By In Stuff

Rather Be Lucky

The Baltimore Orioles are, in most ways, a better baseball team than they were last year.

— Last year, they scored 712 runs. This year, they are on pace to score 760.

— Last year, they gave up 705 runs. This year, they are on pace to give up a similar 712.

— Last year, first baseman Chris Davis came into his own and hit 33 home runs and slugged .501. This year, Chris Davis is one of the best players in baseball — he already has 49 homers and leads the league with 341 total bases. His slugging percentage is 150 points higher. His on-base percentage is 50 points higher.

— Last year, Manny Machado was a 19-year-old rookie who played 51 tentative games in the big leagues. This year, Machado leads the league in hits, doubles and is playing a spectacular third base.

— Last year, starter Chris Tillman made 15 promising starts. This year, he made the All-Star team and you can define his improvement either by his 16 wins or his 3.7 WAR — depending on your statistical preference.

J.J. Hardy is having a better year. Adam Jones is almost exactly the same player. Bullpen pieces like Darren O’Day and Tommy Hunter and Brian Matusz are pitching pretty well. Not everyone is having as good a year as last year — closer Jim Johnson’s quality has dropped a bit as has catcher Matt Wieters among others. But all in all, it seems, the Orioles really are a better team than they were last year.

Last year, they won 93 games and made the playoffs.

This year, they are on pace to win 86 games and miss the playoffs.

What was the one thing statistical analysts repeatedly said about the 2012 Orioles? They were lucky. If I was asked to come up with the most basic way that stats folk and traditionalists disagree about baseball, I’d probably say that it comes down to the role of luck. Stats people might call it the role of randomness. But let’s stick with luck for now.

Take a look at the pitcher win, the contentious statistic of the moment. Everyone would agree, I’m pretty sure, that the pitcher’s win (like the team win) is composed of two parts — (1) run prevention (how many runs the pitcher and defense allow) and (2) production (how many runs the team scores). The pitcher has a huge role in the first part, but little-to-no obvious role in the production part. So what do you make of a halfway statistic like that?

Traditionalists, many of them, believe that good pitchers — that is to say WINNING pitchers — have an ability to prevent more runs when their team is having trouble scoring. That’s pitching to the score. Traditionalists, many of them, think that good pitchers — winning pitchers — inspire their teammates to score more runs when they are pitching. Traditionalists, some of them, will ascribe to certain pitchers an almost magical power to win games because the team needs them to win games.

Stats people, many of them, think how many runs a team scores for a pitcher (and when they score those runs, which matters in a pitcher’s win) is basically random and so the statistic is silly and generally pointless. They don’t believe this because it’s their heartfelt philosophy. They believe it because no matter how they turn the numbers inside and out, they can’t find any consistent evidence that pitchers can pitch to the score or inspire teammates to score more runs on days they pitch. They cannot find this magic in the numbers.

The point here is not the win, but the concept of luck. A lot of people don’t want to believe in luck in baseball. They want to assign meaning to things. This was the thing, I think, that drove people mad about Joe Morgan. In Joe Morgan’s world, a player didn’t succeed in the big moment because of some combination of skill and repetition and sturdiness and luck. It was because he reached deep into his soul and found something inside him that regular people do not have. By any reasonable reading, if a guy bloops a single just over the second baseman’s reach, that’s kind of lucky. But if he did it in the eighth inning, with the bases loaded and the score tied — especially if he was a player who seemed particularly gritty — Joe Morgan (and many others) would chalk it up to the measure of the hitter’s courage and grit. “That,” they would say, “is a ballplayer.”*

*Quick aside: I’m here in Seattle to write about the Seahawks as they get ready to play the 49ers, and yesterday the local media got a few minutes on the phone with San Francisco coach Jim Harbaugh. Apparently, Harbaugh had a legendary session where, in his own inimitable style, he managed in only a few minutes to say absolutely nothing. At one point, a reporter was listening to the tape of the teleconference, he stood up in the room, started walking to the back and and mock-shouted, “Well, I just learned that apparently Colin Kaepernick and Russell Wilson are both football players!”

The thing about luck/randomness is that it generally doesn’t repeat. Anyone who has had an especially good day at the roulette wheel knows that. You don’t want to downplay the role of skill and achievement — in the scenario above, the hitter DID put the ball in play, and some hitters (cough Jeter cough) do seem to have a repeatable skill of blooping a ball into the open space in right field — but the stats tend to show that randomness really is random.

Which brings us back to the Baltimore Orioles. Last year, the Orioles were a staggering 29-9 in one-run games. Going back to 1900, it was simply the best one-run record in baseball history. The 1954 Cleveland Indians, who won 111 games, did not have as good a one-run record. The 2001 Seattle Mariners, the 1998 Yankees, the 1927 Yankees, the 1963 Los Angeles Dodgers, the Miracle Mets, the Maddux Braves — none of these teams had as good a one-run record as the 2012 Baltimore Orioles.

As part of the overall package, the Orioles went 16-2 in extra-inning games, setting records there too.

