Selective Application of Statistics

[Cal Ripken, Jr.]

<tear>

Yes, Virginia, there is a Santa Claus!

Oh, you mean my off-season job?

Speaking of which...

[free2131]

Swoon

[SirKodiak]

I never said its crazy that a defensive player could be rated higher than another in a single season, I said merely that its crazy for a system to rate Camerons 09 performance as equivalent to Bays 09 performance, let alone 20-25% better as some of the values have been posted.

According to fangraphs:
Bay 3.5 WAR
Cameron 4.3 WAR
This does NOT mean Cameron was 20-25% better than Bay because:

1. WAR is Wins Above Replacement, not Win Above Nothing
2. WAR measures observed value, not who is the better player

Almost a full point of WAR difference comes from one playing CF and one playing LF.

So what does this mean? Does this mean that this system claims Ripken, Tejada and Fernandez were all better than Oz over their career?

Basically, it means that they were better offensively against their peers than Ozzie was against his.

Or for the Coles notes version, it means that according to this metric Ripken, Tejada, and Fernandez were better hitters than Ozzie. Once defense is factored in, only Ripken was more valuable than Ozzie over the course of his career. Hope that helps.

Better offensively, not hitting. SB and CS are included.

[dickay]

According to fangraphs:
Bay 3.5 WAR
Cameron 4.3 WAR
This does NOT mean Cameron was 20-25% better than Bay because:

1. WAR is Wins Above Replacement, not Win Above Nothing
2. WAR measures observed value, not who is the better player


Almost a full point of WAR difference comes from one playing CF and one playing LF.

Your issue is purely with FanGraphs WAR. It has Cameron as being about .8 of a win better in 2009. That isn't significantly better. It's 8 runs better overall. CHONE's WAR has Jason Bay being better. BP's WARP has them equal.

my comments were referring to a post filihok made in another thread, which stated something to the effect that Camerons "value" from fangraphs last year was around 19 mil or so while Bay was around 13 or 14mil. That is a substantial difference in terms of dollar value.

So all three graphs supposedly factor in defense and all three are a good deal different. One says equal, one says camerons better, one says bays better?? WTF?

[dickay]

here was fili's post;

I know you're not a big believer, in this, but fangraphs has Cameron being above $7 every year since 2002 and worth about $19 million (4 WAR) the last two.

Jason Bay was worth $15.7 (3.5 WAR) last year

and Manny was worth 5.1 WAR and 1 WAR in 04 and 07

that is a claim that Camerons worth was just about 20% more than Bays in 2009.

[HoustonGM]

my comments were referring to a post filihok made in another thread, which stated something to the effect that Camerons "value" from fangraphs last year was around 19 mil or so while Bay was around 13 or 14mil. That is a substantial difference in terms of dollar value.

The value of a win recently has been about $4.5 million. Cameron's value in 2009 comes to $19.4 million and Jason Bay's comes to $15.7 million, a difference of $3.7 million, which is not a lot. It's less than a win, as I showed - 0.8 wins.

So all three graphs supposedly factor in defense and all three are a good deal different. One says equal, one says camerons better, one says bays better?? WTF?

FanGraphs has Bay at 3.5 and Cameron at 4.3.

Chone has Bay at 5.2 and Cameron at 3.6. Again, as I said, this is because their defensive metric has Bay as a slightly above average fielder in 2009.

BP has Bay at 4 and Cameron at 4.1. Their defensive metric has Bay at -2 and Cameron at +4.

And it's really NOT a good deal different. You're getting down to differences of about a win, which is not a lot.

[gosensgo101]

Good to see that a Dickay/Houston disagreement can withstand even the biggest derailments.

[dickay]

The value of a win recently has been about $4.5 million. Cameron's value in 2009 comes to $19.4 million and Jason Bay's comes to $15.7 million, a difference of $3.7 million, which is not a lot. It's less than a win, as I showed - 0.8 wins.

you do realize that the 3.7 difference is a bit more than 20% correct??

FanGraphs has Bay at 3.5 and Cameron at 4.3.

Chone has Bay at 5.2 and Cameron at 3.6. Again, as I said, this is because their defensive metric has Bay as a slightly above average fielder in 2009.

BP has Bay at 4 and Cameron at 4.1. Their defensive metric has Bay at -2 and Cameron at +4.

And it's really NOT a good deal different. You're getting down to differences of about a win, which is not a lot.

so three apparently respectable stat algorithms place different values on defense, and even measure defensive performance differently as one feels Bay was good defensively and one does not. Why then are people on these forums treated as outcast geezers for doing the same?

my only point here from the beginning was not to say that these stats are rediculous and should never be used....but merely to say that they are imperfect and flawed. I think this pretty much sums that point up. Again, it is my opinion that any matrix that says Cameron was an equal value to Bay, let alone one that says Cameron was 20% more valuable...is flawed and gives way too much credence to defense.

[dickay]

Good to see that a Dickay/Houston disagreement can withstand even the biggest derailments.

i'm confused on our disagreement. I say one claims that Cameron is 20% better. He says no, its not a big difference and then posts the stats that show a 20% favor in Camerons favor??

[dickay]

btw.....a very merry christmas to all!

[HoustonGM]

you do realize that the 3.7 difference is a bit more than 20% correct??

But it's not large in the context of the stat. It's a difference of 8 runs.

my only point here from the beginning was not to say that these stats are rediculous and should never be used....but merely to say that they are imperfect and flawed. I think this pretty much sums that point up. Again, it is my opinion that any matrix that says Cameron was an equal value to Bay, let alone one that says Cameron was 20% more valuable...is flawed and gives way too much credence to defense.

Your free to disagree on UZR's take on their difference - but remember that it's over the course of a year, UZR fluctuates A LOT in one year - as much as offensive stats fluctuate in 2 months or so (but even the long term view of UZR has Cameron being a very good fielder and Bay a very poor one, so that's a moot point in this regard). Personally, I don't find it flawed that a stat sees Cameron as a +10 defender and Bay as a -13 defender. Both numbers line up with the conventional view of their respective defense.

But, in the grand scope of things, you really shouldn't be drawing conclusions about a stat based on what it says about two players in one year. What in the construction of the stat do you find flawed? I went through it step-by-step above.

[dps]

HGM, if you don't think that $3.7million is much of anything, please send me a check for that amount. I'll pm you my mailing address.

[actionjackson]

Better offensively, not hitting. SB and CS are included.

My bad. I did actually know that and what I meant by better hitter was better offensive contributor. I expected you with your awesome powers to be able to read my mind SirK.

[reflections]

So not to smash any sabermetric toes, but when WARP and VORP and UZR start being talked about, I almost expect you to figure out how to put a civilization on the moon. But instead you seem to make a beautiful game with so much history and and stories beyond numbers...into nothing but a bunch of numbers. kinda sad, but I guess that is progress

[200tang]

But instead you seem to make a beautiful game with so much history and and stories beyond numbers...into nothing but a bunch of numbers. kinda sad, but I guess that is progress

lol! I don't even see the connection between evaluating players and enjoying the game. Give me a break.

I appreciate things like hustle and good clubhouse guys, if I didn't I wouldn't consider Jackie Robinson the greatest player to play the game. The only difference is that I don't say things like "Melky Cabrera is a great player and I wouldn't have given him up for Javier Vazquez because the fans here like him and he's a great defensive player because he shows a lot of hustle out there". The stories behind the game are interesting and I love reading interviews with the players on the teams I root for, but like I said, I just don't use the stories to try and make a dumb claim like "Melky Cabrera > Javier Vazquez".