Thursday, November 12, 2015

2015 MLB Postseason: The statistical view

With baseball season having wrapped up, it's time to return to one of my favorite topics: postseason statistics, We'll look at the MVPs and LVPs in both pitching and hitting, then introduce a series that will carry me through the rest of this year.

We started with hitters last year, so let's put the pitchers in first this time. Starting with the most harmful:

Brett Anderson
Addison Reed
Sam Dyson
Jeurys Familia
Yordano Ventura

Brett Anderson was essentially the stand-in for the back end of the Dodger starting rotation, AKA the modern version of "and pray for rain." The rain in Game 3 of the NLDS was purely metaphorical, as Anderson was staked to a 3-0 second-inning lead, then gave up four runs in the bottom of the inning and two in the next, sending the Mets on their way to a 13-7 victory and a 2-1 lead in the series. It was the only game of the playoffs in which the Dodgers did not start either Clayton Kershaw or Zack Greinke - but the Mets managed a win against each of the two aces as well.

In July of this year, Sam Dyson was about as close to anonymity as you can get when it's your fourth season in the majors; he had accrued less than 100 career innings, and was a setup man for a wildly ignorable Marlin squad. Then he moved to the Rangers in a trade that would likely have been virtually unnoticed even if Texas hadn't dealt for Cole Hamels at the same time... and became a significant surprise on a team that was itself a significant surprise. He threw 31.1 innings in the season's last two months and gave up 5 runs (4 earned), good for an ERA of 1.15. In particular, he did not allow an earned run in 16 September innings, helping pull the Rangers to a startling AL West title.

His postseason began promisingly enough, with three scoreless outings, one of them a save. And then came Game 5 of the ALDS against the Blue Jays, in which Dyson simply got caught up in the utter madness. He entered in the bottom of the seventh inning with the Rangers having grabbed a one-run lead an inning earlier when the return throw from Toronto catcher Russell Martin hit Shin-Soo Choo's bat, allowing Rougned Odor to scamper home from third. But the bases were loaded for Dyson, thanks to three consecutive Ranger errors. Dyson induced a force from likely AL MVP Josh Donaldson, which tied the game but also put him one out from ending the inning... and then allowed a three-run homer to Jose Bautista that effectively ended the series.

Commence bat flip.

I skipped Addison Reed earlier, so as to condense the discussion of the Mets' bullpen woes as much as possible. Reed's postseason efforts were pretty solid with the very notable exceptions of his first game and his last game. In the first, he entered Game 2 of the NLDS immediately after Chase Utley ended Ruben Tejada's season with a takeout slide which allowed the Dodgers to tie the game. Reed retired the first man he faced, but then allowed a pair of doubles that gave LA a 5-2 lead. The last game also saw Reed join the proceedings with the score tied at 2. This time, it was the twelfth inning of Game 5 of the World Series, and there was nobody on base when he took the mound - so the resultant rally (single, steal, groundout, RBI single, error, RBI double, intentional walk, Reed leaves the game and his relief gives up a 3-run double) falls more squarely on his shoulders.

Jeurys Familia's October got off to a much better start than Reed's, as he allowed two hits, two walks, and no runs in 9.2 innings over New York's first two series. But "those whom the gods would destroy, they first make proud." (Or, in baseball terms, "they first make the subject of elaborate graphics courtesy of the Fox broadcasting crew.") Familia allowed a game-tying homer to Alex Gordon in Game 1 of the World Series, entered Game 4 with a one-run lead and two runners on and saw an error and a pair of singles turn a 3-2 lead into a 5-3 deficit, and came into Game 5 with a runner on second in the ninth and allowed him to score the tying run courtesy of a pair of productive outs. In two of the three cases, the tying run was put on base by someone else - but Familia still allowed it to score all three times, putting him on the hook for three blown saves in one World Series. That's how you manage to be the second-most damaging pitcher of the postseason despite an 0.61 ERA.

That brings us to the most harmful pitcher of the 2015 postseason - and for the second year in a row, it's a starting pitcher for the World Series champions. Yordano Ventura started 5 games in this postseason, and failed to make it through 6 innings in any of them. More damaging, he allowed 3 or more runs in 4 of the 5 starts, and had negative WPA in all four of those (-.100 or worse in each case). The worst start of the bunch was in Game 3 of the World Series, in which he lasted 3.1 innings and allowed 5 runs, giving the Mets a victory that allowed them to close within 2-1; had Familia been more effective in his role, this might have proved more important than it did.

On the other hand, I doubt Kansas City fans are complaining too strenuously about Ventura's efforts - both for the obvious reason that they won the World Series in spite of him, and because two of his bad starts (ALDS Game 4 and ALCS Game 2) served to set up remarkable comebacks for the Royals. All's well that ends with an epic meltdown by the other team, and all that.

