The ongoing Pythagorean deficit

[Frayed Knot]

Perhaps a more scientific analysis will prove this to be less the case than I'm imagining, but these late-inning pen meltdowns--the ones that not only turn leads or ties into deficits but often small deficits into LARGE ones--have got to be the biggest factor in the RS/RA gap that our W/L record has, from very early in the season to the present, so far managed to overcome*.
There have been a (relative) ton of these this year and I'm hard pressed to think of any that have gone the other way.

It's like the pen has adopted the George Costanza method, the one where he doesn't merely want to get fired, he wants to go in such a way where years after he's gone people will remember the blaze of glory that he went out to as they say; "Now THAT guy got fired!"
Our pen, they don't just let small leads slip away via walks and singles that result in one-run losses. When they blow leads, they BLOW leads and make sure that no comeback is even contemplated!



* Their record currently sits at five games better than the Pythagorean projection as the NYM staff sports a 5.80 9th inning ERA and a 6.00 in extras - and even those numbers were looking better recently seeing as how, prior to last night, they were coming off a Johan CG shutout and three straight Frank-Frank outings where he gave up a total of one hit.

[Ceetar]

It's the bullpens fault. Mostly Manny Acosta but other guys are certainly contributing.

I've run the numbers and even wrote it up a bit but never edited it and the numbers always change, via a game being played, before I do.

It's certainly been very flukey though. I mean, the Mets won pretty succinctly against the Padres. and one game undoes all the good work they did in terms of run differential.

[nymr83]

Fuck this Python guy, we're winning!

The bullpen does seem to get lit on fire a lot though.

And the offense has a tendency to take a lead but then not really expand on it.

[Benjamin Grimm]

Through 51 games, the Mets, when they win, win by an average margin of 2.785714 runs. And when they lose, they lose by an average of 4.434783 runs.

[Ceetar]

via @tpgmets (good follow btw)

Their 41 shutdowns are t-5th best in the NL & t-9th best in MLB. Unfortunately, their 30 meltdowns are 2nd worst in the NL & MLB.

So they've really been more high variance than outright bad (although they have not been a net asset).

[Edgy MD]

Even when they're winning by 2.785714 runs, we really shouldn't feel good about it until they tack on 1.649069 more,

[batmagadanleadoff]

In a warped sort of way, I kinda like that the fringe end of the Mets pen could be so toxic so as to frequently fire-start a close game into oblivion. Because it shows Collins, and any other stupid manager that's hung up on this 9th inning closer guy bullshit that every inning counts. And a run scored in the 7th counts just as much as a run scored in the 9th.

[Vic Sage]

yes, cuz Collins seems to be the kind of manager likely to embrace a radical reimagining of bullpen usage.

[Benjamin Grimm]

Yeah, but I just don't see that it's likely that the conventional wisdom will be overturned any time soon.

[batmagadanleadoff]

Well if Collins doesn't get it, then he deserves the Mets woeful run differential. And if that run differential is not flukish, then the Mets will eventually drop from this pennant race like a sinking stone.

[Vic Sage]

batmagadanleadoff wrote: Well if Collins doesn't get it, then he deserves the Mets woeful run differential. And if that run differential is not flukish, then the Mets will eventually drop from this pennant race like a sinking stone.

this is my thesis, in a nutshell.

[Edgy MD]

Well, it's not like LaRussa didn't win three championships, and Whitey Herzog perhaps should have. There are other virtue to a manager. This one just seems like a brass ring for any manager to grasp.

[Ceetar]

The data doesn't suggest this is a worse method of pitcher use either.

Whether or not it's fluky, it's not relevant to the differential going forward or predictive of any falloff in record, especially since the Mets won't be able to repeat the #1 cause of the blowouts: pitching Manny Acosta.

[Guest The Second Spitter]

Pythagorean Exceptation is not normally distributed (not for wins-loss anyway). There's some evidence suggesting that large differentials early in the season to pervade the course of the season. Iirc the 86 Mets are an example of this.

[Elster88]

+8 today.

[Edgy MD]

A little post-blowout update.

Date	June 1
Runs For	228
Runs Against	244
Pyth. W %	.466
Winning %	.558
Difference	-.092

[Edgy MD]

Always good to deceptively check back after a blowout.

