Why We Feel How We Feel About Clutch
Apologies for walking on trodden ground. None of what’s below is new. Many of you already know everything in here, but I feel like this is a good opportunity to review why our position is our position. I’ll do my best to keep this simple and short. Just like all the world’s best analysis!
Over the last little while, I’ve written a few things about Clutch. The specific stat might be difficult to explain to the average fan, but the idea is a basic one. Teams with high Clutch scores have had really good timing. Teams with low Clutch scores have had really bad timing. Timing is important! This explains a lot of the difference we see between actual wins and BaseRuns wins, which you can just think of as “expected wins.” This year, the five most clutch teams in baseball so far have beaten their BaseRuns win total by a combined 45. The five least clutch teams in baseball so far have fallen short of their BaseRuns win total by a combined 43. It’s hugely important, and this isn’t a one-year phenomenon.
I have BaseRuns information going back to 2002, so let’s plot team Clutch score and the difference between actual wins and expected wins. This’ll cover the completed seasons, between 2002 – 2014.
It’s clearly a strong relationship. Clutch doesn’t explain everything, but it explains an awful lot, and that’s just intuitive. Of course the teams that do the best at the right times will be more successful. If they do better than they usually do at the right times, they’ll look like an over-achiever. That’s how you can get a team to win more games than you’d think just based on the overall statistics.
Clutch can turn a mediocre team into a playoff team. Clutch can also turn a would-be playoff team into a mediocre team. Because it’s so important, it stands to reason teams would try to emphasize clutch performance, if they could. They’d try to gather clutch performers. There’s not a single analyst in the world who doubts the significance of clutch events. But that isn’t the problem. Let me show you some more information. We’ve got team batting Clutch, team starter Clutch, and team reliever Clutch. I decided to look at the window from between 2000 – 2014, splitting seasons by first and second halves. Here’s how batting Clutch has carried over, half to half:
An r-squared of literally 0.00. You can find numbers that aren’t 0 if you go to more decimals, but that doesn’t accomplish anything. It’s a nothing relationship. Here’s how starter Clutch has carried over, half to half:
An r-squared of literally 0.00. You get the point already, but let’s move on to reliever Clutch:
An r-squared of literally 0.00. So this won’t surprise you — putting it all together for team Clutch:
An r-squared of literally 0.00. No observed relationship. No observed hint of a relationship. The only relationship here is the one between Clutch and total randomness, and that’s not a relationship for anyone to rely upon. It would be one thing if there were no relationship between Clutch in Year 1 and Year 2 (and there isn’t). But this is looking at the same teams, within seasons. Even clutch teams are only temporarily clutch. Sometimes they remain clutch, but no less often do they do the opposite.
This is why I’m more down on, say, the Twins than other people might be. Not that it matters at this point, with the season almost over, but there’s trusting the Twins’ record, and there’s trusting the Twins’ other, underlying numbers. The underlying numbers have proven more trustworthy. If you want to argue a certain team is innately clutch or unclutch, that’s fine. Make the argument. You might very well be right. Just, understand what the argument is up against. Understand how hard it’ll be to convince someone of legitimate clutchness. The argument against this stuff is strong, and it’s tough to doubt a 0.00 r-squared. Factors you might think lend themselves to better or worse performance in clutch situations — there’s nothing convincing in the recent history. And teams would have a lot to gain from harnessing this.
The position is one against clutchness, because that’s what all the evidence points to. It’s not coming out of stubbornness. It’s not coming out of closed-mindedness. It’s just that nothing else has been sufficiently convincing. And, this is important — analysts would probably love to be wrong! It would be amazing if real clutchness could be proven. That would be a breakthrough, and if you’re just some person doing research, good research could get you hired by a club. It would be greatly significant if one could demonstrate reasons for under- or over-achieving, ahead of time. I think we’d all love to read that article. It would change the way we see the game.
That just isn’t where we are today. Today, it looks like near or total randomness. So it gets treated as such, and though the actual wins matter more than the expected wins do, as far as the World Series is concerned, one should understand why sometimes analysts think the expected wins are more meaningful, analytically. We’re always trying to drown out the noise.
Jeff made Lookout Landing a thing, but he does not still write there about the Mariners. He does write here, sometimes about the Mariners, but usually not.
I object to characterizing a team as “bad” or “lucky” because they are outperforming their BaseRuns. What you can say is that outperforming their BaseRuns in this way is not sustainable based on the historical data we have. But that’s not how it’s framed. It’s framed as “Team X is a bad or mediocre team that has gotten lucky and will suck going forward.” My favorite thing all the analyst here like to say is that the Dodgers are the best team in the NL and maybe baseball based on their BaseRuns.
Fangraphs loves to just cite BaseRuns and declare the conversation over.
If that were true this post wouldn’t exist
I like the parallel to (a)theism: believing in clutch is like believing in a god. Not believing in it is like being an aclutchist. That can be interpreted one of two ways:
1. Clutch has never been demonstrated to be a repeatable skill.
2. Clutch is not real and teams that have a good clutch score are lucky.
1 is the equivalent to “I don’t believe any gods”. 2 is “there are no gods.” Position number 1 is absolutely supported by stats. So if a team outperforms their Baseruns, you can say “there’s no reason to think they will continue to do this”. You don’t know they’re lucky, but right now luck is probably the best explanation.
“So if a team outperforms their Baseruns, you can say there’s no reason to think they will continue to do this'”
Yes, I agree with this. That is not what I’m arguing. I’m arguing the framing of how a team has performed in the past solely by what their BaseRuns is.
Can you provide an example of what you’re arguing against? I think we are thinking of similar articles, but my understanding is that it’s usually used to evaluate a context-neutral previous performance of a team in order to look forward.
That is, its not used to denigrate a performance that happened, but to strip out the elements that aren’t shown to be repeatable in an effort to estimate true talent level. That estimate can then be used to project moving forward.
“1. Clutch has never been demonstrated to be a repeatable skill.”
This just isn’t true; when it is repeated, we call them an outlier and lucky to be clutch for the duration of their career.
Two use one baseball example – for his career, Joe Crede had a .748 OPS – he also had a .716 OPS with the bases empty.
With RISP, Crede had a career OPS of .792, with 2 outs and RISP .780, with the bases loaded .902. In the postseason he’s a career .949 OPS player. Crede has stated himself that he felt more zoned in and comfortable in big situations; this was supported by his actual production.
He was tied with David Ortiz game winning RBI over a 6 year span and clearly performed well beyond his career norms when the situation was escalated.
On the bell curve, we just consider Crede an outlier because we can’t explain his improvement in play when the situation was bigger. This to me has always been a cop out.
Outliers are an expected and inherent aspect of statistics. While looking at vast amounts of data, each point containing information from a single player’s single season, you see no relationship between at all between clutch over time.
However, you should expect to see individual outliers that are DO seem to consistently outperform or underperform in clutch situations. Does that mean the statistical relationship is untrue? NO! It just means there’s a statistical anomaly. If you flip a coin ten times for a long enough, you will eventually get heads all ten times. Because this happened in one instance out of many does not mean that the coin flips are no longer random.
