Floyd Signs W/Braves (1yr deal)

[Dick Allen]

Then we will agree to disagree.

 

I just find absolutely no meaning in W-L.

 

-ERA tells me how many earned runs a pitcher gives up on average

-WHIP tells me how many base runners a pitcher allows on average

-IP tells me, generally, how durable and/or efficient a pitcher is (though it tells a lot, though not all is incredibly telling of talent)

-GS also tells me about durability; thus, IP/GS is important to me as well

-FIP tells me approximate talent level of the pitcher this year

-xFIP tells me what he should probably be putting up, given his peripherals

-K/9 tells me a little about his stuff and how efficient he is about getting outs (which improves consistency)

-BB/9 tells me how good a guys control is and whether it will ever typically get him in trouble

-HR/9 tells me how many homers a guy gives up

-HR/FB tells me how unlucky the pitcher has been about giving up home runs (typically 10% in all circumstances)

-GO/AO tells me how good a guy is about producing ground balls which, while more likely to become hits, are very unlikely to be anything more than singles

 

I could go on. On the other hand, W-L tells me how good the pitcher's team was, or how the team performed when he was on the mound. A pitcher has very, very little control over that. He can pitch to his defense and in NL parks, he can help try and score runs when up at the plate. Otherwise, all he can do is pitch as well as he can and then hope the team wins the game.

 

Theoretically, Chris Sale gives up a lead off homer every game he pitches, but then retires the next 27 by striking them all out. His opposing pitcher loads the bases every inning but gets double plays and keeps the team off the board. At the end of the year, Chris Sale is 0-32 while opposing pitchers are 32-0. Did Chris Sale really deserve to lose 32 games, or should he be upset at his team for allowing him to lose 32 games?

 

If it is all about how the team performs, can you explain why pitcher's ERAs and all their other numbers are generally better when they gets wins vs. when they get losses and no decisions.

 

Obviously W-L isn't the best, but guys who win a lot of games generally have the other numbers that match up to showing they are pitching very well. While a guy who is 8-12 could have numbers that show he is pitching much better than the record shows, it is less likely a guy who wins 17 or 18 games has numbers that show he actually has pitched horribly.

 

A starting pitcher with a lot of wins is going to pencil out pretty well sabermetrically.

[witesoxfan]

If it is all about how the team performs, can you explain why pitcher's ERAs and all their other numbers are generally better when they gets wins vs. when they get losses and no decisions.

 

Obviously W-L isn't the best, but guys who win a lot of games generally have the other numbers that match up to showing they are pitching very well. While a guy who is 8-12 could have numbers that show he is pitching much better than the record shows, it is less likely a guy who wins 17 or 18 games has numbers that show he actually has pitched horribly.

 

A starting pitcher with a lot of wins is going to pencil out pretty well sabermetrically.

 

If a pitcher pitches well, they are more likely to pitch deeper into the game and their team is more likely to win. If a pitcher pitches poorly, they are not as likely to pitch deeper into the game and their team is more likely to lose. Pretty standard.

 

A lot of times, guys that get "wins" will grade out well sabermetrically, but that is not always the case and that does not make them good pitchers. Chris Sale went 11-14 last year, but he was one of the best pitchers in the league. Meanwhile, RA Dickey went 14-13 with a worse ERA, BB/9, K/9, and WHIP (among other numbers). Who do you believe to be the better pitcher?

 

Paul Abbott went 17-4 with a 4.25 ERA in 2001. He was barely an average pitcher yet was amongst the league leaders in wins. Why? Because he pitched for a team that won 116 games. Jeriome Robertson went 15-7 with a 5.10 ERA in 2003 with the Astros. He was a flat out bad pitcher, but he won games because he played for a team that scored the 4th most runs in the NL, so they could bail him out when he gave up a bunch of runs.

 

In general, good pitchers will win games and bad pitchers will lose games, but a guy winning a lot of games or not winning a lot of games does not say anything about him as a pitcher because, as I've pointed out, good pitchers will lose games if they are on a bad team and mediocre to bad pitchers will win games on good teams. Good pitchers will lose games on good teams if they are unlucky, while bad pitchers will win games on bad teams if they are incredibly lucky. There is no rhyme or reason to W-L and it says nothing about the pitcher himself.

[Balta1701]

If it is all about how the team performs, can you explain why pitcher's ERAs and all their other numbers are generally better when they gets wins vs. when they get losses and no decisions.

Because there is a correlation between a pitcher having a good game and a win and a pitcher having a bad game and a loss. However, that only makes it a useful stat if there aren't already better stats to evaluate whether the pitcher had a set of good games or bad games.

[southsider2k5]

If a pitcher pitches well, they are more likely to pitch deeper into the game and their team is more likely to win. If a pitcher pitches poorly, they are not as likely to pitch deeper into the game and their team is more likely to lose. Pretty standard.

