A shutdown closer might be worth 0.8 of a win over a season

[DragonfromTO]

Closers are commonly a sore spot for sabermetricians. Most people find them misused and underused. But given the current method of use for a closer, just how important is a good one?

I looked at 2011, because that year is interesting (or painful) for baseball fans, with a certain piece of shit blowing a save in the last game of the season for the Red Sox. To evaluate closers, I used the Pythagorean formula's estimates versus the actual records of teams. I then singled out closers with 50 games finished and an ERA below 2.50. Kind of arbitrary numbers, but I wanted to keep it simple- as those closers typically are considered "good." My output is attached.View attachment 137545

The mean of the residuals is only slightly above zero (0.004925), but when you multiply it by 162, you get about 0.8. This means that a team with a dominant closer might add 0.8 wins to his team because of the team's success in one-run games.

I still hate the idea of "closers."

I'm not sure that this study/methodology shows what you think it does.

 

[Omar 382]

I'm not sure that this study/methodology shows what you think it does.

What, then, does it show?

 

[DragonfromTO]

What, then, does it show?

Assuming I'm clear on what you did here, possibly nothing?

 

[Bulletz]

Shitloads?

Try fucktons.

 

[Omar 382]

Assuming I'm clear on what you did here, possibly nothing?

I looked at the relation between Pythagorean residuals and one-run victories, while making the assumption that one-run wins and elite closers are correlated (which I could easily prove if you wanted). Would you like to see my whole line of code?

 

[MilkSpiller22]

I will agree that closers as a position is over valued... But a strong bullpen is not.... Strong Bullpens help the starting pitching... Bad Bullpens force starters to go further into the game and often make those starters have a higher ERA than they should...

"CLOSERS" are generally just your best bullpen arm... and i understand why they end games, even if the end is not necessarily the most important or difficult inning...

 

[DragonfromTO]

I looked at the relation between Pythagorean residuals and one-run victories, while making the assumption that one-run wins and elite closers are correlated (which I could easily prove if you wanted). Would you like to see my whole line of code?

Probably doesn't hurt. Or you could talk me through what you did in a little more detail.

 

[DragonfromTO]

I looked at the relation between Pythagorean residuals and one-run victories, while making the assumption that one-run wins and elite closers are correlated (which I could easily prove if you wanted). Would you like to see my whole line of code?

Out of curiosity, was there a correlation between elite closers and wins in games decided by more than one run in 2011?

 

[Erie Warrior]

I will agree that closers as a position is over valued... But a strong bullpen is not.... Strong Bullpens help the starting pitching... Bad Bullpens force starters to go further into the game and often make those starters have a higher ERA than they should...

"CLOSERS" are generally just your best bullpen arm... and i understand why they end games, even if the end is not necessarily the most important or difficult inning...

This is where I have a problem a closer. If he his your best bullpen pitcher, why is he not used in the most important or difficult inning? Its always some other reliever who has to come in and get the team out of jam, when starter is yanked.

 

[Omar 382]

Out of curiosity, was there a correlation between elite closers and wins in games decided by more than one run in 2011?

In all the years of baseball, yes.

 

[Omar 382]

Probably doesn't hurt. Or you could talk me through what you did in a little more detail.

> setwd("desktop")
> teams = read.csv("teams.csv")
> myteams = subset(teams, yearID > 2000) [,c("teamID", "yearID",
+ "lgID", "G", "W", "L", "R", "RA")]
> myteams$RD = with(myteams, R- RA)
> myteams$Wpct = with(myteams, W / (W+L))
> myteams$pytWpct = with(myteams, R^2 / (R^2 + RA^2))
> myteams$pytResiduals = myteams$Wpct - myteams$pytWpct
> sqrt(mean(myteams$pytResiduals^2))
[1] 0.02541395
> myteams$logWratio = log(m yteams$W / myteams$L)
Error: unexpected symbol in "myteams$logWratio = log(m yteams"
> myteams$logWratio = log(myteams$W / myteams$L)
> myteams$logRratio = log(myteams$R / myteams$RA)
> pytFit = lm(logWratio ~ 0 +logRratio, data = myteams)
> pytFit

Call:
lm(formula = logWratio ~ 0 + logRratio, data = myteams)

Coefficients:
logRratio
1.88
>onerungames <- subset(results, diff == 1)
>onerunwins <- as.data.frame(table(onerungames$winner))
>names(onerunwins) <- c("teamID", "onerunW")