Honestly, I think there’s a misunderstanding of how statistical analyses are actually concluding, what outliers are, and how an outlier does nothing to disprove the underlying relationships.
I agree that entire teams obviously can’t be “clutch” but to think the human aspect of sports doesn’t allow a player to be better or worse in big moments is naive. Regardless of profession, there will always be people who perform better and worse under pressure. To discredit it as an outlier is lazy in my opinion.
> Regardless of profession, there will always be people who perform better and worse under pressure.
That sentence is talking about a normal distribution of pressure performers in a population. An alternate explanation is players who cannot perform under pressure wash out before they hit MLB.
If you understood why your Joe Crede argument doesn’t make sense, you will have gone a long ways towards a better understanding of the application of baseball statistics. Nowhere in this article will you find an argument that explicitly says Joe Crede didn’t perform better in “clutch” situations. This of it this way: if the distribution of “clutch” ability were perfectly random, would you predict to see the same number of Joe Crede level outliers as you actually saw?
If you want to prove Joe Crede was actually more clutch you have to use different evidence, because randomness explains the existence of Joe Crede really well.
This is why you cannot prove a case with a single data point!
To criticize something based on a lack of understanding of distributions and basic statistics is lazy in my opinion.
It’s certainly not a lack of understanding. It’s differing views on the randomness of human nature.
You believe if you’re not clutch you’ll wash out before the big leagues but I disagree. There aren’t many high pressure situations that make or break your recruit status and your draft stock.
As I noted, i chose to not discard the fact that players like Crede state that they feel more comfortable and concentrate more in high pressure situations. Some pitchers perform better in high leverage situations and on and on. I understand that there will be outliers in statistics, I just am of the camp that thinks there are reasons to some of the randomness.
Pressure impacts people in different ways. All pressure is not created equal.
Just because it isn’t predictive based on a larger dataset does not mean there’s no reason for it.
“There aren’t many high pressure situations that make or break your recruit status and your draft stock.”
Well, this is demonstrably untrue. Plenty of players make or break their careers based on how they do during that small handful of games when pro scouts are present. And the pressure that they’re then under in the minors is immense. 90% wash-out somewhere in the minors.
“Just because it isn’t predictive based on a larger dataset does not mean there’s no reason for it.”
Sure. But it does mean that there’s no evidence of a reason OTHER than randomness. The default assumption should be that he’s right, not that you’re right.
>Just because it isn’t predictive based on a larger dataset does not mean there’s no reason for it.
No, but it does mean that you haven’t presented valid evidence to support your assertion.
You’re free to believe whatever you like. But do not state your opinion as fact and then attempt to back your opinion up with invalid statistical evidence.
There’s nothing invalid about the statistical evidence I provided.
Said player says early on in his career that he feels better when pressure is escalated – whether that be RISP, or just runners on base in general.
Said player goes on to greatly outperform his career averages in high pressure situations.
You view that as a singular sample that simply is an outlier, and not predictive – I disagree.
Someone who is clutch, and responds well to pressure, very well may be an out lier but it doesn’t mean it’s not predictive. You don’t need historical statistical evidence to make a claim or belief valid.
There are scouts who have shown an innate ability to read pressure ability – you would view them as “lucky,” simply picking the right outlier and getting it right… I would say they have an ability to determine the reaction of a player under pressure.
It’s no different than pitchers who struggle to close but can dominate the 8th inning – there is still pressure in the 8th inning, but the 8th is comforting for that pitcher. He may feel the game is not in his hands as it is in the 9th.
My entire problem with the pure-statistics community is if you can’t explain it, and can’t predict it, then it’s just an outlier which occurs in all factors of statistics. That’s a cop-out.
These are human beings; not computers. To think pressure and situations don’t get to them – for better or worse – is simply naive.
I just provided you with 10 years of data from one players career that shows he performed better when the pressure was heightened.
You feel the reason is randomness for Crede’s success; I feel you are wrong, based on evidence to the contrary that extends beyond the actual statistical output.
To the gentleman who said one or two games make or break a draft position – this isn’t 1980 pal. These guys are scouted hundreds of times; they’re the best players in their high school and the game comes easy to them. Pressure isn’t heightened until both the situation becomes difficult and important; the situations don’t become difficult until proball for many of these players. Many succeed more than they fail the majority of their baseball playing life.
Define pressure. I’m confident that the made up definition of pressure (or leverage) in baseball (late innings, risp, 2 out etc.) does not guarantee that each player feels the same amount of pressure regardless of the base/out state and inning. Early in the season a marginal player starting because of injury may feel extreme pressure during every PA because he knows that he still has options left. An established player in a slump may stick with his proven approach no matter what the results because he is confident that things will even out over time. They don’t experience the same pressure
Another way to look at, is that to “legitimate” in Fangraphs’ writers’ eyes you need to be right at your BaseRuns or you are a fluke one way or another.
Right now, there are exactly two teams performing in line with their BaseRuns: Yankees and Tigers. There are 9 other teams with a differential in their BaseRuns and actual record by two wins or less.
But there are 14 teams with a differential of 5 or greater wins between their BaseRuns and actual record. Three of those teams have the best records, respectively, in baseball. So what we have here is Fangraphs saying the three best teams in baseball records wise are all flukes to some degree.
While these teams may be outperforming their expected W/L, I don’t think it’s fair to cast this cloud of gloom around them with BaseRuns.
That article on the Cardinals, for one, failed to mention that by BaseRuns they still have an expected .551 winning percentage, which still has them in the top eschelon of teams. Instead, that article was framed to make the team look like a giant pile of luck that was not worthy of being called one of the best teams in baseball.
As everyone knows, framing is everything in an article. And the misuse of statistics or the omission of others annoys me.
“Yet, interestingly, we can also consider OPS differential. By that measure, the Cardinals rank fifth in baseball, between the Pirates and the Yankees.”
I bet if you did the clutch r-square by team in the last 5 years the Cardinals would come up on top
You’re not wrong, framing is just subjective.
So since we’re airing our opinions, I think Fangraphs frames it appropriately and that Dave did so as well in the Cards article.
I mean, there is a distinct possibility that luck just has that great an effect on baseball games, and no one could ever possibly devise a measure of team performance which half the teams don’t deviate from by 5 games or more. So it’s not necessarily true that these teams are flukes, but at the same time, their record is impacted by something other than how good they are. What is the recommended way to say that?
I would suggest the most troubling aspect of all this is “our estimates are fine, everything else is clutch and randomness”. How close were last season’s BaseRuns projections to what actually happened? If they’re close enough, I’d suggest the “and the rest is randomness” argument is strong; if not, it might be worth looking at how those projections are happened upon.
