What will be the 2017 greatest weakness for the Sox?

[Kimmi]

The problem is that with baseball you can bury so much in randomness.

 

Look at Schilling.

 

Regular season

ERA 3.46

WHIP 1.14

K/BB 4.38

 

Postseason 19 starts

ERA 2.23

WHIP .97

K/BB 4.80

 

He was an excellent pitcher in the regular season but in the postseason he took it to a higher level. Look at those numbers, all against playoff caliber teams, obviously.

 

But what I gather is that the argument is that the excellence of those numbers may be mere randomness - that he could have had a stretch like that in the regular season as well. Well, maybe he could have. But is there really proof that it was all randomness? I don't see how.

 

Baseball is far more random than any of the other major sports, and I think that most people do not realize how much randomness affects the game. In small samples, in short series, and in close games, randomness is king.

 

I suppose there is no definitive proof that Schilling's postseason performance is all randomness. There is no definitive proof that clutch does not exist. Only very, very strong evidence. The correlation between clutch from one situation to the next, be it from game to game, month to month, season to season, postseason to postseason, etc., is nil.

[Kimmi]

Yeah, but the 'performances' are all just randomness in action. The players show up and play and whatever happens happens.

 

LOL This post made me chuckle.

 

Skill is a big part of the performances, but I think you agree that there is a lot of randomness going on in the sport.

[Bellhorn04]

The theory that clutch/non-clutch is strictly attributable to randomness is just that, it's a theory. It may be the correct theory. But it's not conclusive proof, because conclusive proof is simply impossible.

[Bellhorn04]

LOL This post made me chuckle.

 

Skill is a big part of the performances, but I think you agree that there is a lot of randomness going on in the sport.

 

I do agree with that. But the real stat gurus seem to be saying that the skills are all scientifically measurable by statistical profiles, and the clutch/non-clutch doesn't exist, so when you get down to it randomness is the only thing that's left to decide the outcome.

[a700hitter]

The theory that clutch/non-clutch is strictly attributable to randomness is just that, it's a theory. It may be the correct theory. But it's not conclusive proof, because conclusive proof is simply impossible.

in every clubhouse , the players know which of their mates has steel in his spine and who will wilt under the pressure. Just because they are all major leaguers doesn't mean the they all handle the pressure. They don't. As long as clutch exists in the clubhouse it exists and is real. Clutch players don't always put up great numbers, because the game is very hard, but the pitchers know who are the toughest outs and batters know which pitchers never give in in tight spots.

[User Name]

in every clubhouse , the players know which of their mates has steel in his spine and who will wilt under the pressure. Just because they are all major leaguers doesn't mean the they all handle the pressure. They don't. As long as clutch exists in the clubhouse it exists and is real. Clutch players don't always put up great numbers, because the game is very hard, but the pitchers know who are the toughest outs and batters know which pitchers never give in in tight spots.

 

Ask Mariano Rivera about Marco Scutaro. He gave him fits because dude just wouldn't give up an AB. Ever. That's a non-choker if I ever saw one.

[moonslav59]

Ask Mariano Rivera about Marco Scutaro. He gave him fits because dude just wouldn't give up an AB. Ever. That's a non-choker if I ever saw one.

 

Yeah, Scoot never choked up on the bat.

 

I always wondered why,

[cp176]

in every clubhouse , the players know which of their mates has steel in his spine and who will wilt under the pressure. Just because they are all major leaguers doesn't mean the they all handle the pressure. They don't. As long as clutch exists in the clubhouse it exists and is real. Clutch players don't always put up great numbers, because the game is very hard, but the pitchers know who are the toughest outs and batters know which pitchers never give in in tight spots.

 

I'm with you-

It really is all relative to the particular situation. If the best players in the world were competing against minor leaguers day in and day out, I would certainly agree that the likelihood of them choking would be very small and the likelihood of them coming up big in the clutch would be pretty good. Since we are talking about the best competing against the best, there will always be those guys that either get it done or not in very situations. Any player or coach in the dugout knows who the players are that you want on the frontline in the biggest situations.

[Dojji]

LOL This post made me chuckle.

 

Skill is a big part of the performances, but I think you agree that there is a lot of randomness going on in the sport.

 

A lot, yes. A preponderance? That would need to be proven for me to believe it.