Enough blown leads for now; let's move on to the best pitchers of this postseason:

Wade Davis
Jon Niese
Luke Hochevar
Chris Young
Dallas Keuchel

Keuchel made two starts this postseason and pitched really well in both of them - the AL Wild Card game against the Yankees, and Game 3 of the ALDS against the Royals. He also had a very bad relief appearance in Game 5 of the same series, but Houston was already well behind at that point, so it didn't hurt them much.

Despite Houston's early elimination, Chris Young was the only starting pitcher to exceed Keuchel's performance, and even that comes with a caveat, because Young's most important outing by far was his scoreless three-inning relief stint in the first game of the World Series. Keeping the Mets off the board in innings 12-14 of a tie game was much more helpful than any of his three sub-five-inning starts - enough to push him into the top five.

Last year's top five pitchers also included a trio of Royals - but it was exactly the trio you'd expect, Holland, Davis, and Herrera. This year, with Holland out and Davis in the closer's spot, the roles were shuffled in the KC bullpen, and Hochevar was one of the beneficiaries. He made nine appearances in the playoffs this year, threw a total of 10.2 innings, and allowed no runs. Outside of his two scoreless extra innings in Game 5 of the World Series, his contributions were mostly small - but they were all positive, and they added up pretty nicely.

If you were going to see a Met pitcher on this list, you'd probably have anticipated it being one of their vaunted front three - but while Harvey, Syndergaard, and deGrom all had positive postseasons, none of them quite cracked the list. Instead, that honor goes to the man who was displaced from the New York rotation by the end of the year, and who capitalized on his move to the bullpen by... posting a 5.06 postseason ERA.

So what is Jon Niese doing here, exactly?

First, the ERA is 3 runs in 5.1 innings; all three of the runs were allowed in the same appearance, and that came in Game 2 of the World Series, which the Mets already trailed badly. Moreover, only one of the runs scored while Niese was still in the game; the other two were allowed to cross the plate by the aforementioned Addison Reed, and WPA spreads the blame between the two relievers as a result.

Outside of that, Niese put up five appearances in the postseason, all scoreless. The first two were both one-batter, one-strikeout affairs during the NL playoffs. The third, however, was a pair of scoreless frames in extras in Game 1 of the World Series. Niese then helped the Mets preserve a one-run sixth-inning lead in Game 4 (which they would blow later), and threw a scoreless eleventh in Game 5 (ditto); his efforts would eventually be for naught, but he should still get credit for them.

But the cream of this year's October pitching crop is Wade Davis, which is exactly what you'd expect from a Royal World Series victory. 8 games, 10.2 innings, four saves, one win, 18 strikeouts, 3 walks, and no runs allowed. It's a line that would fit in nicely with Mariano Rivera's playoff career, and that is about as good as postseason pitching gets.

On to the hitters, starting with the bottom five:

Troy Tulowitzki
David Wright
Travis d'Arnaud
Yoenis Cespedes
Ben Revere

What a coincidence - it's a list of five players whose teams lost to the Royals.

Tulo hit two homers in this postseason; both came in Blue Jay victories (ALDS and ALCS games 3), and in both games, his RBI totals (4 and 3) were the same as his team's winning margin (5-1 and 11-8). Despite that, his WPA totals in those two games were merely very good (+.235 and +.175), and outside of those two efforts, Tulowitzki had all of five hits in Toronto's other nine playoff games. 11 RBI in 11 games is very nice - but .205/.239/.386 is not.

The next three hitters were all Mets; since the Mets won their first two series and then lost the last one, it's not too shocking to discover that all three of these players were at worst harmless until the World Series. At that point, the metaphorical wheels detached themselves rather abruptly.

Cespedes and d'Arnaud are pretty easy to explain; they combined to go 6 for 41 with one extra-base hit, one run scored, two RBI, and three double plays in the five games of the Fall Classic. Wright's raw numbers were better (although .208/.240/.333 isn't exactly a batting line that will push you toward a Hall of Fame plaque), and he had an excellent outing in Game 3, the lone Met victory (2 for 5 with a homer and four RBI), but that was more than counterbalanced by his dreadful Game 1. Wright did manage two hits in the first contest of the Series, but one of them came with two outs and nobody on, which is not exactly the most useful time to hit a single, and after the other, he was immediately caught stealing. Outside of that, he went 0 for 5, and twice struck out to end innings with two runners on (in the third and eleventh, respectively). That's -.233 WPA, in a closely-contested and fairly heavily-weighted game.

