MLB playoffs, round one - the Division Series

We've talked before about Bill James' playoff predictor, a curve-fitting exercise that he came up with 30 years ago to use in predicting the outcomes of playoff series.  And I like to play with it when the Red Sox are in the post-season, so I've done so for all four of the LDS.


Boston vs. Tampa Bay:
The predictor says that Tampa wins this, largely on the strength of better pitching and worse offense.

1. 1 pt to the lead team for each half-game in the standings (12 - BOS)
2. 3 pts to the team that scored more runs (3 - BOS)
3. 14 pts to the team with fewer doubles (14 - TB)
4. 12 pts to the team with more triples (12 - BOS)
5. 10 pts to the team with more home runs (10 - BOS)
6. 8 pts to the team with the lower team batting average (8 - TB)
7. 8 pts to the team that committed fewer errors (8 - TB)
8. 7 pts to the team that turned more double plays (7 - TB)
9. 7 pts to the team that walked more batters (7 - BOS)
10. 19 pts to the team that had more shutouts (19 - TB)
11. 15 pts to the team whose ERA was lower (15 - TB)
12. 12 pts to the team that has been in postseason most recently or went further (12 - TB)
13. 12 pts to the team that won season series (12 - BOS)

Total - TB 83, BOS 56

We've discussed some of these things before.  For instances, the reason, or at least the theory, that having fewer points and a lower batting average were good is that high doubles and batting average are indicative of a "sequential" offense, requiring a lot of individual successes to score runs and those individual successes are harder to come by against playoff pitching.  Whether there's any truth to that now, or ever, or whether it was just an artifact of the available data, is not something I know. 

But I do know this - if you take the 15 points for best ERA and give it to Boston instead of Tampa, the Red Sox win 71-68.  And, while the Rays' ERA is slightly lower than Boston's (3.74 vs, 3.79), when you adjust for their home ballparks (and competition - the Rays pitchers faced the Red Sox' lineup 19 times, while the Red Sox pitchers faced the Rays' lineup 19 times), the Red Sox adjusted ERA+ (a number that post-dates the playoff predictor) is better than Tampa's, 108-102.

In either case, even if the tool is telling us anything, it's telling us that this is a series between two closely matched teams, that should be very competitive.

My prediction?  They play 3-5 games, and someone wins 3. 

I hope it's the Red Sox.  They have clearly been the better team in 2013.  They finished six games ahead of Tampa in the standings, won the head-to-head series 12-7, had a better ERA+, a better adjusted OPS+, outscored the opposition by 140+ more runs than the Rays did and scored 153 more while allowing 10 more.  There's no question that the 2013 Boston Red Sox are/were/have been a better baseball team than the 2013 Tampa Rays.

That and 99 cents will get them home field advantage and a hamburger at McDonalds.

Anything can happen in a seven game series.  Anything is even more likely to happen in a five game series.  I'll go ahead predict that the Red Sox win, but I wouldn't wager a nickel on it.  To really feel good about predicting short series, there would have to be a serious discrepancy in the quality of starting pitchers, and there just isn't, here.  It's easy to imagine the Red Sox pitching shutting down the Rays offense, but it's not hard to imagine the Rays pitching shutting down the Red Sox offense, too.  And the lower the scores, the closer the games and the more room there is for a bad bounce, a marginal strike call, a bloop hit to make the difference.  There is no outcome to this series, from Boston in three to Tampa in three, that would qualify as a serious upset, a shocker.  There just isn't. 

So, here's the official prediction:  Red Sox in four.  (Because if they don't win it four, they lose to David Price in game 5.)






Oakland vs. Detroit:

1. 1 pt to the lead team for each half-game in the standings (6 - OAK )
2. 3 pts to the team that scored more runs (3 - DET)
3. 14 pts to the team with fewer doubles (14 - DET)
4. 12 pts to the team with more triples (12 - OAK )
5. 10 pts to the team with more home runs (10 - OAK )
6. 8 pts to the team with the lower team batting average (8 - OAK )
7. 8 pts to the team that committed fewer errors (8 - DET)
8. 7 pts to the team that turned more double plays (7 - DET)
9. 7 pts to the team that walked more batters (7 - DET)
10. 19 pts to the team that had more shutouts (19 - OAK )
11. 15 pts to the team whose ERA was lower (15 - OAK )
12. 12 pts to the team that has been in postseason most recently or went further (12 - DET)
13. 12 pts to the team that won season series (12 - OAK )

Total - OAK  82, DET 51

Baseball Prospectus' had the Detroit Tigers as the best team in baseball, based on their third-order winning percentage (component offense & pitching adjusted for competition and park).  Their actual record was 11 games worse, far and away the biggest discrepancy in MLB this year.  Which is more indicative of what they are as a team? 

My Prediction:  Oakland in five




St. Louis vs. Pittsburgh:

1. 1 pt to the lead team for each half-game in the standings (6 - STL)
2. 3 pts to the team that scored more runs (3 - STL)
3. 14 pts to the team with fewer doubles (14 - PIT)
4. 12 pts to the team with more triples (12 - PIT)
5. 10 pts to the team with more home runs (10 - PIT)
6. 8 pts to the team with the lower team batting average (8 - PIT)
7. 8 pts to the team that committed fewer errors (8 - STL)
8. 7 pts to the team that turned more double plays (7 - STL)
9. 7 pts to the team that walked more batters (7 - PIT)
10. 19 pts to the team that had more shutouts (19 - PIT)
11. 15 pts to the team whose ERA was lower (15 - PIT)
12. 12 pts to the team that has been in postseason most recently or went further (12 - STL)
13. 12 pts to the team that won season series (12 - PIT)

Total - PIT 97, STL 36

The playoff predictor thinks that Pittsburgh has a big advantage.  I am not convinced.  (Some of those big differences were mighty small, too.  19 points went to the Pirates for having 16 shut-outs to the Cardinals 15.  12 points to the Pirates for winning the season series 10-9, but the predictor doesn't know that Pittsburgh won four straight back in July and St. Louis has won six of the last 9).

Frankly, I'm not buying it.  Congratulations to the Pirates, because it's been a great story to have them back after 20 years in a competitive waste-land, but the ride ends here.

My prediction:  St. Louis in four




Atlanta vs. Los Angeles:

1. 1 pt to the lead team for each half-game in the standings (8 - ATL)
2. 3 pts to the team that scored more runs (3 - ATL)
3. 14 pts to the team with fewer doubles (14 - ATL)
4. 12 pts to the team with more triples (12 - ATL)
5. 10 pts to the team with more home runs (10 - ATL)
6. 8 pts to the team with the lower team batting average (8 - ATL)
7. 8 pts to the team that committed fewer errors (8 - ATL)
8. 7 pts to the team that turned more double plays (7 - LAD)
9. 7 pts to the team that walked more batters (7 - LAD)
10. 19 pts to the team that had more shutouts (19 - LAD)
11. 15 pts to the team whose ERA was lower (15 - ATL)
12. 12 pts to the team that has been in postseason most recently or went further (12 - ATL)
13. 12 pts to the team that won season series (12 - ATL)

Total - ATL 102, LAD 33

I don't like the Braves.  Never have.  Don't anticipate that I ever will.  But am I rooting for the Dodgers?  For Josh Beckett and Carl Crawford and Adrian Gonzalez?  Uh, no.

My prediction:  Atlanta in five