> gl2011 = read.table("2gl2011.txt", sep = "\t", header = FALSE, fill = TRUE) ###starter file
Warning message:
In scan(file = file, what = what, sep = sep, quote = quote, dec = dec, :
EOF within quoted string
> glheaders = read.csv("retrosheetgamelog.csv")
Warning message:
In read.table(file = file, header = header, sep = sep, quote = quote, :
incomplete final line found by readTableHeader on 'retrosheetgamelog.csv'
> names(gl2011) = names(glheaders)
> BOS2011 = subset(gl2011, HomeTeam == "BOS" | VisitingTeam == "BOS") [
+ , c("VisitingTeam", "HomeTeam", "VisitorRunsScored",
+ "HomeRunsScore")]
> BOS2011$HomeRunsScore = as.numeric(BOS2011$HomeRunsScore)
> BOS2011$VisitorRunsScored = as.numeric(BOS2011$VisitorRunsScored)
> BOS2011$HomeRunsScore = as.numeric(BOS2011$HomeRunsScore)
> BOS2011$ScoreDiff = with(BOS2011, ifelse(HomeTeam == "BOS",
+ HomeRunsScore - VisitorRunsScored,
+ VisitorRunsScored - HomeRunsScore))
> BOS2011Wins = BOS2011$ScoreDiff > 0
> aggregate(abs(BOS2011$ScoreDiff), list(W=BOS2011Wins), summary)
W x.Min. x.1st Qu. x.Median x.Mean x.3rd Qu. x.Max.
1 FALSE 1.000 2.000 4.000 5.514 10.000 12.000
2 TRUE 1.000 2.000 3.500 5.450 9.250 16.000
>identify(teams2011$onerunW, teams2011$pytResiduals, labels=teams2011$teamID)
> top_closers = subset(pit, GF > 50 & ERA < 2.50)[, c("playerID",
+ "yearID", "teamID")]
> names(myteams)
[1] "teamID" "yearID" "lgID" "G" "W" "L"
[7] "R" "RA" "Wpct" "pytWpct"
> myteams$pytResiduals = myteams$Wpct - myteams$pytWpct
> teams_top_closers = merge(myteams, top_closers)
> summary(teams_top_closers$pytResiduals)
Min. 1st Qu. Median Mean 3rd Qu. Max.
-0.048690 -0.011560 0.002544 0.004925 0.023090 0.071400

 

[MilkSpiller22]

This is where I have a problem a closer. If he his your best bullpen pitcher, why is he not used in the most important or difficult inning? Its always some other reliever who has to come in and get the team out of jam, when starter is yanked.

Its a mindset more than anything else...

you leave your best BP pitcher till the end, then there are 3 scenarios

1. Your team scores enough runs to not need him
2. your team loses the lead and does not use him
3. your team uses him...

so there are 2 scenarios where you can save your "closers" arm... and one where you use him...
there are 2 scenarios where you either win or get to the end with the lead...

By pitching him whenever the middle of the lineup or in meaningful times(not necessarily the end) then you MUST use him, and you are stuck using him... there is no chances of not using him...

Then, if you use the closer before the end, then what happens if that moment in the game he is used is NOT the most important time?? then you have to rely on a lesser pitcher later!!!

The whole use the best reliever at the most important times is all hindsight...

Yes, after the game is done I would have probably wanted my closer to pitch at different moment, but really who knows??

by leaving the closer till the end, you are also protecting yourself from the worst case scenario(even if you increase the odds of that happening), you will have your BEST bullpen arm available...

The last thing a team wants is to use their best bullpen pitcher and then NEED him later on...

and how about the intimidation factor against opposing offenses, once the closer comes in the game feels over... those batters feel that....

 

[Omar 382]

Its a mindset more than anything else...

you leave your best BP pitcher till the end, then there are 3 scenarios

1. Your team scores enough runs to not need him
2. your team loses the lead and does not use him
3. your team uses him...

so there are 2 scenarios where you can save your "closers" arm... and one where you use him...
there are 2 scenarios where you either win or get to the end with the lead...

By pitching him whenever the middle of the lineup or in meaningful times(not necessarily the end) then you MUST use him, and you are stuck using him... there is no chances of not using him...

Then, if you use the closer before the end, then what happens if that moment in the game he is used is NOT the most important time?? then you have to rely on a lesser pitcher later!!!

The whole use the best reliever at the most important times is all hindsight...

Yes, after the game is done I would have probably wanted my closer to pitch at different moment, but really who knows??

by leaving the closer till the end, you are also protecting yourself from the worst case scenario(even if you increase the odds of that happening), you will have your BEST bullpen arm available...

The last thing a team wants is to use their best bullpen pitcher and then NEED him later on...

and how about the intimidation factor against opposing offenses, once the closer comes in the game feels over... those batters feel that....

What if the highest leverage situation is the ninth inning? Also, not using him in the 7th inning because you might need him in the 9th is like not taking your only Asprin when you have a bad headache because you might have a worse one later- it doesn't work like that. Sure, there will be times you will use him too soon, but if you average it out, you should use your best reliever at the most important time.

