Should pitchers have won-lost records?

[SWC75]

Pitcher's won-lost records date from an era when pitchers were expected to finish games: they pitched in rotation and you wanted to know a team's likelihood of winning when a particular pitcher pitched. Cy Young started 815 games, completed 749, win 511 and lost 315, a wining percentage of .619. (Clayton Kershaw's is .697)

Eventually the concept of a relief pitcher was born but they were both firemen and closers and expected to take over games. They created rules to allow relief pitchers to get wins and losses. Wilcy Moore was 19-7 with 13 saves for the 1927 Yankees and they went 110-44. the next year he was 4-4 with 3 saves and then went 101-53.

But eventually top relief pitchers became just closers. Starters were only supposed to supply 6 good innings, (a 'quality start'),then turn it over to a 7th inning guy and then an 8th inning guy and then the closer. The combined efforts of the starter, these three relief pitchers as well as the batting, baserunning and fielding of the 8 'position' players determines the outcome of the game. Giving relief pitchers won-lost records now seems pointless as they are no long evaluated based on that. Mariano Rivera's won-lost record was 82-60. he's not in the Hall of Fame for that: he's in the Hall for his 652 saves and 2.21 ERA. The games where he got the decision were not his best games: he's not supposed to get the decision. he's supposed to get the save. it makes more sense to just keep track of the team's won-lost record with each starter.

The Jacob DeGrom's frustrations with the Mets are well documented:
Jacob deGrom is one of the unluckiest aces in baseball history (msn.com)
"DeGrom has often suffered this plight. Since 2018, the year of his first Cy Young Award, deGrom has made 78 starts and pitched 503 innings. Seventy-eight percent (61) of those starts were quality starts, meaning he lasted six or more innings and surrendered three or fewer earned runs. He’s struck out 649 batters, walked 110 and allowed just 115 earned runs (2.06 ERA) over that span. Yet the Mets are 36-42 in those matchups. No, that’s not a typo...A lack of run support is part of it. Over the past four seasons, the Mets have given deGrom 4.1 runs of support per nine innings pitched, the third-lowest of any starting pitcher over the past four years. If he had been given an average number of runs (4.6 per data from TruMedia) in support of his 2.06 ERA, we’d expect the Mets to be 45-10 in those starts. Instead they went 36-19, nine fewer wins.

Jake is almost 33 years old and ahs only 70 career victories, (vs. 52 defeats). His personal won-lost totals over those three years have been 10-9, 11-8, 4-2 and 0-1 so far this year. He's been award two Cy Young awards, the voters having to ignore his won-loss record and focus on his other numbers to give them to him. It's bene reported that he's worried he might not make the Hall of Fame because of the paucity of victories.

The Mets won the first game today, 4-3: Phillies vs. Mets - Box Score - April 13, 2021 - ESPN
Walker pitched 4.1 innings, got 13 of 19 guys out and gave up 1 run. Castro got 5 guys out, put three guys on base and one of them scored. Diaz pitched a perfect inning with a strikeout. May pitched the 8th inning, (of a 7 inning game), got 3 guys out but gave up two base runners, (besides the 'designated runner) and the go-ahead run before the Mets won it in the bottom of the inning. he also had a passed ball. he was arguably the Met's worst pitcher in the game but got the win because he was 'the pitcher of record'.

Why not can the W-Ls and just evaluate pitchers on how well they pitched just as you evaluate batters on how well they hit and fielders on how well they fielded? past pitchers could be evaluated on the same basis. Just compare them to their contemporaries using actual pitching numbers. You're never gonna catch ol' Cy anyway!

teams win and lose games not pitchers.

 

[donniesyracuse]

It’s silly.

 

[jncuse]

It still has limited value or decent value for the rare pitcher, but you must understand the clear limitations and there are better stats out there which the voters now do.

Here is one of my annoyances when they analyze the season performance of a pitcher..

Pitcher A - 2.50 ERA, but a higher FIP, 200 innings neutral park/competition
Pitcher B - 3.00 ERA, but a lower FIP, 200 innings.

Many stat geeks will say Pitcher B had the better season because FIP is a better stat. FIP is a better stat because it is more predictive -- that I absolutely agree with. But at the end of the day you know who had the better season you dolt -- the pitcher who gave up the least amount of runs. Stathead will then claim that the pitcher was lucky or cannot sustain that level of clutch with RISP. Quite possible, but for those 200 innings he gave up less runs -- that pitcher was better because he was clutch this year. Some statheads lose something that is valuable as well -- common sense in interpreting data.

