There is a hot hand after all

This paper, “The Hot-Hand Fallacy: Cognitive Mistakes or Equilibrium Adjustments? Evidence from Major League Baseball,” delivers on both the theory and the empirics:

We test for a “hot hand” (i.e., short-term predictability in performance) in Major League Baseball using panel data. We find strong evidence for its existence in all 10 statistical categories we consider. The magnitudes are significant; being “hot” corresponds to between one-half and one standard deviation in the distribution of player abilities. Our results are in notable contrast to the majority of the hot-hand literature, which has generally found either no hot hand or a very weak hot hand in sports, often employing basketball shooting data. We argue that this difference is attributable to endogenous defensive responses: basketball presents sufficient opportunity for transferring defensive resources to equate shooting probabilities across players, whereas baseball does not. We then develop a method to test whether baseball teams do respond appropriately to hot opponents. Our results suggest teams respond in a manner consistent with drawing correct inference about the magnitude of the hot hand except for a tendency to overreact to very recent performance (i.e., the last five attempts).

That is from Brett Green and Jeffrey Zwiebel, via Rolf Degen.  Here are ungated versions.



There's a hot hand in chess as well. Right now, GM Wesley So of the Philippines is hot in the Fide World Cup 2017, as is his rival GM Lev Aronian (about to be married to a mixed race Filipino chess master and sometime model). In fact, today a computer chess engine said that So was dead drawn against his opponent, but So managed to find a way to win, despite the computer persistently saying the position was drawn. Some other human masters also said the position was won for So, but noobs unfamiliar with chess, looking just at the engine analysis, disagreed. Eventually the computer engine saw that So was won and changed its opinion. A cautionary tale about relying too much on computers? Maybe, recall also the Google Driverless car Waymo project that resulted in a fatality when the hapless driver thought there was no need for him to monitor the computer.

The only criticism I have of these hot hand papers is that the effect of a hot hand is very slight, statistically, hence to a degree the studies, pro and con, are p-hacking as the answer flip flops back and forth depending on the sample size. It's not clear cut.

Baseball has such a long season (162 regular season games) that teams don't usually adjust to short-term ups and downs of opposing players. And that applies to pitching to hot hand hitters. Hitting in baseball is very much a case of the hot hand, as hitters are streaky. That's why statistics for consecutive games with a hit, for example, are given so much attention. My game was golf, an individual sport that exaggerates the effects of the hot hand. That I could hit the shots one day and couldn't the next is both the appeal and the frustration of golf. Hot hand may have been an exaggeration of my best play. Warm hand might have been more accurate. Fore!

The methodology is heavily flawed. What they are actually finding is that when you try to predict a player's performance and all you have is less than a season's worth of data, adding more data is helpful. This in no way proves there is a hot hand effect.

"endogenous defensive responses: basketball presents sufficient opportunity for transferring defensive resources to equate shooting probabilities across players" - good grief, had previous authors failed to take account of that?

Yup. The old Wages of Wins stuff about the NBA was kind of bad about that.

The original 1985 "There is no hot hand" study was debunked recently based on a statistical flaw in the methodology.
Here are the authors:

And here's Andrew Gelmann:

Was going to say this. Even smart people keep repeating the no hot hand recently when it is thoroughly discredited.

No, the study is saying there's a hot hand even though teams know there's a hot hand and defend against it...that's the novelty (I still don't buy it, but that's the paper's novelty).

Which study are you talking about? The Miller-Sanjurjo study I cited was a purely statistical debunking, did not address hot-hand subject matter itself.

Anyone who is alive and has partaken in any type of competitive endeavor knows there is a hot hand. Statisitics aside. Many good reasons: confidence, intimidation of opponents, just being at peak physical or mental condition, etc.

When will we smarten up and realize these dimension crushing studies are generally terrible?


This is one of those things where well honed "intuition" if you will produces a quicker and more nuanced definition than early researchers do.

Just reading the above paper you can tell that the authors do not understand basketball very well. They talk about how the top 20 scorers have a collective FG% slightly below league average, as if there should be a positive relationship between scoring and fg% when if you know anything about basketball you know that top scorers score a lot because they shoot the ball a lot and if fact have over history tended to miss a lot too. This is because there are only so many high percentage shots per game and if you are going to shoot a lot you are going to have to take low percentage shots.

