AL MVP: Ortiz or A-Rod

[stevew]

In the great DH debate, does anyone have any studies handy, showing how hitters hit while they played DH vs when they fielded in the same season? If for instance, you were able to determine that a hitter was on average less efficient when he DH'd vs when he fielded, you could give Ortiz some "defensive" credit for playing DH.

I just wonder if the stress of fielding makes some hitters more focused at the plate(cause they are "in" the game) vs. the relative ease of DH'ing, where you wait to hit again.

Anyways, I think the anti-Yankee anti-Arod bias skews the whole mess, as Arod should win this one since he arguably played gold glove calibre 3b.

[John Galt]

Originally Posted by gstelmack

So with only a small part of the story and the vagaries of human interpretation, I again ask you: how the heck do statistics tell you that A-Rod is a better hitter than Ortiz, or vice-versa? I'll grant that stats can tell you that A-Rod is a better hitter than Curt Schilling (although I could also argue that Schilling's sample size is too small to be meaningful), but how can you really tell that ANY of the top 10 hitters are better than any of the others? There is too much slop in these stats to be anything more than rough approximations.

And I'll also say again: baseball afficianados (I don't put myself in this category, I'm more of a casual fan) argue the meaning of these stats all the freakin' time. They're not even clear to those who allegedly understand them fully. There's a whole thread here on this forum trying to reach a consensus on which stats to use to tell who is a better batter.

A statistic is a tool, nothing more. You use a hammer to build a house, but it's not even a big part of making that house, and if you had to you could use the butt of a screwdriver to knock in a nail. Statistics have their place, and are better than nothing (you've got to start somewhere), but to think they are gospel and not subject to human whims just because there is solid math behind them is just burying your head in the sand to the realities of this world.

P.S. Since I mention it occasionally through here, I guess I'll have my quick say in general on stats. I know that, given a properly distributed population sample, you can mathematically prove that your distribution is representative of the total population within a certain number of decimal places. That's great, but getting a properly distributed population is so much harder than it sounds. I don't trust many statistics that are thrown at me simply because I'm asked to "trust" that they used a proper sample. Frankly, in most cases, I don't. And as talked about above, this is where many baseball stats fall apart when used to compare two players that are at comparable levels.

You are just raising doubt about the reliability of any particular metric in evaluating performance among similar players. That is a farcry from "statisticis can be massaged, twisted, and turned to say anything the person using them wants to say." Even then, I think you are overstating your case. The point with A-Rod and Ortiz is that A-Rod beats Ortiz in virtually every category not largely dependent on luck or small sample sizes. "Clutch" hitting is not a repeatable skill and derived from very small samples. RBI's have a decent sample size, but are dependent on the actions of other players. It isn't that A-Rod's OBP is better than Ortiz that makes him a better hitter - it is that the best of the various metrics all give A-Rod an advantage. Even then, A-Rod only has to be comparable to Ortiz in this debate because of his massive advantage on defense.

[Blackadar]

Galt:

First, I don't conceed that A-Rod has any better batting stats than Ortiz. Their stats are remarkably similar. A-Rod has some advantages, Ortiz has others. You seem to want to discount the ones that Ortiz has and highlight the ones A-Rod has.

Secondly, it all comes down to what your definition of Most Valuable Player is. Is it the best player? The most valuable player in the league? The most valuable player to his team? A combination of the above? I tend to favor the combination method that's heavily weighted towards the latter, which is a bit more subjective. That's the reason I favored Steve Nash for the MVP in basketball last year and thought Irchio was a joke a few years ago over Giambi.

And no, everyone who votes against A-Rod doesn't have some sort of irrational hate of him. Actually, I think he's a very nice guy and a talented player. I simply don't think he did as much for the Yankees as Ortiz did for the Red Sox and both teams ended up with the same record. Without A-Rod, I see the Yankees still contending for the pennant. Without Ortiz, I see the Red Sox as a .500 club.

[Hammer755]

Originally Posted by Blackadar

Galt:

First, I don't conceed that A-Rod has any better batting stats than Ortiz. Their stats are remarkably similar. A-Rod has some advantages, Ortiz has others. You seem to want to discount the ones that Ortiz has and highlight the ones A-Rod has.

Secondly, it all comes down to what your definition of Most Valuable Player is. Is it the best player? The most valuable player in the league? The most valuable player to his team? A combination of the above? I tend to favor the combination method that's heavily weighted towards the latter, which is a bit more subjective. That's the reason I favored Steve Nash for the MVP in basketball last year and thought Irchio was a joke a few years ago over Giambi.

And no, everyone who votes against A-Rod doesn't have some sort of irrational hate of him. Actually, I think he's a very nice guy and a talented player. I simply don't think he did as much for the Yankees as Ortiz did for the Red Sox and both teams ended up with the same record. Without A-Rod, I see the Yankees still contending for the pennant. Without Ortiz, I see the Red Sox as a .500 club.

I have no problem with the majority of your post because it's your opinion and you state as much. But the highlighted part is simply ludicrous.

[Crapshoot]

Originally Posted by Blackadar

Galt:

First, I don't conceed that A-Rod has any better batting stats than Ortiz. Their stats are remarkably similar. A-Rod has some advantages, Ortiz has others. You seem to want to discount the ones that Ortiz has and highlight the ones A-Rod has.

Secondly, it all comes down to what your definition of Most Valuable Player is. Is it the best player? The most valuable player in the league? The most valuable player to his team? A combination of the above? I tend to favor the combination method that's heavily weighted towards the latter, which is a bit more subjective. That's the reason I favored Steve Nash for the MVP in basketball last year and thought Irchio was a joke a few years ago over Giambi.

And no, everyone who votes against A-Rod doesn't have some sort of irrational hate of him. Actually, I think he's a very nice guy and a talented player. I simply don't think he did as much for the Yankees as Ortiz did for the Red Sox and both teams ended up with the same record. Without A-Rod, I see the Yankees still contending for the pennant. Without Ortiz, I see the Red Sox as a .500 club.

I'll try and make this reasonable Blackie - the problem is that this belief of yours comes out of fiction - the idea that without Ortiz, the Sox are a .500 team. Fun fact- the Sox scored 24 runs more than The Yanks this year. Ortiz had a lower RC number than A-rod did. Roughly speaking, if Ortiz was as good as you say he was, he ought to have somewhere in the region of 85 WS (3 Win Shares translates to a win) - which would mean he had the greatest season in the history of baseball, by some factor. Barry Bonds' 2001 season and Ruth's 1923 season end up around 50-55 WS - for some context. Wagner's 1908 was 59. And WS is a statistic that actually gives credit for performance in a higher leverage situation. You cannot make the .500 arguement with any degree of credibility.

Secondly, I'm not sure what evidence is in Ortiz' favor - any statistics that aren't context dependent (like say, RBI) are statistics where A-rod is better or at worse, equal.

