If you’re really interested, do yourself a favor buy the latest Baseball Prospectus and get a subscription to their website. They have multiple articles updated daily there.
I’m a stathead, but no, we don’t need to get rid of wins and losses. Wins and losses are kind of the point to baseball. It’s silly to get rid of numbers we’ve been using for 140 years just because there are other numbers, and I’m not even going to get into what a goddamned marketing disaster it would be.
The purpose of statistics, contrary to what some posters seem to be implying, is not just to rate a player’s objective value. The purpose of statistics is to count things and record facts. We count the number of doubles Jacques Jones hits so that we know how many doubles he hits. It’s not necessary to eliminate the raw number translate “22 doubles” into a value number like Win Shares or WARP3. It’s cool to know that stuff and I love it, but we still record the simple fact that he hit 22 doubles in 2004. That’s the purpose of the statistic; to record that fact.
I see there being value in counting up how many times Javier Vazquez was the winning pitcher in a baseball game. It’s a record of the facts. We’re not gaining anything by getting rid of wins. Anyway, you can’t get RID of the stat; you can’t use a Men in Black flashy thing to make the entire baseball world forget about the concept of pitchers’ wins and losses.
Becaise of that, I’d actually agree that the scoring rules should be tweaked to allow the scorer to reassign the win in the situation described in the OP. Bear in mind that when the statistic was created, relief pitchers were not used, so W/L record gave you a near-perfect record of how many games a pitcher’s team won and lost when he was on the mound. The introduction of the relief pitcher forced the rule to be applied that a pitcher gets the win when his team goes ahead for good. Even that worked for a long time, because for much of the first half of baseball history, “Relief pitchers” were guys who came in when the starter was blown out, so it was fair to hand them the W if they held the fort and won a comeback.
Therefore I would change the rule accordingly;
Add this rule wording to 10.19(4)
“If a starting pitcher pitches more than five innings, but the lead is blown by a pitcher who subsequently relieves the starter, and that pitcher then becomes the winning pitcher under the preceding rule by having his team retake the lead for good while he is still the pitcher of record, instead credit the win to the starting pitcher.”
While blown saves are kept track of (you’ll hear the announcers talking about them anyway), I don’t think they’re an official stat.
Exhibit A: Roger Clemens. This season, I believe he’s had two or three instances of pitching 7-inning shutouts with nothing to show for it but a no-decision (well, and an impressive ERA). The other Astros pitchers have similar woes.
When I read the title, I thought it said “A Believer who blows a save should not get the win!” and had an image of a fundie being denied Heaven 'cause someone told him to shove it about the Good News.
Sorry, carry on.
The problem with all the new sabermetric statistics is most of us can agree they are better than the more basic stats of the past but many people also claim they also are very shaky as to correctly showing the comparative talents of players.
Many people are idiots.
I don’t see many pitchers being hired or fired or traded or minored based on any one statistic–or even a bunch of statistics.
Pitchers–hell, most players*–get and keep jobs based on performance. The Sox organization knows how Clement played, and they know how valuable he is. Whether or not he gets the W is AFAICT irrelevant.
- Exception: “marquee” players who post the right kind of big numbers get a lot of love for presumably bringing in more fans. We fans do still love our numbers, when it’s the Unit getting his 20th W or Bonds getting HR number 73.
The numbers, and wins are a big one, do help at contract time, though. That impulse is behind the promotion and softening of the save rule, for instance - it had no effect on relievers’ value to their teams, and may have damaged it with the designated closer’s insistence on coming into games in save situations even when the middle guy is cruising. But the save rule has made them rich as well as prima donnas. The same thing happened when home runs became fashionable with the dawn of the rabbit-ball era - before that, they were essentially oddities, with games “normally” being won with pitching and baserunning. But the new emphasis on a number that used to be considered contrived changed the game itself.
FTR, Dotel got a blown save as well as the loss for the A’s (I have no problem with that), and for the second consecutive game too.
My problem with saves is that the stat actually has come to affect strategy, and not for the better.
Example: your team has one quality reliever, its closer, and you’re ahead 4-3 in the seventh inning. The other team puts men on base, their best hitters are coming up, and your starter is clearly gassed. In the past, your manager would have put his best reliever in the game at this crucial point, to put out the fire. But as saves became a big issue in contract negotiations, that move became more and more questionable (because to finish the game he would need to pitch the last three innings, causing undue strain on his arm, or he would have had to come out before the game was over, and thus lose the save.) Now it’s at the point where every manager routinely puts a lesser reliever into the game at that point, and sees many leads vanish that formerly would been protected.
Which is why I’d like to see the Save stat gone.
I don’t think it would break anything very valuable if on occasion no pitcher got credit for a win/loss, and I’ve seen games where that would seem the just outcome.