 

[DragonfromTO]

Its a mindset more than anything else...

you leave your best BP pitcher till the end, then there are 3 scenarios

1. Your team scores enough runs to not need him
2. your team loses the lead and does not use him
3. your team uses him...

so there are 2 scenarios where you can save your "closers" arm... and one where you use him...
there are 2 scenarios where you either win or get to the end with the lead...

By pitching him whenever the middle of the lineup or in meaningful times(not necessarily the end) then you MUST use him, and you are stuck using him... there is no chances of not using him...

Then, if you use the closer before the end, then what happens if that moment in the game he is used is NOT the most important time?? then you have to rely on a lesser pitcher later!!!

The whole use the best reliever at the most important times is all hindsight...

Yes, after the game is done I would have probably wanted my closer to pitch at different moment, but really who knows??

by leaving the closer till the end, you are also protecting yourself from the worst case scenario(even if you increase the odds of that happening), you will have your BEST bullpen arm available...

The last thing a team wants is to use their best bullpen pitcher and then NEED him later on...

and how about the intimidation factor against opposing offenses, once the closer comes in the game feels over... those batters feel that....

I think it's worse to lose a close game without even using your best (bullpen) weapon. I'd rather use him at what ends up being a less than perfect time than lose a close game without using him at all.

 

[SlinkyRedfoot]

Closers are overrated. It's a really nice luxury, but if I were building a team, that'd probably be the last position I'd fill.

 

[SlinkyRedfoot]

Joseph F. Borowski led the majors in saves in 2007 with a 5.07 ERA.

There are a lot of ways to fuck a cat. Er. I mean skin. Yeah, a lot of ways to skin a cat. . .

 

[Omar 382]

I think it's worse to lose a close game without even using your best (bullpen) weapon. I'd rather use him at what ends up being a less than perfect time than lose a close game without using him at all.

Do you understand what I did now?

 

[DragonfromTO]

Do you understand what I did now?

I haven't had time to look at it yet.

 

[uncfan103]

Its a mindset more than anything else...

you leave your best BP pitcher till the end, then there are 3 scenarios

1. Your team scores enough runs to not need him
2. your team loses the lead and does not use him
3. your team uses him...

so there are 2 scenarios where you can save your "closers" arm... and one where you use him...
there are 2 scenarios where you either win or get to the end with the lead...

By pitching him whenever the middle of the lineup or in meaningful times(not necessarily the end) then you MUST use him, and you are stuck using him... there is no chances of not using him...

Then, if you use the closer before the end, then what happens if that moment in the game he is used is NOT the most important time?? then you have to rely on a lesser pitcher later!!!

The whole use the best reliever at the most important times is all hindsight...

Yes, after the game is done I would have probably wanted my closer to pitch at different moment, but really who knows??

by leaving the closer till the end, you are also protecting yourself from the worst case scenario(even if you increase the odds of that happening), you will have your BEST bullpen arm available...

The last thing a team wants is to use their best bullpen pitcher and then NEED him later on...

and how about the intimidation factor against opposing offenses, once the closer comes in the game feels over... those batters feel that....

This doesn't compute.

The scenario you gave...you want to pitch the stud pitcher first, it guarantees you use him and it guarantees you use him during an important, high leverage part of the game. Getting through that is important. Saving him for the possibility of another higher leverage situation is risky and doesn't make sense - it would imply there were two high leverage situations. You're going to want to pitch your two best pitchers in those two situations.

I'd rather my best pitcher pitch during a threatening part of the game rather than the bottom with a three run lead and the 6-7-8 hitters coming up. Because the percentage of pitchers likely to get those last threes outs is very high.

 

[MilkSpiller22]

This doesn't compute.

The scenario you gave...you want to pitch the stud pitcher first, it guarantees you use him and it guarantees you use him during an important, high leverage part of the game. Getting through that is important. Saving him for the possibility of another higher leverage situation is risky and doesn't make sense - it would imply there were two high leverage situations. You're going to want to pitch your two best pitchers in those two situations.

I'd rather my best pitcher pitch during a threatening part of the game rather than the bottom with a three run lead and the 6-7-8 hitters coming up. Because the percentage of pitchers likely to get those last threes outs is very high.

Never underestimate the power of the safety net... If you use up your best BP pitcher early, you have lesser pitchers pitching later and you have no safety net...

You can not say i want my "closer" to pitch the inning of highest leverage, because that is all hindsight... If you want to have him pitch on first high leverage, then how will you define high leverage?? Will it be the same vs different teams?? Forget the fact that pitchers are the biggest creatures of habit and most superstitious in sports...



Also, you don't realize how overused they would be if managers put them in high leverage times instead of the end...

What manager would take his BEST BP pitcher out of the game as long as he can get an out?? what this tells me is that these pitchers would go 2 innings more often and maybe even longer... And what that means is that they will NEED more off days...