to be in the playoffs? I will not discuss the first question, but I will point out a problem with answering in the affirmative to the second. Let us say that the 4 and 5 seeds are locked in place, but that the 4 seed is playing a contender for the last playoff spot and the 5 seed is playing non-contender. If there is an advantage to resting players (see question 1) then the 5-seed would have an advantage over the 4-seed. That result does not seem fair to me.
- Is there a way to design a playoff system that does not lead to teams resting players for key games for other teams? Jerome Bettis suggests that the NFL fix this problem but does not suggest any solutions. One reason he does not suggest a solution (and for that matter why the NFL has fixed the problem yet) is that there are no easy solutions to this problem. Pro-football Reference suggests three proposals that would help in some circumstances. Proposal 1 would eliminate the automatic home game for division winners, meaning that relatively poor teams that have clinched the division would still have an incentive to win to get a better seed. Proposal 2 at the same link would have a flexible number of playoff teams, meaning a 2-seed would lose its bye unless it was at least three games ahead of the 7-seed. The third proposal would be to have flex games for the last two games of an 18-game schedule. These flex games would be chosen so that playoff contenders played other playoff contenders. There are a couple of problems with these solutions. Proposals 2 and 3 are pretty revolutionary. Proposals 1 and 2 would not have fixed the Colts' and Bengals' problems this year, not that fixing this particular problem was the only idea behind the proposals. The only way to completely fix this problem would be to have no wild card teams and guarantee that teams only play intra-divisional games the last two weeks of the season. So unless you are willing to suggest that or something truly unusual like the above, it is simply a problem you will have to live with.