I sympathize with ElvisL1ves’ perception that in cases like this there seems to be an unfair reward for the reliever. But I’ve also seen games in which a starter gave up seven runs in the first couple of innings and a reliever pitched six or more, giving up one or two runs – and took the loss. So maybe it evens out. I also think that, you know, go team, all for one and one for all, and all that, and we shouldn’t get so hung up on this particular statistic. It wouldn’t take too much imagination, in fact, to see the win/loss statistic opened up so that a win could be credited to, say, the right fielder who went 3-for-4 with four RBI, or the shortstop who prevented two runs from scoring with a spectacular play in the fifth inning, and a loss charged to the ham-handed, hot-dogging left fielder who thinks that two-base errors, like his many home runs, are just all part of the Divine Plan. I’m not advocating this, I’m just saying I think that the win/loss/save statistics make pitchers seem a little more important than they actually are.
In my opinion, a bigger flaw in the scoring rules is the one that makes runs scored due to an error unearned, even if the pitcher himself makes the error. That thing on your other hand is a glove, son.
Well, I’m not so upset by the reliever getting the win, a stat which is generally agreed to be fairly meaningless for a reliever, but by the starter getting cheated out of it. It’s also true that a starter who gets no run support is similarly cheated by his teammates, but, in contrast, I don’t have a ready remedy to offer for that one.
It’s already the rule that, when the starter would have had the win but didn’t go 5 full, the scorer can give it to whichever reliever he thinks was most effective, and objective criteria aren’t even required. Well, why not expand that concept a bit, set an objective criterion, and give every win to the pitcher who allowed the fewest runs, inning count notwithstanding? Break ties with inning count.
And, for that matter, once an inning has been extended by an error with 2 out, all runs after it are unearned. The batters can hit all they want, bash a few homers even, and the pitcher’s ERA is unscathed - even though it’s largely his own fault for coming unglued and starting to groove fat ones. If a run scores because of hits the pitcher gave up, it’s earned, dammit. Score it that way.
I’m not quite with you there. I think there are enough high-run, multiple-pitcher games to turn this into a scorer’s nightmare. I think I’d rather see the win credited to whoever was on the mound the inning his team scored the winning run(s), no matter what. But I don’t see the Win as being that important, so I’m more willing to make it even less so. I might go for your idea, though, if starters were eligible for the win even without five full innings. In many games, they might well be the most effective pitcher, allowing fewer runs in more innings than any single reliever, so why shut them out in favor of even more mediocre performers?
I’m more in line with you here, but I’d tweak it slightly in the pitcher’s favor by stipulating that runs on or after an error would be unearned (unless it’s the pitcher’s error) until the batter who hit the misplayed ball scores or is put out, at which point status quo ante.
Yes, but that has nothing to do with the pitcher’s role as a pitcher. It has to with his role as a fielder. Thus, fielding errors should have nothing to do with his pitching stats. They’re already recorded in his fielding stats.
I understand your point, and disagree. I think the whole point of ERA is to prevent the pitcher from being blamed for runs that are not his fault, that his errors are his fault, and that you can’t really separate his role as pitcher from his reponsibility to field his position any more than you can separate the catcher’s job to call an intelligent game or not let the pitch sail by his left ear to the backstop from his obligation not to throw the ball into center field on a stolen base attempt. If the rationale for error-prompted runs being unearned is to protect the pitcher’s statistics from poor-fielding teammates, I don’t see why he should be protected from himself.
Your point of view is perfectly valid, and I admit nothing I’ve said really refutes it, given your starting point. We just start at slightly different places, that’s all.
And surprisingly many idiots somehow, despite their clear and abiding idiocy, obtain employment in the front offices of various sports teams that I follow. 
But, as I said before, that is not primarily why we keep statistics. We keep them as a record of what took place.
Simply put, ERA is silly. It should be replaced with RA. Forget the distinction between earned and unearned.
The ERA rule is, again, an artifact of baseball when it was very different from today. In the Olden Days <TM> fielders committed a very large number of errors, many more than they do today, and a very large percentage of runs were unearned. Here are the total runs allowed and earned runs allowed for all the teams in the National League in 1905:
New York: 505 total, 364 earned
Pittsburgh: 570 total, 439 earned
Chicago: 442 total, 319 earned
Philadelphia: 602 total, 436 earned
Cincinnati: 698 total, 457 earned
St. Louis: 734 total, 537 earned
Boston: 733 total, 541 earned
Brooklyn: 807 total, 563 earned
As you can see, teams in those days gave up a staggering number of unearned runs; some of those teams were giving up over a quarter of all runs unearned. That’s because they were making an amazing number of errors; the BESt fielding team in the league, the Cubs, a team famous for its defense, committed 246 errors. By comparison, the most fumbled-fingered team in last year’s NL, the Diamondbacks, made just 139 errors. The Dodgers/Robins/whatever they were called in 1905 made over FOUR HUNDRED errors.
