Article: Big Splash? More Like Big Slash

[old nurse]

 

"Small sample size terrorism" is such a better phrase than "crapshoot." I may start using from here on.

Larger sample size, Atlanta

 

[diehardtwinsfan]

if you look at the larger sample of playoff teams, you'll see far more large salaried teams that small. That's reality.  The same is true with series winners and repeats.  Having money matters. It's not an obstacle that cannot be overcome, but it's silly to think that it can be ignored.

[Mr. Brooks]

If it's just a crapshoot, how do you explain the Twins playoff history in the 2000's?

If it was random, the Twins would have had a mix of outcomes, and would have posted roughly a .500 playoff record over that span.

 

If you roll double sixes every single roll, roll after roll after roll, the dice are not random.

[TheLeviathan]

 

If it's just a crapshoot, how do you explain the Twins playoff history in the 2000's?
If it was random, the Twins would have had a mix of outcomes, and would have posted roughly a .500 playoff record over that span.

If you roll double sixes every single roll, roll after roll after roll, the dice are not random.

 

Did you watch any of those games?  Because is there weren't nearly a dozen times you were flabbergasted by amazing strokes of bad luck, I don't know if we watched the same games.  

 

The other half of those terrible records was a psychological component.

 

But to cut past all this, if you don't think isolated 5 or 7 games samples aren't the definition of small sample size, you just frankly don't believe in small sample size.  And you should.  That's a thing.  Like, an important fact about looking at statistics.  So...you're just plain off base.  (No Punto Puns intended there)

[Mr. Brooks]

Did you watch any of those games? Because is there weren't nearly a dozen times you were flabbergasted by amazing strokes of bad luck, I don't know if we watched the same games.

 

The other half of those terrible records was a psychological component.

 

But to cut past all this, if you don't think isolated 5 or 7 games samples aren't the definition of small sample size, you just frankly don't believe in small sample size. And you should. That's a thing. Like, an important fact about looking at statistics. So...you're just plain off base. (No Punto Puns intended there)

Small sample size is not the same thing as crapshoot.

Calling it a crapshoot implies it is virtually luck, with little or no skill involved.

 

Could you please give examples of all the amazing strokes of bad luck? Aside from the blown call against the Yankees, I don't remember much of that, just poor performance by the Twins.

 

Of course there was a psychological component. That is evidence for my side, not the crapshoot side. Not folding under pressure is a skill, not luck.

 

There is no reason to speak to me like a five year old just because you disagree with me.

[jimmer]

 

Small sample size is not the same thing as crapshoot.
Calling it a crapshoot implies it is virtually luck, with little or no skill involved.

Could you please give examples of all the amazing strokes of bad luck? Aside from the blown call against the Yankees, I don't remember much of that, just poor performance by the Twins.

Of course there was a psychological component. That is evidence for my side, not the crapshoot side. Not folding under pressure is a skill, not luck.

There is no reason to speak to me like a five year old just because you disagree with me.

Does horrible managing fall under bad luck? :-)

[Vanimal46]

Does horrible managing fall under bad luck? :-)

Also, relying on players such as Matt LeCroy, Jose Offerman, Michael Restovich, Jason Tyner... Not enough depth and too many replacement players paired up with the studs.

[TheLeviathan]

 

Small sample size is not the same thing as crapshoot.
Calling it a crapshoot implies it is virtually luck, with little or no skill involved.

Could you please give examples of all the amazing strokes of bad luck? Aside from the blown call against the Yankees, I don't remember much of that, just poor performance by the Twins.

Of course there was a psychological component. That is evidence for my side, not the crapshoot side. Not folding under pressure is a skill, not luck.

There is no reason to speak to me like a five year old just because you disagree with me.

 

Calling it a crapshoot means the outcome is uncertain no matter what you think you see beforehand.  I

 

I could go back and pick out 5 game stretches and argue the Twins deserved to be a 100 win team.  Would that be a valid way of determining their talent or ability? 

 

No?  Then why is it ever?  Small samples are prone to wild swings of luck and demonstrations of play that exceed or slump relative to expectations.  We have playoffs for atmosphere, entertainment, and a variety of other reasons.  It does absolutely nothing for determining the league's best team beyond ceremony.  The 162 game season already pretty much told us who the best team was and that team doesn't win the playoffs with much regularity.

 

So yeah, it's a crapshoot.  

[Mr. Brooks]

Calling it a crapshoot means the outcome is uncertain no matter what you think you see beforehand. I

 

I could go back and pick out 5 game stretches and argue the Twins deserved to be a 100 win team. Would that be a valid way of determining their talent or ability?

 

No? Then why is it ever? Small samples are prone to wild swings of luck and demonstrations of play that exceed or slump relative to expectations. We have playoffs for atmosphere, entertainment, and a variety of other reasons. It does absolutely nothing for determining the league's best team beyond ceremony. The 162 game season already pretty much told us who the best team was and that team doesn't win the playoffs with much regularity.

 

So yeah, it's a crapshoot.

Of course there is more variance in a string of 5 and 7 game series than there is over 162 games.

But there is still plenty of skill involved. Most people believe the Cubs are the best team in baseball and they made the world series, pure coincidence I guess.

 

Would skill matter if the Cubs played a little league team, in a 1 game series? Such a small sample size, I guess we really can't say which team would win.

