WPA: 4/13 vs. Cards

Here are the Win Probability Added totals for last night's game. You can read more about WPA at The Hardball Times. The totals come out to .506 instead of .500. The good news is that I'm getting closer to figuring out why.

Joe Randa was the man, of course,. I gave him a decent amount of credit for the game-ending double play. With one out, behind by a run and the bases loaded, the visiting team has a WP of only .467. So that DP was just huge. Maybe the second-biggest play of the season outside Dunn's home run in the opener.

Eckstein's double was also big, and I thought some leftfielders might have caught it, so I debited Dunn's fielding WPA somewhat.

Team    Player      Off    Pitch   Field     WPA
Reds    Randa      0.218   0.000   0.162   0.380
        Wagner     0.000   0.160   0.000   0.160
        Casey      0.102   0.000   0.003   0.104
        Jimenez    0.066   0.000   0.019   0.085
        Belisle   -0.042   0.103   0.000   0.062
        Freel      0.056   0.000   0.000   0.056
        None       0.015   0.000   0.000   0.015
        Cruz      -0.003   0.000   0.000  -0.003
        Kearns    -0.004   0.000   0.000  -0.004
        Pena      -0.016   0.000   0.010  -0.006
        Dunn      -0.012   0.000  -0.021  -0.032
        LaRue     -0.034   0.000   0.000  -0.034
        Griffey   -0.043   0.000   0.000  -0.043
        Aurilia   -0.056   0.000   0.004  -0.052
        Graves     0.000  -0.062   0.000  -0.062
        Weathers   0.000  -0.122   0.000  -0.122
Reds Total         0.249   0.080   0.177   0.506


Team    Player      Off    Pitch   Field     WPA
Cards   Eckstein   0.316   0.000   0.019   0.335
        Tavarez    0.000   0.096   0.000   0.096
        Taguchi    0.092   0.000   0.000   0.092
        Edmonds    0.097   0.000  -0.006   0.091
        Luna       0.091   0.000   0.000   0.091
        Isringha   0.000   0.053   0.000   0.053
        Molina     0.037   0.000   0.000   0.037
        Walker     0.019   0.000  -0.005   0.014
        Sanders    0.013   0.000   0.000   0.014
        Mabry      0.006   0.000   0.000   0.006
        None       0.005   0.000   0.000   0.005
        King       0.000  -0.010   0.000  -0.010
        Grudz     -0.035   0.000   0.004  -0.031
        Aurilia   -0.037   0.000   0.000  -0.037
        Nunez     -0.026   0.000  -0.014  -0.041
        Flores     0.000  -0.046   0.000  -0.046
        Cedeno    -0.073   0.000   0.000  -0.073
        Rolen     -0.143   0.000   0.003  -0.140
        Mulder    -0.092  -0.343   0.000  -0.435
        Pujols    -0.527   0.000   0.000  -0.527
Cardinals Total   -0.257  -0.250   0.001  -0.506

The Reds pulled five big double plays in this game, which added up to a total WP of 0.784.

[ochre]

I think WPA underestimates the suckiness that is David Weathers at the moment.

[smith288]

Pujols worse than Weathers? Put Pujols on suicide watch!

[SteelSD]

Originally Posted by studes

Eckstein's double was also big, and I thought some leftfielders might have caught it, so I debited Dunn's fielding WPA somewhat.

Did you make the same defensive WPA negative "tweak" on the ball that Pena allowed to fall between himself and Griffey a couple games back (Sunday I believe).

Just wondering, because it appears that you're venturing into the realm of what you think should have happened rather than sticking with what actually did.

Originally Posted by SteelSD

Just wondering, because it appears that you're venturing into the realm of what you think should have happened rather than sticking with what actually did.

I don't remember that specific play, but on every single fielded play I am making a judgement call as to what percent of credit should go to the fielder, and what percent should go to the pitcher. As my guide, I use David Pinto's fielding graphs and data, which are available at Baseball Musings. It's basically an eyeball approach to David's fielding stats, or MGL's Ultimate Zone Rating.

I recognize that I'm setting myself up for lots of second guessing, but I think it's better to try to do this than not. And I hopefully learn more about the pitching/fielding split methodology along the way.

[SteelSD]

Originally Posted by studes

I don't remember that specific play, but on every single fielded play I am making a judgement call as to what percent of credit should go to the fielder, and what percent should go to the pitcher. As my guide, I use David Pinto's fielding graphs and data, which are available at Baseball Musings. It's basically an eyeball approach to David's fielding stats, or MGL's Ultimate Zone Rating.

I recognize that I'm setting myself up for lots of second guessing, but I think it's better to try to do this than not. And I hopefully learn more about the pitching/fielding split methodology along the way.

