All Forums > SimLeague Baseball > MLB > Wins and Losses
12/4/2012 10:20 PM
"If someone has a great ERA, we know they pitched well."

If someone had a great W/L record, we know they pitched well.
12/4/2012 10:34 PM
Posted by bad_luck on 12/4/2012 10:16:00 PM (view original):
If someone has a great ERA, we know they pitched well. They may have been lucky or had great defense behind them, but we know that they were effective in limiting the other teams runs. A win doesn't tell us that. It tells us that they pitched at least five innings, had the lead when they left, and the lead held up. Maybe they were great and pitched 9 shutout innings. Maybe they sucked and gave up 9 runs in 5 innings. We don't know. The stat is useless.
"They may have been lucky or..." - what?!? Its ok to be lucky when looking at ERA, but not with wins?

"...or had defense behind him, but we know they were effective in limiting the other teams runs." - what is the difference between having a great defense that helps you win games or a great offense that helps you win games? Why does the great defense help you when it comes to pitching statistics, but a great offense does not? You can't control how many runs your team scores, as you stated, and you also cant control what your defense does behind you. I believe it's you who has stated that a pitcher has no control where the ball goes once its hit by the batter, which is why I make this argument.

In conclusion, if win-loss record is useless, so is ERA.

12/4/2012 10:40 PM
The problem with W/L record is that is so widely used yet can so poorly describe performance/worth.  Take these two this year:
  IP W L WHIP ERA K/BB OPS RS
Cliff Lee    211.0      6      9    1.11    3.16    7.39    0.690    3.60
Lance Lynn    176.0    18      7    1.32    3.78    2.81    0.728    5.90

(RS - Run Support)  Anything that can be that wrong shouldn't be relied upon.  There are so many stats that do a better job, W/L assigned to a pitcher should be unnecessary.  It's just convention.  There were 88 qualified starters in 2012, if you group them into percentiles by Run Support, the groups look like this:

PLAYER IP W L WHIP ERA K/BB BABIP OPS RS
Top 25%    189.3    15      8    1.26    3.78    2.73    0.280    0.702      5.15
25% - 50%    197.9    14    10    1.24    3.67    2.80    0.278    0.708      4.57
50%-75%    190.6    12    11    1.29    4.21    2.61    0.280    0.738      4.15
Bottom 25%    196.1    11    12    1.25    3.69    2.79    0.281    0.699      3.65

There's very little to explain the W/L difference between those groups, other than run support.  That 50%-75% group would have to be considered the worst but the .5 higher run support is all they needed to out-win the Bottom 25% group, which has a case for the "best" group.  Those pitchers in the bottom group didn't fail as much as their offenses did.  Kershaw dropped from 21-5 last year to 14-9 this year.  I don't think he pitched meaningfully worse or differently - he just lost about .7 runs of support per game.

Good pitching should lead to more wins, even at the individual starter level, but we all know more wins don't mean best pitching. Assigning an entire win or loss to one player per game in a team sport just can't end up being descriptive in a meaningful way often enough to be really useful, can it?   Even if you like W/L record, how many other individual stats would you say do a better job of describing pitcher performance by themselves?  ERA, ERA+ or ERA-, etc?  WHIP?  OPS against?  K/BB?  FIP, xFIP, etc?  WAR? Quality Start %? If you like/know a couple of those, W/L probably doesn't add anything except some evidence for a good guess about the pitcher's run support.
12/4/2012 11:01 PM
Given the choice, I'd take a LOT of individual stats over a pitcher's W/L record. I haven't seen anyone argue otherwise.

The argument is "How important are wins and losses for evaluating pitchers?" (from the opening poll).

W/L depends on too many factors outside the pitchers control to be a very important stat.

W/L does have at least a rough correlation to other pitching stats
(see my post on page 10 at 9:51), so it's not useless.

Conclusion: W/L has SOME value when evaluating pitchers, but there are many other stats that are going to be more useful.

I can't believe I've spent so much time on this thread tonight...