I think the problem here is all of the FG staff and most of the community have an understanding that everything is a bell curve. We don’t expect everyone will perform at base runs, that’s just the most likely outcome. Everything on either side is a bit less likely. Here’s the crux though: performing at expectation isn’t actually particularly likely, per se. Seeing only 2 teams of 30 exactly match that expected value is totally fine. Nobody here is doubting base runs because of that. We can see that a couple are bang on, most are close, and a few are way off on the sides of the curve. It’d be weird to expect it to look much different in a game where no matter the matchup, one team really never has any better than about a 70% chance of a win. Usually it’s damn close to 50/50. There’s going to be a lot of noise and uncertainty. That doesn’t render the predictions invalid – they’re based on stats that normalize much better than wins and losses.
…Did you not read the post?
This is so important. This could be one of the biggest divisions between stats-oriented and non-stats-oriented fans. I feel like 90% of non-stats fans think that clutch is real and some players can be identified as clutch, and like 90% of stats fans think that either clutch isn’t real, or at least that we don’t have to ability to detect it if it is real.
Considering how large of an effect clutch timing has on a game (its effect is large by definition since it basically refers to performance when it has the greatest impact on game outcome), this issue almost the equivalent of believing in a god – it is so fundamental to how we interpret the outcome of a game that not being on the same page about this can make communication with those that disagree almost impossible.
Agreed, I remember when a friend first told me “clutch isn’t real” after reading Cameron or James or someone, years before a first visit to FG or the like, I disagreed vehemently.
I still technically disagree, but the disagreement is just one of definition / nomenclature. Clutch is definitely real in that you can have been clutch in a given situation – its descriptive. Clutch is *probably* not real in terms of being able to consistently perform better in those situations – its *probably* not predictive.
Bryce Harper is the most un-clutch player in the majors this year, which must be baseball’s equivalent of the blind watchmaker in terms of evidence Clutch really exists.
If he wasn’t so un-clutch, I’d be getting that second straight Manager of the Year Award.
I’ve wondered if there is any relationship between super star level performance at the plate (e.g., Bonds 01-04) and “un-clutchness”. My thinking is that pitchers would limit the damage these players could conceivably make in high leverage scenarios by walking / pitching around them. I believe at one time Bonds was even walked intentionally with the bases loaded.
Bryce Harper is producing at twice the level of an average hitter this year. You’d think in any high leverage scenario pitchers would just say “f it” and walk him. Obviously walks still count as good production in wRC+, but this might go some way in explaining his lack of RBIs.
28 May 1998, Diamondbacks at Giants, bottom of the 9th, 2 out, 6-8 score, bases loaded.
Intentional walk of Barry Bonds, so it was done PRIOR to his complete dominance years. Bonds was scary.
Yup, I did that. It worked too. Next guy lined out to RF to end the game. But if that liner fell in for a hit, boy would I have looked silly!
So this is a topic I love to think about because a frequent comment you hear about is “the zone”. Athletes talk about it and I’ve personally experienced it (though for chess rather than team sports). And the best way I can articulate “the zone” is that it’s a complete focus to the exclusion of pretty much everything else.
My hare-brained, off the cuff hypothesis is that human beings aren’t capable of maintaining focus in a concentrated way without snapping back. I’m no neurologist, but the basic principle I wonder about is somewhat like short term memory to long term memory. The brain’s state is always changing and you just can’t focus indefinitely.
I then wonder if it expands to a team environment. I want to say Sam Miller had an interesting article about team chemistry that talked about teams looking for indicators and something along the lines of players who fit into multiple groups to cross boundaries improving overall team performance. I think the 2012 A’s were a model.
Anyway… the 0.00 thing is particularly interesting because that means that clutch doesn’t retain inherent terribleness either. Which surprises me a bit… I would think some crappy teams full of give up would tilt this but perhaps trying to maintain ANY focus or lack of focus isn’t something humans can do.
Entropy or something.
I think if a team is terrible and has no morale, it would be more terrible across the board, not just in “clutch” situations. Focus and team chemistry are not the same thing as clutch. Team chemistry and focus could be totally real, but clutch still may not be, because teams that gel well and players that focus well are better across the board, and so are indistinguishable from a team a that is just generally better.
The one thing I would like to see, that stands out as a place where something that looks like “clutch” but is actually skill, is bullpens. I have seen a lot of analysis as above that looks at bullpen performance as a whole. But, what about bullpen performance of the pitchers on a team’s bullpen that are called upon for the most high leverage situations, vs. low leverage?
E.g. team A could have an average bullpen, and that average bullpen could be made up of some long-reliever, 6th inning types who are performing really well when they’ve been called up, but these are guys who are really getting results above their skills levels or are young, and for whatever reason not trusted in key situations. That team could then have some late inning relievers and a closer who are performing very poorly, but because these guys have been good in the past, and because they’re getting paid to be closers and good relievers, they are still used in high leverage situations.
Team B could also overall have an average bullpen, identical in team ERA, FIP, etc to team A. Team B could have some really badly performing long relievers, 6th inning types who are being used in blowouts when there team had little chance of winning even if the guys had been annihilating the opposing batters, and could have dominant back-end relievers. Now if you looks at team A and B’s bullpen performance as a whole, you might think there is no difference between the two teams’ bullpens, except that Team B is more clutch. But they’re not more clutch, they just have better relievers keeping close games close and they have bad relievers making games that get out of hand early get more out of hand.
Also would be interesting to see about lineup construction, does having more contact-oriented hitters like the Royals, v.s guys who might be deriving their wOBA from more walks, make a team like the Royals more likely to score runs in high leverage situations now that home runs have declined in baseball? While the year to year, half to half relationship is nonexistent, are there certain teams that have bucked that trend and is there any reason why?
There are teams that have bucked that trend. You can see it in the graph. Any team on the top right was clutch in both halves of the year. However, given no objective, causal criteria to distinguish them from the other teams in the sample, you have to conclude that random variation could easily cause that team to be clutch all year. Looking at the teams that appear to demonstrate that relationship and then working backwards is a recipe for a huge selection bias.
Now, if your hypothesis is that something unrelated like contact rate has an impact on clutch, you could take another unbiased sample of high-contact teams and see if that same shows a relationship to clutchness, of just graph contact rate as a predictor and clutch as the variable.
You can show, in theory, that a hitter with a high-contact, low-power hitting profile that performs equally across leverage will have a positive “clutch” score, and a hitter with a low-contact, high-power hitting profile that performs equally across leverage will have a negative “clutch” score. I think there was a Fangraphs article that did this, maybe a year or so ago.
Teams’ hitting profiles aren’t as extreme as individual hitters’, however.
well you have to start with some hypothesis right? I think to just say, clutch is totally random, and just looking at a super macro view of all teams, all players, doesn’t really say much. It’s is almost certainly MOSTLY random, so looking at this macro data, this result is not interesting or surprising in the least. There is psychology at play at times (slumps, players being confident or not confident at certain times), which isn’t a skill and evens out in the long run, in addition to the simply random events, bases loaded and a good player hits a bad pitch really hard but right at someone, or a hitter hits a little bloop that lands on the foul line and scores some runs, etc.