So what is that? Skill? Sure, obviously, there was skill. But statistics show that one-run games — more than any other kind of games — are random. Managers and players and ex-managers and ex-players and baseball analysts have spent millions of hours discussing the strategies of winning one-run games, focusing on countless points like doing the little things right, getting the bunt down, moving the runner over, getting strong bullpen work, getting the sure out, getting the key hit, on and on, and yes, absolutely, in a micro-view, all these play a role.

But the numbers people will tell you: There’s flip-a-coin randomness in there too. I remember having a conversation with a big Orioles fan, and he was challenging me with this question: “Who’s to say the Orioles won’t be just as good in one-run games next year?” I told him it was possible, just like a second straight hot night at the roulette wheel is possible, but I wouldn’t bet on it.

“But roulette is luck,” he said.

“So is wining one-run games,” I said.

We agreed to disagree. He wanted to believe the Orioles — through determination and managerial splendor and the ability to make timely plays — had conquered the one-run game. It wasn’t roulette, he was saying, it was blackjack, and the Orioles were card counters. They had learned how to game the system.

Wednesday night, the Orioles lost to the Yankees 5-4, a one-run game, and New York slipped ahead of Baltimore in the standings. The Orioles’ record in one-run games this year? They are 16-26. It is a worse one-run record than the 50-96 Houston Astros. It is a worse record than then 54-90 Miami Marlins. It is, in fact, the worst one-run record in baseball.

Print Friendly, PDF & Email

68 Responses to Rather Be Lucky

  1. Is it all luck though? Without digging into ANY statistics, I’m struck by the fact that 1 of 2 players you point out as being a disappointment compared to 2012 is Jim Johnson. Is it correlation or coincidence that both the closer’s performance and the team’s record in 1 run games have taken a turn for the worse?

    • Tonus says:

      I don’t think it is all luck. I also don’t think we can separate luck from the performances that affect the difference in one-run performance from one year to the next. Not only would the closer’s performance have an impact, but also the team’s hitting with runners on base and in scoring position. And as far as I know, clutch hitting is as prone to fluctuation as is BABIP, which could explain Johnson’s struggles (his hits/9 is much higher this year).

    • Dinky says:

      Oh, come now. Of COURSE the quality of the year your closer is having affects your record in one run games. But consider that a very good closer will still give up a few runs here and there. If the closer pitches a scoreless inning, but it’s at home, tied in the ninth, that’s not a win and a save, it’s just extra innings. If he gives up one run in a tied game, it’s a one run loss; if he had a two run lead, it’s a one run win. Same with two runs and a three run win. So the difference between an okay closer (80% save rate) and a great closer (90%+ save rate) can simply be which tight games get his rare run allowed. That can be four one run wins into losses right there. If you really believe that closers (or any pitchers) can guarantee the save because it’s a one run game, then you aren’t looking at the same baseball as the rest of us.

    • Bob M says:

      not all 1 run games have a closer component and even when they do they are less involved with the 1 run game then you may think. Not all blown saves are 1 run games – especially when visiting team is trailing and gets a big 9th inning rally or hometeam hits multiple run homer to win game.

      another way to understand this is a closer who has a 3 run or 2 run lead and gives back 2 or 1 runs making a close game into a tight game. Hard to praise a pitcher/team that gives up runs in 9th and hangs on to win. That scenario is not about “clutch” as closer entered a multiple run lead situation only to MAKE it a tight game through LACK of top form.

    • Maybe Johnson’s down year is, in fact, also luck.

    • bluwood says:

      I’d wager that Johnson’s great season last year was the “lucky” one.

  2. civil writes says:

    You should go to the Sounders game while you’re in town.

  3. Kerry says:

    Consider 4 games.
    In one, a team takes a 4-3 lead in the 7th inning and the bullpen locks it down, preserving the one run win.

    In another, the bullpen turns a 4-0 lead after 7 into a 4-3 barnburner.

    In the third, the bullpen turns a 4-3 lead into a 8-4 deficit and the furious 9th inning rally falls one run short.

    In the fourth, the starting pitcher gets bombed, but the bullpen pitches 6 stellar innings, allowing the home team to scrap their way slowly from a 8-1 deficit to a 8-7 loss.

    All 4 are one run games, 2 wins, 2 losses…and the bullpen quality can go either way. This is where the randomness comes in and gives us team with good closers and bad one-run records and team with shaky pens and strong records in one-run games. One run games come in many shapes and sizes…

  4. Unknown says:

    This comment has been removed by the author.

  5. ndazcom says:

    Nine blown saves. At some point those were all one-run games (even if it was only theoretically, in between two runners scoring on the same play). If they won all of those it would make them 25-17 in one run games (not really accurate, just for demonstration purposes). If Johnson blew 3, like last year, that still makes them 22-20. As far as “real” one run games, when he comes in up one run, he’s blown 50% of them. So unless you want to assign all the randomness in your system to Johnson, there might be another cause for the change in outcomes (i.e. that their closer isn’t very good). But that would mean assigning value to the closer, which Joe doesn’t do.

