Indians Go Unconventional, Again
(Photo: Erik Drost)
CLEVELAND — While a lot of talk here and elsewhere has focused on bullpen-ing (not such a bad idea, right, Yankees?) and creatively leveraging pitching staffs early in the postseason, the Indians appear to be pursuing a different kind of unconventional strategy.
The club announced yesterday that, while staff ace and AL Cy Young candidate Corey Kluber will be fully rested for Game 1, he will not pitch Thursday against the Yankees. Trevor Bauer, one of the game’s best starters in the second half, will get the ball instead.
In this unexpected scenario, Kluber will start Game 2 and, if necessary, Game 5. Carlos Carrasco will start Game 3 in New York. Carrasco has been a significantly better road pitcher than home pitcher for his career (3.13 road ERA vs. 4.45 home ERA, 3.86 home FIP vs. 2.98 road FIP). He could face Luis Severino in that game, with Severino on regular rest. Or, since Sevenrino didn’t make it through an inning Tuesday, he could perhaps be available in Game 2. Josh Tomlin is penciled in as the tentative starter in Game 4 for Cleveland
It’s not often you see a team with a fully rested, legitimate ace choose not to use him at the first opportunity in the postseason. Indians manager Terry Francona explained the decision to a crowded media room Tuesday in the depths of Progressive Field.
“We want to keep Kluber on his day [regular day of rest], that was the reason for two and five,” Francona said.
Francona said Kluber was involved in the process. As early as a “couple weeks ago,” he’d sat down with Kluber to talk about aligning the postseason rotation.
“I think on a number of reasons it makes good sense,” Francona said. “Not that you’re going into this game [Game 1 on Thursday] thinking you are going to lose, but if you do, you have your ace coming back. The biggest thing was keeping him on his fifth day. That was really important to Kluber. That was the only way we could do it. Don’t want to put cart ahead of the horse, but if you’re fortunate enough to win in four [games or fewer], you have your ace ready for the next series.”
So the Indians have their reasons. The most prominent being, ostensibly, to keep Kluber on regular rest for his second start if needed. If he started Game 1 and Game 4, the second start would come on short rest. If he started Game 1 and Game 5, the second start would come on extra rest. And if the Indians are betting on themselves being good enough to advance to the ALCS, avoiding a Kluber Game 4 or 5 start would mean that he could open Game 1 of the ALCS on Oct. 13, and potentially start three games in that series when the Indians might become more aggressive with his usage.
Francona said he didn’t want to “overcomplicate” matters regarding the rotation while trying to create flexibility by making Bauer available later in the series after pitching Game 1. While Tomlin is listed as the tentative Game 4 starter, Bauer is available to pitch in that game, as is the wild card that is Danny Salazar, who also made the ALDS roster. Bauer has a resilient arm.
There’s some reward potential here, and it potentially sets the Indians up better for the ALCS should they advance in four or fewer games.
But are the Indians actually the ones overcomplicating matters? You can argue that this decision increases the chances for an upset.
The team that wins the first game of a five-game series has a 68.75% chance to advance, all things being equal in terms of roster strength and home-field advantage — which, of course, is never the case.
Here’s my math …
In this specific case, that percentage would be less severe for the Indians, since Kluber would still make two potential starts and Cleveland has home-field advantage.
But Game 1 could be a more favorable matchup for Kluber and the Indians, as Severino could potentially pitch in Game 2 following his short-lived outing in the Wild Card game. If you believe in things like added player pressure, you could argue this places more pressure on Bauer in Game 1 — and potentially on Kluber in Game 2 — if the team is trying to avoid an 0-2 deficit. There’s also the argument to be made that the team should want Kluber available on three days rest to pitch a Game 4 if the club is down 1-2 in the series. This is a formidable Yankees team that is second only to the Indians this season in run differential in the majors and has built an uber pen.
How many times have we heard coaches and mangers tell us about playing it one game at a time? This runs contrary to that cliche and approach, which is interesting and debatable.
The Indians don’t have any bad options available to them.
Bauer has had an excellent second half and enjoyed a breakout season. Jeff noted recently that Cleveland might actually have the best pitching staff ever. Francona and the Indians pushed many of the right buttons last year to advance to the World Series with an injury-depleted staff. The club employs a bunch of really smart people who have more information than we have on the outside; they have their reasons (do they want Kluber on extra rest to begin October anticipating heavy usage?), and this decision could provide reward.
But Kluber has been the best pitcher on the planet since he returned from the disabled list earlier this season. If the Indians lose in four games or fewer in the ALDS, Kluber will have only pitched once in the series. That will be second-guessed for the entire offseason. If the Indians drop Game 1 without putting their best option on the mound, they’ve increased their chances of being upset.
The Indians don’t have any poor options, really, but this is nonetheless an interesting risk/reward play.
A Cleveland native, FanGraphs writer Travis Sawchik is the author of the New York Times bestselling book, Big Data Baseball. He also contributes to The Athletic Cleveland, and has written for the Pittsburgh Tribune-Review, among other outlets. Follow him on Twitter @Travis_Sawchik.
