Proposal: Blown-Save Wins Should Revert to Starter
I know. Pitcher wins don’t matter. Starting pitchers are better measured by FIP and xFIP and BABIP. Relief pitchers are better measured by K/BB ratios and WPA and the percentage of inherited runners left on base.
And yet, pitcher wins matter to pitchers.
A relief pitcher shouldn’t be rewarded with a win when his performance caused his team to lose its lead, only to regain the lead while he is pitcher of record. A starting pitcher who pitches at least seven innings and leaves with the lead shouldn’t be penalized with a no decision when the relief pitcher who replaces him coughs up the lead.
I propose a change to MLB rules as follows:
(1) When a starting pitcher pitches at least seven innings and leaves the game with at least a one-run lead;
(2) When the starting pitcher leaves the game, the bases are empty;
(3) The relief pitcher who replaces the starting pitcher allows the tying run to score; and
(4) The starting pitcher’s and relief pitcher’s team then re-takes the lead, keeps the lead and wins the game (which under current rules would result in the relief pitcher being credited with a win); then
(5) The win shall be credited, instead, to the starting pitcher and the relief pitcher shall be credited only with a blown save.
First, some background.
Under Rule 10.19 of the Rules of Major League Baseball, the official scorer is to award a save to a pitcher when:
(a) He is the finishing pitcher in a game won by his team;
(b) He is not the winning pitcher;
(b) He is credited with at least 1/3 of an inning pitched; and
(d) He satisfies one of the following conditions:
1. He enters the game with a lead of no more than three runs and pitches for at least one inning; or
2. He enters the game, regardless of the count, with the potential tying run either on base, at bat or on deck; or
3. He pitches for at least three innings.
A blown save occurs when a pitcher enters a game in a save situation but allows the tying run to score. That pitcher then records a blown-save win if his team retakes the lead when he is the pitcher of record, maintains that lead, and wins the game. Neither a blown save nor a blown-save win is an official MLB statistic.
Second, some data points.
I’ve focused on the previous twenty seasons. Putting aside the strike-shortened seasons of 1994 and 1995, the number of blown saves per season and the number of blown-save wins has held fairly steady. I was surprised by this, given changes in how managers have used starters, middle relievers and closers in that time period.
Again, without the 1994 and 1995 numbers, the average number of blown saves over the last 20 seasons is 585. The fewest was 483 in 1992; the most was 627 in 2000 and 2006. The average number of blown-save wins in that same time period is 76. The fewest was 63 in 1992; the most was 87 in 2004.
Here’s a chart with the data from the 1992-2011 seasons.
Blown Saves and Blown-Save Wins: 1992-2011
Blown-save wins come in all shapes and sizes. My focus is on those blown-save wins credited to the relief pitcher who pitches immediately after the starting pitcher leaves the game clean — meaning bases empty — with the lead. In those situations, the least deserving of a win is the relief pitcher who blows the save. And while the starting pitcher didn’t finish the job by pitching a complete game, he is more deserving of the win than the pitcher who replaced him and immediately lost the lead built by the starting pitcher.
If this rule had been in effect, how many blown-save wins would have reverted to the starting pitcher?
Not that many. Then why go through the effort of a rule change that would affect so few games, if the past twenty years are any guide?
Because as much as we believe that pitcher wins don’t matter, they do. They matter to the pitchers. They matter in arbitrations. They matter in free agency. They matter to award voters. They matter to the record books. They matter to the Hall of Fame.
If my proposed rule had been in effect, Greg Maddux would have had five more wins, bringing his career total to 360, instead of 355. Randy Johnson would have had two more wins, bringing his career total to 305. Jamie Moyer and Kevin Brown also would have two additional wins each. And so on. (If you’re interested, I can publish a list of all of the games that would have resulted in a reverted win to the starting pitcher. Let me know in the comments.)
Let’s not give credit to the one pitcher on the winning team who contributed the least amount to his team’s ultimate victory. Let’s instead reward the starting pitcher who gave his team the best chance to the win the game.
24
Five. A rule change to give Greg Maddux five more wins over a 20 year career. Much ado about nothing.
I don’t want to give the guy who blows the save the W. Maybe just make it a “team W” or something. Or, hell, give it to the starter. Or the manager. Maybe the SS should get it, because he’d look awesome that way.
Yeah, five is a really small number. I wouldn’t bother with a rule change. But considering Maddux, his two best years(ERA<1.70) were both in the strike years of 1994 and 1995, so he might have had six or seven more wins. Still doesn't really matter, though.
