The most important games all year?
Posted by Andy on May 27, 2010
I caught a lot of flak for my predictions in the AL East.
It turns out that the events of the last few days have changed the possibilities quite a bit.
Let's see what coolstandings.com has to say on the subject.
Going back 3 days before the Rays played the Red Sox and the Yankees played the Twins:
- After the games of May 23rd, the Rays were 32-12 and had a 74.0% chance of winning the AL East, an 18.3% chance of winning the wild card, and an overall 92.4% chance of making the playoffs. They then lost 3 straight games to the Red Sox. Their odds of winning the division fell from 74.0% to 67.8% after the first loss, then down to 65.5% after the second loss, then down to 55.0% after the third loss. (I'm sure that last drop was so large in part because the Yankees won two games yesterday.) Their overall chances of making the playoffs went from 92.4% before the series to 81.6% after yesterday.
- After the games of May 23rd, the Yankees were 26-18 and had a 19.9% chance of winning the division and a 42.8% chance of winning the wild card (and an overall shot of 62.7%.) Now, 3 days (and 2 games) later they are at 33.1% for the division, 34.1% for the wild card, and 67.2% overall. Their overall chances of making the playoffs haven't changed too much but a lot of it shifted from the wild card to the division.
- After the games of May 23rd, the Red Sox were24-21 and had a 1.6% chance of winning the division and an 8.2% chance of winning the wild card. That's 9.8% total chance of making the playoffs. After beating the Rays three straight times the odds are now 5.8% chance for the division, 14.7% chance for the wild card, and 20.5% chance for the playoffs.
- I will add briefly that Toronto has tread water over the last few days and currently have a 6% chance at the division, 13.9% chance at the wild card, and 19.9% chance overall.
The bottom line is that since I composed my initial predictions, the Red Sox have doubled their chances of making the playoffs although they are still a 4:1 shot against that happening. The Yankees have a decent shot at winning the division.
There is one big thing in the Red Sox's favor going forward. For much of April, we saw 3 big anomalies: David Ortiz's hitting, John Lackey and Josh Beckett's pitching, and Marco Scutaro and Adrian Beltre's defense. With the exception of Beckett's pitching (because he is on the DL now) all of these anomalies have been erased with performances closer to career norms and reasonable expectations, and it's no coincidence that the Red Sox's overall team performance has been vastly improved. This is all a long-winded way of saying that I'm not sure the April performance says a lot about how this team will do down the stretch. The only question is whether the 3 or 4 games they lost instead of winning in April will cost them a playoff spot in the end. My money is still on "yes" on this question.
May 27th, 2010 at 1:15 pm
Good stuff, Andy. That's the weird thing about the importance of early-season games vs. late-season ones -- if you created some kind of leverage index from the playoff probabilities at CoolStandings, you would see that games in August and September of a close division/wild card race were much more important than games in April. But the paradox is that if you don't win those games in April, you don't even get a chance to play those important games in August.
May 27th, 2010 at 1:42 pm
I am not a big fan of coolstandings' regression model. They had some issues in the AFC playoffs last year with the Colts being underdogs each game because of the way they ended the regular season.
That site is great in August/September, though. At that point, so many games are in the bank for 2010 so the model is halfway decent and the way they play out the schedule takes things into account that simple magic numbers can't (schedule strength, teams above you playing each other where someone has to win, etc).
Neil, the leverage comparison is not a paradox. Its completely analogous. You have to play well in the early innings to have a chance in the late innings, but there is less urgency because you have so many remaining chances. Take the Rockies. They have to make up 4 games on the Padres in the next 130-odd games. They do have to do that at some point... but they could do this week or next week or any other week, or even incrementally. Once September hits, it gets to the point where they need to make up those games *NOW*.
May 27th, 2010 at 3:05 pm
Right, good point. I just meant that it sometimes seems strange to assign more importance in retrospect to games later in the season when they all count the same in the standings. I know it has to do with what we know at the time of the game, and like you say, the amount of time they have left to change the outcome. But in the end, a loss in April is just as damaging as a loss in September to a team that missed the playoffs by 1 game.