You keep citing this belief that Ortiz was better in the face of all the evidence, when the basis points for this are lacking - show me some proof for this theory. The main problem with this ridiculous notion of "clutch" per se, is that it pretends everything up to that point doesnt matter - which is absurd.

[gstelmack]

Originally Posted by John Galt

You are just raising doubt about the reliability of any particular metric in evaluating performance among similar players. That is a farcry from "statisticis can be massaged, twisted, and turned to say anything the person using them wants to say."

Would it help everyone if I said "Statistics are a valuable tool that are easily misunderstood, so they are easily misused and abused, and so are often misused and abused to make a point"? And by "often" I mean pretty much every time someone throws a statistic at me. I'm always asking questions about stats that are handed to me and rarely getting good answers.

Originally Posted by John Galt

Even then, I think you are overstating your case. The point with A-Rod and Ortiz is that A-Rod beats Ortiz in virtually every category not largely dependent on luck or small sample sizes. "Clutch" hitting is not a repeatable skill and derived from very small samples. RBI's have a decent sample size, but are dependent on the actions of other players. It isn't that A-Rod's OBP is better than Ortiz that makes him a better hitter - it is that the best of the various metrics all give A-Rod an advantage. Even then, A-Rod only has to be comparable to Ortiz in this debate because of his massive advantage on defense.

The "best" of which "metrics"? I've already thrown one out there that shows Ortiz with a CLEAR advantage over A-Rod. Of course, that one also has him 5th in the AL, so it doesn't necessarily bolster his claim to MVP, but it does back up my gut feeling that when I need a run, I want Ortiz coming to the plate to knock it in. But the key here is: who decides which metrics to use? You've cherry-picked your metrics, I've cherry-picked mine.

Let me try again regarding baseball stats: once you go beyond simple stats like "Batting Average", you practically need a PhD in math to understand what the heck the stat means. Talk to fans that "follow" baseball, and I bet most could tell you what "Batting Average" means, many could tell you what "On Base Percentage" means, and only a few could tell you what "OPS" is (of course, I just threw out a stat with no backing at all, so let me say that is a gut feeling, backed up by There are entire BOOKS devoted to the statistical analysis of baseball. You guys argue the merits of these things all the time. What's a "Win Share" for crying out loud?

A problem I have with this entire discussion is that in order to participate in the "who is the MVP?" discussion, I have to have read all of these books and understand the ins-and-outs of a whole bunch of different statistics, some of which have only entered the modern lexicon recently, and generally be a baseball stat geek. So you've shut out what I believe is the vast majority of baseball fans simply by finding more obscure stats to throw into the mix that have more complex definitions. I hope I've given enough info above to show that I at least understand the nuances of batting average and other "basic" stats to prevent someone from saying "oh, he hates this because he doesn't understand it" and get you to see what I'm actually saying: most of these stats you throw around in this dicussion are only meaningful to those who have spent significant time studying what they mean and represent, and are not meaningful to the much larger population who might actually care who the "MVP" is. And those of you who DO study the heck out of this stuff can't even come to an agreement. But I'm supposed to sit here and take it while you say things like "Trust me, win share shows how valuable so-and-so is to his team".

As said MUCH earlier in this thread, if stats were such a crucial part of this discussion and so cut-and-dried, you guys could come up with a "magic formula" along the lines of the NFL Quarterback rating that would tell you who the MVP is, and allow you to keep track throughout the year of the MVP race. You can't do it, and then you get mad when people try to throw "intangibles" into the mix because those can't be proven. But stats are clearly only a part of this discussion.

[Blackadar]

Originally Posted by Crapshoot

I'll try and make this reasonable Blackie - the problem is that this belief of yours comes out of fiction - the idea that without Ortiz, the Sox are a .500 team. Fun fact- the Sox scored 24 runs more than The Yanks this year. Ortiz had a lower RC number than A-rod did. Roughly speaking, if Ortiz was as good as you say he was, he ought to have somewhere in the region of 85 WS (3 Win Shares translates to a win) - which would mean he had the greatest season in the history of baseball, by some factor. Barry Bonds' 2001 season and Ruth's 1923 season end up around 50-55 WS - for some context. Wagner's 1908 was 59. And WS is a statistic that actually gives credit for performance in a higher leverage situation. You cannot make the .500 arguement with any degree of credibility.

Secondly, I'm not sure what evidence is in Ortiz' favor - any statistics that aren't context dependent (like say, RBI) are statistics where A-rod is better or at worse, equal.

You keep citing this belief that Ortiz was better in the face of all the evidence, when the basis points for this are lacking - show me some proof for this theory. The main problem with this ridiculous notion of "clutch" per se, is that it pretends everything up to that point doesnt matter - which is absurd.

If you don't believe in clutch, then you don't believe in it. I'm a firm believer in it - Michael Jordan, Larry Bird, Adam Vinetari, Joe Montana, etc. There are guys who you simply want on your team when the game is on the line. If you want to believe that raw statistics determine wins and losses, so be it.

Ortiz' RBIs made the difference in 20 of the Red Sox wins. I'm not talking about "game winning RBIs", which is an absurd statistic where someone gets credited for a winning RBI in a 10-1 rout. I'm talking about a game where the RBIs from a player were the difference in the final score of the game. I'n talking about stepping up to the plate and getting someone over home plate to add a run to your teams' total. "Context dependent"? The whole game is context dependent.

Given Ortiz' amazing statistics both in "late and close" games AND the difference-making RBIs, there's a pretty strong circumstantial case that Ortiz made the difference in many of the Red Sox wins.

The challenge is that we're playing "what if" scenarios. But given the disruptions in the Red Sox clubhouse this year, their pitching woes and the like, what happens if Ortiz DOESN'T get those huge momentum-shifting hits? What happens if he doesn't give his team the emotional boost it needs with his timely, game-winning RBIs? What happens to the clubhouse if he doesn't have the calming influence he has? It's my belief that the Red Sox would have folded.

Is there any factual way to answer any of these questions? No. There are no statistics that can prove that someone else wouldn't have performed as well as Ortiz in these emotional situations. There's no way to determine what would have happened to the Red Sox clubhouse if Ortiz wasn't there. However, reasonable speculations can be made by the words of the players, managers and press. Reasonable assumptions can be made by comparing his performance to the performances of other players. Reasonable assumptions can be made from what I've seen on the field.


I guess what it boils down to is that I'm a firm believer in "intangibles" and things that can't be quantified by statistics. Much like the statistics that say a "bullpen by committee" is better than having one closer or the statistics that suggest you should almost never punt on 4th down, there are other things that go into winning and losing that can't be statistically analyized. Both of those items are supposed to be statistically correct, yet the teams that try them lose. You know why? Because statistics don't mean a damn thing when there's two men on, two men out and you're down by a run in the bottom of the 9th inning. What you're suggesting is that the hit I get at that point doesn't mean any more than any other hit that year. What I believe is that hit means everything.