The difference between teams in terms of errors and unearned runs could, as you can see, be very wide, so it made sense to distinguish between pitchers in terms of earned versus unearned runs because the conditions of defense could be so hugely important.
Today, not so much; a good defensive team will give up 50-60 unearned runs, and a really terrible one will give up a hundred. As has been pointed out, even most unearned runs can be in part attributable to the pitcher, so the truth is that the distinction from pitcher to pitcher in terms of unearned runs allowed is largely random and not terribly significant.
The truth is that defense probably affects a pitcher’s EARNED runs more than it does his unearned runs. A team with fielders who have lots of range, savvy positioning, great footwork and good arms can save a pitcher more runs than errors will give back. Highly competent fielders can easily save a team 20 to 30 runs or more despite very little difference in the number of errors they make; Ozzie Smith has seasons where he was probably saving his pitchers 40 to 50 earned runs.
Just weighing in. Wins are the worst stat to judge a pitcher on. ERA is second worst. WHIP, BAA, OPSA, those tell so much more.
Granted. But I think the ERA statistic is not especially useful for that purpose. It may be illustrative of a particular pitcher, but not “as a record of what took place” in a game, season, or career. If anything, it obfuscates, because earned and unearned runs count just the same. Even so, I still think the practice of discounting unearned runs (as opposed to the purpose of keeping statistics at all) is for balancing pitchers’ statistics, not to provide a clear record of anything whatever that actually happened, and in fact that’s the history you provide for the measure. It’s easy enough to count opponents’ runs scored as the result of errors, by team/player/ballpark/etc. without involving the pitcher at all.
Now you’re cooking. I’d be fine with that. After all, if a pitcher’s statistics aren’t adjusted for poor run support, ballpark size, boneheaded management and bad luck, why draw the line at errors? Not to mention the fact that errors can be a poor measure of the quality of fielding support a pitcher gets.
Okay, but do they really? What do they tell you?
We’re now getting into the issue of descriptive stats versus prescriptive stats.
I don’t mean to state the obvious here, but ERA is very probably the MOST accurate commonly used statistic for describing a player’s performance. The correlation between ERA and a pitcher’s actual value is extraordinarily high; the absolute fact is that a pitcher with a 2.20 ERA helped his team and a pitcher with a 6.90 ERA did not. It’s preposterous to state that it’s not smart to judge a pitcher on how many runs he gave up. The entire point of baseball is scoring and preventing runs. As a DESCRIPTIVE stat, ERA is extremely valuable, because it’s an accurate record of what happened, and it goes directly to value.
WHIP, OPSA, and such are certainly very important, but where they defeat ERA is in their predictive abilities, not their descriptive abilities. A pitcher with a good WHIP and a 5.79 ERA was not a good pitcher, period, end of story. Maybe he was unlucky; maybe he gave up too many homers; maybe he got flustered in a few key spots and gave up big rallies. The low WHIP would suggest that his high ERa is flukey, and that he’ll probably do better in the future, but the fact remains that the ERA is relatively accurate in describing how well he did.
Like I said before, I’d rather use RA, but ERA is 90% as good.
Here’a good example: Jeff Ballard in 1989. Ballard started 35 games, pitching 215 innings, allowing 240 hits, only 16 homers, struck out 62 men and walked 57. He allowed 82 earned runs and went 18-8.
Ballard pitched very well in 1989; his ERA is perfectly reflective of his performance. His W-L was a bit lucky but not by much. His WHIP wasn’t bad either, not his OPSA. Of course, what tips you off is the strikeouts; pitchers who only strike out 62 men in 215 innings are doomed. Predictably, Ballard never pitched well again; he was beaten like a rented mule for the rest of his career.
So in Ballard’s case, you have ERA telling you he pitched well in 1989, and K/W ratio telling you his days were numbered. Clearly his ERA was not an effective PREDICTIVE stat; his 3.43 ERA did not accurately foretell how his career would go. The K/W career told you the truth; he was gonna get smoked. But his ERA was quite descriptive of his 1989 performance. You simply cannot deny that he DID get lots of guys out in 1989, and really helped his team. He couldn’t fool the American League for long with his slop, but he did it for a year.
Well, I can certainly understand your point of view, and I agree there’s no use arguing. You see the pitcher’s role as a fielder as part of his role as a pitcher, whereas I see the two roles as distinct. I tend to think that the three stats (pitching, fielding, batting) are all separate and shouldn’t get mucked up with each other. Errors are a fielding stat and ERA is a pitching stat. Never the twain shall meet.
If a pitcher covering first drops an easy toss from the first baseman, allowing the runner to reach safely, that’s an error. But it had nothing to do with his pitching. He pitched in such a way to cause the batter to do hit into would normally be a groundout- he just screwed up the execution in fielding it. Thus, I figure his pitching stats shouldn’t be affected, though his fielding stats should (and, of course, are).
But, carry on. No use arguing, as you say. Thanks for explaining your argument so well.