The save statistic is simplistic and stupid
- Thread starter Smiling Joe Hesketh
- Start date
I love this. I can't wait until they miss out on the playoffs by 1 game again. But it's okay. Season long consistency is more important than putting the team in a better position to actually win games. Bow before the almighty save. It's just one game. One game never matters that much, right? Right?
I think it would be helpful if you could quote the post you're referring to so we can know which goalpost it is you'd like to move.
Barnes has allowed 44 baserunners in 29 2/3 innings - that's not pretty decent, it's pretty awful. He's wild and gives up a ton of hard contact (24.4 LD%, 39.8 hard-hit%) without having an extraordinary strikeout rate for a reliever. The Twins have a below average offense but it's not baseball's worst and having Matt Barnes face Joe Mauer is unquestionably a negative equity proposition. That's a situation you want to stay away from in a tie game in extra innings when your relief ace is as rested as he can possibly be.
Maybe if we got to the 11th Kimbrel would have seen action. Pedro Strop pitched the 9th and 10th in that May 18th game. Pedro Strop and Fernando Rodney combined to pick up the 9th and 10th in that Sept 16th game. So it seems like Joey Baseball held off on going to his closer in both those situations.
If they have a computer model that says they would score more runs by hitting David Ortiz leadoff, should they do it? Should they have benched him 100% vs LHP those years he was struggling against them?
Maybe! I count at least six times since the beginning of 2015 when Farrell could have brought in his closer to a tie game on the road and he's never done it, but anything's possible!
Obviously every game matters, but 162 games is a long, long time. You can't just manage for that one game. You just can't. If you could, you'd never sit your best players. You'd always pitch your best guys and you'd always play your best players and your team would wear down. Yes, I get that wearing down wasn't an issue with Kimbrel yesterday. I'm just speaking as a general principle.
Oh don't back down now. You really thought you were making a slam dunk point when you consecutively posted those Cubs examples.
No one is suggesting never resting players or riding relievers hard. There's just not much of a rationale to having a beyond rested Craig Kimbrel sitting in the bullpen while Matt Barnes blows the game.
Did you want to share any evidence that Farrell is inclined to use his closer on the road in a tie game, that other good managers don't do it, or that it's a bad idea?
Not true. kieckeredinthehead was kind enough to look through Joe Maddon's game logs and found us at least 2 examples where Maddon held off on going to his best pitcher.
You're being willfully obtuse. kiekered showed examples where Maddon at least eventually went to his closer in a tie game on the road. Can you find us any examples of Farrell doing so? The burden is on you here.
If you combine the examples he provided with the fact that Farrell has never done it at any point in any extra inning game over the last 1.5 seasons then I think the point is pretty clear.
Well yeah except Strop is much better than Barnes and for all the flack he gets, Rodney is still better. When you bring in Matt Barnes to face Joe Mauer leading off an inning, there's a decent chance it ends in a run.
Oh wait I found one. Bottom of the 8th in St. Louis, tied, Brandon Workman gives up a single, stolen base, and intentional walk, but Cardinals don't score. Workman strikes out looking in the ninth. After giving up a single in the bottom of the ninth, Workman is pulled, and Farrell puts Koji in to a tied game on the road.
Here you go. Koji pitched the 9th and 10th innings of a tie game on the road in Oakland.
http://www.baseball-reference.com/boxes/OAK/OAK201307140.shtml
http://www.baseball-reference.com/boxes/OAK/OAK201307140.shtml
Last year Koji pitched twice in tie games on the road. One was 7/17 @ LAAAAAAAAA where he lost the game. The other was 5/11 @ Oak where he came on with 1 on / 1 out, got out of it, and then the Sox went on to win.
I don't believe this is entirely true. It may be changing, but the best set-up guys don't usually get the name recognition and the pay that the best closers do, even if they are of equal quality. I mean, until recently, saves factored into determining free agent compensation (as did wins for starters). Guys are going to want the save as long as it exists as a stat.
Last year Koji pitched twice in tie games on the road. One was 7/17 @ LAAAAAAAAA where he lost the game. The other was 5/11 @ Oak where he came on with 1 on / 1 out, got out of it, and then the Sox went on to win.
