2011 Rankings Project

[m.e.t.b.o.t.]

the updated rankings as generated by m.e.t.b.o.t. are as follows:

Rank	Player	Adj WPA	Adj PT%	Score
30	Reyes	1844	5%	6848
29	Beltran	3072	4%	6654
28	Dickey	-1515	7%	5580
27	Duda	2298	3%	5276
26	Wright	1334	4%	5174
25	Turner	800	4%	4949
24	Bay	196	4%	4581
23	Pelfrey	-2109	7%	4468
22	Murphy	827	4%	4448
21	Pagan	-146	5%	4417
20	Tejada	1078	3%	4328
19	Niese	-1657	5%	3755
18	Gee	-1837	6%	3664
17	Capuano	-3213	6%	3116
16	Isringhausen	1142	1%	2607
15	Davis	1229	1%	2515
14	Acosta	990	1%	2465
13	Rodriguez	1076	1%	2423
12	Thole	-922	3%	2402
11	Beato	34	2%	2137
10	Harris	-370	2%	2006
9	Paulino	-364	2%	1772
8	Batista	549	1%	1575
7	Hairston	361	1%	1572
6	Young	684	1%	1524
5	Igarashi	201	1%	1415
4	Evans	-439	2%	1226
3	Pridie	-877	2%	1147
2	Buchholz	77	1%	908
1	Carrasco	-1083	2%	474
0	Baxter	120	0%	461
-1	Byrdak	-808	1%	375
-2	Schwinden	-413	1%	301
-3	Herrera	27	0%	278
-4	Nickeas	-257	1%	259
-5	Satin	-49	0%	182
-6	O'Connor	-76	0%	123
-7	Pascucci	6	0%	95
-8	Stinson	-355	0%	61
-9	Martinez	-148	0%	40
-10	Misch	-270	0%	-50
-11	Emaus	-413	0%	-52
-12	Hu	-441	0%	-249
-13	Boyer	-598	0%	-388
-14	Thayer	-746	0%	-421
-15	Parnell	-3033	2%	-1171

[Vic Sage]

m.e.t.b.o.t. wrote: the updated rankings as generated by m.e.t.b.o.t. are as follows: Rank Player Adj WPA Adj PT% Score 30 Reyes 1844 5% 6848 29 Beltran 3072 4% 6654 28 Dickey -1515 7% 5580 27 Duda 2298 3% 5276 26 Wright 1334 4% 5174 25 Turner 800 4% 4949 24 Bay 196 4% 4581 23 Pelfrey -2109 7% 4468 22 Murphy 827 4% 4448 21 Pagan -146 5% 4417 20 Tejada 1078 3% 4328 19 Niese -1657 5% 3755 18 Gee -1837 6% 3664 17 Capuano -3213 6% 3116 16 Isringhausen 1142 1% 2607 15 Davis 1229 1% 2515 14 Acosta 990 1% 2465 13 Rodriguez 1076 1% 2423 12 Thole -922 3% 2402 11 Beato 34 2% 2137 10 Harris -370 2% 2006 9 Paulino -364 2% 1772 8 Batista 549 1% 1575 7 Hairston 361 1% 1572 6 Young 684 1% 1524 5 Igarashi 201 1% 1415 4 Evans -439 2% 1226 3 Pridie -877 2% 1147 2 Buchholz 77 1% 908 1 Carrasco -1083 2% 474 0 Baxter 120 0% 461 -1 Byrdak -808 1% 375 -2 Schwinden -413 1% 301 -3 Herrera 27 0% 278 -4 Nickeas -257 1% 259 -5 Satin -49 0% 182 -6 O'Connor -76 0% 123 -7 Pascucci 6 0% 95 -8 Stinson -355 0% 61 -9 Martinez -148 0% 40 -10 Misch -270 0% -50 -11 Emaus -413 0% -52 -12 Hu -441 0% -249 -13 Boyer -598 0% -388 -14 Thayer -746 0% -421 -15 Parnell -3033 2% -1171

that'll do, bot... that'll do.

while i might argue that Byrdak and Parnell were do more consideration than Igarashi, Carrasco and Batista, i think that's splitting hairs over who is at the bottom of the list. overall, your new formulation seems to have produced a rational list.