 

A lot of times, guys that get "wins" will grade out well sabermetrically, but that is not always the case and that does not make them good pitchers. Chris Sale went 11-14 last year, but he was one of the best pitchers in the league. Meanwhile, RA Dickey went 14-13 with a worse ERA, BB/9, K/9, and WHIP (among other numbers). Who do you believe to be the better pitcher?

 

Paul Abbott went 17-4 with a 4.25 ERA in 2001. He was barely an average pitcher yet was amongst the league leaders in wins. Why? Because he pitched for a team that won 116 games. Jeriome Robertson went 15-7 with a 5.10 ERA in 2003 with the Astros. He was a flat out bad pitcher, but he won games because he played for a team that scored the 4th most runs in the NL, so they could bail him out when he gave up a bunch of runs.

 

In general, good pitchers will win games and bad pitchers will lose games, but a guy winning a lot of games or not winning a lot of games does not say anything about him as a pitcher because, as I've pointed out, good pitchers will lose games if they are on a bad team and mediocre to bad pitchers will win games on good teams. Good pitchers will lose games on good teams if they are unlucky, while bad pitchers will win games on bad teams if they are incredibly lucky. There is no rhyme or reason to W-L and it says nothing about the pitcher himself.

 

On the opposite side of things, don't forget about Nolan Ryan going 8-16 while leading the NL in era, IP, k/9, k/bb, and h/9.

[witesoxfan]

On the opposite side of things, don't forget about Nolan Ryan going 8-16 while leading the NL in era, IP, k/9, k/bb, and h/9.

 

On a Houston team that really was not too terribly bad.

[Eminor3rd]

Because there is a correlation between a pitcher having a good game and a win and a pitcher having a bad game and a loss. However, that only makes it a useful stat if there aren't already better stats to evaluate whether the pitcher had a set of good games or bad games.

 

This.

 

The problem with W-L is that it tells us nothing about a pitcher's performance that other stats don't tell us with more precision and accuracy. Its day has passed.

[Dick Allen]

Because there is a correlation between a pitcher having a good game and a win and a pitcher having a bad game and a loss. However, that only makes it a useful stat if there aren't already better stats to evaluate whether the pitcher had a set of good games or bad games.

So if there is a correlation between getting a win and pitching a good game, a lot of wins would indicate you pitched a lot of good games.

 

There are also times guys give up 5 runs and pitch pretty well, and times when they give up 2 runs and don't pitch as well. EX. Wind blowing out a Wrigley, it's a bandbox, blowing in, it's tough to get it out of there. I believe it usually evens out, but occassionally it does not.

 

I'll take a guy with 15 wins and a 5+ ERA. Generally that will be a #4 or #5 starter, and it shows you were probably over .500 in their starts. And it shows they probably went a decent length in the game resting the bullpen a bit.That's pretty good.

Edited December 17, 2013 by Dick Allen

[Eminor3rd]

So if there is a correlation between getting a win and pitching a good game, a lot of wins would indicate you pitched a lot of good games.

 

There are also times guys give up 5 runs and pitch pretty well, and times when they give up 2 runs and don't pitch as well. EX. Wind blowing out a Wrigley, it's a bandbox, blowing in, it's tough to get it out of there. I believe it usually evens out, but occassionally it does not.

 

I'll take a guy with 15 wins and a 5+ ERA. Generally that will be a #4 or #5 starter, and it shows you were probably over .500 in their starts. And it shows they probably went a decent length in the game resting the bullpen a bit.That's pretty good.

 

Except when it's wrong about all those things. You can measure how well he pitched more accurately by using almost literally every other pitching statistic. Instead of settling for "probably went a decent length in the game resting the bullpen," for example, why not look at IP/GS and measure "definitely went a decent length in the game resting the bullpen."

 

W-L is sometimes right about guys and it's just as frequently wrong. That would be fine if it was the best way we had to evaluate pitchers, but it isn't. Every single question you want answered by W-L can be answered better by another statistic.

 

 

[witesoxfan]

So if there is a correlation between getting a win and pitching a good game, a lot of wins would indicate you pitched a lot of good games.

 

There are also times guys give up 5 runs and pitch pretty well, and times when they give up 2 runs and don't pitch as well. I believe it usually evens out, but occassionally it does not.

 

I'll take a guy with 15 wins and a 5+ ERA. Generally that will be a #4 or #5 starter, and it shows you were probably over .500 in their starts. That's pretty good.

 

There is a far greater correlation between wins and your team scoring more runs than the other team than pitching a good game. That's the point. W-L takes contributions from the entire team and places it on one guy at the end of the game.