[Dekanth]

If this was Ortiz vs. 'anybody not on the Yankees' there would not even be a debate here. ARod brings the most astonishing claims and the most bizarre and mundane and meanlingless stats to the forefront. What last MVP discussion brought up BA with RISP with 2 outs in the 7th, 8th and 9th innings before? And I love the leadership skills arguements...you mean professionals and well paid guys like Manny, Damon, Varitek, Schilling, etc... need Ortiz to pump them up in the lockerroom in order for the Sox to win? Heh, funny stuff.

[John Galt]

Originally Posted by Blackadar

If you don't believe in clutch, then you don't believe in it. I'm a firm believer in it - Michael Jordan, Larry Bird, Adam Vinetari, Joe Montana, etc. There are guys who you simply want on your team when the game is on the line. If you want to believe that raw statistics determine wins and losses, so be it.

There is a big difference between baseball and other sports. If "clutch" were really a baseball skill, then it should be repeatable. Players who are "clutch" should be so from year-to-year or for their whole career. Unfortunately, when you look at baseball, you see clutch leaders being clutch dogs the next. When you look at career totals, you see a pretty normal distribution curve.

I believe in clutch players in football and basketball. I don't, however, believe in clutch hitters for baseball. I may even believe in clutch pitchers for baseball (not based on W-L like the Jack Morris defenders) - that is still an open question. However, for baseball hitters, the evidence has convinced me that "clutch" is really luck * opportunity and is based on small sample sizes.

[Blackadar]

Originally Posted by Dekanth

If this was Ortiz vs. 'anybody not on the Yankees' there would not even be a debate here. ARod brings the most astonishing claims and the most bizarre and mundane and meanlingless stats to the forefront. What last MVP discussion brought up BA with RISP with 2 outs in the 7th, 8th and 9th innings before? And I love the leadership skills arguements...you mean professionals and well paid guys like Manny, Damon, Varitek, Schilling, etc... need Ortiz to pump them up in the lockerroom in order for the Sox to win? Heh, funny stuff.

Manny? A professional? LOL!

[jeff061]

Wellpaid ≠ Professional

[Blackadar]

Originally Posted by John Galt

There is a big difference between baseball and other sports. If "clutch" were really a baseball skill, then it should be repeatable. Players who are "clutch" should be so from year-to-year or for their whole career. Unfortunately, when you look at baseball, you see clutch leaders being clutch dogs the next. When you look at career totals, you see a pretty normal distribution curve.

I believe in clutch players in football and basketball. I don't, however, believe in clutch hitters for baseball. I may even believe in clutch pitchers for baseball (not based on W-L like the Jack Morris defenders) - that is still an open question. However, for baseball hitters, the evidence has convinced me that "clutch" is really luck * opportunity and is based on small sample sizes.

Oh, so you believe in it for other sports but not for baseball? Why would baseball be any different?

[gstelmack]

Originally Posted by John Galt

There is a big difference between baseball and other sports. If "clutch" were really a baseball skill, then it should be repeatable. Players who are "clutch" should be so from year-to-year or for their whole career. Unfortunately, when you look at baseball, you see clutch leaders being clutch dogs the next. When you look at career totals, you see a pretty normal distribution curve.

I believe in clutch players in football and basketball. I don't, however, believe in clutch hitters for baseball. I may even believe in clutch pitchers for baseball (not based on W-L like the Jack Morris defenders) - that is still an open question. However, for baseball hitters, the evidence has convinced me that "clutch" is really luck * opportunity and is based on small sample sizes.

Or maybe "clutch" in baseball takes so much out of you mentally (because clutch is a very mental skill) that it's difficult to be so over a single 162-game season (+ playoffs), let alone over multiple such seasons? Basketball is half that, and football is only 16 games + playoffs (a max of 20 games). Perhaps baseball hitters have so much trouble maintaining "clutch" because it is so difficult to be at that peak level for so long, day-in-and-day-out?

[John Galt]

Originally Posted by Blackadar

Oh, so you believe in it for other sports but not for baseball? Why would baseball be any different?

A couple points:

1) The most important reason - year to year stats show baseball hitting is different. A "clutch" player one year doesn't repeat the skill over the long term.

2) Baseball hitting is an extremely quick, focused event that is repeated many, many times. It is not a long drive down the football field. It isn't even a shoot at the basket (which doesn't involve someone throwing a ball at you). It is about hitting a ball traveling at 90+ MPH in such a way that it avoids the players on the field (or leaves the park). I think it is reasonable to say Joe Montana saved his best for the end of the game. I think it is unreasonable to say a player at the plate really swings much different in a given situation. The swing is practiced over and over and over again. The ability to hit it is a rare one and I don't think any player will be that much better or worse in a given situation (although they will go on hot and cold streaks over longer periods of time).

2) is just my speculation. 1) is the reason I have to speculate. Stats have convinced me that clutch hitters don't exist. Perhaps someone will show that clutch shooters don't exist - I doubt it, but if they do, I will believe it. I doubt anyone will ever show clutch football players don't exist because you will always have small sample sizes.

[John Galt]

Originally Posted by gstelmack

Or maybe "clutch" in baseball takes so much out of you mentally (because clutch is a very mental skill) that it's difficult to be so over a single 162-game season (+ playoffs), let alone over multiple such seasons? Basketball is half that, and football is only 16 games + playoffs (a max of 20 games). Perhaps baseball hitters have so much trouble maintaining "clutch" because it is so difficult to be at that peak level for so long, day-in-and-day-out?

I have a hard time buying the mental drain explanation. When you look at player's "clutch" stats over the long term, you find an amazing thing - they usually match their stats in "non-clutch" situations. Why would someone save their "mental energy" for clutch situations one year, but not be able to do it another year? Was the mental energy drain so great from 1999, that they couldn't manage it for 2000 (yet they were somehow able to maintain it for all of 1999)? Occam's Razor makes me think this explanation doesn't work.

[Blackadar]

Of course John, some studies suggests clutch hitting does exist.

http://www.sciencedaily.com/releases...0506140903.htm

http://www.dolphinsim.com/ratings/notes/clutch.html

[John Galt]

Originally Posted by Blackadar

Of course John, some studies suggests clutch hitting does exist.

http://www.sciencedaily.com/releases...0506140903.htm

Come on - anyone can do a google search. Without access to his data, it is impossible to trust the conclusions. He used a relatively small sample size (only 3 years). There is no evidence he controlled properly, assumed a normal distribution curve, and/or didn't distort the sample with sacrifice hits (if he included bunts especially, that is a major problem). The best studies I've seen (you can read them at Baseball Prospectus, Baseball Primer, or many other sites) show no repeatable clutch skill. If someone proves otherwise, I'll be happy to change my mind.

edit: I posted when you only included the first article. I'm reading the second article which actually seems to contain data and methodology now.