The odder move in my opinion was Gio Urshela making the roster. That means he’s likely the starting 3rd baseman (if Ramirez was going to start at 3rd, Urshela would have no point being on the roster). That puts Kipnis in center and Greg Allen coming in as a late inning defensive replacement.
But why not just start Allen in center and have Kipnis play second? That puts everyone at their best position, gains a little speed, and sacrifices nothing in terms of offense.
I think having Urshela makes sense, but Gonzalez over Yandy Diaz is super weird. Gonzalez has barely played since they called Diaz up at the end of August. I also disagree with the Kipnis thing but they are going with that for some reason. I guess they think it more likely that Kipnis’ defense will be more costly in the infield compared to Ramirez, that he’ll have more opportunities. Looking forward to easy fly balls dropping in front of Kipnis and guys going 1st to 3rd at will on him
I was surprised by Gonzalez as well. I guess the thinking is that, if you pinch hit for Urshela, you’ll need to replace him at 3rd. And they’d rather have Gonzelez’ defense than Yandy’s offense. Plus, if there’s an injury, Gonzalez can cover more positions.
It seemed like they had been grooming Urshela as the backup infielder, because he got time at 2B and SS in addition to his regular 3B. My expectation was Diaz starting games and Urshela coming off the bench as a defensive replacement at 3B, but also being available for SS and 2B in an emergency.
I think if you were going to put Kipnis back at 2B, you wouldn’t have bothered with him in CF in the first place. He didn’t fail miserably out there, so they’re continuing with that plan because they like Ramirez at 2B and still want Kipnis’s bat in the lineup.
The question is whether or not Urshela can be servicable at the plate. I guess if Brantley is only going to pinch hit, Urshela will be limited to probably 2 ABs per game, depending on the score. They must value the glove at 3B enough to hit a guy who hits basically just slightly better than a pitcher.
Or put Jackson at center? One of the networks put together a montage of Kipnis in CF and it was bad. He could easily cost them more runs then he creates.
The ball almost never makes it to Center, especially with Lindor and Ramirez up the middle. You are looking at probably 1 and up to 3 balls going his way, with the likelihood a superior play is needed on any of them being rather small
Compare that to the number of times the ball will go to 2nd or 3rd and be handled by their best fielding lineup
Kipnis may possibly allow a Hit/Run, but Hits/Runs will almost definitely be decreased with an amazing fielding Ramirez-Lindor-Urshela infield.
Wait, is there a reason why fly balls are unlikely to happen? Did I miss something?
Sure they happen. Not very often, and most all are routine, but they happen
Specifically the team averages more than 10 Ks a game. That leaves 17 outs to the fielders. Of those 17, the team has been giving up a 33% FB ratio. That’s 5.6 a game. And of balls in play they also have seen 35% going to center. So of the 6 FB, maybe 2 of them over 9 innings should go to Center.
That means Kipnis plays 7 innings or so out there and he might see both, maybe just 1, 3 on a busy day
And so far he has played 71 innings in CF. In those 71 innings he has seen 17 balls hit into his zone – or one every 4.2 innings. So that also has been the same 1, maybe 2, 3 on a busy day
Then of those 17 he has seen, 14 have been routine (he caught all) 2 have been remote (he didn’t catch) and 2 have been impossible (he didn’t catch) …the likelihood of something outside routine is very low
So as I said to start, he’s likely getting 1, and up to 3 balls his way, with the likelihood of a spectacular catch being asked of him being pretty slim
It was hardly some remarkable statement, it’s just reality
That’s crazy. Honestly, it sounds like an outlier. But maybe not. Is there a way to check to see whether this was true of Cleveland CFs as a whole?
This argument would support putting EE in center field if so many balls are “routine”. Every single inning that an inferior fielder is in a position for which he’s under qualified increases the probability for him to be asked to make a play he’s incapable of making.
@sadtrombone
It’s actually not even an outlier, the ball doesn’t go to CF often. And yes, most all times they do are pretty routine. Now with Cleveland having such a superior pitching staff they might see an even higher rate of routine plays on them, but as a whole CFers do not touch the ball as much as people apparently think
@Buhners Rocket Arm
That is reaching in a spectacularly pitiful fashion. EE is a poor fielding, now 1B limited player. Kipnis is an athletic, natural CFer who Cleveland decided to stop putting out there about 7 years ago
As Kipnis showed yesterday, he’s even still capable of making the occasional spectacular play. Took him 74 innings before he needed to show much athleticism though; otherwise he has just needed to tap into his history and read the ball halfway adequately.
And although he finally did see (and complete) his first difficult play, he otherwise got the ball just once. The very limited numbers of touches hold again.
Not sure why this has been responded to the way it has. An article on this site was written on this very topic – the lack of chances for Kipnis as he was trying out CF. The team strikes people out at a record pace, therefore there are fewer fielding opportunities for everyone.
https://www.fangraphs.com/blogs/an-update-on-the-jason-kipnis-center-field-experiment/
Not to mention (and this may be arbitrary or the Indians staff may know something more in depth about these numbers) that they have the 7th lowest hard hit ball %, 8th lowest fly ball percentage, but the 9th highest infield fly ball percentage. So maybe with more than a quick glance at that data the Indians can make assumptions that the risk of Kipnis in the field is lower than the reward of Kipnis at the plate.