As an aside I grew up from 9-14 reading Bill James so I was a stat geek in the making, but I just moved on from the stuff before the end of my school year. I am not anti-stathead by any means (nor do I place many merits on the old AVG/HR/RBI), but I think some people lack flexibility in thinking their way through numbers and problems.

I still have the 1985-1988 Baseball Abstract on my shelf next to me -- most of that analysis was more basic stuff than today (Secondary average for example) but the one thing it really stressed that I learned early on was how walks were overlooked. I would get in arguments with other kids at school about ball players and I would get flabbergasted that they did not understand that AVG/HR/RBI's on a standalone basis were sometimes not that great a stat to evaluate between players. My 11 and 12 year old friends would give me strange looks.

 

[donniesyracuse]

It still has value, but you must understand the clear limitations.

Zero value

 

[jncuse]

Zero value

I don't agree on zero value, but not much either.

I still think there is value in effective pitchers that can stretch out longer innings as coaches tire out and over manage their bullpen. Generally over multiple seasons that trait will come out in more wins over other starters.

But then again I would highly prefer WAR to point out the value in the current day starter who works more innings than something as finicky and so context impacted as Wins.

 

[OttoinGrotto]

I like wins and losses as a stat for pitchers. It does tell a unique story, and it captures something that actually happened.

 

[OttoinGrotto]

This is a good article on how I view this: You Don’t Have To Embrace Stats Meant For Sports Executives

Sports statistics, I would argue, are a form of storytelling. The point was never to be a perfect, all-encompassing assessment of a player. What we’re looking for isn’t necessarily the most objective measurement of quality, but to be able to relive those moments of glory. The worst of these Sabermetrics-style stats take sports, a thrilling exhibition of humans transcending the normal boundaries of physical achievement, and applies to them the cold, corporate logic of extracting maximum value for minimum input.

 

[SWC75]

I like my stats simple and logical, with regular fans able to compute and understand them. We aren't trying to land a man on the moon, just figure out why what happened happened. I just don't happen to think that a pitcher's won-lost record means much any more. DeGrom shouldn't be blamed for the failures of his teammates.

 

[donniesyracuse]

I like my stats simple and logical, with regular fans able to compute and understand them. We aren't trying to land a man on the moon, just figure out why what happened happened. I just don't happen to think that a pitcher's won-lost record means much any more. DeGrom shouldn't be blamed for the failures of his teammates.

If I’m trying to assess if a pitcher is good, I’m only going to look at measurables over which the pitcher has control. I want to remove chance/randomness from the equation. ERA can take a hike too.

 

[KaiserUEO]

Wins are not as bad as football.

QB could start the game, throw a pick 6 on 1st play off game, get replaced and the backup wins it 49-7 and W gets tagged to the starter.

 

[Chip]

I've just never understood how you can go 5 innings, give up no runs, and get a win, but not a quality start.

Unless someone has a really lopsided W/L, I never pay attention to it.

That said, I do have trouble keeping up with all of the sabermetrics. I know they tell a good story, but there are so many. I probably would have loved it when I was younger and more of a baseball stat nerd.

 

[MSOrange]

I've just never understood how you can go 5 innings, give up no runs, and get a win, but not a quality start.

Unless someone has a really lopsided W/L, I never pay attention to it.

That said, I do have trouble keeping up with all of the sabermetrics. I know they tell a good story, but there are so many. I probably would have loved it when I was younger and more of a baseball stat nerd.

That's how I feel about the sabermetrics. I definitely see their value, but it's a bit too much for me to keep up with.

 

[Fat tire 13]

This is a good debate

Thankfully the Cy Young voters embrace the advanced stats, and DeGrom has won the awards. Same would hold true for his value if the Mets ever traded him.

I remember Felix Hernandez being one of the first pitchers to win the Cy Young with a crappy W/L record. And lack of run support is not a recent phenomenon. Steve Carlton had a couple of awful years on the W/L side even when he pitched really well. Same deal for Nolan Ryan at times

Overall I would say W/L has some value, but context matters a bunch more. And for better or worse, it is a stat with a long history, unlike some of the more recent advanced measures.

 

[Hoo's That]

If a starter goes 7-ish innings and gives up 3 or fewer runs, he's done his job regardless of whether the team wins or loses. When a team loses 1-0 or 2-1, the loss isn't the fault of the pitching staff.

 

[OttoinGrotto]

If a starter goes 7-ish innings and gives up 3 or fewer runs, he's done his job regardless of whether the team wins or loses. When a team loses 1-0 or 2-1, the loss isn't the fault of the pitching staff.