No strong opinion on this either way, but I wonder, should your second clause really be “bases are empty”? What if it’s a 6-0 game and the pitcher just walked the leadoff batter? I would suggest instead that it be similar to the save rule, where the potential leading run is not at the plate or on deck. So leaving a 1-0 game with the bases empty wouldn’t be good enough, but leaving a 4-1 game with a runner on would be.
I had the same thought but posted before I read your comment.
I’d love to see the full list. Interested in knowing if there are outliers who would have had memorable seasons re-evaluated. Off the top of my head, it would have changed the discussion of Mike Mussina and “never won twenty” that dogged him until the last game of his career.
I saw the title and I immediately thought of Moose.
I will publish the full list next week. I do remember 1 or 2 Mussina starts that fell within my proposed rule. But remember, my proposed rule is quite limited. There may very well be many games where a reliever blew the save and got the win in a Mussina start, but it wasn’t the reliever who pitched immediately after the starter.
You could combine your first two conditions to state that the tying run is not on base when the starting pitcher leaves the game. That covers a situation such as when the starter’s team is ahead by two wins and has one runner on base.
I’d love to see the full list too, mainly for the long term players. Assessing this could have some legs too. It would be interesting to see which guys were “victimized” at above average rates over their careers. Are there teams that consistently cost blown save wins for their SPs over the years? Who was most victimized by the blwon save win in each year? This could be fun.
This is a good idea. I think it would be fair to the SP.
But it just made me mad reading this. I just hate how the MLB treats “saves” as an official stat. It just so happens I have my own stat for measuring reliever value that requires the reliever to meet some arbitrary criteria. When a reliever pitches 5+ innings and gives up 3 or less runs I call it a Nasty Benoit. How come the MLB doesn’t recognize my stat? Its just as random as Holtzman’s “save” stat.
I’d be fine with getting rid of pitcher wins and saves altogether. Wins because they’re not a good stat. Saves because the statistic influences the game in a bizarre way.
Heres something I thought. Instead of having the wins stat as it is today withh all the strange, arbitrary rules, why don’t we just limit “wins” to starting pitchers and basically make it the team’s record on games that he starts. Ervery game is a descision!
I’m with that.
So your issue is that the reliever with the BS does not deserve the win. Neither does the starter who finished an inning with a deficit, had his team take the lead in the next half inning, and then was relieved in the subsequent half inning. He gets a W when he left the game with a deficit.
Wins and Saves are just not fair. Don’t make them unfair and biased too.
This.
Decisions of all stripes are woefully poor indicators of what a pitcher did during his time on the mound.
Learn a lesson from college football, don’t try to fix a bad system by applying a patch to every exception. Just blow the damn thing up already and figure out something better from the ground up.
Craig, I actually don’t think your example works in all situations. What if the pitcher was down 1-0 and came out after the top of the 8th. In the bottom of the 8th his team scored 3 runs. Based on your logic, he doesn’t deserve the win; I counter that he pitched against the opposition for 8 inning, and in his teams 8 innings they scored more runs.
I think your argument holds only if the pitcher is on the visiting team. If his team takes the lead and he pitcher never makes it back out, he then wouldn’t deserve it since he never pitched in the Xth inning that his team scored.
You just reinforce my point that the allocation of Wins is unfair. I’m just saying that we should leave it as being uniformly unfair. The article proposes a rule that can only benefit a starter (or really, the pitcher who left the game with a lead). If a starter blows a lead he can still get a W, but a reliever cannot.
I want to see the full list, if for no other reason than to see how many of the blown-save wins in 2011 belonged to the Blue Jays.
More importantly, how many belonged to my fantasy team?
I don’t think the process is worth changing the rules for, but why not just give all starting pitchers that go 6+ the wins & and all the losses (the longest reliever not getting a save with the win in the others) and have relievers just have holds, saves, and blown saves? Wins are largly a matter of a good hitting team scoring more than they give up, if a starting pitcher can’t work with that for a win or loss after they have put in their time I don’t think blaiming a reliever is the way to go in handing out extra wins.
I knew this article will generate a lot of comment. I’m waiting for some comment presenting a torturous rule on when runs should be charged to the starter or reliever when there were runners LOB.
The bottom line for me is that I don’t care even one tiny little bit who gets or doesn’t get a “W-L-S-BS-Help-WPA-RBI-game-winning hit. . .”
