MLB has talked a lot aout the supposed 'parity' of baseball in recent years, the go-to stat: 7 teams have won the world series in the last 7 years. At first glance, this is an impressive number; no other of the big 4 sports have matched that run of new title holders. However, baseball is different than other sports. It is more random, in any given day a worse team is more likely to win than in other sports. This is reflected in the gruelling 162 game schedule, twice what any other league plays and 10 times the number comprising a football season. the longer season works out random fluctuations and, in the end, does a pretty good job of picking the best team. The playoffs, however, are short. Anything can happen in a 5 or 7 game series. At the most, a team plays 19 playoff games (with a 11-8 record). This is not enough to work out an average. Noone is worried if thoeir team is in last place after 19 games, and noone is particularly excited if their team starts 11-8 (granted, there are many games to go, but if the first 19 games were a reliable predictor of the rest of the season, this dismissal would not be justified). So if we do not pick the best team based on the first 19 games, why should we base our choice on the last 19?
Now, even if we accept that the amount of turnover among the best teams in the league will tell us much about parity (a complete analysis of this issue would have to also include turnover amung the worst teams, and the difference in records between these extremes), we would be much better off looking at season record than playoff results. So, here are the top 5 teams by wins over the last 7 years (from ESPN.com):
2006 - (WS: STL)
NYY-97
NYM-97
MIN-96
DET-95
OAK-93
2005 (WS: CWS)
STL-100
CWS-99
NYY-95
BOS-95
ANA-95
2004 (WS: BOS)
STL-105
NYY-101
BOS-98
ATL-96
LAD-93
2003 (WS: FLA)
NYY-101
ATL-101
SFG-100
OAK-96
BOS-95
2002 (WS: ANA)
NYY-103
OAK-103
ATL-101
ARI-98
STL-97
2001 (WS: ARI)
SEA-116
OAK-102
NYY-95
HOU-93
STL-93
2000 (WS: NYY)
SFG-97
STL-95
ATL-95
CWS-95
NYM-94
(I wish Blogger had a better spreadsheet option)
I included the World Series winners each year to demonstrate my point about the randomness of the playoffs. Notice that not once in the last seven years has the World series winner been the winningest team in the league. In fact, in the last seven years, only twice has the WS champion even been in the top 5! Indeed, the playoffs are a very poor indicator of the best teams. Only 8 teams even make the playoffs. It is really remarkable that those 6th-8th teams are responsible for 5 of 7 titles - much worse at picking out the best teams than one would expect even from simple randomness. (What can or should be done about this is another issue entirely. Perhaps it would be best to change the format somehow - lengthen series? play a round-robin round? Even if these were viable options, nothing will be done since this would appear to reduce the parity, as shown below.)
Getting to the meat of the analysis, we can see that 6 different teams have occupied the 9 (ties included) possible spots as the winningest team in the game. This is not too bad, but looking at the top five totals, only 14 teams have occupied the 35 potential spots. The big winners are the Yanks with 6 appearances in the top 5, St. Louis with 5, Oakland and Atlanta with 4, and Boston with 3. These five teams alone count for 22 appearances, almost 2/3 of the possible spots. In a field of 30 teams, this does not seem to represent a great degree of parity to me.
It is kind of hard to know what this really means without the full analysis I mentioned above, and a complete comparison to the same results for the NFL, NBA and NHL, while also balancing for the number teams in the league (i.e. this analysis would appear to signify more 'parity' in a smaller league, as teams are more likely to appear in the top 5 by law of averages). This, of course, is more work than I am willing to do, as I am not paid for this. However, it can at least be said that the annual turnover among the best teams in the league is significantly less than the 7-in-7 championship story implies.
3 comments:
I have two comments about this piece. The first one is that in response to your question of how we can change the playofs to more accurately reflect the "best team," while we could never really do so without rendering moot the idea of playoffs as distinct from the regular season, I think that the best and most obvious option is to lengthen the first round to a best of 7.
Secondly, I would be horribly against this if they ever actually tried to do it since I am emotionally attached to the traditional nature of baseball (I still don't really like interleague play, for example), but sometimes the intellectual side of my brain wonders if it wouldn't be better to adopt a more soccer-style format. This would work in the sense that the regular season and the playoffs would be played at basically the same time with both competition interspersed throughout the course of the traditional season. In addition, if we adopted the soccer format, every team, including minor-league and/or independent teams would have a chance for the "MLB Cup" while the top 4-6 teams from each league (American and National, that is) would qualify for the "Champions Playoffs" which would function very similarly to our current system except it would begin with a round robin first round. Anyway, I think it is an interesting idea but one that will never really happen.
Your idea raises two really interesting ideas to me. First, people seem to take the antitrust exemption for MLB for granted. If Bud allowed anyone to start a team, and then introduced it to the league the way you mentioned (that is, best minor league and worse major league swap), the landscape would be very different. For instance, in a real free market, there would very likely be another team in NY and maybe one more in the LA area. This, of course, is not the only way for such a free market of teams to develop, but it's an interesting argument for revenue sharing: that is, teams like the Yankees and Dodgers get huge markets with exclusive markets protected by MLB, so they are handed an advantage. They are not richer because they are smarter or more shrewd, they are literally richer because MLB puts them in that position. The other side of this idea is the possibility of a form of revenue sharing based on an on-the-fly sort of arrangement of teams, allowing them to change cities and enter and leave the league more freely
The other point is that this kind of system would seem to work against the control MLB teams have over top prospects and rookies. That is, if AAA teams have a chance to enter the ML, then it doesn't make any sense to tie them to a ML team the way they are currently. This would, of course by a bonanza for the player's union and agents like Scot Boras would love it. I'm not sure what it would mean for team salaries, but there would liekly be a lot more money and long term deals out there for young players, and the older 30+, more traditional free agent would probably not get as much. Just specualtion, but I think it's interesting.
I don't know if you were going to post on this, but these are two things to think about if you do.
On a somewhat related note, here is an article with a very interesting idea for a kind of revenue sharing: http://www.baseballprospectus.com/news/20020815zumsteg.shtml
One other thing, as to your mention of "rendering moot the idea of playoffs as distinct from the regular season," I agree that this is a concern. However, if that very basic analysis is right, we could make it a lot more reliable without getting near that point. I think an interesting study might be to compare these correlations in different eras with different numbers of playoff teams and different playoff structures. The one point that immediately comes to mind is that entrance into the playoffs has historically been much more restricted than it is now. This guarantees a better correlation between season and playoff success. However, it's just not very much fun. I do think that the current 8 playoff team setup is good for baseball, increasing interest nationally with better pennant races, and locally with more teams having a shot (Hell, the Sox were wild card champs in '04). I don't know how one would statistically get around that point, but it still might be worthwhile.
Post a Comment