 

A guy that wins 15 games but has an ERA over 5 doesn't last long. Take Jeriome Robertson - he didn't make the Astros out of Spring Training the next year, was traded to the Indians, and pitched 8 more games before his MLB career was over. If a guy is putting up an ERA over 5, they get booted from the rotation in short order, even if they're "winning."

[TaylorStSox]

I think W/L still has some value. More often than not, due to rotations, good pitchers face other good pitchers. Despite overall numbers, some pitchers have the propensity to give up big hits. Some pitchers rise to the occasion better than others. Some guys are just "winners." Not everything in the game is tangible. There's no doubt in my mind that momentum exists in sports and there really isn't a stat to explain it. Occasionally you'll see pitchers that win a lot of games without the best peripherals. Some of those guys just know what it takes to get a win.

[Eminor3rd]

I think W/L still has some value. More often than not, due to rotations, good pitchers face other good pitchers. Despite overall numbers, some pitchers have the propensity to give up big hits. Some pitchers rise to the occasion better than others. Some guys are just "winners." Not everything in the game is tangible. There's no doubt in my mind that momentum exists in sports and there really isn't a stat to explain it. Occasionally you'll see pitchers that win a lot of games without the best peripherals. Some of those guys just know what it takes to get a win.

 

You can measure a pitcher's propensity to give up big hits much more accurately by looking at his performance in high leverage situations. The problem with "just knowing how to win" is that it doesn't correlate year to year. Guys that get a lot of Ws despite pitching badly do not continue to do so, implying that their success was incidental. This is not surprising considering that W-L includes offensive and defensive performance that cannot be controlled by the pitcher.

 

There are absolutely intangible qualities to good baseball players that cannot be measured. Meaning W-L is among that stats that do not measure them.

[Texsox]

It's one more piece of evidence. Not as valuable as others, meaningless when evaluating some pitchers as people have pointed out. But I wouldn't reject any facts. Especially when the wins do not flow with the rest of the stats.

[Eminor3rd]

It's one more piece of evidence. Not as valuable as others, meaningless when evaluating some pitchers as people have pointed out. But I wouldn't reject any facts. Especially when the wins do not flow with the rest of the stats.

 

I think your premise is spot on, Tex, but the problem with W-L as a piece of evidence is that it tells you more about the team performance than the pitcher's performance. And if you spend time trying to analyze it with other numbers to try to strip defensive and opponent context, you quickly realize that the W-L number itself isn't adding anything to the conversation -- the other numbers are. It's more of an obstacle than anything. If every time you talk about W-L you immediately have to look at everything else to decide how 'accurate' an indicator of performance it is for a certain pitcher, by the time you contextualize the metric, you've done most of the work evaluating the pitcher -- you're really spending time evaluating the number rather than using the number as a tool.

 

There's a ton of value in using all available information as evidence, but there's also plenty of bad information that translates to bad evidence. W-L is misleading, to some degree, almost every single time. It's bad information.

 

It's easy to overlook, but remember that although a Pitcher Win is called the same thing as a Team Win, they are completely different metrics.

 

Team Wins: "Number of times a team outscored it opponent over the course of a game. Ultimate decider of the champion"

 

Pitcher Wins: "Number of times a specific starting pitcher, having pitched at least 5 innings in a particular game, left the game while his team was in the lead and able to ultimately win the game without subsequently giving up and regaining the lead later"

 

To disdain the Pitcher Win is not to make an argument that other metrics are more important than Team Wins. Criticizing it feels like garbage because they are branded with the same word, but they are really entirely different things.

[GreenSox]

I always liked Floyd.

He has one beautiful curveball.

Good luck to him.

[bmags]

Best of luck to Floyd. He was frustrating but he gave us a lot of innings and came through in 08 pretty well.

[Steve9347]

Gavin Floyd. I am glad we are done with him.

[Rowand44]

Besides Sale being under .500 and Scherzer having an insane record I couldn't even tell you what other pitchers W-L looked like. It's absolutely meaningless to me.

 

Oh and good luck Gavin. I feel that you became an extremely underrated pitcher around these parts and hope you do well in Atlanta.

[greg775]

Besides Sale being under .500 and Scherzer having an insane record I couldn't even tell you what other pitchers W-L looked like. It's absolutely meaningless to me.

 

Oh and good luck Gavin. I feel that you became an extremely underrated pitcher around these parts and hope you do well in Atlanta.

 

It's funny how W/L used to be the sole barometer of excellence before the days of Sabes.Think about it. You didn't have a chance in hell to win a Cy Young for decades w/out a good win loss record.

Edited December 18, 2013 by greg775

[iamshack]

I absolutely adore this debate whenever it comes up.

[Eminor3rd]

http://www.fangraphs.com/not/gavin-floyd-m...vyawn-floyawnd/