    • Dinky says:

      Johnson has a career ERA+ of 137; this year it’s 130, which is poor for a closer, last year it was 168, which is good for a closer. BUT, if you forecast out the rest of the year, it looks like he will wind up allowing around 26 runs, which is only five more runs than last year. Five runs at the worst possible times (which we don’t know happened) causes 29-9 to become 24-14, far better than the Orioles’ actual record. Last year he had only three blown saves; this year he has nine, which still only turns 29-9 to 23-15 and competing for the best record in baseball. So given their much better runs scored/allowed ratios, what is the difference? Last year the O’s were all-time lucky. This year, they are unlucky.

    • The Rays disprove your theory completely. Rodney is 2nd in MLB in blown saves this year, with an even better year than Johnson in 2012

      Rays 2013 record : 22-21
      Rays 2012 record : 21-27

  6. I am intrinsically a “stats guy” and believe in randomness and regression toward the mean … Totally, I’ve been a researcher for more than 40 years. So I also agree that when teams win (or lose) a lot of close games, one has to know there could be a regression toward the mean the next season (of course, injuries, the draft, whatever, could easily make the team better or worse the next season). However, regarding your discussion of pitchers’ wins here maybe a bit of both is going here … That is some luck and something about some pitchers that enables them to be clutch? I recall reading a research study demonstrating that there are indeed clutch pitchers. A link to the study was incorrectly was incorrectly posted in the comments section to a Nate Dunlevy article about there not being clutch QBs … The commentor thought it supported the thesis but it didn’t. When I pointed this out, Dunlevy said that everyone knows that there are pitchers who do better in those situations … The notion being, I believe, not some magical psychological trait but that pitchers can pitch to the score, focus more or whatever when they need to. Anyway, what do you say to the study supporting the notion of clutch pitchers and Dunlevy supporting that? FYI, I really am not sold on the idea of clutchness … Really interested in more evidence, etc.

  7. Jake Bucsko says:

    I would be interested in a piece about what the Yankees’ one run record has been in the Rivera era, since a few commenters seem content to point to Jim Johnsons poor performance this year. Also, how about the Braves the last few years, since Kimbrel is generally acknowledged to be the best in the game right now?

    • Robert says:

      I quickly went back the last 11 years and the Yankees are 252-213 in one run games, which sounds like maybe Rivera gives them a definitive advantage until you realize that that win percentage (0.54) is noticeably worse than it is in all games (0.59).

      So, just looking at the Yankees, we might conclude that good bullpens/good teams should be better than .500 in one run games, but those games are much more random than larger differentials. That makes intuitive sense, a good team is fairly likely to lose some coin flip, 3-2 games on “lucky” bloop hits but are relatively much less likely to lose 6-0 games.

      There is probably SOME correlation in closer quality and one run wins, but I’ve never seen any evidence that it’s particularly strong.

    • Robert says:

      This comment has been removed by the author.

    • James Martin says:

      I can’t find it but someone like Baseball Prospectus did a study of what factors led to World Series success. I believe one of the key factors was having Mariano Rivera or rather a pitcher with a 0.53 postseason ERA.

    • David says:

      Yes, the incomparable Mariano Rivera, who is so clutch that he would NEVER give up a World-Series-winning blooper in game 7 of the… oh wait. Oops. Maybe even the greatest relief and postseason pitcher of all time sometimes falls victim to bad luck, too.

    • Dinky says:

      Most of the past 11 years the Yankees have had a great offense. They were going to score a lot of runs, meaning if it was a slugfest they were favored to win. Thus, I’m not surprised that their record in one runs games was worse than overall.

  8. David Barry says:

    The Yankees are 26-15 in one run games this year. By far, the only reason they are even in the wild card race (looking at their run differential). Last night was a great comparison against the Orioles. This is in a season where Mariano has been semi-mortal (seven blown saves but then they win for him anyways, 5-2 on the year) and they trot out Hughes and Chamberlain in important roles. Maybe it is because of Mariano, mystique, something else but they scare me in a one-game playoff as they seem to have luck on their side this season (cough Soriano).

  9. invitro says:

    Wonderful article. I did not know their 1-run record this season. I suppose some piece of the team’s 2012 chemistry, or will to win, got lost before opening day…

  10. bigsteveno says:

    Aren’t the odds of a team having the best record ever in one-run games followed by the worst record ever in one-run games astronomical?

  11. Robby says:

    I teach college statistics (and also an Orioles fan), so one thing that really bothered me about 2012 was all these articles asserting that the O’s did what they did by either luck or randomness. The stats people make a good argument that records in 1 run games do not often repeat year to year and seem like random fluctuations, but they make a critical mistake in using statistics.

    People take things that they can’t understand or explain and call them random or lucky. Statisticians model these things as stochastic processes (the fancy term for randomness) because they can’t be explained within a model, but they do not think that these things don’t have reasons. This is why I don’t agree with the roulette wheel example. Roulette rolls are independent. They don’t influence each other. Skill in 1 run games is not independent. There are reasons a team wins or loses games. We just don’t understand these reasons fully. As we can better model why pitchers (and hitters) do well or do poorly, we can understand why these records fluctuate.

