Yesterday, Bob R., TTNorm, Rome, LJK, and I had an interesting discussion in the comments section about the value of “closers.”
Remember when Esty was our "closer?" Good times.
This morning, through either luck or divine providence, Baseball Prospectus adds meat to the bones of our speculation. In this piece, Colin Wyers breaks down the value of using a team’s best pitcher as its “closer” in great detail.
It is a detailed, dense piece (he references Archimedes…seriously…that is intense math) that I will summarize with two major conclusions:
1) While using your best pitcher to close games does increase your chances of protecting 1-run 9th inning leads, the increase is so minimal that it amounts to about 1 win every two years (because you are going to win a huge percentage of games when you lead in the 9th inning no matter who is pitching); and
2) Waiting to use your best pitcher until the 9th inning of a 1-run game reduces the chances that you will have a 1-run lead in the 9th inning.
Another great "closer" from Rays past. This guy was an outfielder when we signed him.
The first point is surprising but the second, I think, is the most overlooked part of this whole discussion. Upon further review, however, it makes perfect sense. Our starter is tired and we are winning by <3 runs, let’s use pitchers that are not that good until we get to the 9th when we will use our best pitcher. Under this scenario, your other relievers (we will call them the bridge since I like that term) need to get 6-9 guys out, and your best pitcher only needs to get 3 guys out. If you flip the logic, and use your best pitchers in the 6-7-8 innings, then your best pitcher is responsible for 6-9 outs at a point in the game when the other team still has some hope (because they aren’t down to their last 3 outs) and the bridge are only responsible for 3 outs when the other team’s back is all but broken.
Devil Ray fans have some experience with this phenomenon. From 1998-2000 Roberto Hernandez was our best pitcher. He earned 101 saves in 103 save situations in those three seasons while the team only won a total of 201 games. (Let’s just agree, again, that Hernandez was incredibly good for incredibly bad teams.)
Before Devil Ray fans got to see Rubber-toe protect a lead, they had to suffer through relief appearances from future Hall of Famers Esteban Yan, Travis Harper, Rick White, Doug Creek, and others.
What if Larry Rotshchild and Hal McRae had the confidence to break the mold? What if they didn’t save Hernandez for “save” situations and used him earlier in games to protect mid-late inning leads? Based on the research above, those Devil Ray teams were going to win a huge portion of the 101 games Roberto earned saves in anyway. Let’s estimate conservatively, based on BP, that those 101 wins are 90 wins. But, the Devil Rays certainly would get those 11 wins back by using Roberto in the 285 games they lost without getting him the ball. Even if Roberto’s contribution helps them win 30% of those games, that is 85.5 more wins (which is a total win increase of 74.5 after subtracting the 11 wins we took away from Hernandez’s save total). That is 28.5 more wins per season turning some 60-win teams into 80 win teams. Even half that effect would have been monumental.
This kind of logic is right down Joe Maddon’s alley but, it isn’t clear that the Rays are ready to totally dynamite the mold. If they do, we’ll be comfortable even if the bandwagon starts getting light.
A hypothetical question. Which bullpen construction would be most efficient? The one the Brewers had in 2010 which cost over $19 million or one in which their 7 relievers were essentially 4 Parra(s) and 3 Villaneuva(s) which cost less than $5 million?
Let me set the criteria. I do not mean which would be cheaper but which would give the better performance over the course of a year. In the first, with Hoffman, Riske and Hawkins pitching at their accustomed level, the relief staff has defined roles. Hoffman is held until the 9th to protect leads of 1-3 runs; a set-up man is ordinarily held until the 8th to hold similar leads. They rarely pitch more than 1 inning.
In the second arrangement, the pitchers also pitch to their accustomed level, but they are used interchangeably and are often used to throw 2 or more innings at a time.
The closer will never go away. The media won’t let it happen. If you use you best reliever before the ninth and then lose the game in the ninth, all heck will break loose. The other flaw I see is exactly how do you know when the key moment in a game occurs? When you play the Yankees EVERY inning is a time for your best pitcher. Just saying.
You may have a fair point regarding the media but, if there is any skipper that doesn’t care what the media thinks of his tactics, it is Joe Maddon. (Otherwise, he wouldn’t have used Shields last fall).