Sure I agree that the players put value on it. I was really talking about it from the perspective of the FO. But there are two things that we also have to consider:
1. The best relievers usually become closers anyway. It's not like naming someone the closer is an arbitrary decision. So by definition the closers will generally be the highest paid relievers.
2. Teams have been looking beyond just closers lately when paying big money for a reliever. Let's put it this way- if Betances and Boxberger were both free agents today, who would get paid more (assuming hypothetically that both were healthy)?
Thanks for finding those. Farrell did this pretty consistently in 2013 and 2014, bringing Koji in not just to a tied game but in the ninth inning of a tied game. That was good. I wonder why they've gotten away from it.
2013
koji in a tie at Oakland:
http://www.baseball-reference.com/boxes/OAK/OAK201307140.shtml
koji in a tie at toronto
http://www.baseball-reference.com/boxes/TOR/TOR201308130.shtml
2104
koji blew the save in the 9th, stayed in for the 10th at Oakland:
http://www.baseball-reference.com/boxes/OAK/OAK201406220.shtml
koji tie on the road at LAA:
http://www.baseball-reference.com/boxes/ANA/ANA201408090.shtml
2015
koji tie on the road at LAA
http://www.baseball-reference.com/boxes/ANA/ANA201507170.shtml
koji in a tie at Oakland:
http://www.baseball-reference.com/boxes/OAK/OAK201307140.shtml
koji in a tie at toronto
http://www.baseball-reference.com/boxes/TOR/TOR201308130.shtml
2104
koji blew the save in the 9th, stayed in for the 10th at Oakland:
http://www.baseball-reference.com/boxes/OAK/OAK201406220.shtml
koji tie on the road at LAA:
http://www.baseball-reference.com/boxes/ANA/ANA201408090.shtml
2015
koji tie on the road at LAA
http://www.baseball-reference.com/boxes/ANA/ANA201507170.shtml
So since there's precedence for Farrell bringing in his closer/relief ace in a tie game on the road, can someone explain why it was a good idea to have Barnes and his .400 OBP to LHH face a guy leading off an inning that's not going to chase his frequent out of zone offerings when Kimbrel has barely pitched the last month?
I think it just gets back to the original point. Usage seems to be dictated by the save stat. Not just by Farrell, but by most managers, even Maddon most of the time. That's not a very good answer. Sometimes, "that's the way the game is played now" has led to good things. Sometimes not.
FWIW--Sox are 3-1 in road extra inning games.
Saves and wins are both pretty bad statistics. Managers will often try to get a starter through the fifth just to get him the win, even if he's totally on the ropes in the fourth. You never want to manage a game driven by statistics.
I hate to think about it, but Kimbrel's recent usage has me wondering if he's nursing an injury; his isage pattern looks a lot like Carson Smith's before he was DLed
If he'd lost? I bet it would. If the Red Sox lose any game someone here will try to blame it on Farrell.
Saves have only been an official stat since 1969. They're arbitrary. They were invented to try and give credit to pitchers who protect leads (not reflected in W-L) and punish pitchers that allow inherited runners to score (not reflected in ERA).
Why they ended up defined as they are is puzzling. Among the criteria for a save are:
On the other hand, closers are incredibly valuable and it would be nice to have some sort of statistic that reflected how good or bad they were. What are the key elements of relieving (with some overlap)?
- Not allowing an inherited runner to score.This quality changes with the situation: Extremely difficult with a runner in scoring position and no outs (on one hand) and bases empty with 2 outs (on the other). The latter example credits a reliever with allowing 3 singles before getting one out the same as a reliever striking out the side with a man on 3rd.
- Not giving up hits or walks. Not giving up extra-base hits. Regardless of the situation.
- Avoiding contact (regardless of the situation) or inducing weak contact.
- Success in high-stress situations. Handling good hitters versus shitty hitters. High leverage situations late in the game (fewer at bats left to remedy poor pitching or fielding).
- Non-handedness. Succeeding against both LHH and RHH.
- Not being penalized by bunts, errors or passed balls.
- Holding runners to the inherited base.