It would be more interesting to see if there are teams that are exceptions to the randomness over longer periods of time, and then try to test some hypotheses to see if there is something real in team-building. While individual player clutch is probably not anything “real”, I can think of things on a team perspective that might look like “clutch” but are actually “team-building” and “better allocation of resources” and “lineup construction.” The bullpen thing is one. A team with better back end relievers, and relievers who are being used properly by their manager, is almost certainly going to look more “clutch” but this is a very real thing. There probably wouldn’t be a huge year-to-year, or at least multi-year correlation here on teams because relievers are fickle and get injured and just have wilder performance variations than, say, position players, but that doesn’t mean the “clutch-ness” of a bullpen in a given year wasn’t a real talent difference. Whether contact-oriented hitters, or hitters who hit more line drives, or a pitching staff that pitches to contact vs one that doesn’t, or if having better team defense might make you more “clutch” than teams with poor defense, would be interesting things to look at.
@isavage30: This article looked at “First and second half reliever clutch”, which has r-squared of zero just like the other ones.
This doesn’t mean that teams should stop trying to manage their bullpens to put their best pitchers in the most important situations. It does mean that in MLB, where every team tries really hard at this, everyone does about as well at it as everyone else. If a team hired a dart-throwing monkey to manage its bullpen it would probably perform worse than real managers do, but no team does this.
Clutchness variations among players are much more interesting than clutchness variations among teams.
My point in all this is just that looking at the apparent exceptions to the rule and finding out what they did is the wrong way about it. It’s a good way to mistake noise for signal. You have to do it like you say in your last sentence.
But that’s the point, how can you determine that a team is an exception to the randomness. Randomness can produce a team that is clutch for 2, 3 straight years, given a large enough sample of teams. To look at that team as some example of a team that is actually clutch and not lucky is to assume your conclusion.
To the extent this effect is real, you would still expect it to show up in that bullpen clutch rating graph, and the effect is still minuscule.
You have to make sure your causation is going in the right direction. I believe some neurological studies are dedicated to whats causes what in this type of situation. Is it the in the zone feeling that causes the success or is the success releasing chemicals such as endorphins which cause the sensations of feeling like your in the zone or have good team chemistry.
What’s a transient, randomly distributed performance impacting factor: bacteria. The early research/fantasy about gut biomes is that they have outsized effects on things like metabolism and mood. And what happens when you get a bunch of guys living together for a six months a year? Their biomes converge temporarily (maybe). If a performance enhancing biome equilibrium (the hypothesis proposes hundreds or thousands of kinds of bacteria all either competing or cooperating) colonized a team temporarily, would they become short term clutch? What’s the sample size over which observed team clutch is likely not to be random? Is it 15 games?
Clutch performance over 81 games is not predictive, so why would it be over 15? I don’t think there is any sample size of games where clutch performance is meaningful and predictive, and to discover one would be to discover something very non-trivial.
So clutch and lucky go hand in hand?? Further proof that you guys have never actually played baseball.
The world looks and feels flat too. I don’t care what those pesky scientist say with all their facts about the earth being curved. Have they every walked on the earth? They say its curved? HA! If that were true I’d just being tripping & falling non-stop. Stupid fools
Isn’t the entire idea of the closer ultimately based on the concept of clutch? A save situation is a clutch situation, and some pitchers are believed to have “the closer mentality” while others don’t. And of course some closers are quite steadily good, while others aren’t. Is this all nonsense?
My theory is that while being clutch is not a repeatable, predictable skill, being UN-clutch certainly is. However, being phenomenally un-clutch reliever generally limits the amount of appearances you make in a clutch situation. So it wouldn’t appear in a statistically significant number of instances to show up.
I agree that “unclutch” players probably exist, but there’s a big question of whether any really “unclutch” players make it to MLB.
It seems likely that any player who struggles with performing under pressure washes out before the majors, so the players who are left are clustered in such tight band of “clutchness” (or “mental toughness”, if you prefer) that they are essentially equal. Or, said a bit differently, if a player can handle the pressure of playing baseball at the highest levels in front of thousands of fans, then a player isn’t going to be predictably better (or worse) in a “clutch” situation in a game.
A commenter below references exactly this survivor bias.
I would think that some players do exist though. Sometimes a player might be so talented that they’ll get plenty of opportunities, so even if they’re held back from a mental perspective, they’ll still make it to the big leagues.
Didn’t Grienke miss time earlier in his career due to depression? Pete Harnisch went on the DL due to depression in the 90’s as well.
When Lucas Duda initially came up, there were articles about how he lacked confidence. He mashed in AAA and has a ton of power, but when things went bad he had trouble getting back on track. Because of how hard he hits the ball, he still made it to the majors.
While MLB is generally the cream of the crop, I don’t necessarily think everybody is weeded out before they get there.
Interesting point. In the examples that you’re citing, however, weren’t the players struggling in essentially all game situations, not just “clutch” ones?
It would seem like a very narrow band of mental toughness for a player to do just fine in a “normal” situation – which, even in the first inning of a game, is playing in a huge stadium in front of tens of thousands of fans – but struggle if he’s in that same stadium in a tie game in the 9th inning. I suppose that’s possible, but it seems unlikely or at least extremely rare.
And the extreme examples that we have of true psychological inability to perform from MLB players – I’m thinking pitchers like Steve Blass or Rick Ankiel, or Steve Sax not being able to throw from 2nd to 1st consistently – wouldn’t just manifest in “clutch” situations. They were problems at any point in a game.
Correct me if Im wrong but since Baseball is a 0 sum game with the success/failure of the pitcher vs. batter directly inverse wouldn’t the exist of unclutch players automatically create the existence of clutch players? If a batter “chokes” in the clutch wouldn’t that make the pitcher he faces “clutch” and how could you figure out whether one guy was a choke or whether the other guys was clutch or whether it was some degree of both. How can you tell whether some guys get better/more focused in clutch situations or whether other people get worse and they simply maintain?
The pitcher would be clutch in that situation, but it would be spread out over every single pitcher that the batter faces in high leverage situations. It wouldn’t produce a meaningful effect in the “clutchness” of any pitchers.
I think the problem is the fans that take it as some insult when you point out that their team has outperformed its Base Runs expected record. It’s not an insult! No one is saying that your team getting it done when it counted doesn’t have merit, bc it does. They’re only saying that this state of affairs is unlikely to continue, something that every shred of objective data points to. There are three responses responses to those who say your team has been “lucky”: (1) Be happy your team has done so much better than expected. You might make the playoffs bc of it. You may already have made the playoffs bc of it. (This is the easiest response, requiring little more than some perspective and a bit of gratitude for the blessings the baseball gods have bestowed upon you). (2) Prove that your team is some sort of exception to everything we think we know about “clutchiness” (w something other than wild speculation or thinly supported subjective theories) (3) Rant incoherently. Please pick either (1) or (2). (3) is just no fun.
WE CHOOSE OPTION 3!!
Get more sophisticated in your approach. Take a nuanced look with an advanced analysis. Piling on data points to find league averages will get you nowhere.
