EconoBall: Kevin Kouzmanoff, Greatest Clutch Player In the Last 20 Years?

EconoBall: Kevin Kouzmanoff, Greatest Clutch Player In the Last 20 Years?

Is there such a thing as a clutch player?  Sabermetrics says no but I've already found a couple problems with commonly accepted Sabermetric theory, so I thought I'd let the data speak on this.

First, it's important to define what I mean by clutch.  I'm going to define "clutch" player as someone who performs better in high leverage situations.  That is, a clutch player is more likely to get a single with a runner on third than a non-clutch player.

I'm going to use RE24 to get a better look at this.  The idea here is that a clutch player will have a statistically higher RE24 than a non-clutch player because he will get more hits than average with runners on base or in two-out situations -- both of which are likely to drive RE24 up.

I started by collecting data on every hitter in the game for the last 20 years.  (Pithcers were dropped.)  I then regressed RE24 on PA, singles, doubles, triples, home runs, sacrifice flies, sacrifice bunts, double plays, walks, and strikeouts.  The idea here is that if a given player is as likely to, say, hit a single with the bases empty as he is with a runner on third, these statistics should define much of the variance in RE24.  The regression results are quite good.  My regression explains 85% of the variance in RE24 and every individual statistic has the expected sign and is significant.  (That is, strikeouts detract from RE24, home runs increase it, etc.)

This regression allows me to predict what their RE24 would be if there was no luck involved in when they happened to get their hits.  I define "clutch" as the difference between their actual RE24 and this predicted value.  It's important to note that, over the course of a season, some players will be lucky and have RE24s well above the predicted value and others well below the predicted value.  However, in the normal course of events, a player will have both good luck and bad luck over the course of his career.  Some years will be well above average, others significantly below average.  So, to identify a clutch player, we need someone who put up an RE24 significantly above the predicted value over a career.

To test this, I average the "clutch" statistics for every player who played four or more seasons over this stretch.  This comes out to 1293 different players.  Only five of them had statistics significantly greater than zero.  The list is a surprising one: Brandon Larson, Yamaico Navarro, Andrew Romaine, Peter Bourjos, and Kevin Kouzmanoff.  Of the five, Kouzmanoff has by far the best record, outperforming his predicted RE24 by almost 7 runs per season.  On the flip side, four players have RE24s statistically below their predicted value: Jason Bates, Matt Walbeck, Mike Gallego, and Adam Piatt.  Bates is the worst of these, averaging an RE24 almost 4.5 runs a year under the predicted value.

When dealing with statistics, it's well known that sometimes statistical noise causes "false positives."  Looking at the data, I'm pretty convinced that's the case for all 9 of the players mentioned above.  Thus, it turns out that Sabermetrics is right: there are no clutch players.

Filed under: Uncategorized


Leave a comment
  • Does Bill Mueller buy that?

  • fb_avatar
    In reply to 44slug:

    Heh, yes, I should probably include a disclaimer to that effect ;-)

  • While it may be statistically true that there is no such thing as a clutch player...

    The Cubs have still cornered the market on non-clutch players. I'm pretty sure those exist.

  • fb_avatar
    In reply to HackWilson09:

    Actually, no. The data suggests that there are neither clutch nor non-clutch players. I would suggest the Cubs seem worse because scoring opportunities are so rare, it's memorable when they don't capitalize on them.

  • In reply to Mike Moody:

    Pat Tabler would like to have a word with you. Seriously, Some people seem to focus better when the tension mounts. He may be an outlier, but still would have been nice to have had him for his career.

  • In reply to IVYADDICT:

    Better focused and relaxed. 'Relaxed intensity'

  • In reply to 44slug:

    I define a clutch player as one who can maintain proper mechanics while under pressure.

    And you see guys who can't get the "yips".

  • In reply to IVYADDICT:

    Run the Stats for Ken Oberkfell when he was with Atlanta. He was one of the best hitters in the game with nobody on or the game was out of reach, but was horrendous if it mattered.