12/4/2012 11:31 PM

If you start with what you think is the most important individual stat, than add the 2nd, then the 3rd, etc... by the time you get to W/L, what is it adding?  Could it change your mind by then?  Good pitching should lead to more wins so of course there is going to be correlation.  IP is also going to correlate to pitching success.  Heck, I think I want to know IP more than I do W/L.  Here are top 25% of qualified pitchers by IP in 2012 vs top 25% by Wins:
 

  IP W L WHIP ERA K/BB OPS RS
Top 25% IP    217.5    15      9    1.16    3.28    3.47    0.673      4.28
Top 25% Wins    204.0    17      8    1.16    3.22    3.31    0.665      4.79

There's some overlap in those groups, of course (50%).  I've always liked to know W/L but the more I look at everything else, the more I realize how little it adds.
 

12/4/2012 11:37 PM
I think I see part of the disconnect here.

A pitcher's win/loss record doesn't directly measure anything quantifiable about a pitcher's performance the way that ERA, OAV, and WHIP do. If you're looking at a single pitcher and trying to evaluate him by W/L, I agree that it tells you very little.

However, using W/L as a comparison among a large sample group of pitchers gives you a pretty decent idea which pitchers are better. 
12/4/2012 11:38 PM
Posted by tecwrg on 12/4/2012 8:52:00 PM (view original):
Posted by inkdskn on 12/4/2012 6:33:00 PM (view original):
ERA is an advanced stat? That's what most people use to evaluate pitchers with. I know it will sound all newfangled and far out there and ****, but it tells how many earned runs a pitcher gives up per 9IP, on average.

Are you really so profoundly ignorant that you have difficulty in separating pitcher performance (how effective they were at preventing runs) from team performance (how many times the team scored more than their ooponent)?

Or are you in first grade?
ERA+, numbnuts.

Reading comprehension.  Try it sometime.
You're aware that ERA+ is a simple function of ERA and league ERA, right? Or do you think it's ******* MAGIC?

Heads out of *****, people, please.

Please read trsnoke's post. He says the same **** badluck & I have been saying (but without being an *******).

Read, think, and learn. If you still think W-L gives any indication of actual pitching performance (preventing runs), please kill yourself.
12/5/2012 1:49 AM
You can't tell people to kill themselves and then expect them to remember/care about anything else you said.  You and bad_luck can yell and scream all day that wins and losses don't give any indication of pitching performance, and I'm guessing that means you've never pitched.  Stat guys - and I'm pretty much of a stat guy - tend to work off the belief that every plate appearance can basically be taken in a vacuum and the same statistical probabilities of events should exist.  This just isn't true.  You and bl constantly ignore or insult the people who have repeatedly mentioned that certain elite pitchers are capable of throttling back when their teams put together decent leads, at which point 6 innings of 3 or 4-run ball might be more valuable than 4 innings of shutout ball.  Nevertheless, this is an absolute fact, and can, in fact, be statistically proven.  You say pitching performance consists 100% of preventing runs.  To an extent this is true, but only to the extent to which it maintains a lead or at least an opportunity to win a game.  Let's go right back to the example I alluded to but didn't specifically outline above.  Say your team scores 6 runs in the bottom of the third.  Now all of a sudden it's better for your team if your starting pitcher finishes the game and gives up 4 runs, saving your entire bullpen, than if he gives up no runs but only gets through the 6th or 7th inning.  So now pitching performance isn't just about preventing runs, it's also about eating innings and saving your own and your teammates' arms.  You can pretend Roy Halladay and Cliff Lee and CC Sabathia don't pitch to contact more when they have a 4+ run lead, but statistically it's true.  If you don't believe me, you can work through their stats and watch their K and BB rates as a function of lead/defecit - both go down as the lead gets bigger, in either direction actually.  If you're probably going to lose the game anyway it's also better to just try to finish some innings.  You can also look at ERA as a function of team lead and you will find that the 3 I mentioned, as well as a number of other pitchers both modern and historic, give up more runs when their teams score more runs.  There is a meaningful correlation between runs allowed and run support on a game-by-game basis on a yearly and career basis if you pick innings-eating aces.  You can ignore these numbers forever if you want to, but when other people aren't willing to and you shout how stupid they are and that they should kill themselves, it's not them that wind up looking like idiots.
12/5/2012 6:05 AM
Posted by inkdskn on 12/4/2012 11:38:00 PM (view original):
Posted by tecwrg on 12/4/2012 8:52:00 PM (view original):
Posted by inkdskn on 12/4/2012 6:33:00 PM (view original):
ERA is an advanced stat? That's what most people use to evaluate pitchers with. I know it will sound all newfangled and far out there and ****, but it tells how many earned runs a pitcher gives up per 9IP, on average.

