Comments on: The New Yankees WAR Pitching Leader This and that about baseball stats. Tue, 16 Jul 2013 17:01:55 +0000 hourly 1 By: Johnny Twisto Sat, 03 Sep 2011 03:31:15 +0000 Also, just checked to see Rivera's career (E)RA based on career Rreps = 792*9/1202 = 5.93. No way the next best reliever on the Yankees is expected to have a 5.93 ERA.

I think you're misunderstanding how it's supposed to work (though I can't blame you, since I'm misunderstanding how the numbers are presented as well).

Even though the closer would be directly replaced in his role by the second best reliever, he is still ostensibly being replaced in the bullpen by a replacement-level reliever. So that's the baseline he's measured against. Then the value of his runs saved are multiplied by the leverage of the innings he pitches. Except that that leverage is reduced because of the chaining effect.

Anyway, yes, a 5.93 RA is much higher than his direct replacement would put up, and it's also higher (I am pretty sure) than a replacement reliever would put up. I just don't know if that number is already adjusted for leverage. I have to try to study this to see if I can figure out what's going on, and then try asking B-R for some guidance.

By: Johnny Twisto Sat, 03 Sep 2011 03:25:16 +0000 Michael, you are correct, but it sort of reads like you're trying to reinvent the wheel. Did you read the article I linked in #47?

As for why to account for chaining only with relievers....Actually, I thought starters were adjusted by their leverage as well (even though their leverages are almost always close to 1.0). But there shouldn't be any chaining there, since their spot would be directly replaced by a AAA SP or the long reliever, who in theory is about replacement level.

Position players don't have any effect over the leverage of the situations they face. Relievers do (indirectly, by virtue of pitching at a level so their manager decides to use them in those spots).

By: Michael E Sullivan Sat, 03 Sep 2011 02:46:00 +0000 Also, why would you account for chaining with relievers but not with starting pitchers or position players.

It's true that if you lose Rivera, you don't replace him with some scrub, but with your next best short reliever, which over his career with the Yankees has usually been an excellent pitcher. But that pitcher must be replaced in long relief with some other pitcher, and so on down the line until you get to the last bullpen guy who may well be replaced at replacement level. Losing your closer started that whole chain, and all of it works out to be sort of like replacing him with a replacement player.

Here's my guess at the justification for taking LI and averaging it with average leverage for adjusting WAR for relievers:

Because of chaining, you aren't putting in a replacement guy into the high leverage innings of a closer, fireman or great setup guy -- those innings get the next best pitcher. Then that pitchers former innings (probably slightly lower leverage) get a next best, etc. down to the guy who actually gets replaced with a replacement pitcher, who probably had average or even less than average leverage. So on average, the innings replaced will be about 1/2 as intense as the closer's actual innings.

That's the only way I could make a coherent argument for the way B-R calculates WAR for relievers.

By: Michael E Sullivan Sat, 03 Sep 2011 02:33:04 +0000 topper: dividing RAR (runs better than replacement) by LI, doesn't make sense unless you want to *penalize* pitchers for facing higher leverage situations and I don't see why you would want to do that.

Whether it makes sense to *credit* relievers who face on average higher leverage situations, with essentially extra innings at the same value depends on what you think about their value.

By: topper009 Fri, 02 Sep 2011 18:06:40 +0000 Also, just checked to see Rivera's career (E)RA based on career Rreps = 792*9/1202 = 5.93.

No way the next best reliever on the Yankees is expected to have a 5.93 ERA. It cant be accounting for chaining.

By: dennis Fri, 02 Sep 2011 18:03:30 +0000 @51.

hartnett was a better catcher then Johnny bench??????????????????????

MHO, Whitey Ford may be oneo the most underrated pitchers in the history of baseball. Certainly, if you take into account ALL of baseball history, he isnt an All time Top Ten pitcher.

But if you look at pitchers in the modern era.....I tink he he s quite high on the list. His winning percentage he is the highest of the modern era....and yes he pitched for the Yankees, but i dont buy the argument that becase he pitched for the Yankees, he should be discounted

He lost two complee years to military service (52 and 53) so there s some credit there. he was carefully spotted by Stengel, pitched out of turn and held for games aginst top pitchers he had seasons with as few as 29 starts and with exception of 1955, his career was typcila yas was equivalent to a top fight starter of current times on 5 man starter staffs.....Like Maddux or Halladay

I fhe had had more starts...he might have have had 15 or more wins....

It was only in 1961 when Ralph Houk pitched him every fourth day did Ford have 20 win seasons (61 and 63),,,,In 61 , he went 25-4, won the Cy Young, and in 63 he was 24-7.Read Jim Boutons hilarious account of how Ford used craft and guile and ellie howard s help to hang on in 64 and 65.

In his HOF induction speech , Mickey Mantle called Ford the true leader of the Yankees....paid that tribute to Ford.....with Ford there...they both went in the same year of 73.

In the modern era, here have only been 13 puitchers who won 100 games or more then they lost, i think it s a telling statistic... Hubell a great pitcher. had a differetntial of 99 and Roy Hallady as of the last start is at 94.....

Ford is on that list with a differential of 130......he went 236 and 106.....

So heres my list of top starters who pitched exlusively in the live ball era. of the top ten, 12 of the 16 won at least 100 more then they lost.

1. Grove
2. Clemens
3. R Johnson
4. Spahn
5. Seaver
6, Feller
7. Martinez
8. Maddux
9. Ford
10 Palmer
11. Koufax

12 Marichal
13. Carlton
14. Gibson
15, Glavine
16. Mussina

I placing thse guys ahead of Ryan, Sutton, Niekro and Perry.....they were all 300 winners but they also lost a LOT.....I ve got Carlton ahead of them, its my preference. .

By: Johnny Dunce Fri, 02 Sep 2011 04:03:14 +0000 As I go through all the guys on the Yankee staff, I see that if you calculate an (E)RA on all their Rreps, the SP are all around 5.7, and the relievers are all around 5.2. (The slight differences among pitchers of the same group are probably based on strength of competition, for which there is supposed to be an adjustment, as well as rounding.) So that half-run must be the difference in replacement level for SP and RP. But since RAR is based on actual runs allowed minus Rrep, this indicates RAR does not include the leverage component (otherwise the RP would show up with very different Rrep ERAs). This is consistent with how the site describes its process, but contrary to what I conclude at #69. So, I am again confused. I have to look at all of this closer when I have some time and then perhaps send a note to the site if I can't figure it out.

By: topper009 Thu, 01 Sep 2011 18:05:43 +0000 Johnny can you post any response you get from the site if they get back to you?

By: Lawrence Azrin Thu, 01 Sep 2011 17:59:36 +0000 @19/ Whitey Ford's all-time ranking -
Looking at his record more closely, I now think Ford definitely belongs in the all-time Top-30, probably higher.

@81/ Cameron - "...Dennis Eckersley ( possibly the worst color commentator ever)..."
Cameron, there's a very large pool of candidates for this "honor"; and while Eck can be annoying at times, I don't think he's close to the worst.

By: Cameron Thu, 01 Sep 2011 12:57:28 +0000 With Mariano at 55 WAR and Dennis Eckersley ( possibly the worst color commentator ever) at 58 WAR and Trevor Hoffman still the all time saves leader an interesting question presents itself, simply if Rivera retired today who would be the greatest reliever ever. My vote goes to Rivera over Eckersley by a small margin.