  • Completely ignores whether or not he number of runners on base and other opportunities to be "clutch".

    Check out the section of Cabrera, where it talks about his OPS with runners on base for a nice way to look at clutch:

    I think that sabre heads tend to think of RBIs as a random thing, Miggy's stats say the opposite, and underline just how great a hitter the man is.

  • fb_avatar
    In reply to SFToby:

    Actually, the reason I used RE24 here is because it explicitly takes into consideration the situation when a batter comes to the plate. Failing to get a hit with runners on base hurts your RE24 much more than failing to get a hit with the bases empty. A strikeout with a runner on third and one out is much, much worse than striking out with a runner on third and two out, etc.

    If you're curious, Cabrera was simply outstanding last year, pretty much at his expected value for 2012, and well below average in his first two seasons with Tigers. He's been very good player but, statistically, that appears to be because he is the best pure hitter of his day; his "clutch performance" looks to be distributed around zero.

  • In reply to Mike Moody:

    So essentially, if you have a high BA, and power, then your good statistics imply you are going to perform well in clutch situations, and if you have a low BA, low power, such as Barney last year, your performance in clutch situations is also going to be what is expected by your statistics? Therefore, Cabrera has a neutral RE24 and Barney may have a neutral RE24, or about "0" because their stats predict their expected results?

    A better indicator of clutch isn't an innate talent for performance under pressure, but instead your stats are an indicator of expected results in all circumstances?

  • What does re24 do for sac flies?

    It would be exhaustive to figure both ops and RBIs relative to pitch counts/men on base, combined with game situations (games tied or one-run games), but I think that may be a better way to measure clutch.

  • fb_avatar
    In reply to SFToby:

    A sac fly is a net positive. Score the run, which adds +1, but drop a runner from base AND add another out, which makes it more difficult for the next guy to plate a run, and thus is negative. On balance, a sac fly is better than a strikeout and worse than a single -- which is exactly what you'd expect.

    A decent chunk of what you're asking for is actually captured in RE24. It analyzes every one of the players at bats and gives him a score based on the runners on base and outs when he stepped to the plate, and the runs, runners on base, and outs after he leaves. It's a great (and underappreciated) statistic.

  • In reply to Mike Moody:

    Thanks Mike. I learned something new today. But I'd much rather have a Cabrera on my team than a Kouzmanoff, despite what you're saying.

    Defining clutch is like defining porn - you can't define it, but you know it when you see it.

  • In reply to SFToby:

    Mike would agree with you. Kouzmanoff was an outlier. Cabrera was just what he is expected to be.

  • In reply to Quasimodo:

    If you look at that link I posted earlier, you'll see that Miggy gets better as the count gets deeper and/or the runner is closer to scoring. If doing that consistently means he isn't clutch because he does it consistently, then the whole exercise is pointless.

  • In reply to SFToby:

    Do not all hitters hit better the more pitches they see? I am not talking 3-2 but how about 3-1? I think it is possible that some seasons some hitters exceed their career averages at hitting with runners in scoring position. But then again as Mike mentioned, getting a hit with runners in scoring position is better than a SF. It is more of a function of how much you produce per out you create. (Either you get out or you hit into a force out). But does that BA with runners in scoring position stay consistent over a career? And obviously if a guy has a .330 BA and hits .330 with runners in scoring position, generally he is going to score a "0" on that RE24 scale. In other words, he doesn't exceed his predictability.

  • Could you factor BABIP into these analysis on being clutch? How about at a team level? Who's the "clutchiest" team?

    I think an interesting analysis could be done on hitting strategy here, where if you have a guy on third and fewer than two outs, a lower-risk single or flyout is perhaps the better route to take than a higher-risk swing for the fences.

    If I recall correctly, I remember hearing something about the Cardinals preaching the safe, measured hits in order to simply get the runs across, where the Cubs seemed almost the opposite last year (lots of slugging, few runs).

