Thursday, November 10, 2016

2016 MLB Postseason: The statistical view


It's been just over a week and it still hasn't fully sunk in. But I'll try to cut through the haze of post-title bliss and be as rational as possible in breaking down what the numbers have to say about the highly memorable 2016 postseason.


There were 35 postseason games played this year. Of those 35, only five were 75th percentile or better among all postseason games by WPL, with three rating 90th or higher. Those three were:

1. NLDS1 Game 3: Giants 6, Cubs 5 (13). Cubs break the October spell of Madison Bumgarner early on a 3-run homer by Jake Arrieta. Giants rally in the eighth to take a 5-3 lead. Cubs tie in the ninth on Kris Bryant's 2-run homer. Giants eventually walk off in 13 to keep the series alive. 5.49 WPL.

2. NLDS2 Game 5: Dodgers 4, Nationals 3. Washington sent potential Cy Young winner Max Scherzer; LA sent their entire pitching staff, beginning with short-rest Rich Hill. The Nats pushed a run across in the second, but failed to extend their lead; Joc Pederson then homered in the seventh to tie the game and chase Scherzer, and the Dodgers piled on against the Washington bullpen, scoring three more in the inning. Washington responded with a two-run homer by Chris Heisey, but Kenley Jansen and Clayton Kershaw then combined to record the last nine outs (several of them with runners in scoring position) to clinch the series. 4.64 WPL.

3. WS Game 7: Cubs 8, Indians 7 (10). We'll break the individual plays of this game down in more detail as we go on. But here's the game considered as a whole: It was an excellent baseball game (91st percentile among playoff games), involving an extra inning that came after a late rally to tie. But at the same time, the Cubs never trailed, and only four half-innings out of 20 ended with the game tied, compared to ten that ended with Chicago holding a lead of at least two runs. It was a very memorable game for reasons that are extremely obvious, and there was plenty of drama included in it - but evaluated purely on swings between winning and losing, it stands as "merely" excellent. 4.35 WPL. (That's #131 all-time among postseason games - but if you want to account for heightened stakes, it's #5 among winner-take-all games, trailing 1924, 1997, 1912 [Game 8], and 1960.)

Outside of those three games, it was a fairly tepid postseason. The WC games were both good, and the LDS round was solid (in particular, all five games of the Dodgers-Nats series were above the median in WPL), but the games in the LCS round were disastrously dull (only two of 11 above the median), and four of the first six games of the World Series were 30th percentile or lower.

But nobody will focus on that - which is appropriate, because Game 7 was excellent, and at the end of it, THE CUBS WON THE WORLD SERIES!!!!!

Sorry. I'm trying.


As a reminder, these lists will be based on Championship Probability Added, which is Win Probability Added adjusted for the amount that the game changes the team's championship odds.

Best 5 pitchers:
Cody Allen (CLE) +.268
Kyle Hendricks (CHC) +.259
Jon Lester (CHC) +.198
Mike Montgomery (CHC) +.148
Kenley Jansen (LAD) +.132

Three Cubs. Over 70% of Montgomery's contribution was the last out of Game 7; Hendricks and Lester were the other major Chicago pitchers in that game (at least the helpful ones), and both also pitched well throughout the postseason. (Lester, in particular, has done that throughout his career, so much so that this postseason pushes him to +.518 CPA in total, one of the top 50 pitching figures ever.)

On Cleveland's side, Corey Kluber and Andrew Miller got most of the attention (and CPA does agree that Miller deserved his ALCS MVP). But Cody Allen was the pitcher among Cleveland's big three who never actually cracked during the playoffs. Kluber and Miller both had bad days in Game 7 of the Series, relegating them to #11 and #7 on this list, respectively.

That's going to be a theme here. When the rest of the important games in the postseason are pretty mediocre, and Game 7 is a very good one, Game 7 dominates the CPA numbers like you wouldn't believe. To wit:

Worst 5 pitchers:
Grant Dayton (LAD) -.055
Joe Blanton (LAD) -.057
Jeurys Familia (NYM) -.058
Aroldis Chapman (CHC) -.105
Bryan Shaw (CLE) -.228

Had Chapman not pitched in Game 7 at all, his score was good enough to be #4 on the best pitchers list. Had he pitched well, he could have moved up. Instead, he gave up a two-run homer in the eighth to tie the game. And then Bryan Shaw, who also had a solid positive score going in, allowed the Cubs to break that tie in the top of the tenth, absorbing the Series-ending loss.

