PING: Baseball Guys (Re: The Bill James/Sabermetrics Revolution)

[albionmoonlight]

This one could go in several directions. Take your pick:

Are some of the new stats being recorded and followed so complex as to be pretty much meaningless? Are they there more to show that the guy computing the stats took some calculus classes than to help us understand baseball?

Is it borderline negligent now for a manager/GM to "go by his gut" rather than the numbers when making roster decisions?

Does Bill James as an individual get too much/too little credit for the "revolution" in a game that was always more about numbers than other sports?

For those who have read Blink--isn't there something about taking a guy who has been a successful manager in baseball for decades and TRUSTING HIM to trust his gut? This is a complex game played by human beings. And sometimes human beings are chaotic creatures who cannot be properly represented by WHIPs and ERA+s and the like.

For those that don't trust things like ERA+ and WHIP, but do trust things like BA and ERA--is it just tradition that you like, or is there something more to your objections?

Is sabermetrics good for baseball? Bad for baseball? Make it more fun? Less fun?

[oykib]

People who don't like them just don't seem to understand Sabermetrics.

Basically, we all make composite stats in our heads. None of us takes a single rate or counting stat and rates players based on that.

What VORP, RC/27, EQA, Win Shares or whatever do is standardize it. Some people don't like the revolution because they are not into clearly justifying their analyses.

Gut feelings are bullshit. People who say they have gut feelings are really making quantitative or qualitative analyses based on their experiences.

[Ajaxab]

Interesting question. I'm actually doing a project that, in part, involves looking into the history of baseball statistics. The following is from John Thorn and Pete Palmer's "The Hidden Game of Baseball: A Revolutionary Approach to Baseball and Its Statistics" (1985)

"So why do we need new stats? Don't we have enough ways to measure performance? Don't we have too many? Why subject every incident on the field to such maniacal ledger-book accounting?

How can baseball's beauty fail to wither under the glare of intense mathematical scrutiny? For those of an antistatistical bent, baseball, like a butterfly, is poetry in motion and a cold, dead thing when pinned to the page. If we subject the game to ever more intricate analysis, in hope that it will yield up its mysteries, are we not breaking the butterfly upon a wheel, in Pope's phrase?

For the statistician, too, baseball is indeed like a butterfly, whose grace can be glimpsed while it is in flight...but then it is gone, having scarcely registered upon memory. One doesn't truly know any longer what it looked like, where it came from, how it vanished in an instant. The butterfly's coloring, its detail, cannot be absorbed while it is in flight; it must be examined to appreciate its complexity. One may love its simplicity in flight as one may love the simplicity of baseball while standing in the outfield or sitting in the grandstand. But the complex texture of the game, which for many is its real delight--the thing that pleases the mind as well as the eye--cannot be fully grasped while the game is in progress.

And that's what statistical analysis allows us to do. Statistics are not the instruments of vivisection, taking the life out of a thing in order to examine it; rather, statistics are themselves the vital part of baseball, the only tangible and imperishable remains of games played yesterday or a hundred years ago.

Baseball may be loved without statistics, but it cannot be understood without them."

So there you have an obviously biased opinion from a couple of statisticians--an opinion more than twenty years old long before some of the increasingly obscure numbers have appeared.

[John Galt]

Gut feelings are bullshit. People who say they have gut feelings are really making quantitative or qualitative analyses based on their experiences.

I think that is overstating things a bit. "Gut feelings" are gap-fillers, not bullshit. Stats, at their current state, only show so much. They sometimes struggle to assess interactions between events. They struggle when sample sizes are small. And they only offer a part of the picture when it comes to predicting the future.

I think appealling to a "gut feeling" in the face of clear and convincing statistical evidence is problematic. However, it is important for any stat user to recognize limitations and be prepared to fill gaps with non-conventional techniques.

As to how this relates to Blink, I think the best source to turn to is Sports Guy's Curious Guy with Gladwell. They talk about how a lay person could do a better job than Isiah Thomas as GM of the Knicks. The reason for this is because a lay person would just use the wisdom of crowds and consensus to avoid the idiotic mistakes of Thomas. The problem there is not that Thomas is using stats and ignoring his guts - the problem is that he is a moron, but believes himself to be smarter than the crowds. I think Blink as a concept is actually indifferent to a "gut" v. stats choice. While stats don't allow us to form the split-second decision, our previously existing knowledge can be heavily informed by stats. That backdrop changes our decisions and allows intuition to serve its function. However, that doesn't mean we are strictly following some primordial gut. Rather, IMO, it means our instincts have been shaped over time by whatever knowledge we have. If you fill that void with stats, that changes your approach. If you leave that void empty, you are Isiah Thomas.