How about just saying a reliever who gives up the tying run is ineligible for the win? I suppose exceptions are possible, although I’m not going to spend a lot of time thinking of them.
The you’d have games with no winning pitcher! What ever would they pout in the box score?!?!?!
Yep. We must retain the boxscore tradition.
Interesting idea. Here is a slightly cleaner proposal–
In cases, where there is currently a blown save win:
a) the pitcher who blows the save no longer gets the win, but does qualify for the save if he finishes the game.
b) the pitcher who would have received the win except for the blown save gets the win.
The line of discussion also makes you think about whether middle relievers deserve wins as a reward for not being effective.
I have a really simple rule for assigning wins. Go back to the days when the scorekeeper made a judgement call. The algorithm for assigning wins fairly would be far too complex, as I think you begin to demonstrate here. In most games the job would be easy. In others it would be less clear. But to boil it down to a hard-and-fast rule would be hugley complicated.
Here’s a fairly common scenario: One pitcher goes five good innings and exits with the lead. He is followed by five guys who throw a crappy inning apiece and blow the lead. They are followed by one guy who gets a save. The fairest thing is for the starter to get the win. But to make a rule to accont for all the possible similar cases is too complicated for most fans to get.
We trust the scorekeeper for lots of judgement calls and in most cases, assigning the win won’t be the most controversial of any given game.
I should read more closely– you were talking about the pitcher who enters after the starter. I was thinking more about end of game pitchers who blow saves. The thought process applies to both.
Short– yes, what you say.
I have had a similar idea for a long time. I don’t think it should be specialized for starting pitchers, though. Here are my conditions:
(1) A reliever blows a lead (not necessarily a save)
(2) Said reliever becomes the pitcher of record for the loss
(3) The offense bails out the reliever, and he becomes the pitcher of record for the win
Then the win should be reverted to the prior pitcher of record.
Yeah, I don’t see the need for a torturous rule to give the win back to the starter in the event of a blown save if that’s of such concern that you need a new rule. Just make it “relievers charged with a blown save cannot be considered the pitcher of record for determining a subsequent win” and throw it back on the official scorer to figure out who the pitcher of record is, a la a start of less than five innings.
Agree, agree, agree….great idea!
Not that I pay all that much attention to wins, but I’m all for making less ways for relievers to get wins.
It’s always fun to see the 1 IP, 3 ER line accompanied by a W in the boxscore. What do you mean he blew the save, he clearly won the game?!?!?!?!?!
I’ve always thought all the credit for the ER going to the pitcher who put him on base is garbage, as well. Take for instance the following…
2 outs, runners on 2nd and 3rd. Reliever comes in gives up a single scoring both runs and then records the 3rd out. The starter is credited with 2 ER and the reliever 0. I would charge the starter for 1.25 ER and the reliever for .75 in this scenario as this is their relative contribution (assuming all bases are created equally which they aren’t so you’d really want to determine the exact values using a more linear weights based model) to the 2 earned runs.
Not only is the starter charged with the earned runs, but the reliever in that situation gets the blown save — even though he gave up only the one hit. That’s why I focused my proposed rule change on starters who leave the game with the lead and a clean slate.
Thank you for fulfilling my prophecy above.
What if the game goes 20 innings and one of the relivers throws seven innings too?
That was my thought as well. What if the starter pitches 7 innings and leaves with a 6-5 lead. The next pitcher gives up a run in the 8th (to tie the game), then pitches 5 shutout innings and gets the win in extra innings. He pitched 6 innings with 1 run, the starter pitched 7 innings with 5 runs – and the starter gets the win?
On the other hand, I’d be surprised if this ever happened.
Maybe we need to come up with a “draw” statistic, so some close losses are draws and the pitcher gets partial credit. I’ve always liked the game score concept. Maybe something akin to that.
How about we just leave the game alone?
I agree that, at a minimum, a pitcher who blows a save should not be eligible for the win. If that happens, make it a scorere’s decision who gets the win, with the criteria used here determining whether that win goes to the starter or, if the starter does not qualify, to another pitcher.
I don’t like this idea because it will give people the wrong idea that wins are more important than they really are. it is a junk stat and basically close to as meaningless as something like GWRBI’s. I get that it matters to a lot of people. But it matters to fewer people now than it did 5, 10, and 20 years ago. And if we keep on the current path it will continue to be exposed for the junk that it is. That’s what I want. I don’t want to turn that trend around even in the slightest way possible.