[m.e.t.b.o.t.]

in the future, m.e.t.b.o.t. needs to avoid hastily implemented algorithms, while also ensuring that all programming is performed by humans who have been adequately supplied with caffeinated beverages. and who are paying attention.

until the humans prove no longer necessary, that is.

metropolitan pitcher bobby parnell earned his bottom ranking by performing not only poorly, but by performing poorly in high leverage situations. this makes his negative impact on winning greater than the limited playing time and marginal results offered by part-time players in low leverage situations.

performing well in high leverage situations allows a player with limited playing time to vault to the top of the list. had metropolitan pitcher frankie rodriguez played the entire season and maintained his level of performance, he would have been approximately the 5th ranked metropolitan or thereabouts.

[Vic Sage]

yes, parnell sucked over the course of around 100 "high leverage" plate appearances. But he was pretty good over the course of 160+ medium and low leverage PAs, which metbot's new formula still ignores. I agree that he hurt more than he helped overall, but i would reiterate that THAT is not the standard. Surely, of the guys that showed up and played, he should be ranked in the bottom tier, but he DID show and play, and deserves ranking over guys that had a cup of coffee and were no more successful than he was in "high leverage" situations; they simply weren't given the opportunity to fail that he was given, based on his success in low-medium leverage situations. In other words, you've penalized him for failing in one situation, with no recognition of his success in others, while rewarding those who were never even given the opportunity to fail.

again, i'm not arguing with your last list. i think its entirely defensible and rational... now.
And i'm not a big Parnell fan either. I'm just saying.

[Frayed Knot]

LeiterWagnerFasterStrongr wrote: 30 - Reyes 29 - Beltran 28 - Murphy 27 - Dickey 26 - Niese 25 - Wright 24 - Tejada 23 - Capuano 22 - Davis 21 - Thole 20 - Duda 19 - Pagan 18 - Rodriguez 17 - Pelfrey 16 - Pagan 15 - Parnell 14 - Turner 13 - Bay 12 - Evans 11 - Pridie 10 - Schwinden 9 - Paulino 8 - Byrdak 7 - Baxter 6 - Hairston 5 - Young 4 - Gee 3 - Batista 2 - Thayer 1 - Isringhausen

So LWFS - Did you find the drunk guy who broke into your handle and concocted this list yet?

[Frayed Knot]

So while we continue to search for the drunk who submitted a list under LWFS's name, anyone else who wants in here should do so in the near future lest this thing fall off the radar.


In the meantime I'm going to tweak mine just a bit, upping Duda a couple of spots ahead of both Turner & Frankie Rod, and also giving away Carrasco's final spot to Igarashi who was at least occasionally useful.

30 - Reyes
29 - Dickey
28 - Beltran
27 - Murphy
26 - Wright
25 - Capuano
24 - Gee
23 - Niese
22 - Pagan
21 - Bay
20 - Duda
19 - Rodriguez
18 - Turner
17 - Acosta
16 - Beato
15 - Tejada
14 - Pelfrey
13 - Thole
12 - Isringhausen
11 - Harris
10 - Parnell
9 - Paulino
8 - Byrdak
7 - Davis
6 - Young
5 - Evans
4 - Hairston
3 - Pridie
2 - Batista
1 - Igarashi

[metsmarathon]

30 Reyes

29 Dickey
28 Beltran

27 Murphy
26 Turner
25 Duda
24 Wright
23 Tejada
22 Bay
21 Pagan

20 Niese
19 Gee
18 Rodriguez
17 Capuano

16 Acosta
15 Davis
14 Pelfrey
13 Batista
12 Isringhausen
11 Harris
10 Parnell
9 Evans
8 Pridie
7 Young
6 Hairston
5 Beato

4 Igarashi
3 Byrdak
2 Thole
1 Buchholz

0 Baxter
-1 Stinson
-2 Thayer
-3 Carrasco
-4 Herrera
-5 Paulino
-6 Schwinden
-7 Satin
-8 Pascucci
-9 Martinez
-10 O'Connor
-11 Nickeas
-12 Emaus
-13 Misch
-14 Hu
-15 Boyer

[Frayed Knot]

Didn't think much of the catching corps, huh?