Date	June 13
Runs For	273
Runs Against	283
Pyth. Win. %	.482
Act. Win. %	.532
Difference	-.050

[Ceetar]

I wrote about this some yesterday. http://www.ceetar.com/optimisticmetsfan/2012/06/12/run-differential-and-the-mets/

The Mets are now +10 swapping Acosta for a 2011 version of Acosta.

53:39 over the last 11 games has them at what should be a .637 winning percentage, or 7-4 when they're actually 5-6.

swapping out Acosta (which I do to make the point that Acosta will not continue giving up runs at the rate that he did by virtue of either improving or not being here) gives the Mets a Pythagorean record of 32-30 which is actually only one game worse than the 33-29 they're at.

[Vic Sage]

I wrote about this some yesterday. http://www.ceetar.com/optimisticmetsfan/2012/06/12/run-differential-and-the-mets/ The Mets are now +10 swapping Acosta for a 2011 version of Acosta. 53:39 over the last 11 games has them at what should be a .637 winning percentage, or 7-4 when they're actually 5-6. swapping out Acosta (which I do to make the point that Acosta will not continue giving up runs at the rate that he did by virtue of either improving or not being here) gives the Mets a Pythagorean record of 32-30 which is actually only one game worse than the 33-29 they're at.

but you don't get to swap him out for a mythical Acosta from the past; you get to swap him out for the Schwinden (12.86 era) and Elvin Ramirez (10.80 era) that replaced him. There's always a guy at the back of the pen who sucks, on every team with a negative run differential. that doesn't mean much to me.

[Ceetar]

Vic Sage wrote: but you don't get to swap him out for a mythical Acosta from the past; you get to swap him out for the Schwinden (12.86 era) and Elvin Ramirez (10.80 era) that replaced him. There's always a guy at the back of the pen who sucks, on every team with a negative run differential. that doesn't mean much to me.

Elvin Ramirez is better than that and Schwinden threw...what, 1 inning? (And they're still better. That's how bad Manny was)

but that's not the point. The data was created with Manny Acosta as an input. He's no longer 'putting. Therefore it's hard to say the Mets will suck because run differential says so, when a big reason for that is gone.

If you want to tell me the guy that gets Acosta's innings will suck, well sure, that's a possibility. I disagree though, I think Ramirez is good, I think Mejia and Beato are pretty good as well. Acosta was on pace to give up like 60 more runs. That's a lot.

[batmagadanleadoff]

Swapping Acosta for MikePiazza2000 would also improve the run differential.

[Edgy MD]

Here's a thought: Let's swap Acosta for R.A. Dickey

Date	June 13 PM
Runs For	282
Runs Against	284
Pyth. Win. %	.496
Act. Win. %	.540
Difference	-.044

[Ceetar]

gimme a 3 run win tomorrow!

[Guest Swan Swan H]

...is, as of this moment, a surplus.

[Edgy MD]

Well, not really. While we have more runs gained than yielded, a Pythag. surplus will come when our Pythagorean winning percentage exceeds our actual winning percentage.

[Edgy MD]

In the black, gents.

Date	June 14
Runs For	291
Runs Against	290
Pyth. Win. %	.502
Act. Win. %	.547
Difference	-.045

The difference falls in the other direction, though. Big whoop.

[Edgy MD]

Date	June 18
Runs For	301
Runs Against	304
Pyth. Win. %	.495
Act. Win. %	.529
Difference	-.034
Pyth. Rec.	34-34
Act. Rec.	36-32
Difference	2 games

[Edgy MD]

Definitely living within the margin of error.

Date	June 19
Runs For	306
Runs Against	304
Pyth. Win. %	.503
Act. Win. %	.536
Difference	-.033
Pyth. Rec.	35-34
Act. Rec.	37-32
Difference	2 games

[duan]

Edgy DC wrote: Definitely living within the margin of error. Date June 19 Runs For 306 Runs Against 304 Pyth. Win. % .503 Act. Win. % .536 Difference -.033 Pyth. Rec. 35-34 Act. Rec. 37-32 Difference 2 games

that's called the Manny Acosta deficit

[Edgy MD]

Acosta's gone, but his spirit haunts the pen.

Date	June 20
Runs For	310
Runs Against	307
Pyth. Win. %	.505
Act. Win. %	.543
Difference	-.038
Pyth. Rec.	35-35
Act. Rec.	38-32
Difference	=#FF0000]3 games