Are you really so profoundly ignorant that you have difficulty in separating pitcher performance (how effective they were at preventing runs) from team performance (how many times the team scored more than their ooponent)?

Or are you in first grade?
ERA+, numbnuts.

Reading comprehension.  Try it sometime.
You're aware that ERA+ is a simple function of ERA and league ERA, right? Or do you think it's ******* MAGIC?

Heads out of *****, people, please.

Please read trsnoke's post. He says the same **** badluck & I have been saying (but without being an *******).

Read, think, and learn. If you still think W-L gives any indication of actual pitching performance (preventing runs), please kill yourself.
I'm well aware of what ERA+ is.  It's an advanced stat.

"Please kill yourself".  Is that a tactic you learned o the high school debate team?
12/5/2012 6:18 AM
Posted by dahsdebater on 12/5/2012 1:49:00 AM (view original):
You can't tell people to kill themselves and then expect them to remember/care about anything else you said.  You and bad_luck can yell and scream all day that wins and losses don't give any indication of pitching performance, and I'm guessing that means you've never pitched.  Stat guys - and I'm pretty much of a stat guy - tend to work off the belief that every plate appearance can basically be taken in a vacuum and the same statistical probabilities of events should exist.  This just isn't true.  You and bl constantly ignore or insult the people who have repeatedly mentioned that certain elite pitchers are capable of throttling back when their teams put together decent leads, at which point 6 innings of 3 or 4-run ball might be more valuable than 4 innings of shutout ball.  Nevertheless, this is an absolute fact, and can, in fact, be statistically proven.  You say pitching performance consists 100% of preventing runs.  To an extent this is true, but only to the extent to which it maintains a lead or at least an opportunity to win a game.  Let's go right back to the example I alluded to but didn't specifically outline above.  Say your team scores 6 runs in the bottom of the third.  Now all of a sudden it's better for your team if your starting pitcher finishes the game and gives up 4 runs, saving your entire bullpen, than if he gives up no runs but only gets through the 6th or 7th inning.  So now pitching performance isn't just about preventing runs, it's also about eating innings and saving your own and your teammates' arms.  You can pretend Roy Halladay and Cliff Lee and CC Sabathia don't pitch to contact more when they have a 4+ run lead, but statistically it's true.  If you don't believe me, you can work through their stats and watch their K and BB rates as a function of lead/defecit - both go down as the lead gets bigger, in either direction actually.  If you're probably going to lose the game anyway it's also better to just try to finish some innings.  You can also look at ERA as a function of team lead and you will find that the 3 I mentioned, as well as a number of other pitchers both modern and historic, give up more runs when their teams score more runs.  There is a meaningful correlation between runs allowed and run support on a game-by-game basis on a yearly and career basis if you pick innings-eating aces.  You can ignore these numbers forever if you want to, but when other people aren't willing to and you shout how stupid they are and that they should kill themselves, it's not them that wind up looking like idiots.
This is true. And the odd part is, the people arguing that "w/l record is meaningless" don't actually dispute this fact, which makes it all the more strange.

There's also the idea that the best pitchers might be less likely to choke in high leverage situations, resulting in a better win/loss record.

