Wednesday, April 27, 2005

ERV Boxscore for Wednesday, April 27

Well, one RV stat reveals all too clearly what kind of game this was for the Nats: Cristian Guzman was the ONLY Nat with a positive batting RV, with his whopping 0.03. The silent bats spoiled a great pitching performance by Esteban Loaiza, who was by far the best Nat with his 2.92 RV, which means that without him, this game is a 6-0 blowout, at least.

The ERV Win/Loss stat seems to work very well here. Loaiza is not on the hook for the loss, rather it is shared by the quietest bats (Guillen, Castilla and Vidro -- often our big three) and Ayala, whose meltdown in the ninth sealed the deal. Myers is the clear and sole owner of the ERV win, as his 3.71 RV (which is 77 percent of the max RV a pitcher can get in a game) is the difference between victory and defeat.

Game Number: 22 , April 27, 2005
Final Score:
Philadelphia 3
Washington 0

ERV Win: Myers
ERV Loss: Guillen, Ayala, Castilla & Vidro (What is this?)


Batter Pos PA Bat Field Run Total RV
RollinsJ SS 4 0.52 0.00 0.00 0.52
LoftonK CF 4 (0.29) 0.00 (0.51) (0.80)
AbreuB RF 4 0.47 0.00 0.00 0.47
ThomeJ 1B 4 (0.86) 0.00 0.00 (0.86)
UtleyC 2B 3 (0.68) 0.00 0.00 (0.68)
MichaelsJ LF 4 (0.27) 0.00 0.00 (0.27)
PolancoP 3B 4 0.34 0.00 0.00 0.34
LieberthalM C 4 (0.75) 0.00 0.00 (0.75)
MyersB P 2 (0.35) 0.06 0.00 (0.29)
OffermanJ PH 1 (0.46) 0.00 0.00 (0.46)
BellD PH 1 1.00 0.00 0.00 1.00
Totals 35 (1.33) 0.06 (0.51) (1.79)

Pitcher BFP RV
MyersB 27 3.71
CormierR 3 0.54
WagnerB 3 0.54
Totals 33 4.79


Batter Pos PA Bat Field Run Total RV
WilkersonB 1B 4 (0.28) 0.51 (0.64) (0.41)
ChurchR CF 4 (0.03) 0.00 0.00 (0.03)
VidroJ 2B 4 (0.82) 0.00 0.00 (0.82)
GuillenJ RF 4 (1.05) 0.00 0.00 (1.05)
SledgeT LF 3 (0.03) 0.00 0.00 (0.03)
CastillaV 3B 4 (0.85) 0.00 0.00 (0.85)
SchneiderB C 3 (0.77) 0.00 0.00 (0.77)
GuzmanC SS 3 0.03 0.00 0.00 0.03
LoaizaE P 3 (0.80) 0.00 0.00 (0.80)
BlancoT PH 1 (0.17) 0.00 0.00 (0.17)
Totals 33 (4.79) 0.51 (0.64) (4.91)

Pitcher BFP RV
LoaizaE 29 2.92
EischenJ 2 (0.55)
AyalaL 4 (1.04)
Totals 35 1.33

Notes: Fielding credit given to Brett Myers for his pickoff of Wilkerson in the first. Note that that it was worth so much (0.64) that his subsequent error and balk (which I account for in his fielding RV) do not bring him into the negative RV for the game.

Also, Wilkerson given fielding credit for his throwing Lofton out in the ninth at third base. I judged that a typical play would have resulted in the batter being out at first and runners left on second and third (in other words, Lofton would have been held at third). It was worth 0.51 RV.


At 12:19 PM, Blogger arrScott said...

You need to pick one winner and one loser each game. In this case, Guillen did the most to lose the game, failing to produce twice with runners on and leading off the ninth by striking out looking. No need to share the honors, and it makes sense to stick with the one-winner, one-loser format of traditional scoring. Crediting lots of guys with a win or a loss is redundant with telling us which team won the game.