But see, I like that the win or loss is a record of did the pitcher do what was needed in that particular game for their team to win?

That's different than whether in theory the pitcher did enough that the team should have won.

 

[Hoo's That]

But see, I like that the win or loss is a record of did the pitcher do what was needed in that particular game for their team to win?

That's different than whether in theory the pitcher did enough that the team should have won.

I have no problem with having W-L record because we've "always" had it. I'm sure the pitchers and their agents keep meticulous records on my "should have won" statistic for contract negotiation time.

 

[OttoinGrotto]

I have no problem with having W-L record because we've "always" had it. I'm sure the pitchers and their agents keep meticulous records on my "should have won" statistic for contract negotiation time.

I definitely think W-L becomes more insightful over longer time periods. It does tell a story about a season. Definitely tells a story about a career.

 

[donniesyracuse]

If I am a NL Cy Young voter this year and I'm deciding between deGrom and Bauer, win-loss record will not influence my decision.

 

[KaiserUEO]

If I am a NL Cy Young voter this year and I'm deciding between deGrom and Bauer, win-loss record will not influence my decision.

but its also why no pitcher should really ever be considered for MVP, i hate when they are.

at the end of the season, starters are only in about 30 games and relievers about 60-70.

stay out of the MVP award.

 

[OttoinGrotto]

To me the conflict is that somewhere along the way, predictive statistics somehow became a substitute for what actually happened, or predictive statistics were misapplied to make an argument that what the numbers say should have happened mattered more than what actually happened.

This is overly simplistic, but it's like saying a team had a better season if their pythagorean record was 95 wins but they actually won 85 games, than a team that won 95 games with a pythagorean record of 85. The actual winning is the point.

I think baseball has swung too far away and devalues traditional counting stats too much. RBI for example tell a limited story, but you don't win games without getting runners in, and every RBI is an event that actually happened, and actually matters for winning. Imperfect as it may be, it needs to be recognized.

I view Ws and Ls similarly. Winning is the point, not projecting wins.

 

[donniesyracuse]

To me the conflict is that somewhere along the way, predictive statistics somehow became a substitute for what actually happened, or predictive statistics were misapplied to make an argument that what the numbers say should have happened mattered more than what actually happened.

This is overly simplistic, but it's like saying a team had a better season if their pythagorean record was 95 wins but they actually won 85 games, than a team that won 95 games with a pythagorean record of 85. The actual winning is the point.

I think baseball has swung too far away and devalues traditional counting stats too much. RBI for example tell a limited story, but you don't win games without getting runners in, and every RBI is an event that actually happened, and actually matters for winning. Imperfect as it may be, it needs to be recognized.

I view Ws and Ls similarly. Winning is the point, not projecting wins.

I understand and respect your point. When it comes to making playoffs and winning divisions then actual results are what matters. For subjective awards (e.g., Cy Young, MVP), I believe it is worthwhile to attempt to control for noise and fortune to decide if a player actually performed better than his peers.

 

[SWC75]

But see, I like that the win or loss is a record of did the pitcher do what was needed in that particular game for their team to win?

That's different than whether in theory the pitcher did enough that the team should have won.

Doesn't "what was needed" = "did enough"?

 

[SWC75]

I definitely think W-L becomes more insightful over longer time periods. It does tell a story about a season. Definitely tells a story about a career.

What story does it tell for DeGrom? Ans is it his story or that of the team?

 

[SWC75]

To me the conflict is that somewhere along the way, predictive statistics somehow became a substitute for what actually happened, or predictive statistics were misapplied to make an argument that what the numbers say should have happened mattered more than what actually happened.

This is overly simplistic, but it's like saying a team had a better season if their pythagorean record was 95 wins but they actually won 85 games, than a team that won 95 games with a pythagorean record of 85. The actual winning is the point.

I think baseball has swung too far away and devalues traditional counting stats too much. RBI for example tell a limited story, but you don't win games without getting runners in, and every RBI is an event that actually happened, and actually matters for winning. Imperfect as it may be, it needs to be recognized.

I view Ws and Ls similarly. Winning is the point, not projecting wins.

I agree with this as regards the team. But looking at how many baserunners a pitcher allowed and how many scored isn't "predicting wins". it's "evaluating the pitcher. Looking at a batter's stats doesn't predict wins and losses either. It evaluates the batter's performance.

By the way, why aren't the pitcher's numbers the same as the batters? Hits, walks, doubles, triples, home runs, runs allowed, etc.

 

[OttoinGrotto]

Doesn't "what was needed" = "did enough"?

No they're different. What was needed is what happened. Did enough is theoretical.