    Also, just because we can’t measure concepts like “clutch” or “choking” doesn’t mean they don’t exist. Players believe in them and it affects their psychological state. It would be absurd to assume that players’ psychological state doesn’t affect how they perform. I think this partially explains why the 2012 Os were so good at close games and why the 2013 team isn’t. When things are good, confidence breeds success. People are more comfortable and tend to let their internalized skill training play out without overthinking things. When things are bad, the opposite happens. People tend to overthink and start to engage more executive control, which causes a reduction in performance by reducing the use of internalized techniques. The same people who tend to be most comfortable when things are good are also least comfortable when things are bad. So they tend to swing from being very confident when they believe (like the O’s in 2012) to being not confident at all when they don’t believe (like in 2013).

    • Trent Phloog says:

      I take your point about events influencing each other, but this:

      ‘just because we can’t measure concepts like “clutch” or “choking” doesn’t mean they don’t exist.’

      …seems like an odd thing for a statistics professor — presumably a person of science — to say. How is this different from asserting the existence of fairies, despite the overwhelming lack of evidence?

    • Bob M says:

      This comment has been removed by the author.

    • Bob M says:

      162 Game Avg .312BA .381OBP .446SLG .828OPS
      158 playoffs .308 .374 .465 .838

      Jeter certainly considered “clutch” – perhaps clutch is just ability meets luck at the right time

      Andy Petitte sort of the pitching Jeter – his postseason very much like regular season

      older Redsox fans remember Yaz as clutch – and also final out in 75 WS and 78 playoff game vs Yanks – both 1 run games

    • I agree a lot with Robbie … That’s why I think the verdict is still out on “clutchness” … Sports are so complicated in terms of what influences winning that no one or two or three studies proves that something doesn’t exist. Every study makes assumptions that determine what data they are going to include and what statistical approach they are going to use. Having a null result can come from bad research, or just research that doesn’t include the relevant data or analysis. This is not actually meant as a criticism of the studies … Often you need a bunch of studies to finally figure things out … That’s how research goes. This is not like believing in fairies, because as far as I know people aren’t explaining phenomena by the use of fairies (other that tooth fairies) … We are looking at phenomena that exist, but we’re not sure of the cause … So we’re trying to figure that out. And, like I said, I did actually read one study that demonstrated clutches in pitchers … So it is not accurate to say there is NO evidence of that … There is evidence for that (not absolute proof, but evidence).

    • jkak says:

      Uh-oh, Trent, someone with an education is attacking the citadel. Better toss out the fairies argument.

    • Trent Phloog says:

      Sorry, guess I should have gotten an education first.

      “…people aren’t explaining phenomena by the use of fairies…”

      A lot of people did, for a long time… and then we came up with science. And actually, feel free to substitute “God” for “fairies” — many people still use that to explain a lot of stuff.

      Also, I don’t find “I read this one study one time that said stuff” to be a terribly compelling argument. Could we have a link? Who wrote it, using what evidence?

      And as you yourself point out, we need a LOT of studies supporting each other to feel confident in any result. A single outlier doesn’t support your view.

    • jkak says:

      One major point of Robby’s comment is that it is a mistake to attempt to statistically disprove concepts such as individual and team clutch performance by relying on statistics that measure something else. Statistics do show that over a sufficiently long period of time, results tend to normalize. However, that fact does not eliminate the reality that individuals and teams do perform at unexpected levels for sustained periods of time. And the fact that those individuals or teams do not repeat that performance the next month or season does not mean that the prior performance was the result of pure statistical randomness or luck.

    • Sorry to not provide a link. I have a PDF on another computer and will try to locate it, the link, etc … Hoped that Nate Dunlevy would have it handy and get it here quickly.

      Yes, one can replace fairies with god … But we have not been trying to explain sports phenomena as long as we have been trying to explain a lot of other stuff for which fairies or god might be considered relevant, so I think the verdict it still out on clutchness … Or, as I am more interested in, not being clutch

      I’ll try to locate a link … But won’t be off work for 6 hrs ….

    • Ah, easier to find the link than I thought. Here is a link to the study I was referring to. Again, it addresses pitching to the score, not a psychological aspect of being clutch:

      And here is Dunlevy’s comment on pitchers … As I said, he doesn’t attribute it to clutchness but does think some pitchers do better in these situations for other than psychological reasons.
      Nate Dunlevylong ago-
      1. I agree clutch pitching exists. Everyone does. It’s not a ‘quality’ but inherent to the nature of pitching as I explained. Given the way pitchers pitch, it would be astounding if there wasn’t clutch pitching. Elite pitchers rarely pitch at max effort until they have to, meaning their best stuff is used for high leverage situations. That’s an issue of strategy, not character.

    • FranT says:

      Bob M, I see your point. However, to be fair, guys like Jeter and Pettitte put up those great postseason numbers against much tougher average competition than they faced in the regular season. Perhaps players have the ability to dial up their focus just a bit for big games?

    • I, too, have fancy degrees and teach university students! But anyway, Robbie is either wrong or making dangerous assumptions in a number of places, here. Let me grab a couple:

      Robby wrote “When things are good, confidence breeds success. […] When things are bad, the opposite happens.”