That would be worth doing if the data in this article suggested it was worth doing. But the data here says that a single team’s past clutch performance tells you nothing about their future clutch performance. If the same team cannot even be expected to maintain their performance, where do you go from there? It’s like looking for explanation for an effect that as far as you can tell doesn’t even exist.
Yeah, Jeff is certainly a small-time nobody. You sure told him.
It seems to me that “clutch” either means:
1. Performing better in Really Important situations than in Normal situations
2. Performing no worse in Really Important situations” than in Normal situations
I tend to feel that #1 is closer to how people generally define it.
If it were true that a player performed better in Really Important situations than in Normal situations, it seems like the only reasonable explanation for such a difference would be that said player employed a different approach under Really important situations than under Normal situations in order to improve his/her performance.
But if that were true, then I’m left to wonder why said player didn’t simply use his/her Important Situation approach all the time, given that it produces superior results…
Yes. Yogi Berra apparently had a consistently positive clutch score for most of his career. But this leads me to ask, was he not trying as hard when the leverage was lower? And aren’t players supposed to be giving it 100% all the time?
If you assume no one can really give 110%, but only 100%, if a player really is repeatably clutch, it means he’s not giving 100% much of the time.
I suppose someone who wants to believe in clutch could argue it’s our nature that we can’t give 100% all the time, and some players manage to make those times when they’re needed least.
I agree with this logic for batters and relief pitchers, but a starting pitcher might have a different approach in Really Important Situations.
It’s at least plausible that a starting pitcher could do something – probably ramp up velocity – that leads to better performance against 1 or 2 hitters in key situations but isn’t sustainable over the course of a full game throwing 100+ pitches.
That said, I agree we haven’t seen evidence that starting pitchers are reliably “clutch” in this way, just that it’s plausible they could be. Maybe their ability to do this – if any – gets swamped by offsetting factors like fatigue within games and within innings, so we don’t see an overall benefit. Or maybe they can’t actually make such changes – or don’t benefit from them – because repeatable mechanics matter more than an adrenaline-fueled temporary uptick in velocity or “stuff”.
I agree that it’s most likely with a pitcher. It seems like there are some pitchers that cruise with nobody on base, but ramp it up when they need a strikeout. I have no idea if that would show up in the numbers though, because it can only be a few pitches per game where it matters.
I do think it can play a factor in hitters too though. It’s easy to say “if they’re capable of doing it in clutch situations, why don’t they just play that way in every at bat?”. But it’s a lot of work to focus that much for 600+ plate appearances. I saw a tweet from Eno Sarris yesterday about Joey Votto, and how he thinks about so much during an at bat that sometimes he has to call timeout to gather his thoughts. I’m sure that can be exhausting. If they’re up 6-1 in the 9th, does he really give it as much thought as if it’s 1-1 in the 9th? Or is it possible that he just lets it rip because they’re going to win anyway?
i’m not a believer in clutch at all, but every once in a while I come across some pretty strange numbers.
Michael Young averaged a 98 wRC+ with bases empty, then a 116 wRC+ with men on. the 116 is over 3500+ PA’s. that doesn’t seem attributable to just luck.
if you go by leverage, he had a 102 wRC+ in low leverage situations, a 109 wRC+ in medium, and a 114 wRC+ in high leverage situations. The PA’s in that case are more strongly skewed towards low leverage, but still over 700 PA’s in the high leverage and over 3200 in the medium.
it’s kinda hard to argue he wasn’t a clutch player.
Are you familiar with the normal distribution? Take a look at it. It is also referred to as the Gaussian distribution.
what of it? Young is an outlier?? that’s kind of exactly what i was suggesting
I believe that his point is that you would expect outliers like Young even if how clutch is completely random. Large samples that are completely random fall on a normal distribution and the existence of data points that are not the mean is what makes it a distribution.
Most players have better wRC+ with men on than with bases empty. If you think real hard about it, it becomes pretty obvious why.
How does runners being on base create a better wRC+ for the batter? wRC+ gives value to the event in isolation, regardless of whether it occurred with a runner on or not. The only outside factor it accounts for is Park factor. That this comment got 7 thumbs up makes Siskel and Ebert roll over in their graves. Good choice on remaining anonymous.
Runners on base creates a better wrc+ because the pitcher has to throw from the stretch dufus.
Also with guys on first for instance it opens up a larger hole on the left side of the infield for a lefty. Depending on where runners are on it can also limit shifting.
And the pitching from the stretch thing.
Batters usually hit better with runners on (e.g. 2015 MLB splits). This seems to mostly be a result of fielder positioning (also because this page is b-r intentional walks count for OPS just like regular walks, and essentially all IBB occur with runners on base). Young appears to have a wider split than average — maybe he was good at hitting behind the runner…
What’s funny is that by FG “Clutch” Young is only +.7 for his career. I think FG career splits don’t quite cover his whole career; b-r splits look somewhat less obviously clutch (they cover his first couple years and split at different points). “Clutch” score treats leverage continuously rather than discretely, and accounts for context more completely than using context-neutral stats broken down by leverage — for example, WPA “expects” hitters to hit a little better with runners on base, and it penalizes double plays beyond normal outs, which wRC+ doesn’t. Maybe that’s part of the answer. Young was a ground-ball hitter, and wRC+ splits give him credit for higher ground-ball BABIP with runners on but don’t penalize the double-plays; “Clutch” expects the higher BABIP and penalizes double-plays.
There’s also some selection biases in the data with runners on. Most PA with runners on base are against the pitcher who put those runners on in the first place, which means you’re more likely to be facing a poor pitcher (or a pitcher who isn’t pitching especially well that day) and/or a poor defense.
Likewise, good hitters tend to be clustered together at the top of the lineup, meaning they get more plate appearances with runners on. Granted, that doesn’t make much difference on an individual level (only on a team or league level), but if you’re looking at someone’s career, it might matter. Michael Young, say, may have batted higher in the order during his heyday, and he may have been dropped when he started to decline or taken more days off – meaning he’d have a disproportionate number of PA with men on base during the years he was hitting well, and fewer during the years he wasn’t hitting as well.
number one is a fair point. number two is a little shaky, to my memory he seemed to hit high in the order right to the end of his career, since most of his managers seemed to have an irrational love for him.
that would be my guess too. he was pretty apt at dumping singles the other way. maybe he tried to maximize that when the 1B was holding the runner on or something. that strikes me as possible.
Here’s the thing. I think clutch is verifiable in hindsight, but it’s not predictable and that’s what has most statisticians irritated when discussing the benefits. You can’t say, there’s a runner at 2nd, so he’s going to get a hit. However if you look at those instances for a long enough period of time, you may see he benefits from that scenario over the course of thousands of at bats. Furthermore, we aren’t talking about the eating your vegetables scenario when you were a kid. Eat vegetables = get ice cream, don’t eat veggies = no ice cream. Even a non-clutch hitter gets lucky, comes through in the clutch. If you are clutch and hit .300/.350/.450 in those situations you aren’t comparing 800 OPS vs 0 OPS, you are comparing it vs a .600 OPS or something of the like, which means it’s a .200 OPS benefit across 160 games or a .00125 advantage per game or .0003125 per at bat. Which basically averages out to something equivalent of an extra foul ball every 5 games or some other near worthless event.
