Master Index of Archived Threads


The Value of Pitching Wins (Split from Zee-Toe)

iramets Jan 02 2007 02:24 PM

I'm sure someone has done a study of this (with so much money at stake, how could someone have not?) but I haven't seen it. Here's an [url=http://www.nytimes.com/2007/01/02/sports/baseball/02chass.html]article[/url] on the cost of some high-priced pitching--my idea would be simply to study the results of the 10 or 20 most expensive, longest term contracts for pitchers over the past 10 years or so, and figure out what the average return has been. Computing how much a single victory, for example, cost on average for these contracts would be appalling, i'd think. Just for fun, care to guess (before we learn the facts) what the cost of a single victory would be for the ten most expensive pitching contracts over the last ten years?

Nymr83 Jan 02 2007 05:06 PM

since wins are as much dependant on run support as they are on how the guy pitches i'd find that to be a pointless exercise.

iramets Jan 02 2007 10:03 PM

Nymr83 wrote: since wins are as much dependant on run support as they are on how the guy pitches i'd find that to be a pointless exercise. Only in small groups. It's idiotic to say that a guy who won 16 games is twice the pitcher as a guy who won 8 in a particular season, but in big samples such as we're discussing here, wins become significant indicators of success. If when Pedro signed a four-year deal, I'd told you that he would win over 70 games in his four years, you wouldn;t have told me "That's meaningless. He could have great run support, and win 70 and I'd say that he sucked." You would have said "From your mouth to Gods ear."

iramets Jan 02 2007 11:24 PM

Do you really want to argue that the ten ERA-qualifying pitchers in the NL with the most wins were no better than the ten ERA-qualifying pitchers with the fewest wins last season? If not, why not?

Nymr83 Jan 03 2007 12:51 AM

Wins are a very poor indicator of individual success or talent. Here are the top 10 winners in 2006 along with the top 10 in other pitching categories. Wins- Santana, Wang, Garland, Garcia, Johnson, Rogers, Verlander, Beckett, Blanton, Halladay. ERA- Santana, Oswalt, Carpenter, Webb, Halladay, Sabathia, Arroyo, Zambrano, Young, Smoltz WHIP- Santana, Carpenter, Mussina, Halladay, Webb, Young, Bush, Oswalt, Sabathia, Arroyo IP- Arroyo, Webb, Harang, Santana, Smoltz, Willis, Haren, Carpenter, Capuano, Zito K/BB ratio- Schilling, Santana, Mussina, Bush, Oswalt, Carpenter, Lieber, Haren, Sabathia, Halladay I'd take the 10 guys on any of these other lists over the ten on the wins list. not suprisingly, the Wins list is the one most dominated by players on winning teams, thats because WINS IS A TEAM STAT and not very indicative of an individual's performance.

iramets Jan 03 2007 08:58 AM

Oh, for fuck's sake. I didn't claim that wins were the best measure of ability, just that they are NOT without significance. Big difference between "over-rated" and "sucks." Now you've done half the work I asked you to do. If you would compare the list of ERA qualifying Wins leaders to the 10 ERA-qualifying pitchers with the fewest wins, we'd be able to see how idiotic it is to claim that Wins are entirely meaningless (and if you'd get the NL leaders, I think we Mets fans would be better able to judge their genuine abilities).

MFS62 Jan 03 2007 09:28 AM

iramets wrote: Just for fun, care to guess (before we learn the facts) what the cost of a single victory would be for the ten most expensive pitching contracts over the last ten years? After the Zito signing, one of the NY papers (Post?) had a side bar that talked about the cost per win of the two "worst" (their words) expensive long term pitcher contracts - Kevin Brown and Mike Hampton. After they signed those contracts, they combined for far fewer than (IIRC) 50 wins. Those numbers could skew the totals, but I guess that's your point - its a crapshoot. Just thinkin'. If that award for "best pitcher" is named after Cy Young, because he is the all-time leader in wins, why not name the award for the "worst" pitcher after him, too? After all, he is also the all-time leader in losses. Ahhhh, baseball. I love this game. EDIT: Who would you name the award for "worst pitcher" after? Later

Nymr83 Jan 03 2007 03:24 PM

the award is named for Young because he died shortly before it was first awarded, it should have been named for Christy Matthewson.

Edgy DC Jan 03 2007 03:29 PM

It should have been named for Charles Radbourne.

metsmarathon Jan 03 2007 03:49 PM

iramets wrote: If when Pedro signed a four-year deal, I'd told you that he would win over 70 games in his four years, you wouldn't have told me "That's meaningless. He could have great run support, and win 70 and I'd say that he sucked." You would have said "From your mouth to Gods ear." well, the point there is not that those 70 wins would necessarily mean that pedro had done well, but rather that the mets had done well in winning those games. by the same token, if you had been told prior to this season that pedro would win only nine games, you'd likely think that our season was sunk, mired in mediocrity or worse. but more to the point - here ya go top 10 ERA bottom 10 ERA Smoltz 3.49 Maholm 4.76 Lowe 3.63 Jennings 3.78 Penny 4.33 Cook 4.23 Zambrano 3.41 Duke 4.47 Harang 3.76 Morris 4.98 Webb 3.10 Hensley 3.71 Maddux 4.20 Young 3.46 Glavine 3.82 Capuano 4.03 Carpenter 3.09 Peavy 4.09 Oswalt 2.98 Fogg 5.49 Ortiz 5.57 Davis 4.91 Schmidt 3.59 Batista 4.58 AVG 3.58 AVG 4.40 less than a run difference in ERA. also worth noting, is that if you plot wins versus ERA you find that in the NL last year, a 1-run increase in your ERA would on average result in a decrease of less than 1.25 wins hardly a huge difference, and therefore, not a great indicator of the quality of a pitcher. but, on the whole, a pitcher with more wins is likely to have pitched better than the pitcher with fewer wins.