Any other team that Cleveland faces and I’d be for the move. But the Yankee bullpen is a major X factor in the playoffs – especially in a shortened series – that Francona might be underestimating.
not sure how the other team’s bullpen affects how you manage your own pitching staf
Especially when the Indians have a stellar pen themselves. It’s not like they rely solely on their starters and have a shaky pen.
Great article and analysis. Couple of typos here, just mixing up home and road for his career ERA numbers.
“Carrasco has been a significantly better road pitcher than home pitcher for his career (3.13 home ERA vs. 4.45 road ERA, 3.86 home FIP vs. 2.98 road FIP). “
Fixed. Thanks!
BTW, one theory is that this would allow Francona to bring Bauer back for game 4 (if the Indians are down 1-2), given the fact that he has a “rubber arm”.
But that raises the question of what the heck Tomlin is doing on the roster in the first place. If he’s not the game 4 starter, then what role does he have? He’s thrown 25.2 innings in relief in his career and has an 8.06 ERA. Yeah it’s limited data but that’s not good. As for Tomlin starting game 4, he has an 8.57 ERA in 21 innings in Yankee stadium. Again, that’s not even close to good.
Meanwhile, three relievers who pitched 50+ innings and had ERA’s under 3.00 were left off the roster (Otero, Goody, and McAllister).
I know Francona’s won two World Series, and I haven’t, but I’m very confused.
he’s the game 4 starter (less important of a role as Clevinger’s relief role). He can be used in an emergency in Game 1-2 as well since he’s really only expected to go around 3 innings in Game 4 anyway
Yep. I really don’t think Clevinger is going to be a long reliever – that seems like Salazar’s role since they were still trying to stretch him out at the end of the season. I think they’re going to utilize him in higher leverage situations against big time righties like Judge and Sanchez because of his splits. Maybe they can use him to help keep Miller’s usage down a little bit. For example, if Judge leads off an inning you use Clevinger to (hopefully) cut down the two righties and then bring in Miller afterwards. You’d still have Smith and Olson as matchup guys as well for lower leverage situations. If you do this over the course of 3 series, you have a much fresher Miller.
I assume they plan on going with Tomlin, to be followed by Salazar and/or Clevinger, if they are up 2-1 going into game 4, and that way if they won they’d make it through the series with their top 3 starters well rested. If they are down, then they may go with Bauer instead. But I agree, I wouldn’t have Tomlin on the roster and would’ve taken Otero along. Tomlin, who is likely to give up a home run at any moment, would have no business pitching in relief in any close game.
I would prefer Otero to Tomlin. I think this is a case of Francona being loyal to his friends, and I hope it does not cost them a series.
Francona cited rest, and if you look at the World Series last year, where Kluber pitched 3 games, it’s clear rest was a factor. He was dominant Game 1, then OK Game 4, then not very good in Game 7. Francona may feel Kluber on short rest is not a good idea, which is why he is going this way….
who else thinks if this is anyone but Francona, they’re roundly criticized and said that they’re outthinking the situation?
I think it’s a good move. Uninformed fans would criticize any manager
This is going to sound weird, but Francona’s ability to sell totally unconventional moves to the players is the reason why he’s a good manager. That doesn’t mean we have to agree with it (I buy it, no strong opinions either way), but I doubt he’s getting second-guessed in the clubhouse.
and it seems everyone *is* criticizing Francona and saying he’s outthinking the situation, this fangraphs article aside. I guess I’m not following your implication that somehow Francona is above reproach, that isn’t what I’m observing at all.
I think you’d have a better chance of advancing this theory if it was Maddon, not Francona.
yeah i compare what Maddon and Francona are now to what you see in the NFL with Bill Belichick. They get some pushback- but no where near as much as if someone like Joe Girardi or a number of other managers would be doing it.
I don’t really think it makes much difference if Kluber starts game 1 or 2, actually if Severino is available for game 2 I’d rather have Kluber going against him. If game 2 gets rained upon though this could backfire, as they could either end up with a rain delay after it started and having to pull Kluber early, or if the game got rained out they’d be back to KLuber on short rest in a potential game 5
Weather is looking like a real issue for Game 2
In the playoffs, a forecast of rain often (usually?) causes MLB to postpone the game outright, even before the rain begins. If so, that could prevent the Indians from using Kluber in game 5 if it gets there, thus defeating one of the main purposes of this decision.
If the Indians win in 4 or 3, Francona looks like a genius for having Kluber ready to go in game 1 of the ALCS. If it rains and goes 5 and Kluber only pitches game 2…totally different.
I bet Kluber starts Game 5 no matter what.
Very good point on the risk of a Game 2 rainout, and I think that is the main reason not to use this strategy. That’s the big downside for the Indians, and for teams in general.
I’m not that swayed by the argument of having Kluber potentially available on short rest for Game 4. My take from prior Fangraphs analysis on this question ( https://www.fangraphs.com/blogs/the-good-news-about-corey-kluber-on-short-rest/ and https://www.fangraphs.com/blogs/giving-a-questionable-start-to-baseballs-best-starter/ ) is that pitchers on short rest as a group end up performing about like what you’d expect to get out of Josh Tomlin, even though the short rest starters are by nature chosen to be a group that includes predominantly ace-level pitchers.