[Blackadar]

Originally Posted by John Galt

Come on - anyone can do a google search. Without access to his data, it is impossible to trust the conclusions. He used a relatively small sample size (only 3 years). There is no evidence he controlled properly, assumed a normal distribution curve, and/or didn't distort the sample with sacrifice hits (if he included bunts especially, that is a major problem). The best studies I've seen (you can read them at Baseball Prospectus, Baseball Primer, or many other sites) show no repeatable clutch skill. If someone proves otherwise, I'll be happy to change my mind.

edit: I posted when you only included the first article. I'm reading the second article which actually seems to contain data and methodology now.

I know you're reading the 2nd article, but the first guy did it for 1,000 players over 18 seasons. That's NOT a small sample size.

[gstelmack]

Originally Posted by John Galt

Come on - anyone can do a google search. Without access to his data, it is impossible to trust the conclusions. He used a relatively small sample size (only 3 years). There is no evidence he controlled properly, assumed a normal distribution curve, and/or didn't distort the sample with sacrifice hits (if he included bunts especially, that is a major problem). The best studies I've seen (you can read them at Baseball Prospectus, Baseball Primer, or many other sites) show no repeatable clutch skill. If someone proves otherwise, I'll be happy to change my mind.

edit: I posted when you only included the first article. I'm reading the second article which actually seems to contain data and methodology now.

Thank-you for making my point. Statistics by themselves are meaningless. It requires research to determine if they are any good. And most people simply do not have the time, patience, energy, or skill to do this for every statistic thrown at them. Which makes it very easy for people to throw statistics around and hope most people just "trust" them.

[John Galt]

Originally Posted by Blackadar

I know you're reading the 2nd article, but the first guy did it for 1,000 players over 18 seasons. That's NOT a small sample size.

Ooops. You are right. For some reason, I read it as 1972-74. My other objections still stand.

I realized I've read the Dolphinsim study before when it was on Baseball Primer. I can't remember the objectins other people had, although I expect they are available somewhere. The problems I have are:

1) In the initial study, he counts all "good" outcomes equally. A single, a walk, and a homer are all the same. It is not a surprise then that he initially shows power hitters "choke" because power is correlated with striking out. Under his system, I bet Ortiz is actually a choke player.

2) He then says he gets similar results when switching to SLG. However, I bet who is a "clutch" player is now totally different (Ozzie Smith won't be a leader). However, he doesn't provide us the info here, so the better of the two studies isn't much help.

3) It's not clear SLG helps things because it ignores BB which was part of his initial study (since unintentional walks are said to be "clutch" under his initial definition).

Basically, by the end of the study, he was offered two opposite definitions of "clutch" (one favoring power hitters and one favoring singles hitters) and never done anything like do it for OPS. That makes the study a little suspect.

I'm also concerned that the data still isn't there. It seems odd to think (as his conclusions show) that the best "clutch" players are actually the worst normal hitters. The so-called "good" hitters are almost always "choke" artists. This could be because the other team uses their best pitchers in these situations (and dolphinsim makes no effort to adjust for pitcher quality), but I'm not sure if that would account for the difference.

Basically, this study has got some weird holes in it and was performed rather haphazardly. It was an interesting read when I read it a while back, but it has major problems.

[John Galt]

Originally Posted by gstelmack

Thank-you for making my point. Statistics by themselves are meaningless. It requires research to determine if they are any good. And most people simply do not have the time, patience, energy, or skill to do this for every statistic thrown at them. Which makes it very easy for people to throw statistics around and hope most people just "trust" them.

Premise: Statistics are hard to understand - they may even require a PhD
Conclusion: Statistics by themselves are meaningless.

The conclusion does not follow from the premise (even if we accept it as true). That things are hard to understand does NOT mean those things are meaningless.

[Blackadar]

Well, I've pointed out two different studies that suggests "clutch" players do exist. I'm sure you can point to two that say they don't. Oh well....I guess we'll see how it all turns out.

However, let me leave you with this tidbit. I've looked back at A-Rod's career...and for the last 4 years, the guy has underperformed vs. his normal statistics in these "clutch" situations.

[Dekanth]

Originally Posted by Blackadar

Manny? A professional? LOL!

Professionals by definition are paid to do their job, and I was pointing out that these guys are well-paid. Knowing that their next contract depends on their production is all the motivation they need. I highly doubt Ortiz (or any supposed team leader) makes all that much of a difference. There are some players who take on a mentor to younger players and show them the right way to do things and the right way to conduct themselves. Ruben Sierra does that for the Yankees now (I know, hard to believe) and I don't think he brings anything to the Yankees.

Jeter is the supposed leader of the Yankees, but how many times have you heard that the Yankees would have won just as many or even more WS with an average shortstop?

[oykib]

I think whether or not Ortiz can repeat his clutch ability of this season is irrelevant. He did it this year. He gets credit for that. But his offensive numbers still don't add up to A-Rod's.

I'll reprint the Win Shares listing. Win Shares take into account how well a player did in 'clutch' situations.

Looking at the 2005 Win Shares (which include the value of situational hitting) in the AL:

Rank	League	First	Last Name	Team	Position	Batting	Pitching	Fielding	ExpWS	WSP	WSAB	Total WS	CWS	
1	AL	A	Rodriguez	NYA	3B	33.3	0.0	3.3	19	0.989	24	37	318	
2	AL	M	Ramirez	BOS	OF	30.9	0.0	2.9	17	1.018	22	34	310	
3	AL	G	Sheffield	NYA	OF	30.5	0.0	2.2	17	0.978	21	33	401	
4	AL	M	Teixeira	TEX	1B	29.1	0.0	3.3	18	0.884	20	32	69	
5	AL	D	Ortiz	BOS	1B	31.4	0.0	0.2	14	1.149	22	31	108	
6	AL	B	Roberts	BAL	2B	23.5	0.0	5.0	16	0.873	17	28	63	
7	AL	T	Hafner	CLE	1B	26.8	0.0	0.0	11	1.190	19	27	56	
8	AL	V	Guerrero	LAA	OF	24.5	0.0	2.3	15	0.893	16	27	222	
9	AL	M	Young	TEX	SS	23.0	0.0	4.5	18	0.753	15	27	90	
10	AL	R	Sexson	SEA	1B	25.4	0.0	1.6	17	0.789	15	27	129	
11	AL	M	Tejada	BAL	SS	20.0	0.0	6.1	18	0.717	13	26	188	
12	AL	D	Jeter	NYA	SS	19.6	0.0	6.0	19	0.672	12	26	245	
13	AL	J	Giambi	NYA	1B	24.5	0.0	0.6	12	1.016	16	25	261	
14	AL	J	Peralta	CLE	SS	17.0	0.0	8.1	15	0.813	14	25	29	
15	AL	H	Matsui	NYA	OF	22.6	0.0	2.6	18	0.711	13	25	73	
16	AL	J	Damon	BOS	OF	19.9	0.0	5.7	17	0.739	13	25	195	
17	AL	G	Sizemore	CLE	OF	18.2	0.0	6.4	19	0.663	12	25	30	
18	AL	J	Lugo	TB	SS	18.6	0.0	5.8	17	0.699	12	24	85	
19	AL	P	Konerko	CHA	1B	21.5	0.0	2.9	17	0.712	12	24	113	
20	AL	I	Suzuki	SEA	OF	19.6	0.0	4.3	19	0.631	11	24	135	
21	AL	M	Buehrle	CHA	SP	-0.1	23.2	0.0	13	0.922	16	23	94	
22	AL	J	Santana	MIN	SP	-0.5	23.1	0.0	12	0.917	15	23	80	
23	AL	J	Mauer	MIN	C	13.7	0.0	9.2	14	0.802	13	23	29	
24	AL	C	Crawford	TB	OF	21.0	0.0	2.2	17	0.672	11	23	63	
25	AL	V	Martinez	CLE	C	16.6	0.0	5.8	16	0.679	11	22	47	
26	AL	C	Figgins	LAA	3B	16.2	0.0	5.5	19	0.586	9	22	50	
27	AL	J	Garland	CHA	SP	0.3	21.1	0.0	12	0.914	14	21	60	
28	AL	M	Ellis	OAK	2B	14.2	0.0	6.8	13	0.809	12	21	53	
29	AL	C	Crisp	CLE	OF	16.3	0.0	4.4	17	0.600	9	21	45	
30	AL	E	Chavez	OAK	3B	16.0	0.0	5.2	18	0.584	9	21	143

