First One To 85 Wins It

Mediocrity is the new black, at least on the west coast. After last year’s debacle of a division called the NL West, where the Dodgers won the division by finishing 84-78, the parity-by-lousiness looks to be shifting to the American League West this year. The Replacement Level Yankees blog just finished up their simulation extravaganza for the American League, and the aggregate of the CHONE/PECOTA/ZIPS/THT/CAIRO/MARCEL simulations (6,000 in all) has the Angels leading the division with an average of 85.4 wins. The A’s come in at 81.1, the Mariners at 77.8, and the Rangers at 72.1.

That’s not good, and it gets worse – these projections are likely too optimsitic for the top three teams.

The Angels are going to open the season with John Lackey, Ervin Santana, and Kelvim Escobar on the disabled list. There are legitimate questions about how well and how often those three will be taking the hill, and an Angels rotation without those three isn’t much of a rotation at all.

Justin Duchscherer is undergoing arthroscopic surgery next week – how long he’ll be sidelined is up in the air, and it wouldn’t be surprising to anyone if his innings pitched for the season was something approximating zero. With Duchscherer out indefinitely, the A’s are essentially counting on five starting pitchers who have never pitched a full season in the big leagues.

The Mariners ability to take advantage of the injury bug hitting the two division favorites took a hit yesterday when Brandon Morrow announced that he wanted to head back to the bullpen and the team shipped Jeff Clement back to Triple-A to start the season.

The only team that hasn’t experienced a recent bout of bad news is the Texas Rangers, but they’re also the only team who the projection systems didn’t think had a real shot at winning the division this year – their simulated playoff odds were 3%.

This division is bad. Four flawed teams fighting for one playoff spot that none of them will deserve. Bad rotations, bad line-ups, bad defense… there’s no shortage of problems floating around the American League West this year. The only thing the division currently has an abundance of is injured pitchers.





Dave is the Managing Editor of FanGraphs.

28 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
MattS
14 years ago

You’re right that the talent is poor, but keep in mind that teams will rarely actually win exactly as many games as their talent level projects. Confidence intervals are about 15 games wide on either side, based on the binomial distribution. That means that an 81 win team will have more than 96 or less than 66 wins a total of 5% of the time. As a result, the Wests are probably likely to be won by teams with about 90 wins, even if there were only two teams in each and each were projected to go 81-81. I know that’s not what you meant, but the title is a bit misleading, since it’s unlikely either division winner would win less than or equal 85 games.

vivaelpujols
14 years ago
Reply to  MattS

First, I though that the confidence intervals were around 10 games for most projections systems, but that doesn’t really matter.

The opposite is also true though, there is a chance that the division winner could have LESS than 80 wins, simply because of how strong the rest of the AL is.

MattS
14 years ago
Reply to  vivaelpujols

I didn’t check the confidence intervals for the projection systems recently (though a got a standard deviation of around 7 a couple years ago for PECOTA, IIRC). I meant even if the projection systems were 100% perfect at projecting the talent level, the binomial theorem projects a standard deviation of sqrt((w%)*(1-w%)*162) which is about 6.4 for a .500 team, so I guess the confidence intervals are about 13 games, not 15.

You make a good point about covariance, though. I’m not sure if that widens or tightens the intervals. It could widen them because AL West teams play each other a lot and their performances negatively covary, or they could be tighter because (as you mention) the odds of each game turning up a win vary by opponent difficulty.

Jim
14 years ago
Reply to  vivaelpujols

Remember the year not too far back when the Padres won the NL West with a record of 82-80?

SG
14 years ago
Reply to  vivaelpujols

FWIW, in the 6000 iterations I ran, the AL West winner finished with fewer than 81 wins 231 times. So it’s possible, but there’s around a 96% chance it won’t happen.