    I understand that the vast majority of the measures say there's really no such thing as "clutch", but realistically, these guys are human. Players are going to be more or less anxious, and they're going to be more or less selfless. There's got to be something, somewhere.

  • In reply to Matt Mosconi:

    Good point. One would almost have to defy the human condition to disregard 'clutch'.

  • In pro sports where everyone is highly skilled mental toughness is the often the difference between the winners and the losers.

    Make the other guy choke and you win.

  • If there is such a think as a clutch hitter, it would be because of mental toughness or discipline or what have you that allows performance in high stress situations. It is very difficult for me to believe that there are not players who are better at that than other players.

    Maybe the problem is that anybody who achieves an MLB career of any length (and, at least in part, base on his bat) has that ability.

    All the non clutch players simply fail to become pros.


  • In reply to Richard Beckman:

    Still some would be more professional than others. Maybe it's like creativity, a gift that comes and goes, but men are not machines. Agree though, those who perform at the highest level have the most consistency.

  • In reply to Richard Beckman:

    I remember one year where Derrick Lee may have had a high RE24. But the next year ............ ? Perhaps it all averages out. Also, if you have Jacoby Ellsbury or Ricky Henderson batting in front of you one year and not the next? Well that could affect opportunity. So I think a lot of this averages out over a career.

  • Off-topic, looks like Homer Bailey is going to re-up with the Reds so that leaves James Shields, Max Scherzer and Jon Lester as the only "ace" types going into the winter.

    Lester has already said he's willing to take less to stay with Boston and I cannot envision him leaving that team to come to a Cubs team that will in all likelihood be coming off three straight 90 loss seasons.

    Scherzer is a big question mark to me. He had a great season, if he duplicates it he's going to be looking for $150+ million. If he has a good season, that leaves him with only one elite season and only two above 200 IP (assuming he passes that threshold this upcoming season), and a question of who is the real Max Scherzer? With his price tag still being around $100 million.

    Shields will be 33 by the time Opening Day comes along. I don't know if the investment will be worth it for a pitcher inching towards his mid-30's. You have to expect competition to be high on him because not everyone will be able to go after Scherzer and I expect Lester to resign with the BoSox. After those three, there's a significant drop-off in starting pitching talent so the price tag can get high very fast.

  • In reply to Jimmie Ward:

    Yes, Shields is likely to be the only FOR pitcher out there if he remains an FOR pitcher. Schertzer has said he wants to stay in Detroit and they are working on an extension, so he likely will not be available either.

  • In reply to krn99:

    I guess we will just have to develop our own. Or perhaps trade for one. Theo is pretty good at trading. I think Starlin will be our most valuable trading chip if he returns to his normal self.

  • In reply to John57:

    I think if he returns to his normal self, keeps improving defensively, he might be harder to let go. But what if he is a .280 hitter and projects to make 25 errors a season? Then I can definitely see him being a trade chip. Even more so if the Cubs draft Trea Turner. And it seems likely to me that Tyler Kolek and Trea Turner are going to be what is left of the top 5. What is Beede slips due to a lack-luster season? What would the Cubs do if Turner and Beede were available. I think it would be about 90% certainty that they would draft Turner. Then it would be easier to let Castro go.

  • In reply to Quasimodo:

    I am leaning that way too. I think Turner would be a fantastic player. Extremely high OBP with blazing speed on the base paths. Oh and GG caliber defense at SS. Would see why the FO would like him. On top of that Baez will be a better SS than Castro too. He would be third in the pecking order. Still though Castro is a very talented and valuable player.

  • In reply to John57:

    Well - eventially - one or more of Olt, Baez, Alcantara, Bryant, Villanueva, Castro, Barney, Valbuena and Watkins will have to be released, traded, and/or moved to a different position.

    Personally,.... I think the IF by June of this season is 3B Olt, SS Castro, 2B Barney/Watkins, 1B Rizzo with Bonifacio, Vitters and Valbuena getting significant at bats spelling these guys.