      According to what… your gut? There’s a certain common-sensical logic here, but there’s nothing to support it. There have been plenty of looks, for instance, at the predictive value of a hitting streak. It shows that, for instance, if you pull together a set of hitters who have a 20 game hitting streak and then look at game 21, the hitters aren’t unusually good – they perform to their career norms.

      I’d also point out that “confidence” is often ascribed post-hoc. We look at a 28 game hitting streak and say ‘wow, he must have felt confident’ or maybe the hitter himself says that. But it’s unlikely that he felt any more or less confident in game 29. So why don’t we say he felt confident for 29 games, at least? No good reason.

      Robby wrote “People tend to overthink and start to engage more executive control, which causes a reduction in performance by reducing the use of internalized techniques. The same people who tend to be most comfortable when things are good are also least comfortable when things are bad.”

      Again, even if this is true, (and I’m not certain that it is) you can’t really prove that it produces a meaningful effect. It also implies that the player is performing independent of all other players on the field. And should, say, the effect of the pitcher’s confidence be mitigated by the effects of batter’s confidence, the fielders’ confidence, and so on? I would think so.

      But now we’re talking about effects that are so small, so numerous, so contradictory, so hard to isolate, and so hard to prove that it really isn’t worth the effort, is it? Just call it luck.

    • Just another take on why the issue of being clutch or not isn’t easily given up (and my greater interest in the not clutch aspect.) I think most of you would agree that, in real life, there are people who are anxious, have little confidence, or whatever, so don’t do well … In school, at work, in little league, whatever. I work with anxiety disorders .. I haven’t worked with professional athletes, but have worked with amateur athletes, on air newscasters, actors, singers, businessmen, whoever, who have problems with anxiety that limit their performance at work and interfere with their personal lives in many ways. Some look nervous, others seem OK to observers, but don’t perform up to their capability. I really don’t think there is any doubt that when they become less anxious, they perform better. The issue seems to me to be, have all those with anxiety or other similar issues been weeded out by the time you are looking at professional athletes? Maybe. I do know world class athletes in Olympic sports who believe their performance has been helped by sports psychologists. This is definitely not proof that there are clutch or non clutch athletes in professional football or baseball. Anecdotes about clutch or not clutch athletes do not prove that the phenomena exist. However, non clutch, at least, exists among people in general, and we all probably see it … So it is not crazy to keep looking for proof of it in pro sports even if a few studies have not confirmed it.

    • Trent Phloog says:

      Thanks for the link, Andrea — that is an interesting study, and pretty convincing on this particular point. However, it also seems kind of beside the point w/r/t “clutchness,” as people generally talk about it.

      I agree with you that “strategy, not character” means that pitchers may pitch better/tougher/stronger in certain high-leverage situations. But that doesn’t really apply to batters — or teams as a whole — since we assume they are ALWAYS trying their hardest to win.

      The question seems to be, are some players able to find another “gear” when the moment demands it? Can players really give “110% percent” sometimes? That is what “clutch” believers seem to think, and I think that is pretty obviously untrue.

    • Trent,
      Actually, a single study would support someone’s point of view … It just wouldn’t prove it.
      Plus, the point of view I am putting forward, is that I don’t think we know yet one way or the other … So, I guess I am opposing the point of view that the verdict is in and the discussion over on whether some athletes can raise their performance level (or perform worse) in clutch situations … And if that happens is it skill or is it a psychological factor.
      I do think it hasn’t been proven one way or another … And feel strongly that we don’t know! :-)))
      I do not feel strongly at all that either clutch or non clutch exist among pro athletes.

    • Trent,
      I don’t think it addresses clutchess per se (though it is relevant … They just chose a different and perhaps more reasonable explanation). I actually brought up this paper initially because Posnanski seemed to be dismissing pitching to win as well as psychological clutchness. So was wondering what he thought of this and of Dunlevy’s comment. Maybe I misunderstood Posnanski’s point?

    • Robby says:

      I’m glad I got to put a bit in for the idea that confidence breeds success and I want respond to a few points by neilshyminsky. I mostly wanted to put a bit about how there is evidence that, but I meant to say that confidence may lead to success for many people. For instance, Dennis Proffitt has work that shows that success leads to differences in even basic perception, such that people who make more free throws or score better in a golf round actually see the rim or hole as bigger (an example paper is here: , his website is What this shows is that an athlete’s performance is affecting things as basic as visual perception.

      I agree that the statistical evidence for the idea of “clutch” players is low. But I also think that just because it’s hard to measure it doesn’t mean that clutchness does not have an impact or that there is no such thing as hot streaks or momentum. And given that in lab studies, we can show that pressure and past performance can cause changes in future performance, I think it’s hard to completely deny the concept of clutch.

      Second, the reason I think clutchness or hot streaks matter and don’t want to let the concept die is this: athletes believe it matters. That belief makes people do different things to try to break their streak. Baseball players play differently in the 9th inning of a close game. When players get in cold streaks, they do something about it as well. They might tinker with their swing or look at their mechanics or watch more film, or anything else.

      Those changes likely affect performance. We can’t statistically measure that because all the variability is on the individual. One person may tinker with things they shouldn’t and make what is essentially a random fluctuation in performance become a big problem. Another might study film and find out that he/she has a mechanical flaw that can be fixed and get back performing well. When we look at the aggregate, this looks like it is random data, but on each individual case, there may be reasons that clutch or hot/cold streaks matter.

