I like the Clean Innings and IRS stats, but the RBI simplification is incoherent as proposed.

Also, we already have Runs Allowed to go along with ERA.

hate the concept of removing rbi's from homeruns, it's the most pure form of rbi out there. It's the only measure of rbi that is more or less 100% accurate.

Clean up pitchers' wins a little more. The stat that matters isn't wins, because you also need to the know the losses stat as well to make any sense of things. Instead, go with a singles stat: wins above/below .500.

All-time leader is Cy Young at 195, followed by Al Spalding 187, Christy Mathewson 185, Roger Clemens 170, Pete Alexander 165, Lefty Grove 159, Kid Nichols 153, John Clarkson 150, Walter Johnson 138, Randy Johnson 137, Eddie Plank 132, Whitey Ford 130, Greg Maddux 128.

Active leader is Andy Pettitte at 102, followed by Roy Halladay 83, Tim Hudson 165, C.C. Sabathia 69, Roy Oswalt 67, Johan Santana 64, Jamie Moyer 63, Chris Carpenter 50.

1. Are you going to go back and re-award wins and losses to all the starting pitchers of previous generations? Otherwise, I don't see the point.

As a few posters noted here when Poz mentioned flaws with wins before, pitching decisions track nicely with IP/starts, so there's really not that big a problem with it. If you have to make a change, give the official scorer more leeway to award the win/loss to the pitcher most deserving of said outcome in each game.

The better solution is simply to track wins and losses for fun and history, but not put a whole lot of weight into the end result. Which, slowly, is kind of what's happening.

2. At bat for sac flies, no at bat for sac bunts.

I don't think of the term sacrifice to mean the individual is sacrificing his batting average, but the team is consciously sacrificing an out to advance a base runner.

At the same time, I'm pretty sure in most cases sac flies are efforts to get hits that happen to result in fly outs (or, if you will, there's no way to really tell), whereas a sacrifice bunt is clearly an attempt to bunt that produces the desired result. But if you're trying to bunt for a hit and happen to move the runner over and get thrown out at first, that's not a sacrifice.

3. Nah. As with pitcher wins, the problem is in the valuation, not the measurement itself.

4. All for it. I don't mind keeping ERA around, but RA and RA+ should be used just as much and be far more easily accessible.

5. An entirely different matter, dealing with advanced stuff rather than basics. If someone can develop a good metric, by all means use it.

Especially if the way we get #4 is to get rid of errors completely.

I would hate that. Error totals may not be terribly useful when it comes to determining which defender is better, but they play an integral part of telling the story of a baseball game. Billy Buckner's play was an error, and similar plays just like that from now until the end of time should be judged the same.

Stats aren't just a means of divining value.

They should give errors on all the errors that they don't currently give them on. How about a compromise, and baserunners resulting directly from errors are unearned, but nobody else? (Although I think that a pitcher gets tired over the course of an inning, and if there should have been three outs already and he gives up hits, they're less his fault than they would have been otherwise.)

I think a modified earned run should definately be implemented. I think any runs that are a direct result of the error should be unearned, and to a lesser extent runners on base because of an error should have the potential(I think it's unearned if the runner can score based upon the rest of the inning being routine 'groundouts') to be unearned. But every runner after the error should still be considered the responsibility of the pitcher. It's ridiculous that a guy can allow a homerun and not get charged anything for it.

I also have no problem with implementing team errors for communications issue(in fact I think team errors could be a moderately useful stat for grading coaching) Of course I'm also a fan of tracking "running errors" on the offense side(not talking about failed steals, but just boneheaded decisions)


as to the last point, I'm a huge fan of a better system for relievers, as a general rule I start with whip and opponent ops, but neither of them have the feel that you get with era(which is a horrible stat for middle relievers, and a poor stat for closers)

I like #2 because even when a batter is clearly trying to get a hit via the bunt he's always credited with a sacrifice if a baserunner advances. That's bullbleep.

If you have to make a change, give the official scorer more leeway to award the win/loss to the pitcher most deserving of said outcome in each game.

Yes - get rid of the "whoever pitches the fifth inning with a lead gets the win" garbage. When's the last time the scorer actually used their discretion to award the win to the "most deserving pitcher?"

Also, I would allow the final score to take a pitcher "off the hook" who was in line for the loss. Example: Zack Grienke pitches 7 innings, gives up 1 run, Royals are held scoreless through 7 (shocking, I know). Garbage time reliever comes in and gives up four more runs, for a 5-0 game. Royals somehow break through for four runs in the ninth and are held there, for a final 5-4 loss. Grienke gets the loss because he gave up the first run to make it 1-0, but I think once the Royals score a run that should take Grienke off the hook for the loss.

—A Setup. The setup stat could be simply be how often a pitcher hands off the lead to the ninth inning. You might even do this with the same rules as a save, only for the eighth inning.

Don't they already track "holds"? I suppose the "setup" would also apply to starter when he makes it through eight innings, but otherwise, how is this different?

I don't know if it's possible to come up with a stat for pitcher wins that will please everybody. The system currently in place, even though it's Byzantine and unfair, has held up fairly well over the years, actually. The better pitchers tend to win more games.

I think they need to start dividing the earned runs on inherited runners. If you inherit a runner on first and let him in, you get .75 runs and the previous pitcher gets .25 runs, this way ERA automatically accounts for allowing inherited runners to score.

1. Team wins = pitcher wins

The good thing about that is Bert Blyleven is a 300 game winner. The bad thing is that so are a bunch of other people. I wonder who the worst pitcher would be who is a 300 game winner by that definition?

It would mean minimal change for the 1800's guys though, they pretty much pitched complete games all the time.