While Travis’s analysis is interesting, I don’t see “odds of winning the series after winning or losing Game 1” as a relevant metric to use for this decision. In either of the scenarios that we are considering, the Indians are going to start Kluber and Bauer (in some order) in Games 1 and 2. Logically speaking, the probability of any outcome for the Indians after those two games (up 2-0, tied 1-1, or down 0-2) should be independent of the order in which those two pitchers start Games 1 and 2.
So the decision comes down to pitcher availability later in the series for potential Games 4 and 5. Any potential for a Game 2 rainout pushes toward using Kluber in Game 1. While I don’t think that starting Kluber on short rest in Game 4 would be wise, a differing view on that question would also push toward using Kluber in Game 1.
The Yankees may be the best team in the postseason not named the Cleveland Indians. And in a five-game series there is zero room for error.
I don’t like this move from the Indians at all.
I can only make sense of it if they think Bauer on short rest will be more effective than either Kluber or Carrasco on short rest.
If the ‘reward’ side of this risk/reward play is anything other than winning the ALDS then it’s not worth it. Setting Kluber up for the ALCS is a terrible reward if it makes you worse in the ALDS.
The main argument that I see is that maybe they prefer having Bauer potentially available in the bullpen for games 4-5. So whether you’re up or down in game 4 you can still potentially roll Bauer for a couple innings in relief if you need to, whereas you’re not going to bring Kluber in relief in game 4 had he started game 1, since you’d probably rather just have him start game 5.
The problem is that game 4 is likelier to happen than game 5. If you’re down 2-1 wouldn’t you like to have at least the option of Kluber on short rest, knowing you still have Carrasco for game 5 if required?
The other question is why they prefer Bauer on short rest to Carrasco on short rest. Carrasco is a substantially better pitcher. The Indians have two really good starters (Kluber and Carrasco) and the way they’ve set up the rotation those two will only affect three games in the series when it could be four.
Presumably it’s because Bauer is considered to have a rubber arm and can be more effective on short rest (relative to Carrasco).
Also, assuming the Indians advance, this guarantees they’ll have either Kluber or Carrasco available to start game 1 of the ALCS.
The goal is to maximize your chance of winning the World Series. A trade-off between substantially improving your ALCS odds is definitely worth a slight decrease in your ALDS odds, because your overall Championship odds go *up*
This sounds more like a Maddon move. He’s so smart!
Scherzer may start Game 2, too, because of concerns over his health. If Kluber and Scherzer win the CYA, as seems likely, would this be the first time that both CY winners started game 2 of their team’s first postseason series?
You can’t use the team wins first game argument cuz usually that’s their best pitcher
except they’re usually also facing the other team’s best pitcher…
The argument that the team that wins first greatly increases their chances of winning doesn’t really work. You have the same players playing the same number of innings, it doesn’t really matter what order you do it in. Fangraphs always preaches that situational things don’t matter, so why are we starting now? I have no problems with this move, Kluber will still make two starts this series if necessary, and this way he’s on normal rest.
If you’d ever played sports you’d understand the difference between playing down and playing up, it affects strategy on the field, player mindset, and preparation; as well as tension, anxiety, and other human physiological responses that all affect performance. These guys are not robots running through a simulation, everything that happens in a game affects everything that happens later in the game and series. It is always best to start off strong in any competition. Always.
I get that, that’s the best argument for starting Kluber 1st (or perhaps the possibility of rain on Friday). The part I had the problem with the part of the article stating that if they lost game 1 they’d have a big disadvantage. Why not assume they lose game 2 if Bauer were to pitch then? It’s the same result with the math.
The problem with the claim that psychology should drive this decision is that a fair number of people have in fact studied the impact of momentum in baseball – both between games and during a game – and found that the impact of momentum ranges between “none” and “very, very small”. That’s a summary conclusion from Russell Carleton writing on this topic a few years ago – http://www.foxsports.com/mlb/story/do-we-give-momentum-too-much-credit-in-baseball-061714
Mentally it’s much easier going into game 2 after winning game 1. It’s not a must win game if you won the first one. It might work out that Bauer wins, but Kluber seems like the most sure bet. Also if you want Kluber to be available for game 5, then don’t pitch him in game 4. Throughout the playoffs you want Kluber to start as many games as possible. Starting him in game 2 reduces the chance for him to start game 1 of the ALCS
Drone Boy was not good in last year’s playoffs. And, besides Kluber, the other options to start are all wild cards to varying degrees, and Carrasco, who was much better away than home this season.
This feels like Francona has decided the reward of getting Bauer off to a good playoff start is more valuable than any risk associated with Kluber not starting Game 1.
I get that pitchers are creatures of habit, but I don’t understand why Kluber is seemingly so averse to an extra day of rest. It only makes sense to me if he’s one of those guys who loses effectiveness on 5+ days of rest.
If you were Corey Kluber right now, would change anything about your routine?
It just seems to me like most every pitcher alters their routine in the postseason. Why should Kluber be any different?