 

The same pitcher can show up one day and get lit and show up the next day and mow the opposition down. Does the fact that we don't know which pitcher we'll get in any given start mean the results are random? Hardly. What it actually comes down to is a multiplicity of tiny factors we can't know or account for. Being intellectually lazy and unable to dig further, we dismiss that as "randomness" but it really really isn't.

 

The fact that we need large sample sizes to sort out the noise is exactly because what happens on the baseball field is NOT random, and it takes the X factors a large group of responses to even out and give a real statistical perception of a person's average skill level. If the distribution were "random" we might expect true data in a much smaller dataset.

 

Is it possible that Schilling had a personal X factor that made him far more likely to perform at a high level in a playoff situation? Yes, because we can't know or document all possible x factors affecting a ballplayer we HAVE to concede that possibility.

 

And once you concede that possibility you CANNOT pretend that clutch is impossible, that there aren't x factors that may exist and impact a player in a way to make them more likely to be effective (or on the other side of the same coin, less likely to fail) in critical situations. And of course the counterpoint is also true, and x factors make a player less likely to succeed in those same situations.

 

Baseball is driven by variables, many of which we still haven't identified. That's why the game is still entertaining after all these years. It's lazy to shorthand that into "driven by randomness" because the randomness isn't truly random. All factors of talent and skill are weighted by attitude and psychology and that will create individual variations in either direction in key situations. Ergo, clutch IS a thing, at least potentially.

 

In fact before I doubted the existence of "clutch" I'd seriously doubt the existence of "random" as the concept of truly random numbers is still a source of serious debate in mathematical circles and has been for a very long time now.

Edited February 5, 2017 by Dojji

[notin]

Baseball is far more random than any of the other major sports, and I think that most people do not realize how much randomness affects the game. In small samples, in short series, and in close games, randomness is king.

 

I suppose there is no definitive proof that Schilling's postseason performance is all randomness. There is no definitive proof that clutch does not exist. Only very, very strong evidence. The correlation between clutch from one situation to the next, be it from game to game, month to month, season to season, postseason to postseason, etc., is nil.

 

Actually beyond just the randomness, baseball is about a very limited range that most fans don't seem to think about.

 

A player who hits .280 is considered successful, while a player who hits .240 is considered ineffective and certainly not someone you want up at a key moment.

 

Yet that is a whopping 4% difference in success. Where else do people put the difference between good and bad in such a tight range. The 570 some hitters in MLB will see an overwhelming majority succeed between 20% of the time and 30% of the time. And those outside this tight range won't be outside it by much. No one is going to hit .684 this year.

 

So with this tight range in mind, not to mention the drift we are seeing in the definition of "clutch", it is looking more and more like an illusion.

[Bellhorn04]

Actually beyond just the randomness, baseball is about a very limited range that most fans don't seem to think about.

 

A player who hits .280 is considered successful, while a player who hits .240 is considered ineffective and certainly not someone you want up at a key moment.

 

Yet that is a whopping 4% difference in success. Where else do people put the difference between good and bad in such a tight range. The 570 some hitters in MLB will see an overwhelming majority succeed between 20% of the time and 30% of the time. And those outside this tight range won't be outside it by much. No one is going to hit .684 this year.

 

So with this tight range in mind, not to mention the drift we are seeing in the definition of "clutch", it is looking more and more like an illusion.

 

So does that mean the difference between a good hitter and a bad hitter is an illusion, because the range is so small?

[Bellhorn04]

Also the difference between .280 and .240 is 16.67%.

[Bellhorn04]

And don't the ranges get bigger with pitchers? If we say a good pitcher has a 3.2 ERA and a bad pitcher has a 4.8 ERA, that's a 50% difference. Not small at all.

[Dojji]

So does that mean the difference between a good hitter and a bad hitter is an illusion, because the range is so small?

 

No, but the range is so small that factors that seem like they should be far too small to make a difference wind up manifesting in surprising ways at times.

[moonslav59]

Also the difference between .280 and .240 is 16.67%.

 

If you're at .240 and you want to get to .280, yes, you'll have to improve on what you are doing by 16.7% (40/240), but you could also say .280 is 14% better than .240 (40/280).

 

However, it is clear that hitting .280 is 4% better than hitting .240

 

280-.240=.040 (4%)

 

The difference between 28% and 24% is 4%.

[Bellhorn04]

If you're at .240 and you want to get to .280, yes, you'll have to improve on what you are doing by 16.7% (40/240), but you could also say .280 is 14% better than .240 (40/280).