As for the key moment, I think the BP article addresses this. It is not a “key moment” you are looking for, it is a leverage point. Under this logic, the 6th inning is more important than the 7th and the 7th is more important than the 8th. The earlier innings are important because zeroes in those innings increase the likelihood of a victory significantly. And, a team leading in the 9th is going to win almost no matter what, so it isn’t that important.
Look at it like this. If you have a three run lead in the 6th and use so-so pitchers for nine outs, that gives your opponent the ability to chip away at the lead with a run here or there. If you have a three run lead in the 6th and use your “closer” they get 0 runs. That 0 in the 6th is more important than a 0 in the 9th because it reduces the number of outs for your opponent and requires them to put up crooked numbers to tie or win.
I agree with you Mark, and I think I go further. My notion is that you do not sign people who have the reputation for being the “best” or even good relievers and are therefore expensive. Rather, if relievers are really fungible, why not look for pitchers with talent who cannot start.
I begin with an assumption that any pitcher in a major league camp has talent. He can throw hard or has excellent breaking or off-speed stuff. He may get lots of ground balls or strike people out. He also may have flaws which limit him in the majors-home run prone, lack of command or control, et al. The key is to unlock the potential while minimizing the negatives.
Why can’t a pitcher succeed as a starter? Stamina? Lack of a wide enough repertoire? Dependence on deception which doesn’t work the second time through an order? Poor command or control? Inability to make proper pitch selection which is less important if the repertoire is kept simple? Age has sapped some power or bite but in short outings it is fine or can be compensated for? Any number of other possibilities, but many of those problems are minimized in short outings while the positives are magnified.
In addition, I think it important to stretch them out so they can manage 2-3 inning outings. That would make it less necessary to carry 12 pitchers and the remaining spot(s) can go to deepen the bench.
I submit this view very tentatively subject to research that confirms, disproves or modifies it. But if the research indicates it makes sense, I think the Rays are exactly the team to try it. It fits with their budget constraints. They have a front office and manager willing to try new things. They rely on depth, not star power for success. And it would allow them to avoid having large arbitration awards as the standard criteria would be less relevant.
Additionally, if there are indeed many pitchers who fit the bill, there is less urgency to convert promising starters like McGee and Torrez into relievers while perhaps they can make such decisions quicker on other prospects.
One downside is that it also would mean they might not get draft choices for relievers who become free agents as they might not become A or B rated.
To me, this point hits on one of the most often unasked questions in baseball:
We accept, almost as natural law, that it is important to use multiple relievers in an inning or limit relievers to 1-3 hitters at a time.
Randy Choate will make $3MM this year based on that very premise.
That means one or two ( or even three ) of your 25 guys can only be used for 1/54-3/54 of the game. That is an absurd allocation of resources that we just accept.
Building on your idea Bob, why not let future starters (like McGee, Torres, and even Archer) break into the big leagues by being 2-3 IP relievers. That makes use of their talents before they are able to crack the rotation (because they are blocked by better SPs) and keeps their arms stretched out for emergency starts. It also lets them get acclimated to big league hitters without letting the opponents see them for more than one at bat in a game.
This is genius. I hope someone important is reading it.
I think Earl Weaver used that approach to break in young pitchers. He used them in long relief and as spot starters both to get them experience and to shield them from the greater pressure of starting.
I didn’t know that but, I think it is always good to agree with Earl Weaver. Right?
Yes, I consider him one of the greatest managers in baseball history. Brilliant at focusing on player strengths and using them rather than harping on what they could not do. Led to some exceptionally productive platoons. Proponent of the 3 run homer rather than the sacrifice bunt. Treasured defense, especially up the middle, hence Belanger at SS but also outstanding defensive players at 2B (Grich) and CF (Blair).
R.J. Anderson had a post up a few days ago that listed 5 quotations. He asked who said each, Weaver or Maddon. They all sounded like Maddon, and were all Weaver.
Of course Weaver is remembered for his run-ins with umpires and his ongoing “feud” with Jim Palmer who repeated the old joke that the only thing Weaver knew about pitching was that he couldn’t hit it. I think Palmer was wrong.