    • Trent,
      You say: “The question seems to be, are some players able to find another “gear” when the moment demands it? Can players really give “110% percent” sometimes? That is what “clutch” believers seem to think, and I think that is pretty obviously untrue.”
      I think you may be right, but I don’t agree that “it is pretty obviously untrue.”
      I think it may be an illusion, but sometimes it does seem a player wills his team to win, or really comes through … Yes, it may be an illusion, but I don’t think it is obviously untrue and the studies I have seen are fine and are evidence but to me they are not conclusive. They make assumptions I wouldn’t make, they operationalize clutchness in ways different than I do, etc. I do think there is more evidence that clutchness isn’t real than there is evidence against the existence of players who are not clutch. Studies don’t necessarily address both.

    • I think clutchness can be real without being statistically demonstrable. What if clutchness isnt constant but rather situational?

      What I mean is maybe players are paying closer attention in the dugout during close games. Maybe they are watching more scout videos during the playoffs. Maybe they notice one particular pitcher is tipping his slider when pitching from the stretch.

      That hitter comes up in a high leverage situation and he has noticed the pitcher doing something. So he sits on that scrap of info and goes deep to win a walkoff. Or whatever. Was that clutch hitting? Did he reach back and hit an extra gear? I would argue that he did. But I would also argue that the skill would not always be repeatable in all situations because he isnt always paying that much attention. Or maybe he is paying that much attention but the pitcher only tips his pitch when going from the stretch and in a non-clutch situatiin the pitcher is not going from the stretch and the opportunity never comes.

      I assume players are paying more attention during these situations. I hope they are anyway. Maybe A-rods clutch struggles are because he goofs off in the dugout. Remember his phone number trick last year? He isnt choking so much as failing to focus… And focusing is what gets called clutch.

      Maybe players watch their diet and their sleep better during the playoffs. Maybe some start paying attention to their hitting coaches. Maybe their hitting coaches start paying better attention. In short I believe in LOTS of clutch-related things but I am not surprised that it cant be measured.

    • Trent Phloog says:

      There are without question players who are NOT clutch… They are low minor leaguers, company softball players, you and me, etc. But to think that Derek Jeter (for example) can become significantly better when he really feels like it… as a fan, I would ask why he’s not trying harder the REST of the time.

      Also, to Stephanie and others who say clutchness is real but not repeatable or measurable… Well, I contend that means it is NOT real. What you are seeing there is luck, or randomness, and only your desire to tell a story about it creates the illusion of specialness.

    • Trent you can contend whatever you want. The statistics so far as I understand the, support you. However virtually all players and coaches believe in clutchness. I tend to believe them and their arguments do make sense

    • Stephanie & Trent,
      I think that if there is such a thing as being clutch, or not being clutch, it needs to be demonstrable … But I’m not sure it has been proven yet one way or the other. The one study that demonstrated that some pitchers can do better, didn’t discuss it as a psychological trait but they did not actually present data that would allow us to decide clearly what caused it or didn’t cause it. Maybe there are players who perform better in key situations because they focus more (try harder) or have particular skills that are more needed in those situations or maybe it is something more sychological. Maybe we will end up showing that what coaches are seeing as clutch is explainable by something other than luck but that something is not the psychological “clutchness” that players and coaches think.

    • A couple quick responses to Robby:

      Robby wrote “For instance, Dennis Proffitt has work that shows that success leads to differences in even basic perception, such that people who make more free throws or score better in a golf round actually see the rim or hole as bigger…”

      I don’t dispute whether this is true or not. It may very well be a skill that we can measure – these might be why some players are good over the course of a whole career. But to explain a relatively short hot or cold streak? The evidence just isn’t there, and the first-person account just isn’t very reliable. (Jack Morris was convinced that he pitched to the score. Joe Posnanski, on this blog, showed that this simply wasn’t the case.)

      Robby wrote: “But I also think that just because it’s hard to measure it doesn’t mean that clutchness does not have an impact or that there is no such thing as hot streaks or momentum.”

      It’s worse than that, though, because there *is* plenty of evidence that it’s almost entirely random. Hitters who are “clutch” in one year will be below-average in high-leverage at-bats the next year. Pitchers who are unusually reliable in close games one year will be awful in those same situations the next. And it’s often the case that people described as “clutch”, who do well in high-pressure situations, are simply good players who do well in *all* situations. So, there’s plenty of data. And it just doesn’t support the concepts of clutch or momentum.

      Robby wrote “Baseball players play differently in the 9th inning of a close game. When players get in cold streaks, they do something about it as well. They might tinker with their swing or look at their mechanics or watch more film, or anything else.”

      That’s happening all the time, though, and I think that what you’re describing is susceptible to confirmation bias and selection bias. Logically, we know that hot and cold streaks end simply because a player can’t continue playing above (or below) his talent-level.

      Robby wrote “When we look at the aggregate, this looks like it is random data, but on each individual case, there may be reasons that clutch or hot/cold streaks matter.”