Ah. Ok. I guess I didn't realize how truly subjective your method was for assigning Defensive/Pitching WPA "points".

And I just checked your WPA thread from Sunday's game and it doesn't appear that Pena was "debited" for a far more aggregious error than having a ball smoked over your head while being positioned in due to the hitter's lack of pop.

Originally Posted by SteelSD

And I just checked your WPA thread from Sunday's game and it doesn't appear that Pena was "debited" for a far more aggregious error than having a ball smoked over your head while being positioned in due to the hitter's lack of pop.

Yes, I'm sure I'll miss some plays. Important to keep in mind when interpreting the fielding/pitching splits. Also, I don't watch every game in its entirety. I usually try and catch up via the "condensed games" on MLB.com, and they don't show every play. They like to show strikeouts but not fielded outs. Very frustrating.

[RedWilly]

Studes
Have you posted a season WPA total thread yet?
I know you said you were planning on doing so, just wondering if I missed it or if it hasnt been put up yet?

[BadFundamentals]

Originally Posted by RedWilly

Studes
Have you posted a season WPA total thread yet?
I know you said you were planning on doing so, just wondering if I missed it or if it hasnt been put up yet?

I'm going to compile each game into spreadsheet. Below are results thru the first 8 games. Also here is link to my spreadsheet. Feel free to take copy for yourself to update or check back for my updates......

http://www24.brinkster.com/michb/baseball/default.asp

player	WPA	OPS
Randa	1.076	1.391
Dunn	0.669	0.758
LaRue	0.265	0.862
None	0.191	
Ortiz	0.132	
Valentin	0.076	0.75
Belisle	0.066	
Weber	0.037	
Lopez	-0.003	1
Wagner	-0.003	
Pena	-0.009	1.071
Cruz	-0.023	
Griffey	-0.028	0.693
Graves	-0.036	
Jimenez	-0.049	0.699
Kearns	-0.08	0.646
Wilson	-0.091	
Harang	-0.114	
Mercker	-0.152	
Freel	-0.155	0.988
Casey	-0.24	0.58
Weathers	-0.291	
Aurilia	-0.549	0.333
Milton	-0.69

[OldCat]

You have Aurilia playing for the Cardinals in your list.

Now that is the one claim nobody on RZ has used to bash RA so far!

[OldCat]

That "None" guy needs to see more playing time!

[Reds Nd2]

Originally Posted by SteelSD

Just wondering, because it appears that you're venturing into the realm of what you think should have happened rather than sticking with what actually did.

Let me get this straight. It's value relies on game state situations, it's highly subjective, and now, it may not even be accurate. Oh boy. Something to remember when someone actually wants to use this.

[MWM]

With runners on first and second and Eckstein being the hitter, all of the outfield was playing WAY in. The ball bounced off the warning track. Dunn's not slow and the ball was straight over his head, so it's not like he could have taken a better angle. Not too many left fielders getting to that ball, IMO. Giving up a shot like that to David Eckstein was all Danny, at least in my opinion.

Just curious, and this isn't meant to start trouble, but on the Clemens play against the Astros when they scored two runs, did you "debit" Aurilia for the fact that most shortstops probably get to that ball and at the very least hold it to one run? Again, just a curiousity.

I know you readily admitted that a lot of subjectivity will go into it, but it seems that's dangerous that things like "reputation" could play a huge role in assigning blame or credit. Ten different people watching the game could score it differently.

[SteelSD]

Originally Posted by Reds Nd2

Let me get this straight. It's value relies on game state situations, it's highly subjective, and now, it may not even be accurate. Oh boy. Something to remember when someone actually wants to use this.

Well, here's the thing...studes already noted that it's not a player performance metric and it's not a predictive metric (although that line isn't quite that defined when I read his articles).

I think studes knows what it measures. I know that I know what it measures (although now I have realistic accuracy concerns). I just worry that other folks are going to glom onto it claiming that it's supposed to be a way to actually evaluate players. WPA isn't a player evaluation tool, but I can guarantee that there are those who will attempt to use it as one.

Originally Posted by Reds Nd2

Let me get this straight. It's value relies on game state situations, it's highly subjective, and now, it may not even be accurate. Oh boy. Something to remember when someone actually wants to use this.

Sorry, but this just isn't fair. It's a measure of how each player affected the game, given his game-state situation. That's what it is supposed to measure.

It's not "highly subjective." I make calls regarding the pitching/fielding split. That adds an element of subjectivity to the pitching/fielding split. That is all I do. IMO, Dunn was playing Eckstein too shallow, particularly given the base/out situation. so my allocation reflected that.

And, given what it is measuring, it is very accurate. I've found a couple of bugs that affect the third decimal a little bit in total. The bug has been fixed.