THE BOOK--Playing The Percentages In Baseball

So a 97-game team (.600) vs. an 89-game team (.550)?

I’d want the 97-game team to win maybe 75% of the time.  Over the 162-game season, it proved itself superior by over a standard deviation.

Honestly, I’d prefer only one team in each league making the playoffs.  I think it makes a mockery of the regular season, how the post-season is just a series of coin flips.

I guess you can do it like England does with soccer where there is no playoffs.  Obviously in that format, the best team wins the title more often. 

Personally I like it when the best team doesnt win.  Overall I’d want the best team to win more often then not but occasionally i like seeing teams that prehaps do not have the best talent end up winning.  Im not sure what the break even point I’d want it set as tho.

I like it when the best team in terms of *talent* doesn’t win.  But I don’t like it when one team dominates over 162 games and then loses to a team that won a lot fewer games.

In the World Series, you can at least make the excuse that because the teams are in different leagues, the regular season record shouldn’t matter.  But in the CS and DS, when a marginal wild-card team wins, it bugs me, even if they have more talent.

Just to be clear, when I say a .600 team, I mean a true talent .600 team.  How they play in 162 games is not the question.

I was not asking how often you’d like a team that sampled .600 you’d like to see win the World Series over a team that sampled .550, each over 162 games.

Ah, okay, sorry.  I misunderstood.

I have a preference for how often I want a .600 team to finish ahead of a .550 team *during the regular season*—and the natural probability, over 162 games, is about right for me.

As for the World Series, my preference relates to how well the team did in the regular season only, and not to how much talent they have.

That is: if the .550 team fluked out and won 105 games, while the .600 team won only 90 games, I want the .550 team to win the Series most of the time.  And vice-versa.

Since MLB can’t (realistically) make each round of the playoffs a 25 game series, maybe the answer for fans like Phil would be to give the team with the better record the home-field advantage in all 7 games. You wouldn’t get to an 80% chance of a .600 team beating a .550 team, but I think you’d get past 70%.

Of course, if we’re going to put so much emphasis on the regular season, you’d probably want to return to the balanced schedule and either eliminate or drastically expand interleague play.

I want to see the Pirates win the World Series (of course, it’s not going to happen this year).

Point is, I root for my favorite team. The rest, I could care less, just have some entertaining games.

Brian, you mean you could not care less!

MGL, you force me to agree with you.

I always want the best team to win the championship.

I don’t like it when the “champion” is the 5th best team.

If you are the Champion, in my mind, you should be the best.

I don’t see how you can always want the best team to win.  Doesn’t that seem predictable?

Anyway, to answer my own question:
Q1: 60%
Q2: 70%
Q3: 85%

If this is the case, the question is: how many games do we need in the Series for this to happen?  Before I answer that, let me tell you what happens in a 7-game series for those 3 combination of teams:
Q1: .600 v .575 = better team wins 56% of the time.
Q2: .600 v .550 = 61%
Q3: .600 v .500 = 71%

As you can see, a 7-game series doesn’t suffice for me.  I get too many upsets, relative to my expectations.

In order to get the number of upsets I can live with, then I need a best 13 out of 25.  Under those conditions, this is how often the better team wins:
Q1: 60%
Q2: 70%
Q3: 85%

So, that’s why I think it’s pure b.s. about the best team winning.  The best team doesn’t win nearly enough for my tastes.  But, for the best team to win to satisfy my tastes, you have to lose the entire drama of a best-of-7 series.

You have to choose.

In theory, I would always want the best team to win. But the key is that I wouldn’t want to know who the best team was prior to the World Series. If we’re sticking to the current number of playoff teams, I would want the regular season to be long enough to differentiate the top four teams in each league without making it clear who the very best was. I would want the playoff rounds to be long enough to determine who the best team was in each league. And I would want the World Series to be long enough to determine which of those two teams was the best. This would be easier if there were no interleague play, as that system more fits the idea of having two champions who cannot be pre-judged relative to another battle it out to determine who is the best.

Like I said, if you KNOW you have a .600 team facing a .500 team, you would need a best of 25 series just to get the better team to win 85% of the time.

So, to have your two preconditions: a) NOT know how good each team is as a prior and b) to always have the better team win would require thousands of games.  You can’t say always.  You need to put a limit, like 95% or 90%, otherwise, you may even need millions of games.

Shorten the regular season, expand the playoffs to 16 teams and four rounds, and lengthen each round.

Again, I’m trying to get the focus on what the final objective is.  Throwing out things about changing the length of the season or playoffs without knowing WHY you want to do that or the EFFECT of that would be pointless.

This is why I am asking specific questions as to what you would like to see as the end result.  I’m saying that if I know I have a .600 team facing a .575 team in the World Series and I would like for the better team to win 60% of the championships, then I need to have a 25-game series.

So, exactly what it is that you want as the objective?  Once you give me that, then we can establish the means.