iramets Jan 03 2007 06:36 PM

Is there a person on this planet who would consider trading the staff on the left for the staff on the right?

metsmarathon Jan 03 2007 07:38 PM

is there a person on the planet who thinks that because a pitcher is on the left, he must therefore be better than any pitcher on the right? or that because a pitcher is on the right, he must therefore be worse than any pitcher on the left? because the whole point is that while a pitcher who has more wins is likely to be better than a pitcher with less wins, it isn't necessarily the case, even when the difference in wins is fairly significant. or to put it another way, you'd surely take chris young or jason schmidt or even jason jennings over some of the pitchers on the left, right? wins may be used to indicate individual success or talent. however, they do not perform this task as well as other readily available statistics, and are also more likely to be influenced by factors outside the domain of the subject pitcher. so while, fine, looking at dollars per win wouldn't be meaningless, it also wouldn't be particularly meaningful. it would just be easy. and that's the kind of thing that gives statistics a bad name.

iramets Jan 03 2007 08:02 PM

As a group, the pitchers on the left are miles more accomplished than the group on the right. Can you find an exception? Is there a single pitcher on the left less talented than a single pitcher on the right? Sure. But overall, it's not even close. What gives stats a bad name is a pigheaded refjection of easy--to-understand stats (Ws. HRs. RBIs) in favor of arcane impossible-to-understand without a slide rule and a sextant in reaching a general conclusion. W/$ ratio is fun and meaningful. What more can you ask for? You're totally full of shit if you think you'd have responded to Pedro winning 70 games as merely significant of the team's success. Leaking out your ears full of shit. "Yes, I'm taking 70 wins from Pedro in the next 4 years because that means he'll be healthy and get a lot of run support. But 70 wins doesn't really speak to the quality of his pitching at all." Yeah, right. That's the ticket.

Nymr83 Jan 03 2007 09:06 PM

]What gives stats a bad name is a pigheaded refjection of easy--to-understand stats (Ws. HRs. RBIs) in favor of arcane impossible-to-understand without a slide rule and a sextant in reaching a general conclusion. in other words, we should only use numbers that are dumbed down enough for people like you? got it. as for Pedro and 70 wins over 4 years its still a TEAM stat, Randy Johnson won 17 games this year and pitched like a total piece of shit, but by your logic he must have done great.

metsmarathon Jan 03 2007 09:50 PM

last year, adam dunn hit more home runs than david wright. who was the better hitter? last year, ryan zimmerman had more rbi than jose reyes. who was the better hitter? last year, brad penny had more wins than jason schmidt. who was the better pitcher? the easy to understand stats give you easy to understand answers. who hit more home runs? adam dunn. who drove in more runners? ryan zimmerman. who won more games? brad penny. who was the better hitter? pitcher? the easy stats don't tell enough to answer the question. 70 wins from pedro would indicate that he pitched better as opposed to worse. most likely, id' be thrilled with 70 wins from pedro. but there's a difference between "doesn't really at all" and "doesn't necessarily" in 2005, pedro won 15 games. was that more indicative of the quality of his season (livan hernandez also won 15 games), or was his 2.82 ERA? heck, roger clemens won 13 games with a 1.87 ERA. which is more indicative of his performance? would you rather have pedro pitch 4 years, throw 200 innings in each of them, and win 70 games with a 3.99 ERA, or win 50 games with a 2.99 ERA? you're full of shit if you take the 70 games because it means pedro pitched better. now, i've already conceded that a pitcher with more wins is likely to have pitched better than a pitcher with less wins. are you ready to tell me that brad penny, ryan zimmerman, and adam dunn had better years than jason schmidt, jose reyes, and david wright? getting back to the notion of wether or not $/win is meaningful and fun, i do think its meaningful, but not because it means the pitcher did better or worse, but because the team did better that signed a pitcher to a better $/win ratio. meaning that a bad team that overpays a pitcher will not score enough runs to drive his win total up, and therefore his $/W down, and is therefore getting a worse return on their investment than the team that scores enough runs to drive the wins up, thereby reducing hte amount they are "paying the pitcher for each win" it has nothing to do, really, with how well the pitcher performs with that team, but rather how well the team performs with that pitcher. and that's really what you'd be finding out with the $/W ratio, and along with that info, you'd better identify which GM's more wisely spent their money, as opposed to the GM who wasted his money by overpaying a pitcher who, no matter how well he pitches, just isn't going to win many games. case in point, again, roger clemens, 2005. he won 13 games, and was paid, what, $18M? that's $13.4M per win. but clemens pitched incredibly well! too bad the team threw their money away...