It can’t be the real reason anyway given that Kluber will have five days of rest going into game 2 (he pitched Saturday). If he pitched game 1 he’d be on his normal four days rest.
I love the fact that Francona is openly admitting a truth that other people (with their ‘we just go one day at a time’ cliche) won’t: you don’t win the World Series with one win. You win it with 11 (or 12).
This isn’t an optimal situation for winning one game. But it IS for them winning 11. If you can’t win a game Bauer starts – particularly against a Yankee team with a tired/depleted bullpen – you aren’t winning the World Series anyway.
I’m not an Indians fan, but I love the thought process here.
“If you can’t win a game Bauer starts – particularly against a Yankee team with a tired/depleted bullpen – you aren’t winning the World Series anyway.”
Thanks Grandpa. And don’t draft that Trout kid, he’s got an ugly girlfriend!
I think its more acceptable for a team that just went to the World Series last year to have a more long term approach. On top of that, I think its even more acceptable for a team that lost the way the Indians did last year (guys like Kluber and Miller running out of gas) to have a long term approach.
If the Yankees or Astros tried to do this, I wouldn’t like it as much because neither team (in this current version, at least) has established itself as a team that can win a World Series. Does that make sense?
I ran some numbers too. In the math box above, we’re assuming that every game is 50/50. But the whole point of this article is that pitching Kluber shifts the odds in favor of Cleveland, which is said in words but not in math.
Numbers are (CLE wins series / CLE wins in 4 / Kluber pitches game 1 ALCS)
If the other games remain 50/50, but the Kluber games go to 70/30:
Assuming New York wins Game 1, as in the above math box:
– If Kluber pitches 2 & 5: 47% / 18% / 18%
– If Kluber pitches 1 & 4: 39% / 18% / 0%
Assuming Game 1 is 50/50 (or 70/30 with Kluber) instead of a Yankees win:
– If Kluber pitches 2 & 5: 65% / 39% / 39%
– If Kluber pitches 1 & 4: 65% / 47% / 18%
– If Kluber pitches 1 & 4, but he’s a worse 60/40 on short rest in Game 4: 61% / 43% / 18%
Assuming Game 1 is a Cleveland win:
– If Kluber pitches 2 & 5: 83% / 60% / 60%
– If Kluber pitches 1 & 4: 76% / 60% / 25%
This isn’t comprehensive, of course, but it seems to me like a solid idea to let Kluber pitch his comfortable 2 and 5. The “Must Win Game 1” narrative is tempered by Kluber pitching Game 2, and offset by his availability for a winner-take-all Game 5 on regular rest. A five game series will reach Game 5 about 35-40% of the time, depending on how one-sided the match-up is.
This exercise is making me wonder if this is not just Kluber’s preference, but a truly advantageous way to deploy a #1 starter in a 5-game series.
Nice work. I think your middle table is the important one: what are the chances if we let the first game go by chance instead of assuming it won or lost.
So if Kluber is hurt at all by short rest, this becomes a tradeoff between the ALDS and the ALCS. Now we need to run through how much Kluber’s availability affects the next series’ odds, and chain them…
Where are you pulling this 70% win probability from? That has to based firmly on something for the rest of these numbers to have any meaning.
Cleveland went 20-9 in Kluber starts this year, which works out to a .690 winning percentage. Yes, that includes a lot of games against teams worse than the Yankees. In the 2016 postseason, CLE went 4-2 in his starts (.667): 3-0 on at least full rest and 1-2 on short rest.
So it’s based on that, and partly on what I pulled out of… the air. We are having this article and discussion on which games he starts because we/I think CLE becomes a favorite when he starts.
Cleveland went 82-51 in games that Kluber did not start this year, which works out to a .582 winning percentage. So when Kluber started, the Indians won 18% more games (.690/.582) then when he did not start.
But your math assumes Kluber boosts the Indians chances of winning each game by 40% (70%/50%).
Short story even shorter: If you are going to run numbers (and turn a blind eye to other variables and white noise in the data), you should at least employ math correctly.
This:
Assuming Game 1 is 50/50 (or 70/30 with Kluber) instead of a Yankees win:
– If Kluber pitches 2 & 5: 65% / 39% / 39%
– If Kluber pitches 1 & 4: 65% / 47% / 18%
– If Kluber pitches 1 & 4, but he’s a worse 60/40 on short rest in Game 4: 61% / 43% / 18%
is the only relevant part.
In the 2nd scenario (– If Kluber pitches 1 & 4: 65% / 47% / 18%), you can sit Kluber in game 4 unless you’re facing elimination, which ups the Kluber pitches game 1 ALCS $ back to 39%. You can dock Kluber for short rest but the fact that you can pitch Bauer a second time over Tomlin slightly more than cancels that out in my opinion.
I’m ok with the idea in a vacuum, but it does look like it may rain Friday. If rain is a possibility, I’d start Kluber game 1 and Bauer game 2 for two reasons:
1) Bauer induces more contact, and if the poor weather produces a home run dampening effect, this would play into the Indians favor. Kluber’s strikeouts travel the same distance, regardless of weather.