Another person referred to baseball as a whole, as a zero sum. Well here’s the problem. If you have a team of nine guys with a career of Michael Young and put them on a team together. While they would perform well as a whole with runners on. One of them would have to first get on and that would require them to depend on the 98 wRC+ version of Michael not the 116 wRC+ version of Michael.
The statistical benefit to the first Michael getting on-base would only slightly influence the second and third Michael in their at bat, which in turn only very occasionally would lead the fourth and fifth Michael to bat when they otherwise wouldn’t. It would take too many games and while 1 out of every x games would prove beneficial. The reality is that even a perfect lineup construction will only yield a marginal benefit of all the dominoes falling in sync in an insignificant amount of games.
The more realistic fallout would be that in most games going from a .300 to a .320 OBP in clutch situations only brings the 5th Michael to the plate 1 additional time per 13 or so games. Which in turn means you get to benefit from the extra 18 wRC+, 1 every 50 at bats, per Michael in your lineup and then that particular Michael will still fail 2/3 of the time. So the benefit will only be seen 1 out of every 150 at bats more or less (approximately 4 per year per player x 9 spots in the order = 36 beneficial occurrences per 5400 plate appearances)
That’s why Brandon Phillips wasn’t clutch when he drove in 100+ RBIs, two years ago, it was because he had 2 guys in front of him that got on base at a .420-.450 clip for most of the season. Essentially getting RBIs for Brandon, was no more difficult than the so-called “targeting of aquatic creatures using lead projectiles within an enclosed space made by Coopers.”
All things being equal, you would rather have guys that get on base because that is more repeatable than the skill of producing with runners on base. However, if someone shows a complete lack of ability to drive in runs, probably you are better batting them at the top of the order where they are the base runner rather than the hitter in trusted with the task of being “clutch” and/or batting at the bottom of the order where their “sins” are lessened by lowering the number of potential opportunities, where said player will “screw the pooch.” These are things you hope your manager understands.
At the end of the day, the guys here aren’t wrong, they just don’t really say what they need to say. Clutch does exist, it just doesn’t matter. The only exception is with say a team like the Mariners in 2015, which had a team full of jackasses who couldn’t hit with runners on (save Cruz and maybe a couple other guys), you can’t win if you have 10 guys that can’t drive in a run with ducks on the pond.
Homework: Chris Davis (clutch + 40 HRs) vs Joey Votto (.450 OBP + baserunning). What’s more important to a team’s success? What if both bat 1st? What if both bat 5th? What if both bat 8th in the NL? If you need a 5th hole batter, better to go with Davis, need a #2 or #3, get Votto. Looking at an 8th hole hitter, get a clutchless low OBP masher like Trumbo. OBP doesn’t slump and even if you hit 40 HRs, Votto still hits 30 HRs, so is a clutch hit and/or an extra HR every 16 games worth more than being on base 7 more times over the same stretch of games?
Looking at an entire series of all baseball players, you can expect the clutch index to be very random or, in fact, meaningless. It doesn’t matter if a poor performing ballplayer is a better in the clutch: he’s still going to get released. Average ballplayers don’t control their own destiny – they’re clutch rate will bounce around, even though they will stay in the league.
The place to look are the relief pitchers. Clutch is only a part of it. (WPA is the other part.) In any case, it’s the time of the game that the manager frequently imposes on game sequence. The late innings, in close games, also feature the best relievers. There is a both a relationship and reason for further study.
One could postulate that both WPA and Clutch are relevant to relief. WPA may be clouded by a bullpen that has a strong frontline but not much depth.
In any case, the Royals bullpen is “lucky” again this year. If that’s what you want to believe. Or that the Athletics are extremely “unlucky”.
I tend to think “clutchiness” is a real thing but probably immeasurable currently. We also have to be aware that by looking at Major league players we are in fact dealing with some level of survivor bias. Someone on this site wrote an article showing how pitcher velocity spikes in their debut. I think it’s reasonable that different people would handle certain situation better than others. There may be a true talent difference between major league players in this skill but we are unable to measure it currently.
Well I think literally you’re showing that team clutch across baseball is evenly distributed over a season when looking at very large sample sizes. Not sure what that proves though. You’ve intentionally mixed in “lucky” teams with “clutch” teams. Then made the sample large enough for the “randomly did well in clutch situations” noise to drown out the “I have some skill in clutch situations” signal.
Since “Clutch” isn’t well understood, I doubt any entire teams have it as a skill. Even if they did, it would be 23 players w/o the skill creating noise for the 2 that do. I thnk you might have better information (trying to understand any “Clutch skill”) if you looked at player years. I don’t know that any GM has tried to assemble an all-clutch-team, of if a successful coach has ever managed to teach it.
I see the point of your criticisms of these specific charts, but the topic has been studied a lot.
Just Google something like “does clutch skill exist in baseball”, and you’ll find plenty of studies where people have looked at player comparisons from year-to-year and found no evidence that performing better in the clutch is an identifiable skill.
Before I reply in full I either need an actual definition of “high leverage” as well as a better definition for clutch than “teams/players with high clutch scores have really good timing.”
I think it’s silly to assume that entire teams will be clutch/unclutch year to year. That’s common sense. Aside from say, the Giants, teams acquire/lose players quite often.
I do however believe in clutch players. I also believe that any formula we currently have to describe “clutch” falls WWAAYY short. To measure REAL clutch, one would need to lay out every possible scenario for every at bat (even every pitch) for every game, and I don’t see that happening any time soon. I.E:
Hypothetical Wild Card Game, 8th inning, 0-0 game, Runner on 2nd, 0 outs.
Batter A is facing Kershaw. Kershaw’s pitch count is 115. Batter A has an epic 15 pitch at bat in which he finally hits a slow dribbler to the second baseman, advancing the runner to third, but out at 1B in the process. Kershaw gets pulled, so Dodgers are forced to bring in a worse pitcher. Infield is in now. Then Batter B, of all things, hits an even slower dribbler to the second baseman. It’s so slow that the runner scores. Batter B is a hero because he drove in the run and is seen as “clutch.” Yet any real baseball fan knows that Batter A is also “clutch” because 1) he advanced the runner and 2) he ran Kershaw from the game after he bared down and had a 15 pitch AB. However, Batter A is not AS clutch because he was not presented with the same opportunity as Batter B. If only the manager had flipped flopped Batter A and B in the order, batter A would be gettin all the babes tonight. Poor Batter A :'(
“CLUTCH” as most people view it IS NOT AN EQUAL OPPORTUNITY STATISTIC.