      When we look at the aggregate, even when the individual has made some sort of dramatic and meaningful change, it’s *still* a lot of randomness. That’s because the hitter or pitcher is never acting on their own – their performance is still influence by the quality of the opponent, the field, the situation, and so on. That’s what it takes so long for us to figure out how much of a player’s BABIP is luck and how much is skill. And that’s why a 10 game hot or cold streak is still effectively useless at telling us whether those changes had real effects or not.

    • jim says:

      jkak wrote: “Statistics do show that over a sufficiently long period of time, results tend to normalize.”

      You should read up on punctuated equilibrium, chaos theory, or so imply non-normal distribution.

      You have a common (if fundamental) misunderstanding of statistics.

    • Luis says:

      So, it is ridiculous to call the “unexplained” random or luck, but it is perfectly fine to call it clutch or winning ability or grit?

      At least statisticians looked at the data and emitted an opinion after finding nothing, and by “looked at the data” I mean hundreds or hours of research. I trust this a lot more than a 2 seconds assessment of an event as clutch.

    • Neilshiminsky, you said:
      (Jack Morris was convinced that he pitched to the score. Joe Posnanski, on this blog, showed that this simply wasn’t the case.)
      How exactly did he do that? He certainly put together a logical argument about regression to the mean, which I buy. He put together a logical argument about pitching to the score and made clear he doesn’t believe in it and said why. But I don’t see where he showed it wasn’t the case in general or for Jack Morris. There is at least one study, which I posted a link to, and one other sportswriter who has said that belief in pitchers being able to elevate their performance in clutch situations (by something like focusing more) is widely accepted.
      You seem to think that it’s proven that pitchers can’t pitch to the score. I’m curious why? Or perhaps I misunderstood your point.

  12. Bob M says:

    God bless Bill James:

    1986 Abstract – Angels section p73 in original Ballantine Books edition – I will quote:

    “What anyone with a statistical analysis would assume…is that just the opposite is true: that the smaller the margin of victory, the more likely it is that the better team will lose.”

    Basically it is common sense – the Astros will need all their resources to beat the RedSox, but the RedSox don’t as they have superior pitching, hitting and defense. A .600 team will win a higher percentage of blowouts and generally lose a higher % of 1 run games than their overall record. Good teams have the ability to score lots and prevent many runs – poor(er) teams do not.

    Mr James takes the illogic of superior teams winning highest % of 1 run games out to season conclusion – who are the GREAT teams – winners by 10 or more games or the teams that squeak in by 1 game?

    To add to this discussion, Bill James 1987 Abstract spoke about high liklihood that a team, like the O’s, who experience dramatic improvement of 20 or more games (2011 won 69 games; 2012 won 93 – 24 game improvement) will decline in the following season. As typical he supports his position, but luck clearly is part of the equation.

    On a positive note for O’s fans – after the dropoff year a team that really is good and young (as these O’s are) will bounce back

    • Bob M says:

      some examples from the extremes
      1998 Yankees .717 overall %
      1-Run Games (31) 21 10 0.677
      Blowouts (55) 42 13 0.764

      2001 M’s .717 overall
      1-Run Games (38) 26 12 0.684
      Blowouts (44) 34 10 0.773

      1962 Mets .250 overall
      1-Run Games (58) 19 39 0.328
      Blowouts (45) 8 37 0.178

      2003 Tigers .265 overall
      1-Run Games (37) 19 18 0.514
      Blowouts (47) 7 40 0.149

      Both Mets and Tigers won nearly 50% of total victories in 1 run games – Tigers even had a winning percentage!

      records in 1 run games just not a good indicator of a teams overall quality

  13. Phil says:

    This is why I settled on run differential as the best measure of team quality in my SABR By the Numbers article: good teams win blowouts, lucky teams win one-run games. See pages 15-18:

  14. nscadu 9 says:

    I think we grant clutchness to a player that maintains composure under pressure. So called clutch players perform as they do in all situations. Players that are seen as choking make poorer decisions or fall prey to nerves under pressure. It is easy to do your job or perform in a comfortable setting, but people react differently, when the pressure is on or the boss is watching over your shoulder or when they are taken out of their comfort zone. Some can handle it and some cannot. Think Ryan Howard watching a close pitch for 3rd strike to end the NLCS. Luis Gonzalez or Edgar Renteria may have won World Series on lucky bloop hits, but at least they maintained enough focus to put bat to ball.

    • “Luis Gonzalez or Edgar Renteria may have won World Series on lucky bloop hits, but at least they maintained enough focus to put bat to ball.”

      But they did it badly. If you bloop a ball, you hit a pop-up – basically the worst kind of contact possible – and got ridiculously lucky by hitting it to a spot where no one was playing. Had those balls followed a slightly different path, we’d be talking about how they duffed the pitch in the most horrible way possible. Who does that demonstrate “focus”?

    • nscadu 9 says:

      I agree that there is some luck involved and I don’t believe there is some magical part of someone’s soul that comes out only in clutch opportunities. My point being that nothing happens if you’re so unnerved you don’t take a hack at it or miss it completely. Or if you’re the pitcher can you maintain mechanics and location with the game on the line. In that situation with opportunity to win the WS and a single wins it against a lights out closer, I’d say just getting a bat on the ball will do. It takes focus. Luis muscled an inside pitch from Rivera out of the infield, some luck but wouldn’t say he did it badly. Renteria’s was a slap hit where they ain’t. Pop ups are something else and are worse contact.