[SackAttack]

People who don't like them just don't seem to understand Sabermetrics.

Basically, we all make composite stats in our heads. None of us takes a single rate or counting stat and rates players based on that.

What VORP, RC/27, EQA, Win Shares or whatever do is standardize it. Some people don't like the revolution because they are not into clearly justifying their analyses.

Gut feelings are bullshit. People who say they have gut feelings are really making quantitative or qualitative analyses based on their experiences.

My issue with VORP and Win Shares, in particular, is that they're overly complex formulas for which there isn't a lot of documentation. So what happens is if we expect GMs to use those as qualitative analysis tools, we're essentially asking them to use it as a tool for player analysis without understanding what makes it such a valuable number.

I'd compare it to driving a car without knowing how it operates, but I think it's really closer to rebuilding the engine without knowing how the engine works.

[gstelmack]

I've said this in other threads, but pretty much I agree with Sack. It's impossible to have a baseball discussion these days if you aren't keeping up with every statistical treatise that has been written on the game. I just started getting used to what OPS was a few months back, and now all of a sudden I see the term "VORP" getting thrown around. What the heck is a "VORP"?

I know what batting average is and can easily compare two guys. I know how much better a ".300" is compared to a ".200". OPS is okay, but still harder to tell what exactly an ".800" means. I can use it to compare two guys, but if I told you "player X has an OPS of .400, while player Y has an OPS of .100", you have to know OPS well enough to realise that player X still sucks. It's like farenheit vs. celsius: you have to be used to the scales to know that 50 is cold in one and darn hot in the other, as opposed to using a scale where "0" is really cold and "100" is really hot (well, that's farenheit, but I digress).

Win Shares, VORP, and some of these others are the same way. They are getting so complex that it takes more than 60 seconds to tell someone what a good number is and what a bad number is. ERA, BA, OBP, even WHIP are fairly straightforward with obvious good/bad ranges. None of these others really are. And with the rate new ones are coming out, how the heck are you supposed to have a conversation about how much of a choker A-Rod is anymore?

[st.cronin]

I like to watch baseball.

[TroyF]

My issue with VORP and Win Shares, in particular, is that they're overly complex formulas for which there isn't a lot of documentation. So what happens is if we expect GMs to use those as qualitative analysis tools, we're essentially asking them to use it as a tool for player analysis without understanding what makes it such a valuable number.

I'd compare it to driving a car without knowing how it operates, but I think it's really closer to rebuilding the engine without knowing how the engine works.


I would expect a GM making what they make to understand sabermatics and the new concepts. This includes both the strengths and flaws of those concepts.

I don't think they need to base every decision off of it, but I do expect them to learn Sabermatics, understand the basic principles of them, and make it a part of their continuing education.

[SackAttack]

Troy, the problem isn't the GM's. The problem is Bill James has been very secretive about some of his more recent stats and their formulations.

I expect them to know, but how can you use a stat you don't understand because the information isn't there?

Anybody can calculate batting average and OBP. VORP and Win Shares? You can approximate them, kind of, but nobody knows but Bill James the exact calculation, and so it's tough to tell how widely results vary with a DIY effort.

[Crapshoot]

Troy, the problem isn't the GM's. The problem is Bill James has been very secretive about some of his more recent stats and their formulations.

I expect them to know, but how can you use a stat you don't understand because the information isn't there?

Anybody can calculate batting average and OBP. VORP and Win Shares? You can approximate them, kind of, but nobody knows but Bill James the exact calculation, and so it's tough to tell how widely results vary with a DIY effort.

Actually, he laid out the entire Win Shares methodology, as well as a short-form WS. The Hardball Times has adapted his win share formula to account for "loss shares", and Tango Tiger has done some good work wih linear weights, which aren't exactly evil.

[Huckleberry]

I understand both sides of the issue, and the complaint about the new stats I agree most with is gstelmack's. But I believe that the sabermetric stats will eventually stabilize and be standardized. I'm not sure how many years it took for BA, HR, RBI, W, L, and ERA to be the accepted stats but it certainly wasn't overnight.

Since my kids were born I haven't had as much time to do the baseball reading I used to do, so the current calculations for VORP, etc. are foreign to me. But I still appreciate the work being done. Baseball is unique in American sports because of its sequential nature. It is by far the sport best suited for statistical and individual analysis. I hope it keeps progressing.

