Bill James Answers readers questions.

[J Brew Crew]

http://freakonomics.blogs.nytimes.com/2008/04/01/bill-james-answers-all-your-baseball-questions/

 

I found this on the Freakonomics blog (free, no registration). Some interesting stuff like this answer. He answers to many Red Sox/ Yanks questions for my taste.

 

Q: Has sabermetrics pretty much squeezed the last drop of new insights out of traditional counting statistics? If so, what data ought to be collected to improve our understanding of the game? If not, where can the boundaries be pushed?

A: We haven't figured out anything yet. A hundred years from now, we won't have begun to have the game figured out.

 

I'm not sure if this should be here or in the Stat area. Mods, please move as you see fit.

 

Regards

JJ Brew

 

 

[Tbadder]

As a total outsider, and some one who is wrong on a consistent basis, I'd agree with James simply because there are more variables than can be accounted for in each situation. At best science offers a thumbnail sketch of the universe (a very important thumbnail sketch). I can't believe science and mathematics are done changing.

[PlayerHader]

Really, would the game be any fun we ever "figured it out." All the fun is in the unknown.

[rluzinski]

I wouldn't say he's wrong but James is being a bit more dramatic than he needs to be. You'll never be able to model the game of baseball (or any sport played by people) perfectly of course but with each passing year they seem to be doing a better job at it. This is not quantum physics we are talking about. The more things are studied, the more they are understood and the less breakthroughs that will be found. Diminishing returns.

 

I don't see football fans being threatened by people trying to figure out the nuances of football through statistical studies. It's odd to me that the average baseball fan is so protective of their sport.

 

Really, would the game be any fun we ever "figured it out." All the fun is in the unknown.

 

Most of the things studied works best in the macro view anyway (value of a player over a season and the like). If you are talking about a single AB, even if you know the batter has a 35% of succeeding or whatever, we all know that anything can happen. Hell, Kendall almost had a HR yesterday! There's plenty enough of randomness and human element in the game to keep everyone guessing.

[endaround]

Russ--Not only are they not threatened by it, a former player (Aikmen) is on the forefront of it (ok his money is)

 

Actually the reason for the resistance is because there has been resistance from many in baseball and fans get their knowledge from former players (see how many people repeat Rock's calls for "consistency" and "manufacturing runs"). Baseball is much more guild like than football. You'll have football coaches who have very little playing experience even at the college where a baseball manager without major league playing experience is almost unheard of. And so stats were seen as breaking down the guild.

[RightFieldCoder]

Here is the difference between football stat guys and baseball stat guys. Football stat guys usually end up with stats that agree with the non stat guys. There might be disagrements of a small degree (ie who was better Marino/Elway) but rarely do I see a football stat say a guy is bad when everyone thinks he's good. The same is not true in baseball. There are players that some think are good, like Eckstein, Glaus, and Piere. The stat guys say they are not good players. This causes the rift. I bet if football stat guys came out and said Farve, Manning, and Ray Lewis were not good players and were worse than league average, there would be as big as rift over there as there is in baseball.

 

Another difference is the quickness to change. When one football team figures out something that works, usually there are 31 other teams that have incorporated it within a year. When one team figures out something in baseball (ie OBP is better than AVG for evaluating talent), there are years after a book is written about it there are still teams, announcers, and fans that don't believe it.

[Robideaux]

Stat guys say that Glaus isn't any good? That doesn't sound right. Low batting average, decent walk rate, power--that seems like the type of guy who is generally undervalued by fans who mostly use triple crown statistics as a gauge of talent. With Eckstein, its not that people say he's bad, rather that the glorification of his scrappyness/grittyness/intagibles is a bit overblown.

Pierre? Yeah, he was actually pretty good there for a few years in Florida and Colorado, but he's now a singles-hitting out machine.

 

I think your on to something with the quickness to change. I think baseball is so steeped in tradition that a certain segment of fans, managers, and execs, are simply resistant to change (regardless of the form).

[Tbadder]

One of the reasons I love science is that it's the only place where being wrong counts for something important. Good science is when important truths are right one day, and wrong the next day. Statistics as a tangental branch of science works really well when it disproves itself. We make decisions based on what we know at the time is incomplete knowledge (if we're honest). Will that bring us closer to an ultimate truth--I'm not so certain, I'll leave that to preachers and mathematicians and other charlatans (insert smiley face, com'on guys only kidding face). And I don't agree, baseball is quantum physics since nothing escapes physics--I'd say it's pretty close to infinitely faceted. But as always, proviso, apology, I'm not worthy, just one man's opinion, etc. etc. etc.