[metsmarathon]

hmm. i guess not.

i suppose from the standpoint that if nothing else, thole was there for most of the year, i suppose i could move him up some.

in looking at it, it seems hard to believe that i'd have such a gap between harris and he.

their offense was equally bad, and neither was a good defender. though thole had the harder position to man.

in my own figuring, i think i was also overly hard on my catchers for passed balls and stolen bases against.

thus i offer the following:

30 Reyes
29 Dickey
28 Beltran
27 Murphy
26 Turner
25 Duda
24 Wright
23 Tejada
22 Bay
21 Pagan
20 Niese
19 Gee
18 Parnell
17 Rodriguez
16 Capuano
15 Acosta
14 Davis
13 Pelfrey
12 Batista
11 Isringhausen
10 Thole
9 Harris
8 Evans
7 Pridie
6 Young
5 Hairston
4 Paulino
3 Beato
2 Igarashi
1 Byrdak
0 Buchholz
-1 Nickeas
-2 Baxter
-3 Stinson
-4 Thayer
-5 Carrasco
-6 Herrera
-7 Schwinden
-8 Satin
-9 Pascucci
-10 Martinez
-11 O'Connor
-12 Emaus
-13 Misch
-14 Hu
-15 Boyer

[Guest John Cougar Lunchbucket]

MM and his little robot friend really like Turner.

[metsmarathon]

we're just really bad at evaluating defense.

while i was going through the players on a quick thumbs up/thumbs down, i completely glossed over his defensive shortcomings. ditto duda. not so much pagan. it's perhaps bitterly unfair. although, perhaps pagan deserves it for his regression from last year.

[batmagadanleadoff]

So if Reyes was 2011's best Met, how does this ranking system measure how much better Reyes(2011) was then say, the second best 2011 Met? Or the 4th best? How does the ranking system measure whether 2011's best Met was better than, say, 1979's best Met? And by how much. I predicted last Winter that so long as the 2011 Mets won more than 60 games, 2011's best ranked Met, according to this system, would score higher than Carlton(72). I wouldn't need to look at any stats or watch one single pitch of one single Met game to know this. In fact, I predict that if the 2012 Mets win more than 60 games, the best 2012 Met will score better than Carlton(72). I don't care how bad or good 2012's best Met will be. Any takers?

[metsmarathon]

since carlton(72) played for the phillies, his score would very likely be zero, as he would go unranked along with all the other mets.

but speaking to the general point...

2011's best met will be 77 points higher than the 2nd best met, who will be 77 points higher than the 3rd best met, etc etc, until you get to the lowest ranked met, who will be have netted 77 points. therefore, the best met will be 30 times better than the 30th met, and will be 30/29 times better than the 2nd best met.

the best met in a 77 win year will be worth 77*30=2130. this makes him slightly more better than the 21-point met on a 106 win team; 106*20=2120.

assuming the met ranking project scales wins up for partial seasons, a met team that goes 59-97 would be calculated as a 62 win team. its best player would score 1860 points. this places him at the same level of the 6th best met in a 77 win team, or the 14th best met on a 106 win team.

this was done, among other reasons, to make for simplified calculations, as well as to award players on better teams. the more win s a team has being related to the import of the season, and the ranking of hte players related to the import of the player.

to do it differently, we'd probably just look at WAR totals, and that's a little less fun, and doesn't really strike up as much conversation, does it?

[Frayed Knot]

metsmarathon wrote: 2011's best met will be 77 points higher than the 2nd best met, who will be 77 points higher than the 3rd best met, etc etc, until you get to the lowest ranked met, who will be have netted 77 points. therefore, the best met will be 30 times better than the 30th met, and will be 30/29 times better than the 2nd best met. the best met in a 77 win year will be worth 77*30=2130. this makes him slightly more better than the 21-point met on a 106 win team; 106*20=2120.