As you can see, Win Shares has A-Rod significantly ahead of Ortiz. As a matter of fact, Ramirez is much closer to A-Rod. Why? Because what he contributes in LF (even though no one says he's Roberto Clemente) is valued against the 0 Ortiz gives you on defense.

I just can't understand how some of you can't see that gigantic hole in your argument. Yes, clutch ability and even 'leadership' may have some value. But we know that defense has significant value.

BTW, I also highlighted Hafner and Figgins because they indicate where the next best production was at Ortiz and A-Rod's respective positions.

But if I wanted to make the contrast even more stark, I'd point out that There are three guys in the top 30 that play third and seven that would be capable of it-- assuming that all the shortstops could-- seven of the top thirty. There are thirty of the top thirty that could play DH.

[gstelmack]

Originally Posted by John Galt

Premise: Statistics are hard to understand - they may even require a PhD
Conclusion: Statistics by themselves are meaningless.

The conclusion does not follow from the premise (even if we accept it as true). That things are hard to understand does NOT mean those things are meaningless.

I believe you are mis-wording my premise:

Premise: It requires a PhD or equivalent to determine if a statistic presented to you is valid.
Conclusion 1: The vast majority of the population do not have PhD or equivalent knowledge of math, let alone statistics, and therefore cannot determine if any statistic presented to them is valid.
Conclusion 2: Because the vast majority of people they are presented to cannot determine their validity, statistics are an easy tool to use for convincing others of a point-of-view regardless of their applicability to the situation at hand.
Conclusin 3: Statistics are meaningless to the general population because the good, solid, valid numbers are lost in a sea of misinformation.

This is unlike most scientific theories / presentations where there is at least the concept of "peer review", where we're told we can trust the theory because a preponderence of people who do understand the material have investigated and found it credible. With most "statistics" we are presented with, no such review occurs until long after the announcement. And quite often, the research and methodology used is not available to even those who would know enough to critique it.

Heck, the first paper above you're debunking comes from a known academic institution (University of Pennsylvania), and yet you're saying we can't trust it.

[John Galt]

Originally Posted by oykib

I think whether or not Ortiz can repeat his clutch ability of this season is irrelevant. He did it this year. He gets credit for that. But his offensive numbers still don't add up to A-Rod's.

I'll reprint the Win Shares listing. Win Shares take into account how well a player did in 'clutch' situations.

I did say I give a bonus to Ortiz for his clutch/luck hitting, but I think it is a stat that should not be overvalued in MVP determination because it is mostly a product of luck and opportunity (which aren't really things that I think should be rewarded). But I agree that it still counts for something.

[John Galt]

Originally Posted by gstelmack

I believe you are mis-wording my premise:

Premise: It requires a PhD or equivalent to determine if a statistic presented to you is valid.
Conclusion 1: The vast majority of the population do not have PhD or equivalent knowledge of math, let alone statistics, and therefore cannot determine if any statistic presented to them is valid.
Conclusion 2: Because the vast majority of people they are presented to cannot determine their validity, statistics are an easy tool to use for convincing others of a point-of-view regardless of their applicability to the situation at hand.
Conclusin 3: Statistics are meaningless to the general population because the good, solid, valid numbers are lost in a sea of misinformation.

This is unlike most scientific theories / presentations where there is at least the concept of "peer review", where we're told we can trust the theory because a preponderence of people who do understand the material have investigated and found it credible. With most "statistics" we are presented with, no such review occurs until long after the announcement. And quite often, the research and methodology used is not available to even those who would know enough to critique it.

Heck, the first paper above you're debunking comes from a known academic institution (University of Pennsylvania), and yet you're saying we can't trust it.

Statistics are meaningless is vastly different than statistics are meaningless to general population. The reason they are meaningless to most people is they don't understand them. That doesn't mean they shouldn't be used - it means people should learn to understand them.

As for my dismissal of the first study, it was because it didn't contain data or methodology. Shoddy works comes out of every institution (especially by graduate students) - that is why people need to learn math and stats so that authority of an institution doesn't hold force.

I said

[gstelmack]

Originally Posted by John Galt

Statistics are meaningless is vastly different than statistics are meaningless to general population. The reason they are meaningless to most people is they don't understand them. That doesn't mean they shouldn't be used - it means people should learn to understand them.

I would argue that this is an unrealistic expectation. Statistics require enough knowledge that I would put it above the level of "every driver should know how to change their own oil and how to change a flat" to the equivalent of "every driver should know how to install a new radio in their car" or something similar. It just won't happen to the general population.

Instead, you need something like the peer-review system applied to scientific articles that lets people know that a certain amount of effort went in to having someone else verify the methodology, someone who had the time and knowledge to do so. I mean, I'm someone who has the math background to understand statistics, but since most aren't presented to me with methodology or data backing them up made available, I still have to not trust them. Or in other words, the vast majority of statistics presented to me on a daily basis are presented like that University of Pennsylvania study you so quickly dismissed above.

[rkmsuf]

popsicle headache

[Swaggs]

If Ortiz is so clutch, then shouldn't his clutch numbers raise his overall statistics well beyond those of Rodriguez. To me, by looking at the end of the season numbers and arguing that Ortiz is "that" much more clutch than Rodriguez, you are also arguing that Rodriguez is "that" much better than Ortiz when there are fewer than 2 outs, or before the 9th inning.

