Back to the predictions: Last year, using 3 different methods (projected team wins based on a particular starter’s starts from 2007, projected individual starting pitcher record and potential changes on close game outcome based on roster changes), I calculated, amidst great ridicule, that the 2008 Twins will win 89 games and predicted that they will win the AL Central Division with and 89-73 record. The fact that I was close (but, unfortunately, no cigar) makes me a bit confident about my methodology.
This year, I have a couple of new calculations to add to the previous ones. Also, I am dropping the close game calculation, because, simply the roster has not changed that much from last year to warrant closer look. What I am basing my prediction will be by looking at:
- Projected wins based on projected team OPS (as I looked at the subject this post-season, it became apparent that OPS is the single best measurement that correlates with these Gardenhire Twins winning, leading to the development of a predictive formula of wins based on team OPS)
- Projected wins based on projected wins above replacement for the Twins’ players
- Projected team wins based on projected starter win-loss record (I used that last year)
- Projected team wins based on the team’s record on a particular starting pitcher starts in 2008 (I used that last year)
There we go:
This spreadsheet contains the data for #1 and #2 above
I am using 2 OPS projections (Bill James’ and Phil Mackey’s) and average them to arrive to wins. I am using the CHONE projections for WAR and set the replacement squad wins to 48, which is the average of what a replacement Twins’ team in the Gardenhire era (2002-2008) would win. Please ask questions in the comments area about the different columns, if you have any; as explaining it on a narrative will make this post the size and readability of Magna Carta.
The bottom line is: Using the OPS-based calculations, the 2009 Twins are predicted to win 89.5 (+/-3) games and using the WAR-based calculations, they are predicted to win 89.5 games. These methods seem to agree; but, again, these are new this year.
For #3 above, here are my projected wins by starter for 2009:
Bullpen 25 (average of last 4 years)
Total: 92 wins.
Also, if you look at the Twins’ record in games individual starters started in 2008 and projected for 2009 (#4) using 33 starts for Liriano, Baker, Slowey and Blackburn and 30 for Perkins, the number projects to 92 wins.
#1 89.5 +/-3
The average is 91, but I will be conservative and predict that the Twins will win 90 games (and the AL Central division) ; however it is not out of the realm of possibility that they can win up to 92. If Mauer was lost for the season, using the OPS method, the Twins’ predicted wins drop to 86.5 (-3), which makes me feel is a good reason to lower the pitching-based expectations to mitigate for Mauer’s potential lack of play.