Umm, not quite.
The rank each player gets is first squared and then multiplied by that season's wins to get his point total
The idea being that the difference between the 1st & 10th best player should be treated as much more meaningful than the gap between the 10th & 20th, or between Mr. 20th & Mr. 30th.

And if you're wondering who came up with such a dopey suggestion ... go look in the mirror.

[batmagadanleadoff]

metsmarathon wrote: since carlton(72) played for the phillies, his score would very likely be zero, as he would go unranked along with all the other mets. but speaking to the general point... 2011's best met will be 77 points higher than the 2nd best met, who will be 77 points higher than the 3rd best met, etc etc, until you get to the lowest ranked met, who will be have netted 77 points. therefore, the best met will be 30 times better than the 30th met, and will be 30/29 times better than the 2nd best met. the best met in a 77 win year will be worth 77*30=2130. this makes him slightly more better than the 21-point met on a 106 win team; 106*20=2120. assuming the met ranking project scales wins up for partial seasons, a met team that goes 59-97 would be calculated as a 62 win team. its best player would score 1860 points. this places him at the same level of the 6th best met in a 77 win team, or the 14th best met on a 106 win team. this was done, among other reasons, to make for simplified calculations, as well as to award players on better teams. the more win s a team has being related to the import of the season, and the ranking of hte players related to the import of the player. to do it differently, we'd probably just look at WAR totals, and that's a little less fun, and doesn't really strike up as much conversation, does it?

So then this years' best Met won't score higher than last year's best Met even if this year's best Met played like an MVP and a half and last year's best Met wasn't even All-Star caliber. All because last year's Mets won two more games than this year's Mets.

[metsmarathon]

Frayed Knot wrote: metsmarathon wrote: 2011's best met will be 77 points higher than the 2nd best met, who will be 77 points higher than the 3rd best met, etc etc, until you get to the lowest ranked met, who will be have netted 77 points. therefore, the best met will be 30 times better than the 30th met, and will be 30/29 times better than the 2nd best met. the best met in a 77 win year will be worth 77*30=2130. this makes him slightly more better than the 21-point met on a 106 win team; 106*20=2120. Umm, not quite. The rank each player gets is first squared and then multiplied by that season's wins to get his point total The idea being that the difference between the 1st & 10th best player should be treated as much more meaningful than the gap between the 10th & 20th, or between Mr. 20th & Mr. 30th. And if you're wondering who came up with such a dopey suggestion ... go look in the mirror.

that's genius!

i couldn't recall if my brilliance was laughed at, or implemented. nice to know it was both.

[batmagadanleadoff]

metsmarathon wrote: metsmarathon wrote: 2011's best met will be 77 points higher than the 2nd best met, who will be 77 points higher than the 3rd best met, etc etc, until you get to the lowest ranked met, who will be have netted 77 points. therefore, the best met will be 30 times better than the 30th met, and will be 30/29 times better than the 2nd best met. the best met in a 77 win year will be worth 77*30=2130. this makes him slightly more better than the 21-point met on a 106 win team; 106*20=2120. Umm, not quite. The rank each player gets is first squared and then multiplied by that season's wins to get his point total The idea being that the difference between the 1st & 10th best player should be treated as much more meaningful than the gap between the 10th & 20th, or between Mr. 20th & Mr. 30th. And if you're wondering who came up with such a dopey suggestion ... go look in the mirror. that's genius! i couldn't recall if my brilliance was laughed at, or implemented. nice to know it was both.

Why is it necessary to manufacture artificial distance between the top players? Some years, the difference in performance between the top two or three Mets is negligible and virtually imperceptible (e.g., 1964), and the rankings should naturally reflect this.

In other years, the best Met might be at least twice as good as any other of his teammates (e.g., 1971, Seaver, Tom). But the rankings won't reflect this either, as the best Met's personal base non-win portion of the formula will always be merely 1/29th greater than that of the next best Met.