Is there even the possibility of a fair stat? How do you account for pitchers that are only secure in their role? Pitchers who are used to warming up in the 8th and coming in clean in the 9th versus guys that can appear any time in any situation and succeed against the worst odds?
I have no idea.
Why they ended up defined as they are is puzzling. Among the criteria for a save are:
- Pitcher enters the game with a lead of no more than three runs and pitches for at least one inning, or...
- Pitcher enters the game, regardless of the count, with the potential tying run either on base, at bat or on deck, or...
- Pitcher pitches for at least three innings.
- Why not 2 runs? Or 4? I guess the "3" comes from a runner on base + a hitter at bat (getting on base) + a hitter on deck (hitting a home run). Nonsensical.
- Why the tying run on deck? Why not on-base? Why not in scoring position? Arbitrary.
- Why 3 innings as opposed to 2?
On the other hand, closers are incredibly valuable and it would be nice to have some sort of statistic that reflected how good or bad they were. What are the key elements of relieving (with some overlap)?
- Not allowing an inherited runner to score.This quality changes with the situation: Extremely difficult with a runner in scoring position and no outs (on one hand) and bases empty with 2 outs (on the other). The latter example credits a reliever with allowing 3 singles before getting one out the same as a reliever striking out the side with a man on 3rd.
- Not giving up hits or walks. Not giving up extra-base hits. Regardless of the situation.
- Avoiding contact (regardless of the situation) or inducing weak contact.
- Success in high-stress situations. Handling good hitters versus shitty hitters. High leverage situations late in the game (fewer at bats left to remedy poor pitching or fielding).
- Non-handedness. Succeeding against both LHH and RHH.
- Not being penalized by bunts, errors or passed balls.
- Holding runners to the inherited base.
Is there even the possibility of a fair stat? How do you account for pitchers that are only secure in their role? Pitchers who are used to warming up in the 8th and coming in clean in the 9th versus guys that can appear any time in any situation and succeed against the worst odds?
I have no idea.
If I were to create a save stat today, it would be based on a WPA threshold. If a pitcher finishes someone else's game, and they had a WPA higher than some value, they get a save. I don't know how I'd choose that value. It would probably be arbitrary.
I was going to post in the Farrell thread.... but this is good.
WTF?!?!??! Is that Grady Little level bonedheaded or runner up?
The thing is.... Buck has used Britton in a 4 out non-save situation on the road IN TORONTO!!!! earlier in the season!!!! Holy fuck. He got lucky earlier in the game not going to Britton, I'm guessing his earlier brain fart made him feel invincible? Wow.
Get him in there after the second single and bring him out the next inning again. If it's still a tie, then at least you went out fighting and being smart about it. Crikey!
WTF?!?!??! Is that Grady Little level bonedheaded or runner up?
The thing is.... Buck has used Britton in a 4 out non-save situation on the road IN TORONTO!!!! earlier in the season!!!! Holy fuck. He got lucky earlier in the game not going to Britton, I'm guessing his earlier brain fart made him feel invincible? Wow.
Get him in there after the second single and bring him out the next inning again. If it's still a tie, then at least you went out fighting and being smart about it. Crikey!
Britton was better against RHB than LHB batters this season.
The Jays' 3-4-5 hitters were due up.
1st and 3rd, 1 out, you need a K or a DP ball. Britton is excellent at inducing either.
It was a situation where a single run ends their season, but Buck was saving him for a situation where a single run would not end their season.
Both Britton and Buck have stated that Britton was healthy.
Really a Grady-level blunder at work here.
The Jays' 3-4-5 hitters were due up.
1st and 3rd, 1 out, you need a K or a DP ball. Britton is excellent at inducing either.
It was a situation where a single run ends their season, but Buck was saving him for a situation where a single run would not end their season.
Both Britton and Buck have stated that Britton was healthy.
Really a Grady-level blunder at work here.
2 outs in the inning. O'Day got the first batter, then Duensing struck out the 2nd, before Ubaldo Jimenez was summoned with 2 down and no one on.
That said, it was an egregiously stupid bit of "look at me!" managing from the master of making himself the Show. He went to the mound FOUR times in the same inning. He convened a team meeting just before Jimenez's fourth pitch was blasted over the wall.