By the way, following this debate and tracking stats, I just noticed what an incredible season Manny Ramirez had.

[Crapshoot]

Originally Posted by gstelmack

I believe you are mis-wording my premise:

Premise: It requires a PhD or equivalent to determine if a statistic presented to you is valid.
Conclusion 1: The vast majority of the population do not have PhD or equivalent knowledge of math, let alone statistics, and therefore cannot determine if any statistic presented to them is valid.
Conclusion 2: Because the vast majority of people they are presented to cannot determine their validity, statistics are an easy tool to use for convincing others of a point-of-view regardless of their applicability to the situation at hand.
Conclusin 3: Statistics are meaningless to the general population because the good, solid, valid numbers are lost in a sea of misinformation.

This is unlike most scientific theories / presentations where there is at least the concept of "peer review", where we're told we can trust the theory because a preponderence of people who do understand the material have investigated and found it credible. With most "statistics" we are presented with, no such review occurs until long after the announcement. And quite often, the research and methodology used is not available to even those who would know enough to critique it.

Heck, the first paper above you're debunking comes from a known academic institution (University of Pennsylvania), and yet you're saying we can't trust it.

Let me at least disagree with the first premise. I have a Finance/BA degree and some statistical grounding, and I understand the majority of the statistics out there. Some, like Win Shares, are a fairly long and drawn out procedure, but their underlying logic is perfectly sound. Most sabremetric proponets I know are at least somewhat mathematically inclined, but they are by no means PhD's. It doesnt take a PHd to understand that OPS correlates better with RC and Runs than batting average does. It doesnt take a Phd to understand that fundementally, there is such a thing as a replacement level talent - this is basic statistical grounding, the kind most people get in high school. While I accept the premise that a lot of statistics are getting more complicated, the underlying premises are still fairly simple to grasp for anyone with the intellectual capacity to do so. I simply refuse to dismiss something that's fairly simply because people have an aversion to mathetmatical grounding/

[Warhammer]

Regarding clutch players: I think we can all agree that there are players out there that are not clutch players. Therefore, since I can tell you someone who lacks certain characteristics, then there must be players who either potentially or actually posses those characteristics.

I can tell you from my own experience that there is such a thing as clutch performances. Why do we value 4th Qtr comebacks in football? Why do we talk about players that can execute the 2 minute drill? Because some players do perform better than others in those circumstances!

Taking this back to the MVP discussion, people were talking earlier about runs in the first few innings being just as important as those in the later innings. That is correct, provided you win the game (i.e. getting beat 16-2 and scoring 2 runs in the 9th is pretty meaningless...). However, scoring runs late in games when you are losing is very important, and the fact that some players perform better in such situations is important as well.

When you are losing 1-0 after the top of the first, it is not a big deal. Why? You still have 9 at bats to make up the deficit. Each player basically has 3 or 4 chances to put the ball in play and possibly make something happen. However, in the later innings, each player may only have one opportunity to make something happen. What happens? They begin to press, and they make mistakes. Imagine yourself at work, working under a deadline performing an every day task, if it takes an hour for you to do this task, and you have 8 hours to get it done, chances are you will perform it well. It is something that you have done countless times before, and you can take your time with it. Now, let's say you have one hour to get that one hour job done, some people give themselves undue pressure to get the job done. They begin to make mistakes. The same thing happens in baseball. Its the bottom of the 9th, you are down a run, two runners are on, and there are two outs. Are you thinking about whether the pitcher is setting you up for his slider? Or, are you thinking that you need to put the ball in play, and you have to put it in play, and make contact, and....

My conclusion is that runs scored in the late innings are more important not because they are more important in and of themselves, but rather in when they are scored. This is because you have fewer chances late in the game to score additional runs. Therefore, you need to make the most of your opportunities. Some people fall apart when they have limited opportunities to do something, others perform equally well, and others thrive in those sorts of situations.

[Dekanth]

Originally Posted by Warhammer

Some people fall apart when they have limited opportunities to do something,

You are absolutly correct and they even have a name for these people: Career Minor Leaguers

[DanGarion]

Originally Posted by Warhammer

My conclusion is that runs scored in the late innings are more important not because they are more important in and of themselves, but rather in when they are scored. This is because you have fewer chances late in the game to score additional runs. Therefore, you need to make the most of your opportunities. Some people fall apart when they have limited opportunities to do something, others perform equally well, and others thrive in those sorts of situations.

You have just as much opportunity to score a run in the 1st inning as you do in the 9th.

[rkmsuf]

Originally Posted by dangarion

You have just as much opportunity to score a run in the 1st inning as you do in the 9th.

Not if you bat 9th!

[DanGarion]

Originally Posted by rkmsuf

Not if you bat 9th!

Ok then 3rd and 9th.

[Mr. Wednesday]

Originally Posted by MrBigglesworth

At the beginning of the season, Ortiz wasn't the most feared in his lineup either. Both Manny and Sheffield (or Giambi) are a big part of the success of Ortiz and ARod. You almost have to pitch to them because of who is in the on deck circle.

And yet, toward the end of the season, I remember Ortiz getting walked to face Manny, which was part of my point. I agree that going into the season, Manny was expected to be the guy other teams worried about more, and I think that it says something that the perspective changed by the end of the season, in spite of yet another outstanding season at the plate by Manny.

[Mr. Wednesday]

Originally Posted by gstelmack

So with only a small part of the story and the vagaries of human interpretation, I again ask you: how the heck do statistics tell you that A-Rod is a better hitter than Ortiz, or vice-versa?

There are two issues here:
1. Since we're discussing the MVP, the issue isn't so much who the better player is, as who played better last year. We have 100% confidence in the actual results we observed in the field, so the issues you raised with error bars and confidence and the like don't apply, IMO.

2. The question you raise seems to be, basically, how much confidence can we have in the descriptive value of baseball statistics as an indicator of the "true" ability of a player? I can't answer that directly, but I can posit that if we have statistics that have been found to have good predictive value for future performance, and we find that one player (let's call him A-Rod, for the sake of argument) consistently outperforms another (say, Ortiz) in those statistics, then even on a slim margin we might have some confidence in saying that A-Rod is a better player than Ortiz.

Having made that argument, I think the margin is slim enough that, given that Ortiz seems to be just peaking, we really can't be sure which player is better going forward, but I don't think you need a margin as extreme as A-Rod vs. Schilling; I'd be comfortable saying that A-Rod is a better player than anybody on the Sox not named Manny or Papi.

[Mr. Wednesday]

Originally Posted by oykib

The point is that whatever the amount of value that A-Rod's defense has over replacement is irrelavent. When comparing him to Ortiz that's not the metric that you use.

Bullshit. A-Rod's defense doesn't occur in a vacuum -- if he wasn't playing, there would be a replacement-level player there (actually, probably better than that considering it's the Yankees, but I'll pass over that). It's theoretically possible to quantify how many runs A-Rod has saved (or extra runs he has allowed) over a replacement-level third baseman, and this is all a positive (or a negative) in the ledger for him compared to Ortiz who essentially hasn't played in the field at all.

