artie_fufkin wrote:

forsberg_us wrote:

I'll be the first to admit I understand little about sabermetrics, but I have a question about WAR.

According to Baseball Reference, WAR stands for wins above replacement, and is defined as: "a single number that represents the number of wins added to a team above what a replacement player (think AAAA or AAA) would add.  This value includes defensive support and includes additional value for high leverage situations."

http://www.baseball-reference.com/leagues/NL/2010-value-pitching.shtml#players_value_pitching::11

The National League pitcher WAR rankings have Wainwright listed as have a 5.7 WAR.  Wainwright won 20 games last season.  Who is the replacement that would have won 14.3 games last year, and why wasn't he in the #5 spot in the rotation last season?

See!!! Just thinking about this crap will give you an ice cream headache!

It's even worse than you may realize. I was reading an explanation on how they calculate WAR. Admittedly I grew tired and gave up after a bit, but one of the things I got out of it was that they use some statistics called Fielding Independent Pitching which is a figure that removes any outcome involving a fielder. In essence it only considers HR, Ks and walks.

I think (although I'm not entirely sure) that you could have 2 pitchers on the same team with the exact same stats (W-L, ERA, WHIP, IP) except than one had significantly more Ks. The one with more Ks would be considered worth more because ground outs and fly outs are apparently worth less than a strikeout.

Max wrote:

APRTW wrote:

CY Young, the guy the award was named after is the all time leader in ........drumroll.....WINS.

We weren't talking about the Cy Young award, but even so, he is probably the career leader in a few things.  There is nothing in the Cy Young award itself that says it goes to the pitcher with the most wins.  If there were, there would be nothing to argue about on that score. 

Bottom line:

Team wins: two enthusiastic thumbs up!
Pitcher wins: thumbs down

He is the leader in wins, complete games,  earned runs allowed, hits, loses and batters faced.

So if you think that a good pitcher faces alot of batters, allows them to get hits, score and loses the game then Cement-head should have won a CY.  Or you can go with the idea that winning games and finishing them is what makes a pitcher the best in the league.

JV wrote:

Max wrote:

Can we please get back now to why JV is defending the indefensible?

Come on, Max; I haven't said ONE THING in support of Artie's position.

Et tu, J?

artie_fufkin wrote:

"one of the things I got out of it was that they use some statistics called Fielding Independent Pitching"

So if your FIP is independent of your WHIP, how can your VORP be dependent of your WAR? And if your WAR drops, you could reasonably assume your Ps and Qs aren't in order, so your Ks will suffer, your ERA will balloon and you stand little chance of winning the CYA, much less being elected MVP. Which means you'll lose your VIP pass to the Circle K, so you'll have to park out back and maybe run into a mob from the USMC that's AWOL and hiding from the MPs because they were put on KP by some officer who is fresh out of ROTC. And those are some mean SOBs, especially if they've been drinking PBR.

REC

Egg head comment: I am not a statistician, but I have used programs like SYSTAT and SPSS and I have friends, family, and colleagues who understand multivariate statistics in a way that I never will.  I know nothing at all of advanced statistics in baseball, but I am assuming they are just someone with a statistics background trying to develop more advanced statistical tools than BA, ERA, etc.  So, on the chance that I might have something worthwhile to add I will mention . . .

when a statistic is said to be independent of this that or the other thing, it doesn't necessarily mean they had to find situations where the condition in question does not occur.  For example, a very common use is in epidemiology where they try to find the effect of, say eating red meat on heart attacks.  One of the biggest issues, of course, is that many things affect the likelihood of a heart attack, the biggest of which we have no control over whatsoever, our age.  So the data must be corrected for age, such that the results are independent of age.  This doesn't mean that they studied ageless people, but rather that they used fancy multivariate statistics to remove the effects of age.  There is no perfect way to do this, of course, so it is an ugly compromise of many subjective decisions, a 'making the sausage' type of process.  But the end results are often much better than simply throwing ones hands in the air and saying, 'it's all subjective, no one can really know anything.'

So my assumption has been that the sabermetrics people have built very large databases that include not only things like plate appearance, at bats, and hits, but also, ballpark, pitcher, etc. and they crunch them in a program like SPSS using some sort of multivariate regression.  This way the statistics account for things that we know, but could not quantify before.  For example, we know that Pujols is a lifetime .331 hitter, but we also know that there are other variables that factor large in whether he is likely to get a hit in any given at bat, in addition to the obvious, is the pitcher a righty or lefty?  These include things like, WHICH pitcher he is facing?  Which park is he batting in?  etc. 

Overtime I predict that, just like Deep Blue beat gary Kasparov, they will develop computer programs that predict an outcome better than the best baseball men, but they are still a long way off.

And FWIW, I don't know the new statistics well enough, but if I were measuring up two pitchers on different teams, I would look first at their ERA, and then there IP, and I might look at the W, but  wouldn't pay it much mind.  To me pitcher W's is one of the stupidest, least helpful statistics, dumber by far than HR's and K's that count something that actually happened on the playing field.

artie_fufkin wrote:

JV wrote:

Max wrote:

Can we please get back now to why JV is defending the indefensible?

Come on, Max; I haven't said ONE THING in support of Artie's position.

Et tu, J?

I was tempted to post that exact phrase and say I'd saved you the trouble.

artie_fufkin wrote:

OK, so we agree that we're going to agree to disagree on this issue. But I still think you're making my point. Good pitchers tend to win more games. Not primarily because of their offense, or defense, or bullpen, but because they're good pitchers.

We're gonna have to agree to disagree about that, too, because I think you have repeatedly made my point for me, and even acknowledged a few years back when you finally wrote that, standing in a whole you didn't know why you kept on digging.  But that's that.  At least we can agree that the Raiders are pigs and Walter Payton is the best running back the NFL has ever known, or is likely to know anytime soon.