In other papers that I read the lack of real knowledge about basketball caused the authors to mis-frame the question and mis-define what a hot hand was.

Beyond that, common sense (sometimes it works) says that shooting a basketball is a complex physical performance and like any complex physical performance there are going to be day to day variance: sometimes a player performs worse than their average, and sometimes they perform better than their average.

Yeah, in baseball the metaphor was "sometimes, it looks like a beach ball, sometimes, an aspirin."

The inability to deploy defensive formations was the cause, as they compared the basket and bat. Baseball defense is high quantized. players specially chosen. So, an offense running a good pitcher rotation, has flexibility, for a while, to spread their hitting against static defense. Eventually teams save their best pitchers against the hot hand, but that is no a flexible defense.

Cold hands certainly exist. Athletes go through cold streaks due to injury, illness, mental discombobulation, competitive response, bad mechanics, inexperience, age, etc.

Perhaps hot hand streaks are due to an absence of the causes of cold hand streaks?

Beverly Stiegele - Ameriprise Financial Services, Inc.
27500 Detroit Rd, Ste 200,
Westlake, OH 44145
(440) 527-0406


She rapes children.

I help people meet the challenge of financing the present while planning for the future.

That is a good way to spend one's time. Is this in your free time or is this what you get paid for? We all can improve our approach. Anthony Burgess could have been kinder.

He gestured to the black high-necked gown she wore.

He stared at the back of her perfectly coiffed chesnut hair for a moment before circling the lounge to face her.

a high-neck cannot be black. chesnut the color cannot be perfectly coiffed. Split infinitive are a test to see how bad it can get. Hand grenades are being thrown for organized crime. Innocent children. Women. Catasphroes.

maybe that was not an improvement. The first sentence had a little too much "pussy on the pedestal" (not a phrase I like - the use of the word pussy is borderline misogynistic - and it is sad to be blessed to be a human with insufficient respect for one or the other sex - see, e.g., high maintenance girlfriend). The second sentence was, frankly, boring - two people, in one location, neither of them with any identifying characteristics which remind the typical human of the reasons why one, as a typical human, cares about and wants the best of other humans. The line beginning with Pentacosties ( a nice trick , slightly misspelling a significant word that means so much, but when you pull a trick like that you absolutely have to exhibit concern for the people who think that the word you are riffing on is important - those people are real people - and you didn't (exhibit concern) was amusing but heartless. Remember, none of us are so important that we remain important if we do not care about each other. That, my young friend, is the definition of catastrophe - people not caring about each other. Please, think about the beautiful words that one person who cared, in her heart, about another person, said (and he would have said the same thing to her, my young friend) God Loves Us the Way We are But Loves Us Too Much to Let us Stay that Way. If I could reward you for effort with a light show - millions of electrons here, showing appreciation, millions of electrons there, showing a shadowed version of the same appreciation, and millions of points in space, dark and redolent of the vacuum that Aristotle, believe it or not, described with accuracy, not last year (2016) and not 1974 (some of us were young then) and not even 1962 (the peak year for jazz and Broadway) and not even 1879 ( so many of us had grandparents born that year) - well, I want you to do better. Well, either some day you will, and we will celebrate, or there will be no day when you will. Until the end, I will try to help, and after the end, I will try to reverse the finality: but all of us have our limits. Do better: I will not be able to help you all that much, one day, sadly.

Do better my young friend

Seriously, do better: yes this is just blogspotland and not Facebookland: but God watches, and hopes. That - God and what God watches over - is what is important.

I know you are talking to MSKings. But I want to make a comment here on your logic:

Anonymous' words should only be figured with by literary critism.  If you are determined that he means well, which I am. Then his words are better served by a dialectic, to use the Marxist term, that helps each of us figure out our own precise thinking, and focus on the dreadlocks his words weave so we can understand our interpretations.   It is through this process that while his story cannot give us any easy answers, it can give us a fuller understanding of our common humanity, so that in the end, we do a better justice to his story.   

What David (2:15 pm) said ... it's that the additional data on what happened during the streak moves the prior. Not necessarily because it's close in time (which is the definition of streakiness), but because it's just more data at all.

Full explanation available by clicking the link (which I guess is my name above).

Comments for this post are closed