Even if we're all hopped up on win shares and VORP someday, the fears that some traditionalists have of the romance being gone are folly to me. Scouting will still be needed to tell me who will be able to achieve those numbers. And the play of the game itself will still be beautiful.

As for gstelmack's concern specifically, that's why I tend to look at adjusted OPS+, adjusted ERA+, and time played. Those are quick easy to compare numbers. Then you add in their position, baserunning, etc., but for the sabermetric numbers if you don't have time to study them all as I don't have time, you can go straight to those.

Then you make time to watch games and get deeper into their numbers for better evaluation.

[st.cronin]

The problem is that when you say something like "player A has 12.6 win shares, and player B has only 11.7, therefore player A is better than player B." No, no, no - these are ALL just approximations, with built-in biases and blind spots. It's no better than saying that since Ted Williams had a higher batting average than Joe Dimaggio, that he's a better player than Joe Dimaggio. We've just made them the numbers more complicated, is all.

[oykib]

The problem is that when you say something like "player A has 12.6 win shares, and player B has only 11.7, therefore player A is better than player B." No, no, no - these are ALL just approximations, with built-in biases and blind spots. It's no better than saying that since Ted Williams had a higher batting average than Joe Dimaggio, that he's a better player than Joe Dimaggio. We've just made them the numbers more complicated, is all.

Well, it's much better than formulating a theory based on Batting Average. Batting Average is one of the poorest but most enduring methods to evaluate hitters. OBP is much better.

The reason: OBP correlates very highly with winning. Branch Rickey knew this and pursued players that were able to get on base by hook or crook. AVG is only weakly associated with winning.

You can't compare Williams and Dimaggio by just using AVG. We have to know who the better fielder was (till very recently mostly subjective info-- ZR, defensive WS, range factor), who the healthier player was (G, GS), which player was more powerful (HR, SLG), who was more efficient (OBP, SB%), faster (SB, 3B, GDP)...

[Anthony]

i know some of the shit that Bill James comes up with is seriously for math geeks. i remember my reaction to one of the new stats he invented (forgot which one) was "huh??".

[Katon]

Well, it's much better than formulating a theory based on Batting Average. Batting Average is one of the poorest but most enduring methods to evaluate hitters. OBP is much better.

The reason: OBP correlates very highly with winning. Branch Rickey knew this and pursued players that were able to get on base by hook or crook. AVG is only weakly associated with winning.

You can't compare Williams and Dimaggio by just using AVG. We have to know who the better fielder was (till very recently mostly subjective info-- ZR, defensive WS, range factor), who the healthier player was (G, GS), which player was more powerful (HR, SLG), who was more efficient (OBP, SB%), faster (SB, 3B, GDP)...

True, but it's still important not to overrate the significance of small statistical differences. James himself is on record as saying that any difference less than 3 win shares should be regarded with real skepticism, so a difference of .9 win shares doesn't say much. If I have to pick I'll take Player A, of course, and be much more confident than if I were choosing off a similar difference in their batting averages, but I'd need a good deal more information before I could make an informed choice.

[oykib]

True, but it's still important not to overrate the significance of small statistical differences. James himself is on record as saying that any difference less than 3 win shares should be regarded with real skepticism, so a difference of .9 win shares doesn't say much. If I have to pick I'll take Player A, of course, and be much more confident than if I were choosing off a similar difference in their batting averages, but I'd need a good deal more information before I could make an informed choice.

I didn't say it would be a good choice-- just a better one.

I agree with you.

[TroyF]

Any stat geek who thinks VORP, Win Shares and the like are the end all of scouting or judgement should never have decision making power.

Any "gut reaction" guy who basis his judgements on his eyes instead of with some number also shouldn't have power. (note here: if I have to choose between moron A and moron B, give me moron A EVERYTIME)

There needs to be a mix of the two. I'm not going to use win shares to determine who is better between two players. I'm taking all of the factors in and will give an opinion based on that. Yet, if I see one guy has a dominant lead in Win Shares, I have to seriously question what I'm thinking if I end up favoring the other player.

I have to look at the statistical evidence before me.

Sack,

James has given full explanations of his win shares theories. I'd be stunned if I found out a Major League GM who wanted the calculation info couldn't get it.