metsmarathon Jan 03 2007 09:52 PM

iramets wrote: As a group, the pitchers on the left are miles more accomplished than the group on the right. Can you find an exception? Is there a single pitcher on the left less talented than a single pitcher on the right? Sure. this is a point that i make myself. iramets wrote: But overall, it's not even close. sure it is. there's less than a run difference.

iramets Jan 04 2007 05:38 AM

metsmarathon wrote: ="iramets"]As a group, the pitchers on the left are miles more accomplished than the group on the right. Can you find an exception? Is there a single pitcher on the left less talented than a single pitcher on the right? Sure. this is a point that i make myself. ="iramets"]But overall, it's not even close. sure it is. there's less than a run difference. In the group we've selected (ERA qualifiers) it's practically impossible (other than by selecting specifically for it) two groups of pitchers who will have a greater ERA difference than we have here. You dont get to qualify for the ERA title without having a half-decent ERA. There are no pitchers with a 6.75 ERA who pitched 162 innings last year, so a run difference is tremendous in this context. Hell, a run difference in ERA is tremendous in any context, and extremely valuable in a negotiating context. Back to my main argument here, the notion that you can pick one or two exceptions to the pattern I'm showing is not the point, which is that as a group, the highest paid pitchers will show a disappointingly low return in terms of wins. In terms of $/W, you may be getting very little when you sign a Zito or a Randy Johnson or a Roger Clemens. I'm pretty sure it's a matter of diminishing returns. That is, because you pay so little for a AAA pitcher to step into your rotation, even if he doesnt win a lot of games, those games are very cheap in terms of $/W, while a 16-mil pitcher is bound to cost you a mil per win even if he has a pretty good year, which is to say that a Brian Bannister cost the Mets "only" a litle under 200,000 for each of his wins, while each Glavine win cost them about five times as much, and they consider that Glavine had a far more valuable year, which he did. I think my main objection to paying top-dollar salaries when you have other options is the lack of flexibility it gives you. If you're paying Johnson umpty-umpth mil to pitch for you, youre committed t o pitching him no matter how badly he pitches. You're allowing his salary IOW to make baseball decisions for you. In fact, my main objection here is to allow ANY non-baseball decisions to take the heat off you for making a bad call, whether it's pitching Johnson no matter how shitty he looks or putting Wagner in a spot where he doesn't belong by any indicator other than his role (and salary) as "closer." Management can always say "Hey, what choice did I have?" when in fact they've got every choice but have restructured the game so they seem to have none.

Nymr83 Jan 04 2007 03:04 PM

]In the group we've selected (ERA qualifiers) it's practically impossible (other than by selecting specifically for it) two groups of pitchers who will have a greater ERA difference than we have here. try WHIP.

iramets Jan 07 2007 08:34 AM

="metsmarathon"]less than a run difference in ERA. also worth noting, is that if you plot wins versus ERA you find that in the NL last year, a 1-run increase in your ERA would on average result in a decrease of less than 1.25 wins hardly a huge difference The more I thnk about these assertions, the more questions I have. I'm not sure what ' a 1-run increase in your ERA would on average result in a decrease of less than 1.25 wins' means exactly-- If a pitcher gave up an extra run per 9 innings it woud have cost him only1.25 wins? So you mean if Trax had given up an extra 4 runs a game, giving him a neat 9.00 ERA or so, he still would have won 10 games? Somehow I doubt that. The way I figure a run difference means a HELL of a bigger difference than you're suggesting. Let's look at Pythagorean projections. Obviously if a pitcher gives up the same number of runs that his team scores for him, he's going to be a .500 pitcher, according to Pythagorus (and common sense). But if those RS/RA proportions change from 3/3 to 3/4 (as occurs on your two lists of pitchers above), his w/l pct is going down to around .360--that's a hell of a lot more than 1.25 wins over the whole season. A run per 9 innings is not "hardly a huge difference"-- it IS a huge difference, and you're either lying to yourself or trying to lie to us by pretending otherwise. In terms of offense, for illustration's sake, 162 runs per season, which you say is "hardly...huge" is the difference between your 3-4-5 hitters knocking in 65 runs apiece or knocking in 119 runs apiece. Would that be a matter of indifference to you? I don't want to insult you or anything, but if you think a run a game is hardly significant, I don't think you follow baseball very carefully. A run per game is--I don't have another word for it--HUGE.

iramets Jan 07 2007 08:42 AM

Nymr83 wrote: Randy Johnson won 17 games this year and pitched like a total piece of shit, but by your logic he must have done great. Actually, it's Arizona's logic, isn't it? They just extended this total piece of shit an extra year on his contract, and gave up valuable players for him--guess someone else sees those 17 wins as having some value. As to WHIP--knock yourself out comparing the top ten in WHIP to the bottom ten in WHIP among last year's ERA-qualifiers. I think you'll be surprised by what comes up.

Edgy DC Jan 30 2007 09:35 AM

I split out this tangent.