Scott Aug 27th:
This is an intriguing idea, Asher, that I'd like to expand upon.
I think a +/- system for relievers, similar to hockey's, would be the simplest way to give the most amount of information about the situation.
It's simple: If your team's game status changes unfavorably while you are pitching (winning to losing, winning to tied, or tied to losing), you net -1. If the opposite happens, you get +1.
A BS would net you -1.
A BSW would net (-1) + (+1) = 0.
A save would be 0.
An honest relief win yields +1.
You can get a +1 without a win if you pitch well enough to let your team tie up the game.
You'd probably only get a small handful or relievers with a positive rating by this system, so it would be easy at a glance to see who are the most effective relievers.
Hopefully this would also demote the Hold to somewhat less useful than the GWRBI (how silly was that?) and still more useful than...ok, I can't think of an example. Holds suck.
Anyway, if we could start scoring games with this system next year and it proved valuable, we could be assured of mainstream acceptance of it by the year 2024, if Peter Gammons' apparent discovery of OPS in 1999 is any indication.
Tom Aug 27th:
Under that system, offense plays a dominant role in calculating the effectiveness of a reliever. In order to earn any + points, your offense has to score while you are pitching.
Conversely, you could amass a season full of two inning, one run saves and score zero.
It's not really any different than saves and holds that we have now, except that you don't get rewarded for pitching scoreless innings unless:
A) you happened to come into a game while your team was tied or behind, AND
B) your team scored while you were pitching.
And that just comes down to luck.
Asher Aug 27th:
Tom is right.
BSW can not be equal to a save (both would be 0). That's the whole point - pitchers getting credit for wins when really they only did bad things to get the win. Where Tom is right is that you have valued a BSW as greater than a BS, which, while the team result is better, the pitcher performance has been the same - he didn't get the job done.
Now, if a closer gets a bona fide win, this is exactly like getting a save. Generally, a bona fide win (BFW) involves the pitcher coming into a tied game, or even trailing, getting the job done, and then his team getting the winning runs in the bottom of the inning. The only difference between a BFW (not a BSW mind you) and a save is the team's offense, like Tom said.
The one thing you overlooked, of course, is the loss. Relievers get losses when they blow saves and the team can't recover, or come in to tied games and give up the go ahead runs. From the pitcher's perspective, the only difference between a blown save loss (BSL) and a BSW is the team offense. Of course, there is also the bona fide loss (not a save situation) which as a reflection of pitcher performance is actually not as bad as a blown save, because the margin for error in a tied game, or even when trailing, is so much less than the margin for error when the team is up by three. You can blow a save and still be tied, but if you come into a tied game and give up even one run, you are sunk.
Here's how I see the situation:
First the fundamental:
BS = -1
S = 1
W = 1
L = -1
Second the bona fides:
BFL = -1
BFW = 1
Third, the blown save results:
BSW = -1
BSL = -1
For each scenario, we add up what happened to the pitcher
Last, the scenarios:
Pitcher enters in a save situation:
---Gets the save: S; score = 1
---Blows the save, gets the win: BS+BSW+W;; score = -1+-1+1 = -1
---Blows the save, gets the loss: BS+BSL+LL; score = -1+-1+-1 = -3
---Blows the save, gets no decision: BS; sscore = -1
(Note: a blown save and a blown save win are equal in value: pitcher gets NO CREDIT for team's offense)
Pitcher enters a non-save situation:
---Gets the win: W+BFW; score = 2
---Gets the loss: L+BFL; score = -2
(Note: a loss in a save situation is worse than a loss in a non-save situation; a save is worth less than a win in a non-save situation: pitcher's performance in tied game or when team is behind is more important than when team is up by three)
The worst thing a closer can do is enter a game in a save situation, blow the save and get the loss. -3
The best thing a closer can do is enter a game with a tied score, or even trailing, hold the other team at bay and give his team a chance to win. +2
The thing the standard closer will do most often will be enter the game and get the save, often not remarkably. 1
Blowing the save and getting the win is exactly equal to blowing the save and getting no decision. -1 A blown save is a blown save.
Entering a game with the score tied and getting the loss is not as bad as entering in a save situation and getting the loss. -2
I like it.
Asher Aug 28th:
Here's the first two guys I have looked at. Danny Graves has 37 saves, which is two more than Eric Gagne. Lets all get us some Danny Graves, right?
Danny Graves 61 games; 1 5, 37/46 Saves
percentage (total score/game) 20/61 = .328
Eric Gagne 55 games; 4-3, 35/37 Saves
percentage = 35/55 = .636
Tom Aug 28th:
Workhorse is, of course, a relative term when discussing relief pitchers, but I like guys that can throw more than one inning.
You could include that by multiplying the result you already have by IP per appearance.
Danny Graves goes from .328 to .332 due to his 1.01 innings per appearance; certainly an immaterial difference.
Gagne, however, goes from .636 to .716 thanks to the 1.13 innings he averages per appearance.
NOTE: The closers that benefit most from this addition are Gagne, Smoltz, Dotel, Foulke, Armando Benitez, & Braden Looper. The pitchers who lose points are Kolb, Herges, & Percival.
Since this evaluation applies mostly to closers, most of them will average almost exactly one inning per appearance, and in that case, there is no change, since you are multiplying by one.
The pitchers who benefit most from this are middle relievers, as they often throw more innings, but maybe that's ok, since they get squeezed out of the decisions. Using innings pitched keeps their two innings every three days from getting marginalized.
Also, it suggests that pitchers who do the Tony LaRussa special (pitch to one batter) are worth very little. Which they are.
I didn't calculate anybody's stats, so this may look rediculous when you try to use it.
Keith Aug 29th:
This is pretty good, but there are still a few problems not addressed:
1) A pitcher who enters a tie game in the 8th inning and pitches 2 scoreless innings before the game goes into extra innings gets zero credit for perhaps the most important performance out of those you've identified. Tom's addition helps some, but it's still inadaquate.
2) There needs to be a tier for saves. A save with a one-run lead is about seven times more valuable than one with a three-run lead. We need to do something to account for that.
Tom Aug 29th:
How about if a pitcher gets bonus points for entering a tie game, or a close one.
He would receive bonus points if he entered a one run game, thus making a one run save more valuable. Also, that makes giving up a one run lead less of a penalty than giving up a three run lead, as you've received bonus points for the close score at the time you entered the game.
You would get the most points for entering a tie game, less for a one run game, and less still if the lead is larger.
Disagree with something? Got something to add? Wanna bring up something totally new? Email us at firstname.lastname@example.org