    While I would love to see Baez and Bryant on this team - I hope to see them getting regular playing time in Iowa first, but one or both of them forcing the Cub's hand before the trade deadline - making the case for significant playing time themselves.

  • I don't think of clutch by that definition. It's easy to say prove it doesn't exist when you get to define it too.

    My definition of clutch is Yadier Molina against the Cubs with a man in scoring position with first base open in a critical situation and I'm screaming at the TV to walk him. Of course we don't and that's the game. ;-(

    Don't get me wrong, I think statistics of all kinds can be indispensable in choosing the guys you want on the field, and evaluating their performance or impact, but they are not the answer to everything.

    Perhaps the definition of clutch in my opinion is just the top 5 or 10% of players in terms of batting average with runners in scoring position. They are better at it and therefore considered clutch in my mind.

  • fb_avatar
    In reply to Bilbo161:

    But, really, your anecdotal story is just a subset of my definition. One of the key parts of it is that there is a runner on third base. If Molina is systematically better in situations where the game is on the line and a tying/winning run is on second or third, he would, necessarily, be systematically better in situations where there are runners on base. (Unless he is systematically WORSE when there are runners on second/third base and it is not a "pressure" situations.)

    The analysis suggests that this is NOT the case. The thing about Molina: he got on base over a third of his times up last year. It shouldn't be a huge surprise when he comes through in such a situation -- there's a better than one in three chance.

  • So RE24 is basically a clutch measuring stat into itself, right? So by saying clutch doesn't exist, are you saying RE24 is irrelevant?

  • fb_avatar
    In reply to Oneear:

    Not really, it's a measure in one number of a player's contribution to his team's offense in a given year. For example, we can see that Miguel Cabrera was an absolute monster offensively last year -- a much bigger contributor than even during the triple crown season -- and the reason for it is, as Toby pointed out, his hits were remarkably stacked to situations where runners were on base.

  • Mike, have you thought of the possibility that your measurements do not reflect true clutch hitting (that is, hitting well under pressure), therefore, the fact that there is no significant variance among players does not in any way mean there are no clutch hitters.

    Pressure is not a binary attribute, which is why it is so difficult to truly measure it correctly. There are many shades of it, as anyone who has ever competed in a sport could attest.

    Try to find instances where the probability of a pressure situation is very high. (Driving in a guy from third base when your team is in last place and you're down by 3 runs is not a pressure situation, sorry you folks who believe RISP has any relevance to clutch hitting.)

    I would propose that any situation where your team is tied or down by 1 run, don't care about the inning, is probably somewhat accurate in measuring pressure hitting, although even here it probably doesn't in many cases. If the outcome of the game is on the line, you have a better chance of it being really a pressure situation.

    The "clutch hitting" bashers tend to come up with meaningless stats to back up their theory. Measuring pressure, and the degree of pressure, is truly difficult. But it exists. Oh yes, it exists.

  • fb_avatar
    In reply to HefCA:

    The truth is I don't have the data to test it at that level of specificity. I made a definition that seemed fair given the data I had and then found the result.

    Also, as I said to Bilbo above, if a player is systematically better in "high pressure" situations with runners on base, his numbers will necessarily be better in ALL situations with runners unless he systematically worse in non-pressure situations.

  • In reply to Mike Moody:

    Ask yourself this question: why do some pitchers thrive in the closer role, while others who may have equal or better stuff/control do not? Is it just that some pithers are 'luckier' pitching in the 9th than others?

    I actually like what you are doing here, trying to find some players who excel under pressure rather than looking at the whole league for general tendencies. Because if pressure does exist, then there will be pitchers as well as hitters who perform better under pressure, so it will balance out as a whole.

    Maybe it's true that clutch hitting is a fallacy. But I'm not convinced yet, and I think many others aren't either, because the measurements that people offer up to dispute clutch hitting make assumptions that are very incorrect. How many at-bats have you seen where there was no pressure on the guy to drive in a runner from 2nd base, when the team is down by 3 in the 7th inning? Thousands, if you watch a lot of baseball. So why should a stat that does not accurately measure pressure performance show no variability with most players.

