The Playoffs Are a Crapshoot: Final Thoughts

If you have missed all 5 pieces in “The Playoffs are a Crapshoot” well…tough! You should’ve been reading Braves Journal more, ya twit!

Nah, just joshing. Here are the links to the first 5 parts:

A Few Final Observations

  • You used Bradley-Terry rankings to figure out who was the best team.  Maybe there’s a better measure of who the best team is which would then improve the probability of the best team winning.

Sure.  Maybe.  But people have been looking for some measure that predict champions for a long time and they haven’t found any yet.  You can see one such effort here

  • You made the probability of winning a game in the playoffs constant.  But it isn’t really.  Pitching matchups are really important, they vary from game to game, and the essence of managing in the playoffs is to get the best matchups for your team.  In the regular season, this isn’t nearly as important, so it isn’t reflected in skill.

It’s essentially impossible to separate varying probabilities of winning from a constant probability of winning plus luck.  And while this new dimension of skill differentials is not really tested in the regular season, no one that I know has actually done the analysis necessary to show that it really makes much of a difference at the margin, especially since the machinations to improve matchups often makes a team worse by varying the expected rest of a starter.

  • Lots of other things are different in the playoffs: pressure, clutchness, etc., etc.

Prove it.  One of the things that makes sports great is projection.  We imagine the stress the players have because we know if we were at the plate in these circumstances we’d be peeing on ourselves.  I’ve stood trembling over a six foot putt with $25 on the line and stabbed the thing almost in the wrong direction.  But PGA pros manage to calmly sink those putts with $1 million on the line.  But we fail to recognize that that’s what makes them pros.  The baseball players whose nerves get to them on a 3-2 count in the 9th with the bases loaded aren’t MLB players.  Those guys got weeded out in A ball, if not before.  While the playoffs may represent a higher level of stress than the regular season, these guys are exactly the guys who are impervious to stress.  I recognize that this is a relative statement, but those who want to defend clutchness have been trying, unsuccessfully, for years.  That doesn’t mean I wouldn’t rather have Soroka on the mound in a tight spot rather than Folty, but the difference in performance is still mostly luck, or at least indistinguishable from it.

  • This focus on the “best team” is misplaced.  The playoffs aren’t about assuring that the best team wins.  It’s enough that the worst teams lose.  You didn’t test that.

I didn’t, but Pete Palmer did. 

In real life, we do not know which team was really the best, but if we assume that it was the team with the best record during the season, then that team has won the World Series five times since 1995 when the wild card was introduced, although in one case there was a tie for the best. That works out to about 20 percent, consistent with the above table. The average rank of the World Series winner was fourth out of eight. If the playoffs were completely random, the average rank would be 4.5. The worst team has won 4 times.

  • You can prove anything with math, particularly when I don’t understand what you did.

No, you can’t.  At least I can’t.

So the playoffs are a crapshoot.  It should be pointed out that a Crapshootiness Index of 0 is almost as bad as a Crapshootiness Index of 1.  In that case, all we’d be waiting for is the inevitable coronation of the best team.  Why bother to hold the playoffs?  To paraphrase Montgomery Burns: “They barely won… I don’t even know why they even bothered to show up.”

And even a Crapshootiness Index of 1 has something to recommend it.  After all, fans come to the games with rooting interests, so even if the results are completely random, we can be amused or horrified by the way that randomness asserts itself.  Just because the result is indistinguishable from a coin flip doesn’t mean you’re indifferent between watching the game played and watching a coin flip.  And as long as there is scope, even at the margins, for skill differences to manifest themselves, there is every incentive to be as good as you can be.

Does the crapshootiness of the playoffs make me feel any better about 1991-2005?  Only a little.  Even under pure crapshootiness, we would expect the Braves to have won two World Series, and, if you think they were the best team in baseball most of those years, their expectation would be closer to 3.  Instead we got 1.  But it could be worse: we could be Cleveland.

Thanks for reading! If you enjoyed this series on “Playoffs are a Crapshoot”, here is a category from the vault that you might enjoy!

Long Live Braves Journal!

Author: JonathanF

Alive since 1956. Braves fan since 1966. The first ten years were pretty much wasted. Exiled to Yankees/Mets territory in 1974 --- bearable only with TBS followed by MLB.TV.

19 thoughts on “The Playoffs Are a Crapshoot: Final Thoughts”

  1. Magnificent work, JonathanF. Thank you so much for doing it. Reading Braves Journal has been making me smarter for 15 years.

    This all makes me feel MUCH better about the Braves definitely winning the World Series in 2019, because we’ll have a much worse team than we did in the ’90s.

  2. I believe I have today finally spotted a ‘c’word, Jonathan in your estimable summary of what you had gathered, sorted, analysed and written up on a topic that clearly fascinates you. A nuance of numbers indeed – alliterate but that’s hardly strong enough for the emphasis you give it. And the number of your followers must be growing with every word – loaves and fishes next?!