Again, the record shouldn't be the only piece of information you look at when evaluating pitcher. I completely understand that the 20-8 pitcher might have 2 more runs of run support per game than the 14-14 pitcher with a better ERA. I also understand that maybe he doesn't have that much. And if you looked game by game, that it's possible that the 20-8 pitcher is doing something better or different than the 14-14 pitcher to get more wins.
12/5/2012 6:39 AM
Posted by bad_luck on 12/4/2012 11:12:00 AM (view original):
This is amazing. I really can't believe how lucky I am to have stumbled upon the last dark corner of the internet where people still argue that a pitcher's win/loss record isn't completely worthless. What a treat. After this are you guys going to argue that the earth is flat? Or make the case for a geocentric solar system?
I want to come back to this.  Where, exactly, do you spend most of  your baseball time?  Because almost all of the paid analysts in the world still at least mention wins and losses.  Even guys on fangraphs do it.  So where is it that you're reading that has you convinced that this is "the last dark corner of the internet where people still argue that a pitcher's win/loss record isn't completely worthless?"  Are you just a subscriber to uber statnerd monthly?  I mean, I spent a week trying to convince anyone on the Orioles website that it actually makes sense to bat a guy like Mark Reynolds 3rd and a more complete hitter 4th, because statistical models say so.  That's one of the highest pinnacles of stat nerdery, 'cause even very smart, stat-oriented guys are still stuck in the "you hit your best guy 3rd" mentality.  But I'm still missing the part of the internet that says you totally ignore pitchers' most meaningful results.
12/5/2012 7:31 AM

A question for the folks arguing that there is no correlation between pitchers W-L record and their "success"?

Do you remember the season in which the Cy Young Award was given to a starting pitcher who went 5-17?

No.  Neither do I.

12/5/2012 8:35 AM
Posted by bad_luck on 12/4/2012 9:15:00 PM (view original):
The entire point is that you have other stats available. If you had to pick a pitching staff using one stat of your choice, you sure as **** wouldn't pick W/L.
No.  The entire point is that W/L is not COMPLETELY WORTHLESS as someone has claimed. 
12/5/2012 8:37 AM
Posted by mattedesa on 12/4/2012 9:57:00 PM (view original):
Posted by bad_luck on 12/4/2012 9:50:00 PM (view original):
What does pitcher W/L tell you about a pitcher's performance?

Does a pitcher even have to pitch well to get a win? Does he have to pitch poorly to get a loss?
A pitcher with a winning record is more likely to be a good pitcher than a pitcher with a losing record. That's what it tells us. How much more likely is the question.

And we all know that sometimes pitchers pitch poorly and win, or pitch well and lose. Any statistic could have its aberrations. A guy could have a great ERA and just be a lucky fool that no one scored the 5 times he loaded the bases. Someone could have a great OAV but walk every other batter. Someone could have a great WHIP but have every third hit be a homer.


We have 100+ years of data to support this.   Maybe we should link a site for him.
12/5/2012 8:39 AM
Posted by bad_luck on 12/4/2012 10:16:00 PM (view original):
If someone has a great ERA, we know they pitched well. They may have been lucky or had great defense behind them, but we know that they were effective in limiting the other teams runs. A win doesn't tell us that. It tells us that they pitched at least five innings, had the lead when they left, and the lead held up. Maybe they were great and pitched 9 shutout innings. Maybe they sucked and gave up 9 runs in 5 innings. We don't know. The stat is useless.
If someone has a great W/L record, we know they pitched well.  Unless, of course, you can find someone with a great W/L record that really stunk it up.

I'll wait here while you find him.
of 28
All Forums > SimLeague Baseball > MLB > Wins and Losses

Terms of Use Customer Support Privacy Statement

Popular on WhatIfSports site: Baseball Simulation | College Basketball Game | College Football Game | Online Baseball Game | Hockey Simulation | NFL Picks | College Football Picks | Sports Games

© 1999-2014 WhatIfSports.com, Inc. All rights reserved. WhatIfSports is a trademark of WhatIfSports.com, Inc. SimLeague, SimMatchup and iSimNow are trademarks or registered trademarks of Electronic Arts, Inc. Used under license. The names of actual companies and products mentioned herein may be the trademarks of their respective owners.