 

However, it is clear that hitting .280 is 4% better than hitting .240

 

280-.240=.040 (4%)

 

The difference between 28% and 24% is 4%.

 

Faulty math in my opinion. The difference between 5% and 1% is also 4% - out of 100%. But 5% is 5 times as much as 1%.

[Bellhorn04]

In 2016 the Red Sox team batting average was .282. Oakland's team batting average was .246. Small difference?

 

Well, the Red Sox scored 225 more runs, or 1.39 runs more per game, so I'm going to say no.

[notin]

Also the difference between .280 and .240 is 16.67%.

 

Only comparatively, which is immaterial and misleading.

[notin]

In 2016 the Red Sox team batting average was .282. Oakland's team batting average was .246. Small difference?

 

Well, the Red Sox scored 225 more runs, or 1.39 runs more per game, so I'm going to say no.

 

You're talking about 600 plate appearances vs 15,000 to make that difference. And over that stretch it was still less than 1.5 runs per game...

[notin]

Faulty math in my opinion. The difference between 5% and 1% is also 4% - out of 100%. But 5% is 5 times as much as 1%.

 

Not funny math. If you were gambling, the 5% outcome is the smarter bet, but would you put 5 times as much on it?

[Bellhorn04]

You're talking about 600 plate appearances vs 15,000 to make that difference. And over that stretch it was still less than 1.5 runs per game...

 

Are you saying that 1.4 runs a game isn't a lot, when the league average is 4.5 runs?

[Bellhorn04]

Only comparatively, which is immaterial and misleading.

 

Except that's exactly what we're doing, comparing the two players.

[Bellhorn04]

Not funny math. If you were gambling, the 5% outcome is the smarter bet, but would you put 5 times as much on it?

 

Sure, why wouldn't you? Your chances are 5 times as good. It's a wash.

[notin]

Except that's exactly what we're doing, comparing the two players.

 

But you changed it to a relative baseline. That's like saying a player batting .010 is twice as good as a batter hitting .005. The reality both are equally ineffective...

[notin]

Are you saying that 1.4 runs a game isn't a lot, when the league average is 4.5 runs?

 

Are you trying to say the 1.4 rpg is a result solely of batting average? And not the 40 more home runs the Sox hit or the higher team slugging and team walks? We are talking about batting average ranges here...

[notin]

Sure, why wouldn't you? Your chances are 5 times as good. It's a wash.

 

No.

 

At 1% and 5% chances of victory, chances are very high you lose both bets...

[notin]

Also the difference between .280 and .240 is 16.67%.

 

 

Just so you know, when you move the baseline like this, you're ignoring that hitters make outs, which is the overwhelming majority of their at-bats and the entire point....

[moonslav59]

Faulty math in my opinion. The difference between 5% and 1% is also 4% - out of 100%. But 5% is 5 times as much as 1%.

 

Is a batter who hits .050 really 5 times better than one that hit's .010?

 

On the scale of .001 to 1.000 I think the more realistic statement is that .050 is 4% better than .010.

 

He gets a hit 4% more than the other guy over 100 PAs (5 hits to 1).

 

I can see how you get the 5 times better number, but I look at it like this:

 

If player A is 5 times better than player B in the above scenario, is that a good comp to these two hitters?

 

Player C: .200

Player D: .400

 

Player A is 5 times better than B, but D is only 2 times better than player C!

 

That's a joke.

 

Player A is 4% better than B.

 

Player D is 20% better than player C.

 

That's seems more telling than your way.

[Kimmi]

The theory that clutch/non-clutch is strictly attributable to randomness is just that, it's a theory. It may be the correct theory. But it's not conclusive proof, because conclusive proof is simply impossible.

 

Conclusive proof may not be possible, but there is certainly strong evidence against clutch.

 

If the performances were completely random, you'd expect more or less the same results that we have from the actual performances.

[Kimmi]

in every clubhouse , the players know which of their mates has steel in his spine and who will wilt under the pressure. Just because they are all major leaguers doesn't mean the they all handle the pressure. They don't. As long as clutch exists in the clubhouse it exists and is real. Clutch players don't always put up great numbers, because the game is very hard, but the pitchers know who are the toughest outs and batters know which pitchers never give in in tight spots.

 

But those players are the toughest outs or the pitchers who never give in in non pressure situations as well.

 

These players perform well in the clutch because they are simply good players.