Sorry if I’m off track. I’m probably not the right person to be answering the kind of questions you’re asking. Numbers like a “.600 team” and a “.575 team” don’t really resonate with me. I have no idea how often I’d want one to beat the other.

My goal in shortening the season, adding a round to the playoffs, and lengthening the rounds would be to [1] shorten what I think is an over-long regular season, [2] increase the number of fans whose teams make the playoffs, [3] make the final champion less flukey by having more and longer playoff rounds.

Ok, but why do you think the regular season is overly long?  What is it that a 162-game season is not providing you that say a 116-game season would?

The longer the regular season, the more meaningless games there are for out of contention teams (though this is obviously connected to letting more teams in the playoffs too). And I prefer the drama of the playoffs to the regular season, so I’d rather have playoff games than regular season ones in that 36-game timespan.

Or 46, even.

The team I want to win is simply the team that plays better in that series. I don’t care about true talent. I care about watching players perform well or poorly, and letting them win or lose based only on that. That’s baseball.

#11 - having the best team always win isnt predictable because I frequently dont know who the best team is.

Some years there is a clear division; many years there is not.

Right.  I have strong preconditions.  I know who the best two teams are, I know what their true talent levels are.  And under those conditions I need a 25-game World Series to crown a true winner.

The reality is that we don’t know the true talent level of each team (you potentially have 4 or 6 or 8 teams that can reasonably be argued as the best team).  So, the idea that the playoffs does anything other than give you a snapshot in time (sample performance) is ludicrous on its face.

The playoffs is drama.  It has nothing to do with crowning the best team.  Merryl Streep is the best actress of our generation, but that doesn’t mean she wins the Oscar for 30 straight years.

Tango, are you also able to calculate the probabilities for other sports?  I would imagine you need less than 25 games to distinguish a .550 NBA team from a .600 team ...

If you start off with the idea that you have a true talent .600 team facing a true talent .550 team, I don’t see why it would matter which sport it would be.

However, if you are saying that you have a team that sampled .600 for a full season facing a team that sampled .550 for a full season, then certainly it matters the sport and number of games.

Are you asking for the 2nd scenario?

Hmmm ... good point.  But I think it still does matter.

In baseball, a .600 team has a chance to lose to a .300 team, or to beat a .700 team.

But imagine that baseball games lasted a million innings.  In that case, a .6000 team would ALWAYS beat a .300 team, and ALWAYS lose to a .700 team.

In million-inning baseball, a team would be a .600 team because it was 5th best team in a league of 11 teams.  It always beats the six worse teams, and always loses to the four best teams.

So in that scenario, the chance of a .550 team beating a .600 team would be zero.

Basketball isn’t that extreme, but it’s more extreme than baseball.  So the chance of a .550 team beating a .600 team should be lower in the NBA than MLB.

But, yeah, maybe I *was* thinking a team that *finished* .600.  If you look at true talent, could it be that the difference between basketball and baseball is so small that you can ignore it?

That’s a good point, and we discussed this somewhere else.

Indeed, your methodology is one that I used in the past to show that the Log5 does NOT work so simply.

Part of the problem with Playoff baseball is that it is different from Regular Season baseball. Player usage is different..

So, you could get a .600 regular season team that is a .550 playoff team and vice versa.

Usage of certain players is much different in the playoffs -

there is no 5th starter.

there is no backups starting games to rest a starter

DH used/not used

Closer pitches much more in playoffs

You only play in certain parks, which helps/hinders certain players

etc.

I think I’d be close to Tango’s numbers, maybe slightly higher.

Phil/#5: I don’t get the logic of your position: “if the .550 team fluked out and won 105 games, while the .600 team won only 90 games, I want the .550 team to win the Series most of the time.” You’re saying that a team that gets very lucky deserves to continue being luck, and vice-versa?  Seems to me that justice argues for just the opposite:  the very lucky team should have their luck run out before it delivers a totally unjustified championship, while the unluckly team deserves some good fortune.

If you are at all interested in reading about 5,000 words on what exactly it means to be the “best” team in a given season, feel free to click my name above. Though I think David Smyth above did it a little more succinctly in comment #20. Hear, hear.

Guy/28: Hi, Guy, haven’t seen you commenting in awhile!

My feeling is that if the team that wins the 162-game season, but a wide margin, doesn’t win the World Series, then what’s the point of the regular season?

If an unlucky team is going to win, it’s much more likely that they win the post-season than the regular season.  So the regular season should be the one that counts the most.

I hear you, Phil, but I think your feeling that “the regular season should count” is in part based on the idea that it’s a far better measure of a team’s true talent.  And that’s usually true, of course.  But not in Tango’s exercise:  we know the teams’ true talent.  And what you’re saying is that the team that got lucky during the season should continue to be rewarded with good luck in post-season, while unlucky teams should continue to suffer.  Philosophically, I guess I’m more of an egalitarian:  I’d like to see the luck even out a bit.  I certainly can’t see any reason to root for a team simply because it’s been lucky.