Ortiz gives you nothing in the field. There's no way you can massage the statistics to say that Ortiz was enough of a better hitter than A-Rod to make up for 160 well-played games towards the difficult end of the defensive spectrum compared to ten games plaed poorly at the easy end of it.

I don't know whether that's true or not. I'm not suggesting "massaging" statistics, either... I'm suggesting that taking context into account may give Ortiz a significant boost in terms of offense vs. A-Rod (compared to context-neutral metrics which assume, incorrectly in this case, that production averages out to be context-neutral), and I'm saying that I don't know if that is sufficient to overcome the defensive difference.

Ortiz does not even give you replacement value defensively.

I think he's played few enough games in the field that this is negligible.

Looking at the 2005 Win Shares (which include the value of situational hitting) in the AL:

I'll defer argument on this basis until I've seen whether there are any rebuttals from people who are knowledgeable enough to rebut. I know enough to set up the parameters of the argument, but I don't know enough or have enough information to carry it through myself. In particular, I'd feel more comfortable working from something like UZR as far as fielding contribution is concerned. Provisionally, I do think that this is a very strong argument in favor of A-Rod.

[Mr. Wednesday]

Originally Posted by John Galt

"Clutch" hitting is not a repeatable skill and derived from very small samples.

While that's obviously an issue in terms of choosing which player is likely to perform better going forward, I disagree with the implied assertion that it has no value assessing what happened during the past season.

[Mr. Wednesday]

Originally Posted by Blackadar

Much like the statistics that say a "bullpen by committee" is better than having one closer...

That's a misstatement of James's argument, which is to say that it's turned into a popular straw man.

[John Galt]

Originally Posted by Mr. Wednesday

While that's obviously an issue in terms of choosing which player is likely to perform better going forward, I disagree with the implied assertion that it has no value assessing what happened during the past season.

I did not say it had no value. I said it had less value because it was largely a product of luck and opportunity. I think some bonus should be afforded to doing well in the right place at the right time.

[John Galt]

Originally Posted by Warhammer

Regarding clutch players: I think we can all agree that there are players out there that are not clutch players. Therefore, since I can tell you someone who lacks certain characteristics, then there must be players who either potentially or actually posses those characteristics.
...
Some people fall apart when they have limited opportunities to do something, others perform equally well, and others thrive in those sorts of situations.

It's as though you didn't read anyone's posts. Your conclusions are the same as your arguments. No one has agreed that some players are anti-clutch at the Major League level (usually the people that can't handle the pressure don't make it to the Majors or last very long). Even if you take that as the case, it does not follow that there are also clutch hitters (both don't have to exist). And you offer nothing to rebut statistical evidence that players don't maintain clutch ability over the long term (which shows it is more a product of small sample sizes in a given season than an actual skill).

[Warhammer]

Originally Posted by dangarion

You have just as much opportunity to score a run in the 1st inning as you do in the 9th.

No argument, I was not talking about your opportunities to score in a SINGLE inning but in the rest of the game. So, in the first inning, you have say 3 or 4 chances to score (based upon 3 or 4 ABs for the game), but once you get into the late innings you only have one, possibly two, ABs. That means fewer chances.

If you are given a 20 pitches to hit the ball out of the park (and you are told this, controlled setting, etc.) the first 15 or so pitches you are going to be loose and taking your swings. However, after that 15th pitch, some people are going to start thinking that they only have 5 more chances and they need to make those chances count.

Just like in basketball, when you are down 7 points with 8 minutes left to play, you have some time to make up the points. But, when you're down 7 points with 2 minutes left, it's a different ball game.

[Warhammer]

Originally Posted by John Galt

Your conclusions are the same as your arguments. No one has agreed that some players are anti-clutch at the Major League level (usually the people that can't handle the pressure don't make it to the Majors or last very long).

Go back and read the first page of the thread, it looks as though you are the one who is not reading posts. The first page is all about whether or not there are clutch players, clutch ability, clutch hitting, etc.

Originally Posted by John Galt

Even if you take that as the case, it does not follow that there are also clutch hitters (both don't have to exist).

I will disagree with that. If there are players who are not "clutch" players, then there are those who perform better in the "clutch". Those players become the "clutch" players for that era. "Clutch" is a comparative description, there is no single item that says if a player bats .395 with players in scoring position that he is clutch, because if he is batting .450 for the year he actually suffers in that situation. Conversely, someone who bats .290 for the year, but bats .395 with runners in scoring position he is a better "clutch" hitter, because he performs better in those situations than normal.

Since it is a comparative term, you will always have "clutch" players and "chokers".

Originally Posted by John Galt

And you offer nothing to rebut statistical evidence that players don't maintain clutch ability over the long term (which shows it is more a product of small sample sizes in a given season than an actual skill).

OK, let's look at a 3 season span for Ortiz:

http://sports.espn.go.com/mlb/player...&type=batting3

For those three years, his batting average is .297. However, with runners in scoring position he is batting .328 over that time frame with 461 ABs. Let's look at the close and late situation, he's batting .326 with 221 ABs. Over this 3 year time frame, those 461ABs is equal to 28% of his total ABs. Those 221 ABs are about 14% of his ABs, both those numbers are a good sample size.

Looking at his 1631 ABs and 119 HRs, he hits a HR in about 7% of his ABs. In the close and late situational he hits a HR in 10% of his ABs! That is almost doubling his HR rate! Additionally, his SLG % is 174 points higher in the C&L situation, but 45 points lower when runners are in scoring position.

Given the overall look at this, he is considerably better in these situations than he is normally. Therefore, he is better in the clutch.

Look at A-Rod's numbers this year and over the last three years in these situations, Ortiz's numbers are better across the board. A-Rod's numbers are lower than what you would expect from a normal situation which backs up the rep he had in Texas of padding his numbers in meaningless situations and games.

A-Rod's stats
http://sports.espn.go.com/mlb/player...&type=batting3

[Warhammer]

Originally Posted by Warhammer

No argument, I was not talking about your opportunities to score in a SINGLE inning but in the rest of the game. So, in the first inning, you have say 3 or 4 chances to score (based upon 3 or 4 ABs for the game), but once you get into the late innings you only have one, possibly two, ABs. That means fewer chances.

If you are given a 20 pitches to hit the ball out of the park (and you are told this, controlled setting, etc.) the first 15 or so pitches you are going to be loose and taking your swings. However, after that 15th pitch, some people are going to start thinking that they only have 5 more chances and they need to make those chances count.

Just like in basketball, when you are down 7 points with 8 minutes left to play, you have some time to make up the points. But, when you're down 7 points with 2 minutes left, it's a different ball game.

Pursuant to this, you could argue that unless you score more runs than the other team, all of those runs scored in a game are meaningless. In baseball a win is a win, and a loss is a loss. Unlike football or hockey where point differential and points scored are tie breakers at the end of the season.