Your ERV box scores are awesome, by the way. Please keep doing them!

At 12:24 PM, Blogger Brian said...

Quick question about the fielding score. How is it Brett Myers made an error yet still warranted the only non-zero fielding score?

At 12:47 PM, Blogger Brian said...

what is the value of the pickoff vice the loss for the errant throw? also, where does the balk get accounted for?

At 1:20 PM, Blogger DM said...

arrscott, thanks for the kind words about the boxscores. I understand your point about one winner/loser, but I'm looking to shed the conventions of the old system, because, just as the pitcher may not be responsible alone, nor should a single batter. Guillen only cost them 1 run, but they lost by 3. I don't think it is fair to saddle him with the total loss, esp. since the others had nearly the same negative RV.

At 1:25 PM, Blogger DM said...


Good questions. I assign balks, wild pitches and pickoffs to the fielding RV for pitchers. Myers got a plus 0.64 (IIRC) for picking off Wilkerson, which overcame his negative RV for the balk and the error in his throw trying to get Church. Which supports a common theme you see with ERV, that avoiding outs is usually more valuable than gaining bases. Getting picked off really sucks, as you lose a base and gain an out.

At 1:30 PM, Blogger dexys_midnight said...

DM, I think you both have points. What about the idea that it can't be more than two players, and if it is two, they are each assigned a true portion (to the tenth). In this case, based on ERV, you could give Ayala .5 of a loss and Guillen .5.
I find it interesting that Ayala could lose a game that he came into with the Nats already down in the 9th, but based purely on ERV, that does appear to be the case. My guess, DM, is that Ayala's "win value" statistic on this game is far closer to zero than his negative ERV? And if so, should we set forth an ERV boxscore, but assign the win/loss on win value?
Like we've said folks, this is a work in progress, but an exciting time, and I think we are there, just tinkering with the functionality like moving from a 2005 car to the 2006 model.

At 1:32 PM, Blogger DM said...

Also, in the ERV win/loss section, I list the player with the highest RV (or lowest for the loss) first, so for those who want to follow arrscott's rule, you can ignore everyone after the first player.

I am in the process of tallying up ERV wins and losses for the year, and will post them tomorrow morning as part of the Friday Morning Figures. For decisions with multiple players, I'm dividing it in equal shares (e.g. the quartet of Nats yesterday each get .25 of a loss).

At 1:39 PM, Blogger DM said...

dexys, our posts crossed. I'm not comfortable enough with Win Value yet to use it in the ERV W/L (plus I think my boxscore spreadsheet has a bug in it when calculating WV, and I haven't had time to fix it).

And yes, you could be proportional in assigning the decision to multiple players. Turns out from yesterday that all four Nats are between 20-25%. But I don't want it to get too complicated, for now, since I have to do ERV decisions "by hand", and have not yet made it automatic in my Excel workbook. (It took me about six weeks to figure out how to make the boxscore spreadsheet -- I think automated ERV decisions might take longer).

At 1:43 PM, Blogger DM said...

Another thing. Like Dexys said, this is a work in progess. For that reason, we welcome as much feedback as possible, esp. questions like arrscott's suggesting different ways to go about it. Please comment!

At 4:00 PM, Blogger DM said...

Yet another thought. Arrscott is right that listing a lot of players defeats the purpose somewhat. In fact, it seems reasonable that in some cases a loss or win is really a "team" loss not ascribable to any player or groups of players. Maybe Dexys's no more than two players rule (like the assist rule in hockey) is a good compromise.

At 4:21 PM, Blogger DM said...

And another. (Sorry for the stream of consciousness, but that's the way I think - like a baseball game, one thought after another).

I think I need a "rule of plurality" here for the case where no one player has enough RV to make the difference. Something like only those with X% of the run difference. What should it be? 50%? 33%? If no one meets that, then maybe it's a team decision.


Post a Comment

<< Home