[lynchjm24]

My issue with VORP and Win Shares, in particular, is that they're overly complex formulas for which there isn't a lot of documentation. So what happens is if we expect GMs to use those as qualitative analysis tools, we're essentially asking them to use it as a tool for player analysis without understanding what makes it such a valuable number.

I'd compare it to driving a car without knowing how it operates, but I think it's really closer to rebuilding the engine without knowing how the engine works.

The problem with Win Shares is that it is stupid. VORP is a legitimate number that gives you a rough idea of a player's offensive value as compared to his position.

[lynchjm24]

Troy, the problem isn't the GM's. The problem is Bill James has been very secretive about some of his more recent stats and their formulations.

I expect them to know, but how can you use a stat you don't understand because the information isn't there?

Anybody can calculate batting average and OBP. VORP and Win Shares? You can approximate them, kind of, but nobody knows but Bill James the exact calculation, and so it's tough to tell how widely results vary with a DIY effort.

I think you might need to do a little more research on this. Bill James has nothing to do with VORP. Win Shares he wrote an entire book on how to calculate it, and it's reproduced at the Hardball Times and updated every day I believe.

I don't see what the big deal is anyway. I'm pretty versed in this stuff and most serious analysts don't take Win Shares very seriously. There are no major league GMs who are using Win Shares as a basis for this decision making, I have little doubt about that.

As for the VORP forumla, it's hardly black magic:

"Therefore, to calculate VORP one must multiply the league's average runs per out by the player's total outs; this provides the number of runs an average player would have produced given that certain number of outs to work with. Now multiply that number (of runs) by .8, or whatever percentage of average the replacement level is designated to be; the result is the number of runs you could expect a "replacement player" to put up with that number of outs. Simply subtract the replacement's runs created from the player's actual runs created, and the result is VORP."

[Huckleberry]

What are some of the primary arguments against Win Shares? I'd be interested in reading those arguments as well as any rebuttals. Certainly it's a very convoluted number, but the building blocks to get there all seem rational.

[lynchjm24]

What are some of the primary arguments against Win Shares? I'd be interested in reading those arguments as well as any rebuttals. Certainly it's a very convoluted number, but the building blocks to get there all seem rational.

Firstly the defensive aspect is not very well done. A lot of people don't like where he set replacement level. My biggest problem with it is that it doesn't really tell you anything. It can give you an idea of who had a good season last year - but there is nothing predicitive about it - I guess that is what I'm usually looking for in a stat. I took a quick look on Google and didn't see anything, but I'm sure that baseballthinkfactory has plenty of well thought out arguments for and against the stat.

[st.cronin]

Firstly the defensive aspect is not very well done. A lot of people don't like where he set replacement level. My biggest problem with it is that it doesn't really tell you anything. It can give you an idea of who had a good season last year - but there is nothing predicitive about it - I guess that is what I'm usually looking for in a stat. I took a quick look on Google and didn't see anything, but I'm sure that baseballthinkfactory has plenty of well thought out arguments for and against the stat.

Actually I think the main argument against is that it "forces the numbers to balance." So if a team doesn't win as many games as you would think, looking at their stolen base total, for example, then it penalizes that team's stolen bases in respect to the rest of the league.

That was my understanding, anyway, but I'm not very good at math.

[lynchjm24]

Actually I think the main argument against is that it "forces the numbers to balance." So if a team doesn't win as many games as you would think, looking at their stolen base total, for example, then it penalizes that team's stolen bases in respect to the rest of the league.

That was my understanding, anyway, but I'm not very good at math.

Yes at the margins it forces results, which seems stupid.

Uberstats in general are stupid, and this is a stat the world would be no worse off without.

[Crapshoot]

One of the biggest problems with WS is the lack of Loss Shares - as James formulated them - a player cannot be worth less than 0. THT's methodology attempts to account for this. And defense is the other canard, per se.

[clintl]

Firstly the defensive aspect is not very well done. A lot of people don't like where he set replacement level. My biggest problem with it is that it doesn't really tell you anything. It can give you an idea of who had a good season last year - but there is nothing predicitive about it - I guess that is what I'm usually looking for in a stat. I took a quick look on Google and didn't see anything, but I'm sure that baseballthinkfactory has plenty of well thought out arguments for and against the stat.

It isn't intended to be a predictive stat, is it? I thought James' entire motivation for creating the Win Share stat was as a tool to compare player value across eras - basically, a new tool to interpret history. So it's meant to be a backward-looking stat.