2) If there is a rain delay for an hour+ the Tribe could lose Kluber which would really put them at a disadvantage.
Good article, but this narrative that Bauer was “awesome in the second half” is just not true. Yes, his ERA improved. Not much else did.
Bauer 2017 first half: .267/.333/.454 for a .337 wOBA.
Bauer 2017 second half: .263/.324/.436 for a .326 wOBA.
He certainly wasn’t “one of the best pitchers in baseball.” The walks came down slightly (from 8.5% to 7.5%) and the Ks went up slightly (from 25.8% to 26.7%). The BABIP actually slightly rose (from .336 to .338) and the HR/FB dipped slightly (from 17.1% to 15.1%).
While you could argue that a bunch of very tiny improvements adds up to a big one, I don’t think that’s right here. The biggest change was his LOB% going from the mid 60s to 87%. The ERA change is mostly a mirage, as his xFIP went from 3.67 to 3.53. His FIP did change, and I’m not saying he didn’t improve at all. But he definitely wasn’t close to being one of the best pitchers in the game in the second half.
In terms of wOBA, Bauer was the 36th best pitcher in the second half of guys who had 80 or more innings.
I mean, he went 10-2 with a 3.01 ERA in the second half. FIP and xFIP are really about what should have happened, but what actually happened (ERA) was a Top-10 pitcher who was tied for 2nd in wins after the All Star break. I think its fair to say he was one of the best pitchers in baseball over the last half of the season without discounting that maybe he was getting lucky on some of that (with the LOB percentage spiking up).
I mean, are we really going to use ERA and pitcher wins, of all things, on this website?
It’s not that he improved a ton AND got lucky; his BABIP didn’t change, his HR/FB barely changed. He improved slightly AND his relievers were better and his wOBA was STILL 36th best among starters in the second half.
He improved, sure, and he became a decent to good pitcher in the second half. But he didn’t even improve to very good.
Unless we are going to say that Ervin Santana is almost as good as Severino because of ERA and wins.
I don’t know what you’re getting at here, but can we dispense with the idea that ERA measures individual pitching performance better than FIP?
FIP effectively isolates the contribution of a pitcher, which ERA and RA9 do not. You can say “what actually happened is RA9” and that’s true, but only at the team level. Same with wins. These are essentially team-level statistics, which are heavily influenced by the pitching performance but not equivalent to it.
At the individual level, FIP does a better job of telling you how a pitcher performed. So if you are trying to figure out if a pitcher was better or worse than his peers/other versions of himself, FIP is what you are looking for.
(xFIP is a slightly different story)
“FIP effectively isolates the contribution of a pitcher”
No it doesn’t. Not as long as catcher framing isn’t included in the equation.
That’s fair. I should have said that FIP is *more* effective at isolating the contribution of the pitcher than ERA or RA9. (I’m a SIERA guy, but even then your point is well-taken)
FIP says that a pitcher can’t control any batted ball that isn’t a homer. But they can at the same time control intentional walks. Sorry, but that’s bogus.
Not quite, the theory that FIP is based on is that the randomness of batted balls is much stronger than any pitcher contribution. Subtle difference–it means that taking into account batted balls introduces so much noise to the measurement it hurts its construct validity.
So the triples that Arizona hit last night for instance had nothing to do with the pitcher? Give me a break. A pitcher can control an IW though. I have no problem with not dinging pitchers for singles. But extra base hits- TOTALLY a different animal.
I never said or implied it measures performance better. I just said that based on the results from when he pitched, he was one of the best pitchers in the league. It doesn’t matter if his defense was helping him or if he got lucky spreading out hits or whatever the case is. There are metrics (wins, ERA, strikeout rate, etc) that have him as one of the best in the league over the second half and I think that can be acknowledged while still appreciating the advanced metrics this site offers.
I don’t know if you meant to do this, but: If you say that someone is one of the best pitchers and you cite a statistic or two to back up your claim, that does seem to imply that you think the statistics you cite actually measures performance well, at least better than the metrics the OP cites.
I think it’s fine to appreciate outdated metrics for historical purposes, but that’s about it. If something is a substandard measure then it shouldn’t be used to determine the “best” pitcher.
You don’t seem to understand what I’m saying, I guess. What you’re saying is that Bauer wasn’t actually one of the best pitchers because his FIP and peripherals are saying he is not. What I’m saying is based on what actually happened (low ERA because he didn’t end up giving up many runs, and wins) his results have him as one of the best in the 2nd half of the season.
The disconnect is, maybe he didn’t “earn” the ERA and wins he got over the last 3 months bsaed on his FIP and other metrics, but the fact is that it happened that way. The Indians opponents left a lot of guys on base when he was pitching and he ended up not giving up many runs and winning a bunch of games.
Is that predictive to future success? Not really. Is it what actually happened? Yes. So maybe a better way to put it is his results are that he was one of the best, but his actual true talent level is more along the lines of just good.
Now I really don’t understand what you are saying. Why would anyone declare a player’s performance “the best” or “good” or “bad” based on what other players on the team do? Cleveland’s success when Bauer pitches =/= Bauer’s performance, although they are partially related.
