Wednesday, October 8, 2014

Clayton Kershaw's Postseason Woes Lie at Mattingly and Colletti's Feet

Mounting Pitch Counts and Fatigue: an Old October Tune

Clayton Kershaw is not the person to blame for Clayton Kershaw's postseason troubles.

To preface the rest of this article, let me say that I am not a Kershaw apologist; he held the ball and failed to execute his pitches.  I am sure that Kershaw feels he is more to blame than anyone else in letting the Dodger's lead slip away thrice to the Cardinals in 2014 and '13 and once to the Phillies in '09.

But there are some that I believe are more to blame for the Dodgers losing when Kershaw has been on the mound in October than Kershaw himself.

Kershaw in spring training, 2010
Going into Tuesday's do or die game for Los Angeles, someone monitoring Kershaw's postseason performances could have told you that near the top of his pitch count, the 75+ range or 6th and 7th innings, the wheels start to come off.  Heck, even someone not necessarily monitoring the games but just casually watching could tell you that near the end of Kershaw's performances things have gotten away from him.  But for the sake of thoroughness, I decided to compare the two time (soon to be three) Cy Young award winner's pitching performance by pitch count in both the regular and postseasons, breaking the numbers down by 25 pitch segments and looking at average against, on base against and slugging against.  This examination forgoes looking into how many times a pitcher has gone through the lineup, but I think that in many case the pitch count segments below go hand in hand with lineup turnover.

Here is what I found going into Tuesday's Game 4 of the NLDS:

Kershaw's regular and postseason triple slash line against (45IP postseason)
Over the regular season, a sample size of 1,378.1IP, Kershaw's triple slash line against doesn't fluctuate all that much regardless of how many times he throws the baseball.  But come October, which admittedly is a much smaller sample size of 45IP, Kershaw's numbers sharply climb once past the 75 pitch count.  This has traditionally been the 6th or 7th inning, the times where the Cardinals have hurt him the most in the past two postseasons.  Over 44 PA, hitters in October were combining for a slash line of .333/.389/.514.  In other words, it's like having Miguel Cabrera standing in the batter's box every time Kershaw pitches beyond 75 tosses.  So why did Mattingly leave Clayton in?  Well, he had up to that point dispatched just about every Cardinal he had faced going into the 7th inning, not to mention that GM Ned Colletti's $33M bullpen couldn't be trusted with a lead until the 9th.  But what history has told us up to this point, including the very recent Game 1 of this NLDS, is that Kershaw shouldn't be relied on so heavily in the postseason past the 75 pitch threshold (or perhaps the 85+ range, as that is where the bulk of the damage has come against the Dodger ace).  Wouldn't it be a better bet to roll the dice with one of your "untrustworthy" relievers, or even one of your starting pitchers that could have been ready out of the 'pen?

Well, Mattingly didn't do that, and this is what happened:

Kershaw's regular and postseason triple slash line against (51IP postseason)

Of course the slugging against skyrocketed in the 101+ pitch count range, as Matt Adams' go ahead home run came on Clayton's 102nd pitch of the game, and the number of PA in that range is a measly twelve... but it's a pitch count threshold that Kershaw has absolutely faltered in during postseason play.  To be fair, in Kershaw's final inning of the NLDS Game 4, the two lead off singles weren't exactly screaming shots off the bat; Matt Holliday reached out of the zone for a hanging breaking ball to hit a grounder back up the middle, followed by Jhonny Peralta's pretty well hit liner on a cutter right down the heart of the plate.  Kershaw then hung a breaking ball to Adams and the left hander's day was done. I think that line of thinking echos pretty well what Jeff Sullivan wrote on Fangraphs last Monday when he broke down Clayton's 7th inning in Game 1 of this NLDS.  He seemed to be fatiguing and missed his spots (a sentiment that Pedro Martinez also shared on the TBS post game show the night of the game).

What the 7th inning says to me is that Kershaw was tired, missing his target and not finishing his breaking pitches as well as he should have.  Is he to blame for the Dodgers losing Game 4 of the NLDS?  I don't think so; a team that has to put all of its stock into one starting pitcher in a decisive game is just trying to outrun the baseball grim reaper in October.  And, I mean, it's not like the guy was pitching on just three days rest...