[Anthony]

can anyone point me in the direction of any article that mentions that you can bat any player anywhere in the lineup and it wouldn't have a difference on his stats? i know someone out there mentioned it (Bill James?), i can't seem to find it via google.

[oykib]

I think that people's biggest problem with WS is the lack of negative players ( no less than zero, no loss shares). Many people also have a problem with his defensive measures. But his formula does seem to have far fewer problems than linear weights as far as WTF moments when looking at player ratings.

WS has Manny Ramirez as the best defensive LF in the AL since his move to Boston. It seems that the strange configuration of the park throws off his WS evaluation. But that's much better than, I believe, linear weights saying that Johnny Bench is the worst defensive catcher of all time.

The main reason I like win shares is that the rankings tend to match up logically with what we subjectively believe. Where it doesn't, we can come up with a logical reason why.

[QuikSand]

Seems to me there are two different arguments being used (not just here, but generally):

- I don't like [insert combined baseball stat here] because I think it's flawed in some way

- I don't like combined baseball stats in principle

Those are very different arguments. If you reject Win Shares because you think it doesn't measure defense well (or something of the sort) I can respect that opinion. If you reject Win Shares simply because it wasn't published in the newspapers way back when, then not so much.

[Huckleberry]

can anyone point me in the direction of any article that mentions that you can bat any player anywhere in the lineup and it wouldn't have a difference on his stats? i know someone out there mentioned it (Bill James?), i can't seem to find it via google.
I'm pretty sure the study implied that moving players around in the lineup had a minimal impact on team performance, not player stats. Obviously runs and RBI for an individual depends heavily on who is in the lineup behind and in front of him, respectively.

[bosshogg23]

Is sabermetrics good for baseball? Bad for baseball? Make it more fun? Less fun?

More knowledge IMO is always great. Hell the more knowledge we have the less we realize we know, particularly in a sport like baseball. I find it great to watch trends in stats in baseball. OBP has already lessened in value, defense has picked up value and bullpen price is next in line after defense.

If you watch Oakland, they value UNDERVALUED players, not OBP or Speed or Defense or Strength. They value less valued ppl ONLY.

[Arctus]

can anyone point me in the direction of any article that mentions that you can bat any player anywhere in the lineup and it wouldn't have a difference on his stats? i know someone out there mentioned it (Bill James?), i can't seem to find it via google.

I'm almost positive its James. IIRC he covered this in his first Historical Baseball Abstract.

And its splitting hairs a bit, but I think his conclusion was that a team would score roughly the same amount of runs regardless of how the batting order was set.

[st.cronin]

I just remembered - doesn't win shares also look at rbi? I think that's another criticism.

[oykib]

Seems to me there are two different arguments being used (not just here, but generally):

- I don't like [insert combined baseball stat here] because I think it's flawed in some way

- I don't like combined baseball stats in principle

Those are very different arguments. If you reject Win Shares because you think it doesn't measure defense well (or something of the sort) I can respect that opinion. If you reject Win Shares simply because it wasn't published in the newspapers way back when, then not so much.

Unfortunately, it's almost always the latter-- at least it seems so. People who dislike WS because of percieved flaws will look for stats that don't have them or use multiple stats and make a composite evaluation-- not say WS is flawed so Bill James and all his disciples are morons.

[clintl]

FWIW, I think Win Shares is flawed precisely for the reason that has been stated - it doesn't account for negative contributions. I think OPS is a flawed stat. But being flawed doesn't mean they're not useful.

[Buccaneer]

For what it's worth, I love modern sabermetrics. They are much better than what we used to have (or not have).

[lynchjm24]

Unfortunately, it's almost always the latter-- at least it seems so. People who dislike WS because of percieved flaws will look for stats that don't have them or use multiple stats and make a composite evaluation-- not say WS is flawed so Bill James and all his disciples are morons.

Just for the record, I love baseball statistics and all the progress that has been made in many areas. I just think Win Shares is stupid. I know it's not supposed to be predicitive, that's just one of many reasons I don't like it.

It was statistics and Rob Neyer that brought me back seriously into baseball after I drifted away in college. Neyer, Prospectus and the statheads brought me back, but I've gotten very into scouting. I'd like to think of myself as someone who thinks like Kevin Goldstein, he's easily my favorite writer at this point.

[oykib]

Well, then, it doesn't apply to you, lynchjm24.

But haven't you become frustrated with people who can't even accept OBP?

[lynchjm24]

Well, then, it doesn't apply to you, lynchjm24.

But haven't you become frustrated with people who can't even accept OBP?