 

The Twins lost 15 of their last 17 playoff games. The odds of that happening based on pure randomness (coinflip), is 1:963.8

 

So, possible sure, but extremely unlikely. The more likely answer is that skill (or lack of) contributed as much or more than luck.

 

Are you suggesting that results are meaningless if there is ANY luck involved?

What is your threshold for how much luck can be involved before it's a legit outcome and not just small sample size noise?

Because if you a need a large enough sample size to weed out all luck, then the regular is also meaningless.

 

It's still a team of humans pitting skills against another team of humans. The results still mean something.

[TheLeviathan]

 

Of course there is more variance in a string of 5 and 7 game series than there is over 162 games.
But there is still plenty of skill involved. Most people believe the Cubs are the best team in baseball and they made the world series, pure coincidence I guess.

Would skill matter if the Cubs played a little league team, in a 1 game series? Such a small sample size, I guess we really can't say which team would win.

The Twins lost 15 of their last 17 playoff games. The odds of that happening based on pure randomness (coinflip), is 1:963.8

So, possible sure, but extremely unlikely. The more likely answer is that skill (or lack of) contributed as much or more than luck.

Are you suggesting that results are meaningless if there is ANY luck involved?
What is your threshold for how much luck can be involved before it's a legit outcome and not just small sample size noise?
Because if you a need a large enough sample size to weed out all luck, then the regular is also meaningless.

It's still a team of humans pitting skills against another team of humans. The results still mean something.

 

1) Most people believed the 2015 Cardinals were the best team.  What happened to them?  Go back and look at the teams with the best records and see what happened to them in the playoffs.  This is what happens over a small sample size.  The better team routinely gets beaten.  

 

Hell, how many play-in teams in just the last few years have made WS runs?  So, c'mon, that argument is nonsense.

 

2)  Of course the Cubs would beat a little league team.  One of them isn't filled with professional baseball players.  Again.  C'mon.

 

3)  1, 5, or 7 game series after a 162 game season are the definition of a small sample size.  Small sample sizes are not nearly as good a measure of quality as a larger sample.  To deny that small sample size terrorizes better teams in the playoffs is to deny a simple reality about statistics.  You are welcome to do so, but you're wrong if you do.

 

This is, in no way, arguing the Twins were a better or worse team.  That'd have to be a year by year thing.  The argument that the playoffs do anything other than show "this team lucked and skilled their way over a ridiculously small sample to a title" is nonsense.  The playoffs are purely for entertainment.  The regular season always give us the best measurement of who the best team is.  What happens in the playoffs is just crazy *@%$ you hope you come out on the right side of.

 

Can you try and have a roster that gives you better odds of surviving all that crazy *@%$?  Sure.  You'll still fall victim.  No matter how you construct your build.  No matter how much more superior you are to your opponent.  No matter how well you manage.  You will still fall victim.  Often.  More often than you'd like.  That's what small samples do, they lie.  The outcomes frequently lie about who is better.

Edited October 26, 2016 by TheLeviathan

[TheLeviathan]

Just a few relevant Fangraphs articles.  

[Mr. Brooks]

1) Most people believed the 2015 Cardinals were the best team. What happened to them? Go back and look at the teams with the best records and see what happened to them in the playoffs. This is what happens over a small sample size. The better team routinely gets beaten.

 

Hell, how many play-in teams in just the last few years have made WS runs? So, c'mon, that argument is nonsense.

 

2) Of course the Cubs would beat a little league team. One of them isn't filled with professional baseball players. Again. C'mon.

 

3) 1, 5, or 7 game series after a 162 game season are the definition of a small sample size. Small sample sizes are not nearly as good a measure of quality as a larger sample. To deny that small sample size terrorizes better teams in the playoffs is to deny a simple reality about statistics. You are welcome to do so, but you're wrong if you do.

 

This is, in no way, arguing the Twins were a better or worse team. That'd have to be a year by year thing. The argument that the playoffs do anything other than show "this team lucked and skilled their way over a ridiculously small sample to a title" is nonsense. The playoffs are purely for entertainment. The regular season always give us the best measurement of who the best team is. What happens in the playoffs is just crazy *@%$ you hope you come out on the right side of.

 

Can you try and have a roster that gives you better odds of surviving all that crazy *@%$? Sure. You'll still fall victim. No matter how you construct your build. No matter how much more superior you are to your opponent. No matter how well you manage. You will still fall victim. Often. More often than you'd like. That's what small samples do, they lie. The outcomes frequently lie about who is better.

I don't think I ever claimed the best team will always win in the playoffs.

But I also don't think it's pure randomness luck.

I'm sorry. Agree to disagree there.

Most of the time if a team gets beat it's because they left pitches up in the zone, made defensive errors, or some other human error. Rarely is it because they got unlucky.

I think if you consistently build one of the 3 or 4 best teams in baseball, with a good manager, you will eventually cash in a title or two.

And that was my point all along. Nowhere did I ever claim you are guaranteed to win every year just because you have the best roster.

[Mr. Brooks]

I'm also genuinely curious on all of the amazing strokes of bad luck that those Twins teams had. I don't remember any, aside from that bad call in NY (which had no effect on the outcome).

[TheLeviathan]

Leaving pitches up, making errors, or having defensive gaffes are not a direct reflection of talent. Sometimes the bad luck is that otherwise reliable players had that happen and the resulting damage is heightened in the playoffs.

 

Call that whatever you want, but talent doesn't determine that either way. That sort of randomness is why the playoffs are a crapshoot.