Don't screw up Game 7 of the World Series. It's all people will remember.

Worst 5 hitters:
Dexter Fowler (CHC) -.120
Jose Ramirez (CLE) -.127
Francisco Lindor (CLE) -.155
Mike Napoli (CLE) -.222
Javier Baez (CHC) -.255

That's a surprising list, especially since the bottom player is the co-NLCS MVP. But in all five cases, a large majority of the score is from World Series play. (In the cases of Baez and Lindor, they actually entered the Series with solid positive scores, then drowned them out.)

World Series batting lines for all 5 men:
Fowler: 7 for 30 with a double and 2 homers; .233/.258/.467, 3 runs, 2 RBI
Ramirez: 9 for 29 with a double and a homer; .310/.333/.448, 2 runs, 2 RBI
Lindor: 8 for 27 with a double; .296/.367/.333, 2 runs, 2 RBI
Napoli: 4 for 24 with no extra-base hits; .167/.231/.167, 1 run, 1 RBI
Baez: 5 for 30 with a homer; .167/.167/.267, 1 run, 1 RBI

Napoli and Baez had obviously bad series. Baez's homer did come in Game 7, but it came when the Cubs already had a multi-run lead, and was countered by the fact that he struck out with a runner on third and one out in the ninth when the game was tied. (WPA doesn't care that he was asked to bunt with two strikes.) But the numbers for Fowler, Lindor, and Ramirez don't look disastrously bad.

The issue, as usual, was timing. Ramirez went a combined 5 for 8 with a double and an RBI in Games 1 and 3 of the Series, but had negative WPAs in the other five. He was 2 for 5 in Game 7, but was picked off after one of the hits; his other single started Cleveland's eighth-inning rally, but there was no particular reason to think that a three-run inning would result from a two-out, bases-empty single.

Lindor was even more surprising, as he had multiple hits in three of the first four Series contests. But even with that, his WPA in Game 1 was mediocre (he was caught stealing once), and in Game 3 it was a large negative (he was picked off to kill a significant rally in the first, and hit into a double play to kill a significant rally in the fifth, both times in a scoreless tie). And as for Fowler, despite three hits and a leadoff homer in Game 7, his WPA was a slight negative; he too was caught stealing once, and made the last out of the top of the ninth with the go-ahead run at third. He had negative figures in five of the other six games as well, and that adds up.

And now, the fun one (kind of):

Best 5 hitters
Rajai Davis (CLE) +.306
Anthony Rizzo (CHC) +.298
Ben Zobrist (CHC) +.203
David Ross (CHC) +.130
Brandon Guyer (CLE) +.121

Literally four of these five players had negative scores entering Game 7. Zobrist, Guyer, and Ross at least all had fairly neutral negative scores; Davis, on the other hand, was 0-for-the-postseason before the Series began, and when he stepped to the plate in the bottom of the eighth with the Indians trailing 6-4, was right on the edge of the bottom 5 hitters list. And then he homered for the first time in over two months, off of a pitcher who hadn't given up a homer in over four months, to tie the seventh game of the World Series. Had the Indians gone on to win, he would be a legend. As it is, his home run is likely to join Hal Smith's in the the 1960 Series on the list of seventh-game eighth-inning homers that won't be as well-remembered as they should.

And that leaves Anthony Rizzo. Rizzo's struggles early in the postseason were well-documented; through Game 2 of the NLCS, he was an ugly 1 for 23. And then, well, he got better. In Chicago's remaining 11 postseason games, Rizzo amassed 17 hits (including 5 doubles and 3 homers), 10 runs and 10 RBI. His batting line was .405/.490/.738 for that stretch. Moreover, his WPA was positive in 10 of those 11 games. It's a startlingly consistent run that easily overcame his early scuffles - and made him the most valuable playoff hitter on the first Cub world championship team in over a century.

By the way, did I mention that THE CUBS WON THE WORLD SERIES!!! Because honestly, that strikes me as the major take-home message here.

No comments:

Post a Comment