But it's still not enough to catch Ben Revere. Like Wright, Revere had raw numbers that would send you to the bench over a full season, but wouldn't usually cripple your team over a three-week stretch (.255/.314/.277, two steals, 7 runs scored in 11 games). But his bad moments were even more poorly-timed. With his team's season on the line in Game 5 of the ALDS, Revere stepped to the plate with the bases loaded, nobody out, and the Blue Jays down a run in the bottom of the seventh - and hit into a force at home. His teammates rather notably got him off the hook for that one, but then came Game 6 of the ALCS; Revere flied out with two runners on and his team down a run in the fifth, then stepped up in the ninth with runners at second and third and one out and struck out. Throw in an 0 for 9 performance in the first two games of the series, and you get easily enough non-production to designate Revere as the most harmful hitter of the 2015 postseason.

Which leaves us with the best ones:

Eric Hosmer
Jose Bautista
Ben Zobrist
Alex Gordon
Daniel Murphy

Murphy has been covered pretty thoroughly elsewhere, right? Home runs in six consecutive games (four of which were an NLCS sweep, and his team probably would have won without him, so CPA is not wildly impressed by that series, though his LDS efforts were highly valuable), then got to the World Series and regressed hard to the mean (3 for 20 with no extra-base hits). Note that we're not penalizing his defense, which was eminently worthy of penalty, so you can give Curtis Granderson (+.146 CPA and no devastating errors) both the fifth spot on the list and the Mets October MVP if you want.

Alex Gordon did not have Murphy's raw numbers (.241/.349/.426 for the playoffs, .222/.391/.444 for the Series), though those lines aren't bad for a bottom-of-the-order hitter. What he did have was (stop me if you've heard this before) good timing, in the form of a game-tying homer in Game 1 of the Series, which makes up just over 90% of his score here.

If the refrain about timing sounds familiar, by the way, it might be because I keep saying it in this post - or it might be because I said it about Gordon last year, when he was (by CPA) the best hitter in the postseason. Between the last two years, Gordon has now edged into the top 20 hitters all time by CPA.

Ben Zobrist hit very well all postseason (.303/.365/.515), but his most important effort was Game 1 of the World Series. He led off the sixth inning with a double that sparked a game-tying rally, led off the eighth with a double and ended up stranded on third with the tying run, and singled to move the winning run from first to third in the bottom of the fourteenth. That's a WPA of +.415, and it includes only two of Zobrist's 10 extra-base hits this October.

Had the ALCS ended differently, Jose Bautista's sixth game in that series would be the stuff of legend; he hit two homers to bring his team back from a pair of deficits into a 3-3 tie. It didn't work out in Toronto's favor, but CPA is still highly impressed by his efforts in both the ALDS and ALCS - and well it might be; .293/.408/.629 is pretty good.

And that's without any extra points being awarded for bat flip.

It's becoming clear at this point that CPA and traditional numbers are never going to agree on how good a postseason Eric Hosmer has. Last year, Hosmer hit .351/.439/.526 in the playoffs, and might have been the World Series MVP if the Royals had won - but he went 0 for 4 with two strikeouts and a double play grounder in Game 7, which knocked his CPA from +.260 down to +.030.

This year, naturally, he hit .212/.236/.288, which would be appalling for a backup catcher, let alone a first baseman and cleanup hitter. But he also turned into the human embodiment of perfect timing:

ALDS Game 4: RBI single in the middle of the five-run seventh inning that saved the Royals' season, followed by a two-run ninth-inning homer to ice the victory. +.198 WPA, which is about what you'd expect from 2/5 with a homer and 3 RBI anyway.

ALCS Game 2: RBI single in the middle of the game-changing five-run seventh inning, which sounds awfully familiar; he scored the tying run later in the inning. +.149 WPA.

ALCS Game 6: Tiebreaking single in the eighth inning that scored Lorenzo Cain from first base, because of course it did, they're the Royals. WPA gives Hosmer full credit because it doesn't know what else to do with amazing baserunning. +.163.

WS Game 2: Two-out, two-run, tiebreaking single in the sixth inning. +.238.

WS Game 4: No actual hits - but he did hit the grounder that Daniel Murphy botched, allowing the tying run to score in the eighth inning. +.218.

WS Game 5: Ninth-inning double to score Cain and chase Matt Harvey from the mound; scored the tying run two batters later courtesy of a Salvador Perez grounder. +.180.

Hosmer went 4 for 21 with a double and two walks in the World Series (.190/.240/.238) - but he had six RBI (which doesn't count the run that scored on the error in Game 4), several of which were of enormous in-game importance.

I'm not sure whether Hosmer just had the worst excellent postseason ever, or the best terrible postseason ever. But whichever it was, he contributed more to his team's chances of winning the World Series than any other player this October.

You may have noticed that a whole lot of the players on these lists had a substantial contribution to Game 1 of the World Series. That is because Game 1 of the World Series was easily the best game of this postseason, and in fact ranks among the top 20 playoff games ever by WPL. Moreover, it has the highest WPL of any postseason game the Royals have won.

Which makes its occurrence a pretty good excuse to start a new series of posts: The best postseason wins by each of the 30 teams (plus a few bonus entries).

No comments:

Post a Comment