Are there many people left besides Michael Kay who don't accept OBP?

[Pumpy Tudors]

Here's a (hopefully) simple question: Why are so many statistics being "invented" for baseball and not so many for other major sports?

[Crapshoot]

Here's a (hopefully) simple question: Why are so many statistics being "invented" for baseball and not so many for other major sports?

Well, mathematically speaking, baseball is far more simple for analytical purposes. You have a series of discrete events that allow one to isolate the effects of a given variable on performance - unlike basketball, football, etc etc, where you have to account for multiple factors simultaneously. That alone leads to baseball being perhaps the easiest sport to model.

[cuervo72]

Along those lines, it should also be the easiest to emulate programmatically...

(I'm hopping between this and another thread :) )

[st.cronin]

Here's a (hopefully) simple question: Why are so many statistics being "invented" for baseball and not so many for other major sports?

What is your favorite flavor skittles?

[st.cronin]

Along those lines, it should also be the easiest to emulate programmatically...

(I'm hopping between this and another thread :) )

And yet, from the evidence I have, it's far and away the hardest.

[lynchjm24]

And yet, from the evidence I have, it's far and away the hardest.

I think for the reason that it's also the hardest to suspend belief during, and also the easiest to pick apart when looking at results.

[TroyF]

Here's a (hopefully) simple question: Why are so many statistics being "invented" for baseball and not so many for other major sports?


But that's just wrong. . .

The Football Prospectus has some incredible theories on statistics in football. The book last year was incredibly thought provoking and really well written. They also have their own stats that they have created specifically for football.

John Hollinger and his PER rating along with 82games.com have worked on creating a new way to look at statistics in basketball.

The thing about baseball statistics that makes them more popular is that you don't have to worry about getting statistics for an entire team most of the time. it's hitter vs. pitcher in a series of one on one battles. Even "protection" in the order can be studied easily.

It's more difficult to do this in other sports because every action requires looking at the team concept. Put Troy Aikman on a team without a good offensive line and a poor running game and how good would he be? What if Jerry Rice would have been drafted by the Arizona Cardinals? Things are more difficult to judge.

But people are trying and working hard to find ways to rate players in other sports. footballoutsiders.com and 82games.com are both terrific sites for stat geeks.

[st.cronin]

Diamond Mind Baseball was a fantastic computer game. It just didn't evolve the way it should have. It's possible to make a good baseball sim. Isn't it?

[Buccaneer]

Diamond Mind Baseball was a fantastic computer game. It just didn't evolve the way it should have. It's possible to make a good baseball sim. Isn't it?

A good baseball sim <> a good career sim

Stuff like financials, player transactions, free agency, drafting, etc. get in the way of a good baseball sim focusing on the sabermetrics and details of an at-bat, an inning and a game.

More on this thought tomorrow...

[cuervo72]

I agree with that. Like I've said many times, the most I've played a baseball game is MicroLeague, which has none of the extra stuff. But as a simming engine - though I'm certain it was rudimentary - I had no problems with it. It worked well in a small league where humans were drafting/trading (with real players), and it worked well if you wanted to take Team A from whatever era and fire off a game against Team B from any era (such an exhibition does not seem easy in OOTP, where you are dealing with a set league or universe). PLUS, the little players ran around and the ball moved realistically!

[Pumpy Tudors]

What is your favorite flavor skittles?
I'll never tell.

But that's just wrong. . .
Well, the fact that I was wrong kind of supports my point. The "invention" of new statistics (and I understand that I'm using "invention" in a layman's way here) may be happening in other sports, but it's not nearly as widespread as it is in baseball. I'd never heard of 82games.com, and I had no idea that footballoutsiders.com was doing the things that you described. With baseball, however, if I go into a bookstore, Bill James' name is all over the sports section. He might as well be standing there, leering at me and hissing, "Let's get together and calculate how many runs Toby Hall created last year." These relatively new baseball stats are all over the place.

Anyway, I think my question has been sufficiently answered.

[Huckleberry]

I think Crapshoot's explanation is on the money. I've made that argument countless times.

We all know formulas are much easier to derive and understand for systems that operate in series. Much more difficult for systems in parallel.

Baseball is played much more in series than any other popular sport.

[Anthony]

I'm pretty sure the study implied that moving players around in the lineup had a minimal impact on team performance, not player stats. Obviously runs and RBI for an individual depends heavily on who is in the lineup behind and in front of him, respectively.

sorry, that's what i meant to say.