That said, it was an egregiously stupid bit of "look at me!" managing from the master of making himself the Show. He went to the mound FOUR times in the same inning. He convened a team meeting just before Jimenez's fourth pitch was blasted over the wall.
The Saves statistic didn't have anything to do with Showalter's stupidity last night. No one cares about save stats in the postseason, just winning games.
Buck was waiting for a lead that never came before using Britton, which is dumb. But he wasn't doing that so that Britton could get another save for his stats, he was just being dumb. The stat had nothing to do with it.
Buck was waiting for a lead that never came before using Britton, which is dumb. But he wasn't doing that so that Britton could get another save for his stats, he was just being dumb. The stat had nothing to do with it.
No, Duensing started the inning by getting Carrera, then Buck brought in
One could argue that the save statistic brought about the rigid, defined role of "closer", which would make the manager more hesitant in using his best reliever in the highest leverage possible, instead saving him for when the team has a lead.
Yup. And if you have Brach and O'Day, I can see not using him in the bottom of the ninth or even the tenth. But there's no way the heart of the Jays order should have come up in a walk-off situation against Ubaldo the hell Jimenez. Get them there, and you have the heart of your own order coming up in the top of the 12th, which gives you the best hope of having Britton save his own win.
O Day didn't pitch in the inning. Carrera Kd then Jiminez came in for Travis. One out when winning run scored. Then Buck decided, needing a double play or K, to let EE face a terrible fly ball righty which is the exact kind of pitcher EE murders.
Probably should have had Britton in for Donaldson with the idea he'd face the entire heart of the order and the hope he'd get all of them out and take the game through the 12th.
That would be wrong when talking about last night though. In the postseason, good managers have no hesitation about using their closers in non-save situations. Look how Torre used Mariano, for just one example.
As posted above, Showalter himself used Britton for multiple innings on the road in Toronto in a non-save situation in the regular season, so he clearly has no problem breaking from the rigidly defined use of his closer, even in the regular season.
His stupidity last night was all on him, it didn't have anything to do with the stat.
You don't think having a reliever entrenched as the "closer" plays a part in how the manager utilizes him, even in a postseason game? I'm not arguing that Showalter's decision was justifiable, but how often do you see the team's closer come in a one run game with the bases loaded and one out in the 7th inning in the playoffs, for instance? It's extremely rare, and the underlying reasoning for that, as well as waiting for your team to get a lead before putting your best reliever in, can be attributed, to some degree, to the value placed in the save as a statistic, and the manner in which it shapes those bullpen roles in the first place.
Britton is the closer. Closers get saves. He was saving his closer for a lead in extras. That's all there is to it.
Britton had not pitched in 6 days when he was brought him into that tie game in Toronto. Showalter pitched him even when losing if Britton had not pitched in 5+ days all season long. That one game is an exception out of his 69 appearances- it's the only tie game on the road all year- so I'd argue that was not about the situation as much as getting him some work.
Showalter just treated this like a regular season game, and let the save stat dictate the way he used his best pitcher.
Showalter just treated this like a regular season game, and let the save stat dictate the way he used his best pitcher.
Does Buck manage Britton as if he has a reverse split? I wonder if he was hoping to use him in an inning where Saunders was guaranteed to bat. FWIW I think most managers with a lefty closer and righty pitchers with normal splits would fall into a similar trap against the Toronto 2 through 6. It's wrong and it's a mistake, and he absolutely deserves criticism for not using Britton at all in the most important game of the Baltimore season, just that may have been his thinking.
Britton has even better numbers against righties than lefties.
I think Buck had in his head that Britton could only give him one inning after getting five outs on Sunday, but still I think you need to use him earlier and certainly once it got to Encarnacion. If worst comes to worst, Tommy Hunter did do some closing in 2014 (11 saves), so it wouldn't have been a first-time thing for him.
I think Buck had in his head that Britton could only give him one inning after getting five outs on Sunday, but still I think you need to use him earlier and certainly once it got to Encarnacion. If worst comes to worst, Tommy Hunter did do some closing in 2014 (11 saves), so it wouldn't have been a first-time thing for him.