Bauer in the 2nd half in the AL:
Era: 3.01 (6th)
FIP: 3.68 (11th)
xFIP: 3.53 (9th)
WAR: 1.7 (9th)
K/9: 10 (10th)
W: 10 (2nd)
So you were able to cherry pick wOBA, which doesn’t really give much credit at all to his K/9 rate, and determine he wasn’t “close to being one of the best pitchers in the game”.
Sorry that him being top 10 in his league in nearly all meaningful stats doesn’t make him even “close” to being one of the best pitchers.
I think your perception is entirely wrong here my friend. In fact your perception is so wrong, you’re attempting to cite his xFIP as evidence of his poorness, when in fact his xFip was AL top 10 in the 2nd half. In case you were wondering, that means his xFIP was pretty good relative to his peers this season. No he’s not scherzer, kluber, kershaw or sale. But he was one of the 10 best pitchers in the AL in the 2nd half pretty much by any metric out there. I think that qualifies the author’s statement.
I sort of agree with this, but the only ones that are definitely “meaningful” to this discussion is FIP (although FIP- would be better) and WAR. ERA is a substandard measure, wins are determined as much by run support as pitcher performance, and K/9 (while interesting and an important factor) is only part of pitching performance. xFIP is borderline, since it has good predictive validity and therefore should help determine if Bauer should start in the second half.
Keep in mind, you’re also looking at it in the AL. It’s fair to say he was good in the second half, but he’s 21st in FIP-, 15th in xFIP-, and 15th in SIERA. That’s very good still, but…perhaps not as dominant as it looks when you think of a “Top 10” pitcher.
Doesn’t a lot of this have to do with the fact that Severino started the Wild Card game?
Normally, you want your best pitcher to match up with their best pitcher, and so on down through the fourth best, just because you want to spread your win probability evenly. Having your ace start against their four concentrates win probability in one game at the expense of the others.
However, this is dependent on the other team having their number one available, and it’s not going to be Severino, but probably Tanaka. So, if you view Bauer and Tanaka as comparable in talent, throwing out Kluber on short rest (or even normal rest) is something of a waste – the Indians don’t need to concentrate win probability because they are the better team.
“Normally, you want your best pitcher to match up with their best pitcher, and so on down through the fourth best, just because you want to spread your win probability evenly.”
Respectfully, I think the opposite is true — teams almost always start their ace in game one if fully rested, with little consideration given to the opposing pitcher. That’s why it’s so surprising that a fully-rested Kluber isn’t starting game one.
As far as I see it, this move is actually the smart way to play it
Right now Bauer is probably equal-to-better than whoever he will face in Game 1 anyway, Carrasco is better than whoever he will face in Game 3, and by far the best the Yankees have is likely to go in Game 2 up against a machine in Kluber
Game 4 is the one which is of question to Cle (and NY), with the Indians anticipating having to use two otherwise shutdown starters as inning eaters (something the Yankees don’t have avail) – two shutdown starters who would not be able to be as ready at full capacity if possibly needed for both Game 3 & 4
Right now this is how they are seemingly lining up their rotation for the first 6-7 innings:
G1 Bauer ->Clevinger ->Salazar(if needed)
G2 Kluber
G3 Carrasco
G4 Tomlin ->Clevinger ->Salazar
…(or likely Salazar ->Clev ->Bauer if they lose Game 1or2)
G5 Kluber
That is the best they can offer in any series, and moving their most questionable of the top-3 into Game 3 jeopardizes game 4 because Clevinger and/or Salazar may not end up as available/as unleashed
Since Bauer is better at Home it increases this efficiency rather than having Bauer in NY right before anticipating Clevinger and Salazar likely needing to go multiple innings
From where I’m sitting, starting Bauer instead in Game 3 kind of eliminates the gain they get by having Clevinger in the Pen in the first place, and decreases the odds they win Game 4.
Pretty sure this move is because they are championship or bust. I don’t think they think it changes their alds odds much. They may even realize that it gives them worse odds. But they had one of the best stretches of any team ever and are likely willing to accept a slightly larger chance of getting upset in order to win it all.
The crux of Francona’s reasoning seems to be that if Kluber started Game 1 and Game 5, the latter would come on extra rest (5 days).
But by saving him for Game 2, Kluber will be starting THAT game on 5 days’ rest. He last pitched on September 30 (and threw just 81 pitches).
So, between these two choices, both have one start on 5 days’ rest — but only the one Terry chose GUARANTEES one such start in this series.
That would make it a very odd choice, if avoiding extra rest were truly the main reason. So I suspect there are other reasons he’s not saying.
FWIW (and with SSS caveat), there’s no meaningful difference between Kluber’s career stats on 4 days’ rest (100 GS) and 5 days’ rest (44 GS). The TTO rates are indistinguishable. The longer-rest starts have lasted almost one out longer.
“So I suspect there are other reasons he’s not saying.”
Not to sound too conspiratorial, but I am also left wondering if there’s some small undisclosed issue for Kluber -a stomach bug, a cold, or minor soreness – that makes Francona reasonably think that the extra day before Game 2 will help out Kluber.
You don’t suppose THIS could have been part of Francona’s rationale? Of the 124 prior best-of-5 series:
— Teams that won Game 1 went 88-36 in series.