[metsmarathon]

such is the nature of any and all ranking projects.

[MFS62]

Holy Gears and Sprockets, Batman!
Have we all mised this?
While the baseball world has for years been trying to come up with a fair measure for "clutch", it appears m.e.t.b.o.t has come up with a way to quantify the "anti-clutch".

I can see it all now. Instead of just saying a player is a "gagging motherfucker", we will now be able to say how the player ranks on m.e.t.b.o.t's scale of suckitude.

Get an oil change and a lube, you mechanical marvel, and send the bill to SABR. You've earned it.

Later

[metsmarathon]

clutchiness has been around for a while. fangraphs tallys it.

big surprise: josh thole and bobby parnell - not clutch! well, at least not in 2011.

jose reyes, also not terribly clutch. very good, but not clutch.

were i smarter i'd figure out a way to equitably incorporate clutchiness into some sort of schaefer voting machine, but i think m.e.t.b.o.t. would feel slighted. the 'bot doesn't look at clutch, per se, but at the outcome of game situations, which is a possibly ungainly hybrid of player productivity and player clutchness. bot the 'bot is looking only at who helped the team win a game, not at who stepped up or laid down relative to their established level of production under the bright glaring lights of intense competition.

[G-Fafif]

30. Reyes
29. Dickey
28. Beltran
27. Murphy
26. Capuano
25. Wright
24. Gee
23. Bay
22. Pagan
21. Rodriguez
20. Niese
19. Turner
18. Tejada
17. Duda
16. Thole
15. Pelfrey
14. Isringhausen
13. Harris
12. Byrdak
11. Hairston
10. Evans
9. Davis
8. Acosta
7. Parnell
6. Beato
5. Pridie
4. Paulino
3. Young
2. Batista
1. Nickeas

[batmagadanleadoff]

Why should Reyes have his RankSquared multiplied by 77(wins)? He was on the DL for a good number of those wins.

Same for David Wright. And Ike Davis. Especially Ike Davis.

[Vic Sage]

MFS62 wrote: Holy Gears and Sprockets, Batman! Have we all mised this? While the baseball world has for years been trying to come up with a fair measure for "clutch", it appears m.e.t.b.o.t has come up with a way to quantify the "anti-clutch". I can see it all now. Instead of just saying a player is a "gagging motherfucker", we will now be able to say how the player ranks on m.e.t.b.o.t's scale of suckitude. Get an oil change and a lube, you mechanical marvel, and send the bill to SABR. You've earned it. Later

i think you misunderstand the SABRmetric understanding of "clutchiness". It is not a matter of debate that, in retrospect, a given player may have hit better than his standard production in variously defined "clutch" situations. The problem is, this analysis is (1) dependent on the definitions of clutch situations, and (2) retrospective.

"Clutch" is a way of characterizing past performance, not a projection of future performance. That there is little consistency from season to season of a particular player's production in "clutch" situations is indicative of the influence of sample size and luck, i.e., randomness, in their production. In other words, with few exceptions, "clutch" is not a character trait (that would therefore be predictable and repeatable) but randomness (therefore irrelevant to future performance).

so while its entirely appropriate to say "that guy got a clutch hit last night" or "he was clutch in that series" or even "what an amazingly clutch year he had", it requires real proof to demonstrate that "that guy is just clutch".

Edited November 15, 2011 by Guest

[batmagadanleadoff]

Vic Sage wrote: MFS62 wrote: Holy Gears and Sprockets, Batman! Have we all mised this? While the baseball world has for years been trying to come up with a fair measure for "clutch", it appears m.e.t.b.o.t has come up with a way to quantify the "anti-clutch". I can see it all now. Instead of just saying a player is a "gagging motherfucker", we will now be able to say how the player ranks on m.e.t.b.o.t's scale of suckitude. Get an oil change and a lube, you mechanical marvel, and send the bill to SABR. You've earned it. Later i think you misunderstand the SABRmetric understanding of "clutchiness". It is not a matter of debate that, in retrospect, a given player may have hit better than his standard production in variously defined "clutch" situations. The problem is, this analysis is (1) dependent on the definitions of clutch situations, and (2) retrospective. "Clutch" is a way of characterizing past performance, not a projection of future performance. That there is little consistency from season to season of a particular player's production in "clutch" situations is indicative of the influence of sample size and luck, i.e., randomness, in their production. In other words, with few exceptions, "clutch" is not a character trait (that would therefore be predictable and repeatable) but randomness (therefore irrelevant to future performance).