For a bit of perspective, I dug through Pedro Martinez's postseason game logs, recording by hand his triple slash line against per pitch count range (much as I had to do for Kershaw; if anyone is aware of where to find this data for postseason pitching, please let me know).  Unfortunately, I couldn't find pitch by pitch tracking for Pedro's '99 and '98 postseason performances, so my data had to start with the 2003 season (which is a shame, because he was particularly dominant in '99).

Pedro Martinez's regular and postseason triple slash line against (2,827.1 IP regular season, 72.1IP postseason)
While Pedro's postseason pitching was fairly consistent with his regular season performance, things did spike heavily after the 100 pitch mark (but here again is a very small sample size of 22 batters faced).  All sets of data point to a need for managers to be a bit less hesitant with taking the ball from their ace.  After riding them for 100+ pitches nearly every time out, and often times 8+ innings, you can't expect the exact same against good lineups in October.  Kershaw threw a whopping six complete games this season... so to the people pointing to his lower than usual IP this year ("only" 198.1), I'd point to how much he threw over a shorter than usual season.  

Clayton Kershaw has thrown just a few too many pitches in postseason starts; had he been pulled when he probably should have been, this whole "Kershaw is not as good in the postseason" idea wouldn't even be on anyone's radar.

Photo courtesy of SD Dirk (Kershaw)


  1. Kershaw looks more & more like Goatboy these days during the postseason.

  2. That's all well and good but this is the problem with pitchers today - PITCH COUNT. They don't know how o pitch out of trouble and rely too heavily on the bullpen. Kershaw may be the best 2/3 of a game pitcher in the last three seasons but he is certainly not the greatest of all-time. When he starts matching Drysdale, Kofax, Bob Gibson, Goose, etc. then you can start talking great. I blame the system of coddling pitchers, an over emphasis on velocity and an extremism on pitch counts for why these guys are getting hurt and cannot pitch an entire game. A pitcher that can only go 6 IP on a regular basis is a called a spot-starter.

    1. I'm certainly not claiming Kershaw is the greatest of all time, just that in the postseason he needs to be managed more aggressively rather than trying to ride him for a complete game like the team does in the regular season. And I think it's a big stretch to claim he regularly goes 6 IP, unless you are just talking about October starts.

      I have to ask though, how does a system that "coddles pitchers" lead them to get hurt? Do you believe that the reduced pitch counts are enabling pitchers to throw harder and thus overexerting over a shorter period of time?

  3. It's not coddling, it's the way they coach pitching today. Starting pitchers used to be taught how to pace themselves to go a full 9 innings. To do that, you can't go all-out on every pitch; you have to be willing to pitch to contact and only go for the K when you're in a jam. But today, they teach starters to air it out on every pitch, which is why batter strikeouts have gone through the roof, and also why pitchers are gassed after six innings and 100 pitches, and also why they blow their arms out and need Tommy John surgery.

    You're right about the over-emphasis on velocity, but as long as they keep that up, the pitch counts have to stay in place. If you let a pitcher throw 140-150 pitches a game the way they make you pitch now, you wouldn't have a single MLB starter make it through a whole season except for Mark Buehrle, and him only because he doesn't have a 95 mph heater and HAS to pitch the old-school way. Which, not coincidentally, is also why he just completed his 14th consecutive season of 200 or more IP.

    1. This is my rationale as well, John. I'd like to look at guys like Buehrle, Vargas, and other soft tossers side by side with Sale, Archer, and other flame thrower starting pitchers, particularly in the 101+ range and 3rd and 4th times through the lineup.

  4. I'm wondering how pitchers like Spahn, Marichal, and the like pitched so many complete games and overall innings, often on 3 days rest. Does anyone have ideas?

    1. One of the things that both guys you mentioned did really well was cause a lot of deception with their delivery... but obviously there is a lot more too it than that. My first guess would be the difference between advanced scouting and video between then and now. The amount of information on a guy, even one of the greats like Spahn and Marichal, would be worse than the report on Kershaw and Felix Hernandez are today. That would allow them to have more of an edge going into the game, I'd think.
      I really wish we had numbers for the velocities that they were throwing at back then, so we could better compare them to today's guys. Maybe Spahn and Marichal's deliveries kept their arms from getting tired as easily, with how full body their motions are?