    So much hard work, such erudition, those skills formed from a lifetime as an economist which taught you trust the figures, ignore the human element. Advise on it, charge for it, but, essentially, it’s a whimsy. You were captured young by this excitement – no need to worry about stuff you can’t put a number to, fuggedaboutit. I suspect as the years advanced the doubts may have grown. But as baseball now assumes a lead role in your life there’s something you can have at it and have fun without a care in the world.

    A question from my end. Who hit what turned out to be the decisive home run in Game 7 last October? How many here saw that happen, live? Who might have anticipated it from someone’s last career at bat? Very few. Clutch can lose its excitement if it’s not your guys and you live in the world of analytics. Leyritz, Lonnie Smith. Jeter’s cut off flip from the first base line. Misery, magic. Choke, clutch. The essence of post season ball.

    Your work is a major asset here and a source of much pleasure, right or wrong. Allow me to place my ladder onto your ramparts and start climbing. You, undoubtedly, will be waiting for me up top,something wicked in your hand, at the ready. Throw/pour it not, it’s fun to engage you. And educational!

    So, to baseball and 1 in 14. Crapshoot? Nah, never! Cheers.

  3. Powers of observation
    do not necessarily involve a gradation
    they can on the other hand
    include identifying emotion, be it brave or bland.

  4. You misunderstand the argument, which is decidedly not that clutch plays don’t happen — duh, we see them happen all the time — but that there is no evidence that any players possess “clutchness” as some kind of repeatable skill.

  5. Absolutely terrific series, JonathanF.

    Stu is right of course in his explication of clutchness. Moreover, even though I’m not nearly as numerate as JonathanF, I’ve found that appreciation of evidence and statistics in no way diminishes my joy in watching the games themselves.

    The finding that “luck” plays such a large role, especially in small sample sizes such as at bats, games, or series, is a large part of what makes baseball so fascinating.

  6. Bill James in one of his Abstracts (1984?) wrote about a postseason series prediction system. For all the postseason series to that point, he looked at how many times the teams that had won more games in the regular season had won, how many times the teams that had hit more HRs had won, etc. If the teams with more wins had won 5 more series than they lost and the teams with more HRs had won 12 more series than they had lost, etc., to predict a future series, you would give 5 points to the team with more wins, 12 points to the team with more HRs, etc., and the team that has more total points would be the one predicted to win the series. Of course the system did a good job of “predicting” the winners of series that had already taken place, since it was built on the results of those series, but I read many years later an article (not sure if by James or someone else) saying that the system had correctly predicted well over 50% of the series in the next however many years after publication.

    A couple of points that I remember were that the teams that had hit more doubles lost many more series than they won (he speculated that since doubles-dependent offenses relied more on sequences of hits, they would be hurt more than HR-dependent offenses by good pitching, which you face more often in the postseason than in the regular season) and that walks and wild pitches/PBs didn’t seem to matter much (perhaps because they came disproportionately from 4th/5th starters who didn’t pitch much in the postseason).

  7. Up there on 7 where I said “@ 4” should be “@ 5.

    I am almost sure when I clicked in, I knew it was 5 I was replying to. Then, when it posted up, I KNOW I edited it to 4.

    Later, I saw it was 5 I was responding to. Weird.

  8. @Cliff – I’m seeing ads counted as comments sometimes, and throwing the count off. I imagine this is temporary.

  9. Thanks to all, even blazon, for reading and liking. The fact that there is a principled way, in the aggregate, to separate luck from skill is something that still strikes me as kind of magical. That it has such immediate application to my favorite sport is particularly satisfying. And there are plenty of applications that haven’t been really explored yet (unlike this one). For example, it could be used to study clutchness. I’ve never really wanted to do it, but blazon is possibly inspiring me at some future point.

    In response to blazon, I agree with Tfloyd: there may be people who think of baseball as some kind of Strat-O-Matic game, but I’m not one of them, even as I can appreciate how closely real life can resemble a simulation of real life. But I am able to admire (or curse) the skills of players without having need to invoke either “clutch” or “choke” as a concept, or even the third “c” — character; happily, we can demonstrate that neither is required to actually understand the game, IMO. Happily, I say, because skill and character are, as near as I can tell, uncorrelated, or even mildly negatively correlated. And character is really hard to observe, while pitifully easy to be fooled about.

    I’m back in the US now, so I can answer questions more directly, if anyone has one…. Thanks again.

  10. back in the USA…

    very fair, thanks…for whatever factual errors/misunderstandings I included, my apologies.

    But I am heartened by your expressed thoughts, looking ahead, and would look forward to them.

    I haven’t changed my base on this subject but take some pride in saying the door remains open. To guys like you. You too I believe!

  11. Keith Law
    now Athletically inclined
    Riley’s bat still slow, or does he really mind?

Leave a Reply

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