    • clashfan says:

      nscadu, the problem is that the group of players who do well in clutch situations varies from year to year. If, say, Chipper Jones had been in the top ten in OPS+ in clutch situations for fourteen of his eighteen* seasons, that would indicate a skill. I don’t think it’s been shown that players do that. If it’s not a repeatable phenomenon, how can we say it’s a skill?

      *–Eighteen full seasons

    • nscadu 9 says:

      I don’t think being clutch is a skill, rather not succumbing to the pressure takes some mental strength. Look at Bonds 2000 and earlier. You could see it on his face in every playoffs and he was nowhere near what he was in the regular season. I think it was Andy Van Slyke who had a good anecdote about Bonds attitude in the playoffs back in Pittsburgh’s last playoff appearance. That unsung guy that suddenly comes through in the clutch is just odds coming around or dumb luck. I think we are talking about guys that seem to consistently perform in the clutch. Guys like Jeter or Ortiz and yes, Renteria. Renteria is consistently very good in the WS, but he is only playing up to his ability throughout a playoff run. They don’t elevate so much as they retain their focus as in all other situations.

    • Yes, the reality may be that clutch players play as well in tough situations as they do all the time. That the difference really is that other players play more poorly in those situations. So most analyses aren’t looking at the right data … It’s not a matter of “statistics” it’s a matter of design analysis. To be simplistic, you need to look at 4 cells at least: non clutch players in normal situations, non clutch players in clutch situations, clutch players in normal situations, clutch players in clutch situations … And you are looking for an interaction. It might be that, for example, the non clutch players do objectively worse than clutch players in both situations, but they will do worse than expected in the clutch situations. Just theory, but an example of a weakness in some studies.

  15. Eric says:

    What was even more astounding, in my mind, than the Orioles record in 1 run games last year was how well they did in those games on the road. I looked up their records last year and this year and expected to see that they played in a lot of 1 run home games last year and maybe a lot of 1 run road games this year. Last year, the Orioles played over 1/2 of their 1 run games on the road and finished 16-5 in those games, good for a .762 winning percentage, the highest such winning percentage in baseball history. In fact, only 2 other teams in baseball history in non strike years finished the season with a winning percentage over .700 in 1 run games on the road. 1970 O’s (20-7, .741) and 1984 Tigers (10-4, .714). Those teams weren’t just good, they were historically good, both winning over 104 games on their way to World Series wins. That’s what made the Orioles last year so interesting was that they were doing things that had up until that point only been done by outstanding teams, and there was nothing particularly outstanding about them last year other than the 1 run game record. Incidentally, the 1971 O’s and 1985 Tigers finished a combined 24-28 in 1 run road games, a .461 winning percentage. The Orioles this year are 5-16, a .238 winning percentage. If they had the same average winning percentage of the other 2 teams, they would be about 9-11, good for an extra 4 wins which would have them tied with Texas for the 1st wild card. None of this means anything, it’s just interesting how they have gone from one extreme to the other.

  16. Sean Roark says:

    Call it what you want, but it is definitely infuriating to watch.

  17. BobDD says:

    For 60-some years Sophia Loren has not passionately burst through my bedroom door searching for me. Randomness – OR – my lack of clutch (wilted under pressure)?

    • nscadu 9 says:

      Have you at least tried to call her or taken a shot at getting her attention. Or have you left the bat on your shoulder while passing on the slightest glimmer of a chance.

  18. Claire says:


    Last year, if Jim Johnson had given up no runs at all, then I think that the Orioles’ record in one-run games would have been 25-9, instead of 29-9. There were a few games when he turned a two-run lead into a one-run win. Not much wrong with that. Of course, if he had given up a few runs more, then they might have been 25-14 instead of 29-9.

    This year, if he had given up no runs at all, then I would call their record 22-18.5 instead of 16-26. The half comes from losing a game that was tied: if he had pitched a scoreless inning, they might have lost it later.

    So Johnson is fairly strongly implicated, but I’m not sure what it all means: he’s been charged with 25 runs this year, compared with 21 last. Bad luck?

  19. Best offers on Health Products Online, best deals and experience at Shop online blood pressure monitors, thermometers, nasal aspirators, ear Wax cleaners, Diabetic Shoes, Diabetic Socks with best Healthy juices online Aloe Vera Juice, Triphala Juice, Amla Juice Giloe Juice. Solomon a GMP certified with its E-Portal Present under offers best healthcare products online.

  20. This is such a great resource that you are providing and you give it away for free. I love seeing websites that understand the value of providing a quality resource for free. It?s the old what goes around comes around routine. Did you acquired lots of links and I see lots of trackbacks??
    Dynamics AX training

  21. Buy diabetic socks online Diabetic Socks Online at with free shipping and COD. Shop online best diabetic socks online from

Leave a Reply

Your email address will not be published. Required fields are marked *