At the beginning of 2008, Nate Silver was fairly anonymous; known only to those with a deep interest in baseball statistics. By the end of 2008, Nate Silver had become a pop culture phenomenon.
Silver has been projecting baseball standings and statistics with a system known as PECOTA (Player Empirical Comparison and Optimization Test Algorithm) for Baseball Prospectus since 2003. PECOTA relies on projecting a player’s performance by comparing him to past players to determine how that player is projected to perform in the future.
2008 started off with a lot of criticism for Nate Silver when Baseball Prospectus published their expected standings for the season. Silver projected the Tampa Bay Rays to finish at 82-80. This was still only good for fourth Place in the AL East, but still a huge improvement over their 66-96 record in 2007. As everyone knows by now, Nate Silver actually underpredicted the improvement of the Tampa Bay Rays. In fact, an evaluation after the 2008 season by the website Vegas Watch showed that PECOTA had the lowest error in predicting team wins for 2008 among the best-known forecasts.
What really made Silver a pop culture phenomenon in 2008 was his website FiveThirtyEight.com. Silver took his forecasting and Monte Carlo simulation methods for predictions and applied them to the 2008 Presidential Election. Silver ended up predicting the winner in 49 of the 50 states in the Presidential Election, only missing Indiana. He has also predicted the winner of every Senate race that has been decided to date (with the Minnesota election still held up in court).
In early 2009, Silver used this same method to attempt to predict the winners of the Oscars. Amongst the main categories (Best Actor/Actress, Best Supporting Actor/Actress, Best Director, Best Picture), Silver went 5 for 6, missing only Sean Penn’s win for Best Actor.
With all the successes that Silver has had over the past 12 months, now it is time to get back to Silver’s first love: baseball. Let’s take a look at Silver’s projected standings for the 2009 Season.
|
American League |
|||||
|
East |
Central |
West |
|||
|
Yankees |
99-63 |
Indians |
86-76 |
Angels |
84-78 |
|
Red Sox |
96-66 |
Tigers |
84-78 |
Athletics |
84-78 |
|
Rays |
95-67 |
Twins |
78-84 |
Mariners |
76-86 |
|
Orioles |
76-86 |
Royals |
76-86 |
Rangers |
71-91 |
|
Blue Jays |
73-89 |
White Sox |
74-88 |
|
|
|
National League |
|||||
|
East |
Central |
West |
|||
|
Mets |
90-72 |
Cubs |
92-79 |
Dodgers |
91-71 |
|
Braves |
86-76 |
Brewers |
85-77 |
Diamondbacks |
87-75 |
|
Phillies |
86-76 |
Cardinals |
81-81 |
Giants |
76-86 |
|
Nationals |
75-87 |
Reds |
79-83 |
Padres |
74-88 |
|
Marlins |
70-92 |
Astros |
69-93 |
Rockies |
72-90 |
|
|
|
Pirates |
64-98 |
|
|
Let’s go over a few questions about the predicted standings.
1) Who is this year’s Tampa Bay? From the looks of it, it would be either the Cleveland Indians or the Detroit Tigers. Both of them suffered through seasons last year that were below expectations. The Indians’ rotation didn’t come through and Detroit had injuries, poor starting pitching, and a lineup that didn’t produce as expected. The Indians added Mark DeRosa and Kerry Wood from the Cubs in the offseason and Detroit got healthy. In the weak AL Central, this could be more than enough to push one of them into the playoffs. As a side note, Silver’s model seems very high on the Washington Nationals. I think many of their fans would be pleased with a 75 win season.
2) The White Sox in last place at 74-88? Really? White Sox fans will not be happy to hear this, but in 2007, Baseball Prospectus predicted a similarly poor record (72-90) for the White Sox. This upset many fans, who felt the White Sox would repeat their good 2006 season, when they finished 90-72. But Silver’s PECOTA was dead on: The White Sox finished in fourth, at 72-90.
3) The defending World Series champs in third place in their division? Sure…Silver’s models can’t account for the September collapse that the Mets seem to have every season. The Braves also acquired Javy Vasquez in the offseason and due to his high strikeout rate and double-digit wins every season in the 2000s, he is a player that prediction models are known to overvalue. If the Braves can stay healthy and get solid pitching from their largely unknown starting rotation, they could be a threat.
4) Does Milwaukee really have enough pitching to win 85 games? Probably not. Yovani Gallardo is coming off an injury, Jeff Suppan has been getting hit hard all preseason, and Braden Looper is scheduled to start the season on the DL. The Brewers will also have the same infield and catcher issues that they had last season. The Cubs should run away with the NL Central again and their biggest threat could actually be their rivals down I-55.
5) What effect will the World Baseball Classic have on their standings? This is a very interesting question, but its effects are unknown. Are players who played in the WBC more likely to have stronger seasons than they otherwise would due to early game action? Or on the flip side, will they wear down earlier from playing in too many games? This is something to keep an eye on as we head towards August and September.
6) How does team defense play into these projected standings? PECOTA does take defense into account; however, due to the lack of defensive statistics kept, it is the most unreliable piece of the equation. So how will Michael Young moving to third base for the Rangers or Skip Schumaker moving from the outfield to second base for the St. Louis Cardinals affect their teams? The answer is that we don’t know; it’s something to watch throughout the season. Will the runs that they create on offense be enough to account for any potential defensive deficiencies they have at their new positions?
7) What are the playoff projections? In the AL, New York Yankees vs. Oakland or Los Angeles and Cleveland vs. Boston. In the NL, Chicago Cubs vs. Arizona and Los Angeles vs. New York Mets.
Of course, there are lots of factors that Silver’s system or any projection system cannot take into account. Every team has to deal with injuries throughout the season. Rookies get called up from the minors and make an impact that no one would have forecasted (Francisco Rodriguez and David Price as two examples). Teams make trades during the season (imagine last year’s predictions not knowing that CC Sabathia would become a Brewer and make every fourth start). This could cause a huge impact in 2009, especially if the economy continues to stay at this level and teams are forced to sell off superstars. That being said, predicted standings are a good water-cooler topic to get everyone talking about the baseball season.
As for me, I won’t claim that I can predict the standings as well as Nate Silver, but I hope he’s wrong about my local team, the Houston Astros (maybe the late Pudge Rodriguez signing will help). It’s going to be a long summer at Minute Maid Park if we have to suffer through a 69-win Astros season.
–J. Gluskin
Tags: Pecota, Sabermetrics, Season Preview

April 3, 2009 at 5:07 pm |
[...] 2009 PECOTA MLB Predictions (SixFourThreeBlog) [...]
April 24, 2009 at 8:53 am |
[...] pitching. The Toronto Blue Jays, either picked to finish fourth with panache in a tough division or floundering to finish fifth behind a much-hyped youth movement, are in first place in the American League East. They have one [...]