Therefore, it is arguable that the most important run in a game is the one that puts the other team in the lead. Since an early lead gives the other team more chances to score runs later in the game, you could argue that an early lead is not that important, but becomes more important as the game goes on (if you are down 1-0 prior to your first at bat, your team still has a minimum of 27 chances to make up that run, whereas in the 7th inning, you only have another 9 chances to make up that same run). Therefore, having a player that has higher batting statistics in clutch situations is preferable to those that don't.

[John Galt]

Originally Posted by Warhammer

Go back and read the first page of the thread, it looks as though you are the one who is not reading posts. The first page is all about whether or not there are clutch players, clutch ability, clutch hitting, etc.



I will disagree with that. If there are players who are not "clutch" players, then there are those who perform better in the "clutch". Those players become the "clutch" players for that era. "Clutch" is a comparative description, there is no single item that says if a player bats .395 with players in scoring position that he is clutch, because if he is batting .450 for the year he actually suffers in that situation. Conversely, someone who bats .290 for the year, but bats .395 with runners in scoring position he is a better "clutch" hitter, because he performs better in those situations than normal.

Since it is a comparative term, you will always have "clutch" players and "chokers".



OK, let's look at a 3 season span for Ortiz:

http://sports.espn.go.com/mlb/player...&type=batting3

For those three years, his batting average is .297. However, with runners in scoring position he is batting .328 over that time frame with 461 ABs. Let's look at the close and late situation, he's batting .326 with 221 ABs. Over this 3 year time frame, those 461ABs is equal to 28% of his total ABs. Those 221 ABs are about 14% of his ABs, both those numbers are a good sample size.

Looking at his 1631 ABs and 119 HRs, he hits a HR in about 7% of his ABs. In the close and late situational he hits a HR in 10% of his ABs! That is almost doubling his HR rate! Additionally, his SLG % is 174 points higher in the C&L situation, but 45 points lower when runners are in scoring position.

Given the overall look at this, he is considerably better in these situations than he is normally. Therefore, he is better in the clutch.

Look at A-Rod's numbers this year and over the last three years in these situations, Ortiz's numbers are better across the board. A-Rod's numbers are lower than what you would expect from a normal situation which backs up the rep he had in Texas of padding his numbers in meaningless situations and games.

A-Rod's stats
http://sports.espn.go.com/mlb/player...&type=batting3

Let me try this last time.

1) No one has agreed that there are non-clutch players. That is your assertion without evidence.

2) Even if there are non-clutch players, it does not follow that there are clutch players. If 2% of the league is non-clutch, that does not mean any number will be clutch. If you say everyone else (98%) is clutch, that means the word has no importance. Your latest attempt to say there are always "chokers" and "clutch" players means the words have no meaning.

3) The grand total sample size for Ortiz over those 3 years is 221 AB. That is not a sufficient sample size. His OBP and OPS are higher in close and late situations, BUT not so much so that a slump in those situations over the next 200 AB won't even it out. 221 AB from ONE player is not sufficient to derive a general statistical rule. It proves NOTHING.

4) 7% to 10% is not "almost doubling."

5) If Ortiz (like almost every player in history) fails to be "clutch" over the next few years, does that mean he forgot how to be "clutch?" Or is it just a regressive to the mean?

[jeff061]

Or what if his numbers persist or get even better. Will we need to give him more time?

[John Galt]

Originally Posted by jeff061

Or what if his numbers persist or get even better. Will we need to give him more time?

I will wager $1000 with you, Warhammer, or anyone else that Ortiz will not perform in close and late situations at the rate above his average stats over the next 3 years that he has over the past 3 years.

[Warhammer]

Originally Posted by John Galt

Let me try this last time.

1) No one has agreed that there are non-clutch players. That is your assertion without evidence.

Go back and read the first page. Seriously, people are (were) arguing that there is no such thing as a "clutch" player.

My assertion is that there are clutch players and chokers. A clutch player plays above his head, performs better than another player in a "clutch" situation. A choker performs far worse than expected in a "clutch" situation. The choker sets the bar, than the other players fill in. So let's say that the average baseball player has a BA 20 points below his normal BA with RISP, the choker would have say a BA 40 points below his normal BA, and in this example, the clutch player would perform around his normal BA.

In another era, a choker might perform only 20 points below his BA, the run of the mill player performs at his normal BA, and the clutch player would perform above his BA. Again, it is compared to how other players are performing during the era.

Originally Posted by John Galt

2) Even if there are non-clutch players, it does not follow that there are clutch players. If 2% of the league is non-clutch, that does not mean any number will be clutch. If you say everyone else (98%) is clutch, that means the word has no importance. Your latest attempt to say there are always "chokers" and "clutch" players means the words have no meaning.

Well, they could all be chokers. However, given a certain population, I am confident that for all practical purposes the players will separate themselves into a group that performs below expectations, one that meets expectations, and one that exceeds expectations in a given situation. Say, 2% (using your number) would be chokers, performing below their normal performance, say 50-75% would perform at the same level, while the rest would perform above average in those situations, those would be your clutch players. In this case, the definition of a clutch player might shift slightly to a player who are a standard deviation or two above the player population for a given situation.

Or, let's use a different example, let's say that in a RISP situation, 95% of the players perform below their average. That means that the other 5% would appear to be clutch players BY COMPARISON!

Originally Posted by John Galt

3) The grand total sample size for Ortiz over those 3 years is 221 AB. That is not a sufficient sample size. His OBP and OPS are higher in close and late situations, BUT not so much so that a slump in those situations over the next 200 AB won't even it out. 221 AB from ONE player is not sufficient to derive a general statistical rule. It proves NOTHING.

Tell that to any manufacturing company in the US. In my industry, the typical sample size for quality assurance is 2%, and we produce far fewer identical parts over a similar time frame. Look at polling numbers, we generate national polling numbers based upon a fraction of 1% of the total population, yet those are accurate to +/-3%.

Also, those 221 ABs are over a 3 YEAR PERIOD! Your earlier assertion was that players do not keep it up over several seasons. Yet, he has amassed roughly 35% of a season worth of numbers in a situation, but this is not significant? How much do you consider a significant sample? A 100% sample? Of course the larger your sample size the more accurate it is going to be, but there is also the law of diminishing returns here as well.

Originally Posted by John Galt

4) 7% to 10% is not "almost doubling.".

Sorry, meant to say a 150% increase in production in the rate, that is a significant increase! You got me there!

Originally Posted by John Galt

5) If Ortiz (like almost every player in history) fails to be "clutch" over the next few years, does that mean he forgot how to be "clutch?" Or is it just a regressive to the mean?

It could be a number of factors. He could be pressing more to get that last ring. He could be trying to recature the glory days of yore. It could be a number of factors, but just as players skills fade, some players ability to come through in the clutch fade.