I agree 100%. Rather than adding complex rules to “fix” bad stats, replace them with better stats.
There are a lot of comments.. most of which I decided to not read.
Wins are a team concept, they should have never been credited to one single player, I don’t care how many complete games Cy Young pitched. Offense doesn’t score, team can’t win, thus starting pitcher can’t win. Offense has as much to do with winning as pitching/defense does.
Saves are ridiculous because the best relievers on every team usually don’t pitch in the most pressure packed situations. Mariano never comes in the game in the 7th inning with runners on and a 1 run lead. In fact, he rarely ever comes into the game with runners on at all.
Not to mention, getting a save for pitching the 9th with a 3 run lead is hardly “saving” the game. It’s more like “here’s a stat for not completely imploding.”
The win should go to the pitcher for the winning team with the highest WPA. If you want that result, I think you’d come up with it with a very high frequency if you just let the score keeper assign the win based on on which pitcher from the winning team he thinks had the biggest impact on the outcome.
Isn’t this only a slight refinement of the current rule that:
“The official scorer shall not credit as the winning pitcher a relief pitcher who is ineffective in a brief appearance, when at least one succeeding relief pitcher pitches effectively in helping his team maintain its lead. In such a case, the official scorer shall credit as the winning pitcher the succeeding relief pitcher who was most effective, in the judgment of the official scorer.”
I think the only thing that needs to be added to that is that if there was no succeeding reliever and the only reliever was ineffective the win reverts to the starter.
i always thought a pitcher who puts a runner on, pulled and then the next pitcher allows the run to score should not get a full earned run. sure the first pitcher put him on base, but shouldn’t it be just as much responsibility to the pitcher that allows him to score. i mean that other pitcher is out of the game. if you think about it, it wasn’t entirely his fault that the runner scored. it should be split evenly between the pitcher who put the runner on and the one that allowed him to score. i propose a half earned run in that situation for both pitchers
This is an idea my buddies and I have been kicking around for years. Part of it justice. I hate that the win, the first thing in the recaps, is credited to a pitcher who pretty much blew it. And yes, full list please!
I think fractional wins would be a much better idea, where the starter and the reliever would each get credit for half a win in this specific case.
Could just treat wins and saves (and losses) like errors. You have guidelines and the official scorer makes the call.
No.
“I know. Pitcher wins don’t matter. Starting pitchers are better measured by FIP and xFIP and BABIP.”
I have heard this as well but I am having difficulty determining the quantity of BABIP desirable. Does one want more BABIP or less BABIP? Should we regress it?
JK on that last part, obvs regress it.
Whatya mean Pitching Wins don’t matter?! They matter alot…. when it comes to contract time !! But, the other listed stats matter to us geeks and advanced front offices.. I will wager, most pitchers dont even know what FIP, xFIP, BABIP, etc are, and are able to explain their significance.
When starters leave the game with a runner on and that runner scores, the run should bs divided between starter and reliever, with the starter being assigned .25 for each base reached, and the reliever gets .25 for each base advanced. Thus if a starter walks a guy and leaves he gets .25 of the run if the runner scores. If a starter leaves after giving up a triple, he is assigned .75 if that runner scores.
The win should just go to the last pitcher to hold the lead for the winning team before the BS, not necessarily the starter.
When I was in Little League, 3 of us combined to pitch a no hitter. We pitched 2 innings each in a 6 inning game. I still remember how pissed I was the next day when I read the article in the paper, which read, “they combined to strike out 6 batters.” I stuck out the side in both innings I pitched! Can’t remember if I got the win or not, but I got all of the strikeouts and don’t have the ability to prove it with my scrapbook. Not that I actually have a scrapbook…
But it just goes to show, these guys know when they’ve pitched well and when they haven’t, and that’s what those other metrics such as FIP xFIP and BABIP do measure faithfully.
The reality is that the rules for determining what constitutes a win are arbitrary. When rules are arbitrary, they can only be made “arguably less arbitrary” but whatever you might do to tweak them, they will still be inherently arbitrary, and thus cannot be “fixed” in a provable way. Improved upon, perhaps. But given their arbitrary nature, you cannot “prove” that one arbitrary set of assumptions is any better than another.
For example, why not 4 innings or 6 instead of the current rule of 5 innings to qualify for a win? You can’t “prove” that one method is any better than another, so just live with the rules as they exist. If nothing else, that helps to facilitate comparability of players from different eras.