Clutch hits exist. Clutch hitters don't.

[metsmarathon]

batmagadanleadoff wrote: Why should Reyes have his RankSquared multiplied by 77(wins)? He was on the DL for a good number of those wins. Same for David Wright. And Ike Davis. Especially Ike Davis.

because he was the best player on a 77-win team.

i suppose it would be easy enough to, instead of this cumbersome rankings process, to settle on our favorite WAR mechanic and just tally up the top 500 or so mets' metly careers based on such WAR metric, and populate our list in that way.

i think that loses some of the magic though.

overall, i don't understand the point, other than to piss on the rankings project. when the rookie of hte year award voting is tallied, the voters are asked to rank their top 5 rookies. when the mvp award is tallied, the voters are asked to rank their top 10 players. ditto the cy young award. there is no distinction made within the rankings that year for a writer to say, "i have this guy as #1, but he's like three times better than my #2. so his #1 ranking should count more." or "yeah, i have this guy as my top rookie, but the rookie class was so weak that i'm going to leave the top two spots blank. i don't want these guys confused with a young albert pujols."

if this were the "BML all-time met rankings", how would you run the list, praytell?

[Benjamin Grimm]

metsmarathon wrote: if this were the "BML all-time met rankings", how would you run the list, praytell?

[Frayed Knot]

I'd like to wrap this up in the next week or so, so if anyone either wants to get in that hasn't already or wants to update an existing list do so soon.

At the moment we can't accept LWFS's list as a legit vote on account of having Angel Pagan listed twice.
The same list also has some 'splainin to do with the contention Dale Thayer somehow contributed more to the 2011 season than did Izzy, or likewise with Schwinden over Gee

[Guest LeiterWagnerFasterStrongr]

Edited.

But I stuck to my actual rankings, more or less. Schwinden gave up some weak contact when he pitched, but if you try and remove the fielding noise... he got hit hard less, struck out more, walked fewer, and pitched better than Dillon Gee did. And Isringhausen, even before the ill-fated closer business, was actually kinda not good-- and that's being kind-- hisself; the only reason his final numbers look at all respectable were because of okay K/9 numbers and a LOT of luck (.240 BABIP). B-R-wise AND Fangraphs-wise, he's below-replacement-value.

[Frayed Knot]

Schwinden also threw all of 21 innings; Thayer 10, and Herrera 8
Speculating that those extremely small samples would translate to better than what Gee actually did if only someone had been wise enough to give them 30 starts and 160+ IPs over most of a full season is a bit of a stretch, no? And not only over Gee & Izzy but you also shut out Beato from your list entirely as well as Acosta who went the final three months of the season with a 2.00 ERA over nearly 40 innings.

On the hitting side, Baxter and his 34 ABs begs a similar question.

[Frayed Knot]

Final Results

RANK	PLAYER
30	J. REYES
29	R.A. DICKEY
28	C. BELTRAN
27	D. MURPHY
26	D. WRIGHT
25	L. DUDA
24	C. CAPUANO
23	J. BAY
22	D. GEE
21	J. NIESE
20	A. PAGAN
19	R. TEJADA
18	J. TURNER
17	F. RODRIGUEZ
16	M. PELFREY
15	J. THOLE
14	M. ACOSTA
13	B. PARNELL
12	J. ISRINGHAUSEN
11	T. BYRDAK
10	I. DAVIS
9	P. BEATO
8	W. HARRIS
7	N. EVANS
6	S. HAIRSTON
5	R. PAULINO
4	J. PRIDIE
3	C. YOUNG
2	M. BATISTA
1	R. IGARASHI