[TerraceReserved]

Q: Has sabermetrics pretty much squeezed the last drop of new insights out of traditional counting statistics? If so, what data ought to be collected to improve our understanding of the game? If not, where can the boundaries be pushed?

A: We haven't figured out anything yet. A hundred years from now, we won't have begun to have the game figured out.

That's a crap statement. He really didn't answer the question, and he's wrong. Stuff has been figured out, but it will always be improved upon and new findings will always come about. Just saying we haven't figured anything out yet is stupid.

[Fatter than Joey]

Just saying we haven't figured anything out yet is stupid.

 

I read what we said to mean in an absolute fashion -- that is to say, we will never figure out any facet of baseball absolutely.

[danzig6767]

Here is the difference between football stat guys and baseball stat guys. Football stat guys usually end up with stats that agree with the non stat guys. There might be disagrements of a small degree (ie who was better Marino/Elway) but rarely do I see a football stat say a guy is bad when everyone thinks he's good. The same is not true in baseball. There are players that some think are good, like Eckstein, Glaus, and Piere. The stat guys say they are not good players. This causes the rift. I bet if football stat guys came out and said Farve, Manning, and Ray Lewis were not good players and were worse than league average, there would be as big as rift over there as there is in baseball.

 

Another difference is the quickness to change. When one football team figures out something that works, usually there are 31 other teams that have incorporated it within a year. When one team figures out something in baseball (ie OBP is better than AVG for evaluating talent), there are years after a book is written about it there are still teams, announcers, and fans that don't believe it.

Football likely will never have a sabermetrics following like we see in baseball for one simple reason, football is to much of a team game compared to baseball. Sure there are plays in baseball that take team play, but it's mostly pitcher vs batter and fielder makes play or doesn't.

In football, nearly every play whether on offense or defense needs almost all or all players to do their job properly or the play can fail. If an offense runs say a sweep play, 10 of the 11 players can do their job well, but if one guy misses his block, the play can be ruined. Football is the ultimate team game, put a running back behind a terrible line and he'll struggle to produce. A hitter in baseball needs only his own skill to succeed.

Crunching numbers has some value in basketball also, but that is too more a team game than baseball which is why baseball works so well with stat crunching. There is no grey area to a batters OBP for a season.

 

[J Brew Crew]

"Q: Is clutch hitting a repeatable/retain-able skill?

A: I don't know"

 

I wish he would have expounded a little more. Or at all. It really interests me on the how and the why of stats he would use to figure that out.

 

Regards

 

J

[rluzinski]

Football likely will never have a sabermetrics following like we see in baseball for one simple reason, football is to much of a team game compared to baseball.

 

I agree that baseball is simpler to model but I'm sure that football is very heavily studied. I've seen some work online and I bet much of it is proprietary.

[danzig6767]

Football likely will never have a sabermetrics following like we see in baseball for one simple reason, football is to much of a team game compared to baseball.

 

I agree that baseball is simpler to model but I'm sure that football is very heavily studied. I've seen some work online and I bet much of it is proprietary.

I'll admit up front that i've never looked into it at all, but i question how a sport that relies so heavily on teammates all performing well during each given play for it to be successful, that a sabermetrics system could evolve anywhere near like it does in a largely individual sport like baseball is.

I could see studying things like down/distance and what plays then generally work best at getting first downs. I also can see teams developing numbers to rate players that fans don't know about, but don't ever see a day that football fans are discussing sabermetric numbers for say nose tackles or guards on the offensive line. Even skill position players like running backs or wide receivers that put up actual stats are so reliant of the play of their QB, offensive line, and even the play calling of their coaches. A good pass rush can make corners look good and no rush can make them look worse than they are. There are so many things that happen in football that are out of the hands of individual players that can drastically skew their stats.

Randy Moss was a perfect example of this. He played in a terrible situation in Oakland, his stats took such a nosedive that many thought Randy was in serious decline. One year later in a much better situation, he shatters records during an amazing year. A good pass rusher could have a great season with 15 sacks. The next season he loses the other good pass rusher next to him, thus faces constant double teams, and his sack totals get sliced in half. You could go on and on for individual players, the situation around them that helps make them look better or worse than they really are. Baseball players face none of that, except to a slight degree pitchers and how good/bad their defense is behind them.

 

[logan82]

danzig6767 wrote:

Baseball players face none of that, except to a slight degree pitchers and how good/bad their defense is behind them.

Unless you believe in protection. Then a player on a team with many other good players could see more pitches to hit and have better stats, in theory.