Thursday, October 6, 2022

2022 stupidstats 1: what a great regular season it has been for new york fans

Baseball's regular season is over, and for New York fans it has been a pretty good one. The Yankees finished in first place in the AL East. The Mets, after leading the NL East for most of the season, got swept by Atlanta in the last weekend, and ended up in second place*. But they did win 101 games, and are in the playoffs.

But was it the best regular season New York has had? Who's to say.

I actually started thinking about it early this summer, when the Mets and Yankees were both doing well, and seemed to have realistic shots at winning their divisions.

Spoiler alert: I decided to look at the product of the two teams' winning percentages. If, for example, both teams have completely mediocre records of 81-81, then my statistic would be .250. That is, .500 × .500.

I wanted a statistic that captured both teams' performances. Obvious candidates are combined winning percentage and total wins. Both have some advantages, to be sure. Combined winning percentage is the better of the two because it normalizes for the length of a season. If both teams do very well in a season that's shortened by a strike or a pandemic, total wins won't reflect how good a season it was.

But both teams fail in one regard. I wanted a statistic that gives a better score if both teams do well than if one does very well and the other is mediocre. For example, I consider it a better year if both teams go 100-62 than if one goes 120-42 and the other goes 80-82. Total wins and total winning percentage doesn't capture that. But mine does. If both teams go 100-62, my statistic is .381. If one goes 120-42 and the other goes 80-82, my statistic is .366.

My statistic doesn't account for post-season performance, which is why I said (above) that this has been a pretty good regular season. If you care about the post-season, and most sports fans do, the 2000 was the best season for New York baseball in the Yankees/Mets era. My statistic also doesn't account in any way for a team's position in the final standings. Both teams finishing in first place is better than one (or both) finishing out of the playoffs -- regardless of record. And I will get to that point in a bit.

Based on my statistic, this has been the second-best regular season for New York since the Mets' inception. See the following table. For statistics that combine the teams' records, each item is shaded in green if it is the best so far.


In 2022, the product of the Mets' and Yankees' winning percentages is 0.381. That's second to 1998, when the product was .382. But 1992's result was driven by the Yankees, who finished at 114-48. The Mets were respectable, finishing 88-74, in second place. But that wasn't good enough to make the postseason, so by any reasonable account, this has been a better year.

So, I am open to hearing suggestions for other statistics to measure the goodness of a regular season for two teams. Of course, the two teams don't have to be the Mets and Yankees. For Chicago fans it could be the Cubs and White Sox. Or it can be any two teams. Someone who likes the Royals and the Marlins can look at the combined statistics of those two teams. And similar analyses can be done for a combination of larger numbers of teams. For 50+ years New York had three teams between the AL and NL. There's no reason one couldn't look at them combined.

For what it's worth, this has been the best regular season in two regards: It's the first time that both the Mets and the Yankees won at least 99 games. And the minimum of their two winning percentages—0.611—is the highest it's ever been.

__________________________________________________________

*They have the same record as Atlanta, but MLB's first tiebreaker is head to head record (which, by the way, is a stupid tiebreaker -- but that's another matter), so they are officially in second place.

6 comments:

  1. Why not create a sharpe ratio to show the impact of the disparity of 1998 season

    ReplyDelete
  2. Now I have to Look up Sharpe ratios. Wern't they on exam 220?

    ReplyDelete
  3. If you go back to pre-1962, you'll find that New York has three teams in 1951 with a minimum winning percentage of 0.618.

    ReplyDelete
  4. Spitballing: How about a statistic to measure final regular season ranking among n teams in the division, rank r from 1 to n, score is q^-r - q^-n (so basement scores 0), q such that the sum of the scores in the division is 1? Then you can take your product of percentages and pick some weighted mean of the two.

    ReplyDelete