So let’s see what all the variables we could change are…
1) It could be the 1st game of the season. Hell, it could even be the 2nd or 3rd game of the season. Maybe even the 114th game of the season. Maybe the 163rd? 163 and Wild Card seem an awful lot alike and should therefore be close on any “clutch scenario.” The later we get in the season, the more “clutch” we get
2) It could be the 1st inning, 2nd inning, 3rd, 4th… Just like the season, the later we get in a game, the more “clutch” it seems
3) The score could be 0-1, 1-0, 17-0, 0-17. The closer the score is, the more “clutch” it seems
4) The Runner is on 1st, not second. Batter A hits the same dribbler to second. Now we get a double play. Batter A isn’t clutch at all! What an asshole. OR The runner is on 3rd. Now Batter A is a hero and gettin all the babes tonight. OR there was no runner. Batter A simply grounded out to 2nd.
5) There is 1 out instead of 0. Batter A advanced the runner to 3rd on his dribbler, but now, with 2 outs, it’s going to require a hit from Batter B to get the runner to score. His dribbler isn’t as clutch as it was with 0 outs. OR there are 2 outs. The dribbler is the 3rd out of the inning.
6) The pitcher isn’t Kershaw. Something about it makes it more “clutch” being that it was against Kershaw. Think Matt Carpenter. Was he more clutch for hitting playoff bombs off a dominant Kershaw? Or would it have had the same clutch factor if he hit bombs off Joe Blow with a 3.90 ERA (or 4.50 FIP for some of you fellas)?
7) The pitch count was 80 instead of 115. Suddenly the 15 pitch at bat doesn’t seem as special or “clutch.” Kershaw gets to stay in the game instead of bringing in Joe Blow.
8) Batter A has a 2 pitch at bat instead of a 15 pitch at bat. Again it’s not as clutch because Kershaw gets to stay in the game.
9) Instead of a dribbler, Batter A hits a single, double, triple, or a jimmy jack. I think everyone can agree that any of those are more “clutch” than a dribbler to 2nd. Batter A is super clutch and for sure gettin those babes tonight.
10) Instead of a second dribbler in a row, Batter B goes ahead and knocks the snot off the ball vs Joe Blow for a 2 run homer. Suddenly Batter A doesn’t seem that “clutch” for advancing the runner.
So really, until there is a clutch statistic that takes into account every pitch and every scenario of every game, there is no way to say “clutch” doesn’t exist, or that it’s irrelevant or whatever some of y’all are trying to say it is. Surely y’all aren’t trying to say it doesn’t exist though. Clutch doesn’t have to mean driving runs in. There is no way to measure all true aspects of clutch. There probably is, but it would take forever.
I’m a Rangers fan. For the last 6 seasons I’ve probably watched 140+ games. During that time I have seen over and over and over Elvis Andrus (and his bad AVG) use an inside out swing to hit the ball behind runners, advancing them to 2nd or 3rd, so that the big boys can drive them in. If you don’t think that’s clutch you’re dumb, and probably don’t really understand the game. He could try to hit gappers, and in turn strike out.
Or what about Colby Lewis in 2010. In the regular season his ERA was 3.72. In the playoffs (4 games vs the best teams in baseball) his ERA was 1.71. WWHHAATT?!?!?! No way
Surely that was a fluke though, so let’s look at 2011. In 2011 his regular season ERA was 4.40. In the playoffs (4 games vs the best teams in baseball) his ERA was 3.04. One of those games was bad. Three were good.
Surely Lewis isn’t clutch. No way. Because clutch doesn’t exist. I bet he has some rare medical condition that makes him better once he surpasses the 200 Innings Pitched mark. I mean that makes WAY more sense than any silly “clutch” thingy ma jig.
I guess I ended up replying in full. So I don’t need those definitions anymore. Just remember gang, as I just outlined for y’all using a simple dribbler to second.. until it includes eevveeryythiingg, CLUTCH ISN’T AN EQUAL OPPORTUNITY STATISTIC. YYYAAAYYYY!!!!
In the heat of the moment I said “you’re probably dumb…” I don’t mean that. I bet y’all are really smart at mathematical equations that don’t necessarily translate to real baseball.
I couldn’t have said it better myself, but I’d like to add a few things:
1) Clutch depends on where your team is at in the standings. A 9th inning walkoff home run in a 9-8 game when your team is 20 games out vs an opponent that’s 25 games out is clearly not the same as hitting the same home run in in a 2-1 victory vs an opponent who you’re tied for the division lead with on September 28th.
2) It seems logical to me that if you look at all these graphs in a macro way like this article does, of course things will balance out and you get near-zero correlation. For example, for every clutch (however you define it) HR a batter hits, the pitcher he hit it off is non-clutch to an equal and opposite degree. It balances out perfectly.
3) What if Batter A in Just Relax’s example hits a 420 foot bomb to CF that Mike Trout leaps and snags as it’s going over the wall? So now he’s not clutch? What if instead Batter A hits a jam shot that bloops of the first baseman’s head for a hit? Now he IS clutch? It would seem that batted ball authority needs to be taken into account somewhere.
What would be really interesting would be to analyze players on the teams in the upper right and lower left quadrants. What characteristics do THOSE players have? Could the answer lead to better roster construction? Better drafts? Trades? Free agent signings?
The problem is that “clutch” is a state in a player’s mind. You know it when you see it. but it reveals itself over a long period of time. David Ortiz has it. Derek Jeter had it. Pete Rose certainly had it. Maybe we should be looking for some innate psychological trait common to these rare players.
Can you be “clutch” on a bad team? Can you be “potentially clutch?
OVER the first baseman’s head… OVER.
Point #2 there, nice.
All of the points are good, but point #2 explains the 0 correlation, rendering the team/macro argument useless.
These charts show separate correlations for batters, starting pitchers, and relievers, and they’re all still 0. It’s not double-counting.
Well that certainly is a lot of words.
Definitions of Clutch and Leverage are in the fangraphs glossary.
what did I just read
I see that some of you silly nugget heads didn’t enjoy my article there. Maybe instead of neg voting you can give a rebuttal. Don’t just neg vote because my valid points disagree with your short, vague definition of clutch. I truly don’t care that it wasn’t a popular piece, but maybe be a grown up and use your words now, children.
For those of you that are good at punching numbers and not so great at understanding the game baseball, I was essentially saying that any formula we have for clutch right now falls WAY short.
Another lad below said something to the same effect. He basically said no Braves player is in a high leverage situation right now because the games don’t matter. To assume a Braves game is equal to a Ranger game right now is false. So again, until there is a better, all encompassing definition of high leverage/clutch, there’s really no reason to try to prove or disprove clutch.
To everyone reading/commenting on this thread, please take some time to read this article on the CommunityGraphs site:
http://www.fangraphs.com/community/stop-thinking-like-a-gm-start-thinking-like-a-player/
Not directly related to Clutch, but relevant to how we’re all discussing it.
Thank you. That’s one of the most interesting pieces about baseball that I’ve read, maybe the most interesting one that I’ve read in all of 2015.
No time to read all the comments, so apologies if this has been addressed already.
If clutch exists, we might expect it to have something to do with the psychological state of the player. If so, we would expect it to change through time for each player. Sometimes their psychological state will be high, sometimes it will be low. If so you would expect CLUTCH to be autocorrelated through time. If the rate of change of CLUTCH is high, then looking for a correlation from the first half of the season to the second half of the season would be expected to find no relationship even when one exists.