    I'd like to see a measurement that calculates avg/obp/slg for hitters when batting with the tying or leading run in scoring position.
    The game is all about the win, and this is the most accurate determination of a "money" situation. Use all innings in this situation, there are many games where the first run on the board was the most critical. If this measurement shows that there are no offensive players with significant differences, then you will convince me there is no such thing as clutch hitting.

  • fb_avatar
    In reply to HefCA:

    I don't have the data to test this, but I've found myself wondering if the closer vs. setup man thing is more a function of how both guys are used. A setup man is usually brought in for specific situations and batters where the numbers say they excel. The closer comes for the 9th inning regardless of who is coming up.

    An interesting test on this would be to see if dedicated 8th inning men, e.g. Carlos Marmol for Kerry Wood, Mariano Rivera for John Wetteland, have more success when promoted to closer than standard "setup men."

    As I say, though, I lack the data to really test that at this point. (Hey, baseball-reference, if you'd like to work out a deal...)

  • In reply to Mike Moody:

    That would be a good test. Remember Latroy Hawkins? He's a guy I think of who had excellent stuff, but just couldn't hack being a closer. He gets unfairly punished for the Cubs 2004 debacle, IMHO. People forget he was one of the best setup men in the league that year, but when Boroski went down, not one but two key relief positions for the Cubs went down with him.

    Guys like Boroski are interesting to examine also, although I don't know how you would do it. He wasn't known for having great stuff, and I don't think he was even considered a good bull pen pitcher, until he started closing for the Cubs. But was he ever good at that role once it was given to him.

    Anyway, I digress. Your comparison of setup vs. closer roles for a pitcher is similar to 'clutch hitting', although you could make the case that both roles involve pressure. I would agree that the closer has more pressure involved, though.

  • It's nice when sabermetrics confirms common sense. If a .250 hitter figured out how to hit .350 in high leverage situations, why wouldn't he use the .350 approach all the time? It always seemed to me that unusually high variances between BA and BA with RISP had to be simply small sample aberrations, and your analysis confirms it. Thanks!

  • I think it would be interesting to find out if the percentage of errors have historically risen during the playoffs. That would tell me a lot more about pressure affecting players than an at bat. I think hitting in any game in front of 30,000 fans and millions of tv viewers is a high pressure situation whether its the first inning or 9th in a one run game.
    Also in a 9th inning one run game at bat the pitcher would be under a similar elevated amount of pressure. Maybe both players are capable of elevating their game in that situation.

  • Interesting article and analysis.

    I would add that this just looks at one dimension of "clutch performances". I would argue that RBI's is a bit too narrow (for offensive performance) and treating all RBI situations equally is too broad in defining "high leverage situations". I totally understand the logic in using narrowly focused analysis as a sampler for time and resource purposes, but it can lead to skewed or incorrect interpretation of the data.

    In fact, an entire research project could be focused solely on defining "high leverage situations" or "clutch play". I would look more at situations in terms of "contributing to wins" and then weight those situations in terms of "contributing to winning championships". For example, an RBI in late innings of a playoff game is weighted more than opening day OR RBI's in tie games weighted more than in blowouts.

    I predict that you would almost certainly see some significant differences for players simply due to "stage fright" or "choking" more than anything else. Part of me always believed that being clutch (in any sport) was less "stepping up" and more about not falling off from your typical level of play in high pressure situations.

    Good examples of this are athletes like Michael Jordan, Tom Brady, and Derek Jeter who didn't so much increase their level of play, but rather did not suffer any significant falloff from their typical performance in high leverage situations or playing on a bigger stage.

    Anyway, I analyze biostatistics all day for our core lab, so I love comparing statistical analysis techniques we use here to the analysis sports organizations and media use. Plus, it gives me a nice break while still being able to claim (somewhat honestly) to be spending time on work related material.

Leave a comment