— Teams that won Game 2 went 94-30 in series.
That looks like noise to me, but who knows. You’ve gotta think he’s at least aware of it.
lol wtf. are you really suggesting its more important to win g2 of a best of 5 series than g1???
I merely posted the numbers. Everyone can draw their own conclusions.
Btw, I heard John Smoltz last night cite the series results for game-1 winners, with no mention of the even better numbers for game-2 winners. It’s funny how often that happens — people see an impressive number and fail to ask the natural follow-up question.
It’s interesting how much gets ignored when the superficial surface at a quick glace seems to support what you desperately want to believe, isn’t it?
Further still, ponder the odds a team wins a 5 game series after being up 2-1 …think it might be even better?
If someone didn’t know better they might even conclude setting things up to be in a position to best win multiple games is much more important than doing everything you can to win solely the first
Once Francona made the determination that he wouldn’t use Kluber on short rest, at that point there was no difference between pitching him in Game 1 vs. Game 2.
1 thing that’s kind of interesting about what Francona is doing. Does it force Girardi to manage in game 1 much differently, knowing that he’s got the Cy Young award winner coming up against in game 2? For instance, if say Gray gives up 2-3 runs early, does Girardi have a quicker hook than he would have otherwise? While it wouldn’t be quite to the WC game level, it would be much more aggressive than normal you would think.
idk i feel like im seeing a lot of fuzzy logic (and even some lolmath) trying to justify what seems like pretty obviously a bad decision. assuming theres no underlying health issue with kluber going on here, theres no reason to not start him game 1. at least not that i can see. even ignoring the weather issues.
youre a much bigger favorite with kluber on the mound than with bauer. i would assume that youd want kluber to start as many games as possible on his regular rest to maximize his effectiveness and your chances of winning. whoever wins game 1 in a best of 5 series has a huge advantage so why not just put yourself in the best possible position to win game 1?
if i came here and said that tito should hit lindor and ramirez down at the bottom of the lineup so that if the worse hitters in front of them happened to get on then youd have the best hitters up in high leverage spots youd tell me that im dumb and that you hit your best hitters near the top of the lineup so they get the most at bats. not sure what makes this any different.
and now once you factor in that game 2 could be rained out and you risk only starting kluber one game in the series, it makes it a slam dunk that he should start game 1.
titos decision has worked out brillianty. bauer was excellent and now cle is a huge favorite up 1-0 and having yet to use the best pitcher in the al. but what if it had gone wrong tonight? or even keeping the results of tonights game the same, what if game 2 does get rained out and the yankees end up winning after having to only face kluber once? then how would cleveland fans feel?
This argument is rather silly as it ignores many facts.
I mean, it isn’t like we are talking about some 21 year old rookie who was bad in the regular season – even with his poor first half, Bauer (3.88 FIP) showed himself a better pitcher than Gray (4.87 FIP as Yankee) this year.
And as you saw in the game, Bauers strength (esp at home) plays to a weakness of the Yankees, where the opposite holds true for Gray.
Plus, as I said above, this meant the Indians had Clevinger and Salazar primed in the Pen ready to pick up the slack if Bauer wasn’t on.
But apparently you think the Indians should have instead played Bauer in NY (despite him being better at home) possibly against Severino (the yanks only really strong pitcher) and in the game right before Tomlin (possibly lessening the availability of Clev/Salazar for Game 4) just so the Indians most dominant pitcher can go up against a guy that should not be expected to pitch that great against the Indians?
You play games to win. The team is instantly more likely to win the games both Kluber and Carrasco start, regardless who they face. Meanwhile the best game for Bauer to win was game 1, and the odds in Game 4 would decrease if Bauer was off in game 3. It wasn’t really unorthodox at all considering the specifics; media talking-heads obsessing on their old-school thinking aside
lol im not really sure how to respond to that nonsense…
i mean first you called my argument “rather silly as it ignores many facts” then didnt really cite any ‘facts’ beyond some fip numbers.
but whatever idc anymore. keep downvoting me indians fans!
I can only conclude your opinion of Gray bares absolutely no resemblance to just how bad his rates have been in 2016, since going to the Yankees or on the Road as an Athletic this season, and instead completely rests on his luck in games at the cavernous Colosseum this season (a location he will not be pitching at this postseason)
Otherwise, as you showed in the other thread too, you apparently have zero grasp of gameplaning and instead insistently believe a series is somehow won merely by shooting your entire load to win 1 single game to start, regardless of the disadvantage it puts you in in the other 4. That’s called idiotic, and it’s exactly how you lose a series you should otherwise win
I’m reminded of a story in racing: Both teams have a top horse, a middle horse, and a poor horse, but each of Team A’s horses are better than their Team B counterparts.
Team B runs their worst horse against Team A’s top horse, their top horse against Team A’s middle horse, and their middle horse against Team A’s worst horse, and takes two races out of three.
Of course, it’s a moot point now that Bauer won Game 1. Guess the Indian knew what they were doing when they commanded their ballpark to swallow up Bauer’s drone.
came here to ask whether or not titos still a genius. thanks in advance.