Given the way hitters talk about hitting, and how it varies through time, I don’t think the methods chosen here are appropriate to test whether it is a real phenomenon. CLUTCH can be a real phenomenon and still not be predictive across a season.
Essentially you’re stating that each player has a clutch + and clutch – state that he can move between, correct?
Well, if he can’t control that and its not a repeatable skill to stay in the clutch + state more frequently than others, then we’ll never be able to find it and its probably not a real thing anyway.
For every player in a clutch situation who succeeds, there is another player in the same clutch situation who fails. Literally. If it’s a clutch, or Really Important, or high leverage, or pressure, situation, there’s a batter and a pitcher. One will succeed and the other will fail. So when looking at aggregate team performance, the only thing we can judge is if a team, in aggregate, has managed to amass a collection of players who stand out in their clutchiness or lack thereof. An R-squared on the population of teams on first-half or second-half performance won’t even tell you that. And “that” is not even what we should be looking for in the first place.
Further, of the two players in a clutch situation, the batter has a low probability of success as a normal course of their work, and the pitcher has little to no control over what happens when the ball is put in play. I would think this has a significant impact on our ability to identify what constitutes success. Bottom of the 9th, 2 out, tie game, runner on 2nd… Is a BB a clutch outcome? And for whom?
That’s not the worst part. The concept of clutch analysis requires that a clutch situation arises. For a pitcher to be in danger of losing a game his team is winning, the team must get into that perilous situation in the first place. That situation arises as a product of prior performance. It seems there’s a dependence within the foundation of clutch analysis, much like the mention above about performance with bases empty vs. performance with runners on. There are two differences. First, there is a clearer link between caliber of prior performance and the categorization into {bases empty} vs. {runners on} than there is between {clutch situation} vs. {normal situation}. A clutch situation can arise for a pitcher because he just turned a lead into a narrow lead, or because the offense just mashed a half-inning ago, or the defense threw the ball away. Second, in the case of bases empty / runners on the game states are objectively defined, while clutch game states are subjectively defined.
Yes, they are. And before you reflexively post, “No, they’re not,” read on.
Ultimately, we’ve tended to attempt to define clutch situations as either (a) dramatic for the emotionally-invested observer or (b) filled with pressure for the participants. There is no objective measurement of these, not even an attempt at an objective measure. Are we measuring pressure felt? Not even close. We can’t, at least not without extraordinary cooperation or a court order. Instead we subjectively determine a proxy among objective classifications. We non-players choose a proxy based on situations in which we as observers or sandlot misfits think these professionals would feel pressure. Any analysis we do is useless if we’re wrong.
And that “if” is on the quicksand foundation of huge assumptions. As the selection-bias hypothesis goes, they wouldn’t even have made it to MLB in the first place if they couldn’t handle pressure as we’d define it. And even then, you have a batter in a clutch situation opposed by a pitcher in the same clutch situation. One will succeed, one will fail, and in the aggregate it will be a wash.
After decades of analysis and debate, I think there are two things I don’t know about clutch performance:
1. I don’t know if there are meaningful differences among players.
2. I don’t know how any sane person could say they know one way or the other.
Seriously, take any analysis of clutch performance and preface its conclusions with a string of conditional statements on each assumption. IF clutch situations are properly classified, and IF there is no interdependence between performance in clutch situations and the performance that allowed the situation to arise (or IF we have controlled for it properly), and IF we have defined success properly as it pertains to clutch situations, and IF teams are attempting to select their roster around clutch performance, THEN this shows they’re not good at doing so. Do we really know anything from this? No. We are left with what we don’t know, and what we want to believe.
ah so many good ideas here to think about. nice.
IMO one of the best posts on the subject I have ever read.
“Bottom of the 9th, 2 out, tie game, runner on 2nd… Is a BB a clutch outcome? And for whom?”
Very, very nice.
I’m surprised that there is no relationship. I would expect that teams with a number of players that are shifted against would be more clutch throughout the season. Teams shift for a reason and they can’t shift with runners on base, so I would expect that a player that is usually shifted against would be better with runners on base and so his team would be better with runners on base. Maybe this effect is simply too small to measure when looking at an entire team.
“Today, it looks like near or total randomness. So it gets treated as such, and though the actual wins matter more than the expected wins do, as far as the World Series is concerned, one should understand why sometimes analysts think the expected wins are more meaningful, analytically. We’re always trying to drown out the noise.”
Drowning out the noise is the problem. Clutch is noise, but the noise consistently exists and can’t be drowned out without missing much of what actually happens. Who actually wins in any given season is, to a large extent, determined by this noise. The problem with Fangraphs’ approach is that you too often conflate — at least in the way you talk about things — two different concepts: (a) how likely an outcome is in light of what we can predict and (b) how likely an outcome in fact is. Because of noise, (a)and (b) are very different. To be sure, you frequently acknowledge this in retrospect, after certain unlikely events come to pass. But in my view, you systematically underemphasize it in your prospective analysis.
more good ideas to ponder.
If you are generally looking to figure out what will happen more than worrying about what has happened, then the noise almost literally doesn’t matter. That heads came up 10 of 14 times won’t change my opinion that [a fair coin] will likely only come up heads half the time going forward.
what if i was especially skilled at flipping that coin a specific number of rotations?
Fallacy. The saber guys always fail to grasp that what is true at the population level does not necessarily at the individual level.
I believe team clutch may in many cases be simply random noise, but I do believe players as individuals can be clutch, and their contributions to the teams clutch scores may or may not be measurable.
I also believe that the definition of leverage is incomplete. A team like the Braves has no high leverage situations regardless of game score. Until leverage takes into account the importance of the game to the team and player, any such studies are meaningless.
Also, why would we think a skill like clutch, rare as it may be, would have any less variance than other skills like BA. We have all seen teams and players have drastic changes from half to half, with perhaps neither representing a players true talent level defined as a players performance the past 3 years . At the team level, the players in the first half and second half can be quite different due to injuries and acquisitions, and like I said before, for many teams, there are simply no high leverage situations since W and L are meaningless to the player and team, and all AB’s are equally important for them since they are only playing for personal stats. For teams in a pennant race, 2nd half clutch scores should be weighted more than 1st half since they are higher leverage, at least until a team clinches a division or WC spot
there is something to this.
i’m thinking of calling it The Tyranny of Sample Size.
i.e. we know that statistically speaking, given a big enough sample size, EVERYTHING regresses to the mean….even things we know exist such as plain old talent.
Is it really true to say that just because something doesn’t exist on average, then it doesn’t exist at all? sounds wrong….and i’m pretty sure that’s kinda what we’re saying here about clutch.
and imo it seems that Jeff’s experiment just assumes its conclusion anyways.
i.e. “a) baseruns is everything we known how to measure and can show exists, b) any variance between baseruns and actual runs we can’t prove exists and call Clutch, therefore c) Clutch doesn’t exist.”
not sure jeff’s experiment actually says anything more than that.