I think a modified earned run should definitely be implemented. I think any runs that are a direct result of the error should be unearned, and to a lesser extent runners on base

I agree with this. And one scoring rule (don't know whether it's written or unwritten) that's always made no sense to me is the one that says "you can't assume a double play," even though Jose Molina has just hit a one-hop scorcher to the shortstop with a runner on first, and the shortstop has then played around with it for about five seconds before nipping Molina at first, while the runner who would have been easily forced out on a cleanly handled ball has been allowed to advance to second. I don't see any good reason why that unmade force play at second shouldn't be considered an error, regardless of whether Molina was thrown out at first.

....But every runner after the error should still be considered the responsibility of the pitcher. It's ridiculous that a guy can allow a homerun and not get charged anything for it.

By logic, it's hard to argue with that, although one might say that a pitcher's job is to induce three outs per inning, not four or more.

-----------------------

I think they need to start dividing the earned runs on inherited runners. If you inherit a runner on first and let him in, you get .75 runs and the previous pitcher gets .25 runs, this way ERA automatically accounts for allowing inherited runners to score.

That makes a lot of sense, and as a bonus the new Retrosheet hires could drop the unemployment rate by a full percentage point.

Don't they already track "holds"? I suppose the "setup" would also apply to starter when he makes it through eight innings, but otherwise, how is this different?

In general, yeah, my understanding is that coming into a save situation and passing it on to the next pitcher is a hold. This presumably would be different insofar as you could only get it in the 8th inning.

The bad thing is that so are a bunch of other people. I wonder who the worst pitcher would be who is a 300 game winner by that definition?

Jack Morris would have 302 wins.

Others over 300 include Tommy John, Jim Kaat, Jamie Moyer, Frank Tanana and Dennis Martinez.

Don't they already track "holds"? I suppose the "setup" would also apply to starter when he makes it through eight innings, but otherwise, how is this different?

Yes, but if a pitcher has 10 holds, it doesn't tell you much. Poz is proposing a ratio instead of a counting stat. If you hold 10 out of 10 leads, your setup ratio is 100%. If you hold 10 of 20, it's 50%.

Since not all leads are equal, I would track it as number of runs given up/number of runs ahead.

So if you go into the 7th up 5 runs, give up one in the 7th and none in the 8th, you gave up one run. You were 5 runs ahead in the 7th, 4 runs ahead in the 8th. So your setup ratio is 1/(5+4) or 11%.

Example: Zack Grienke pitches 7 innings, gives up 1 run, Royals are held scoreless through 7 (shocking, I know). Garbage time reliever comes in and gives up four more runs, for a 5-0 game. Royals somehow break through for four runs in the ninth and are held there, for a final 5-4 loss. Grienke gets the loss because he gave up the first run to make it 1-0, but I think once the Royals score a run that should take Grienke off the hook for the loss.

Is that right? I thought the pitcher who gave up the winning run (in this case the 5th run) would take the loss?

Yes, but if a pitcher has 10 holds, it doesn't tell you much. Poz is proposing a ratio instead of a counting stat. If you hold 10 out of 10 leads, your setup ratio is 100%. If you hold 10 of 20, it's 50%.

I have no problem with this if they add both holds and saves together to get a success rate, the problem I have with the save percentage stat is that it unfairly penalizes setup men and middle relievers. All hold opportunities, by definition of the stat is also a save opportunity, while all blown holds are recorded as blown saves--there is no blown hold stat--so a middle reliever who records 3 saves, 2 blown saves and 30 holds has a save percentage of 60%, to get a true setup ratio using holds all you have to do is add saves and holds divide it by opportunities in my example giving you a 94.2% success.

Is that right? I thought the pitcher who gave up the winning run (in this case the 5th run) would take the loss?

yes his example is 100% correct. It's not winning run it's the go ahead run that determines win/loss

Of course I'm also a fan of tracking "running errors" on the offense side(not talking about failed steals, but just boneheaded decisions)

cardsfanboy, tootblan. tootblan, cardsfanboy.

Last night during the game, my Awesome Girlfriend was rooting for Lincecum to stay in the game. I said, "Even if he comes out, he still gets the win" -- which led to a weird brief conversation where I tried to explain how it is decided which pitcher gets the "win."
Awesome Girlfriend verdict: "So you can pitch really well, and get the 'loss' because your teammates didn't hit? Or get a 'win,' even if you pitch terrible? That's stupid."
Born sabermetrician, right there.

(I also discovered that she rarely remembers ballplayer names, but instead uses vaguely Native American-sounding nicknames like Bloody Sock Pitcher (Schilling) or Really Good Fat Guy Who Had Lots of Sex (Babe Ruth). But, hey, she knows Babe Ruth was a great pitcher. Keeper!)

cardsfanboy, tootblan. tootblan, cardsfanboy.

of course that linked actually was kind to Soriano, a mistake I'll probably never make.. (quite funny stat though-have to remember it)

Especially if the way we get #4 is to get rid of errors completely.

I understand more or less why everyone craps all over errors, but while we're still getting this from the advanced defensive metrics (from the fielding bible thread) -

Anyone know how Dewan's +/- and UZR came up with such opposite results for Cleveland this year?
Dewan: +50, T-2nd best in the AL
UZR: -60, worst in the AL

- I still find the incredible disdain and dismissal very overblown.


I say count 'em. If you bunt over a runner, if you drive them in from third base with a fly ball, you get credited with a sacrifice, you get the appreciation of teammates and fans, you get known as a team player. But your batting average goes down. That's a true sacrifice, my friend.

Wouldn't this create an indistinguishable disconnect between who isn't good and who isn't selfish?

any batted ball that results in a run scoring should get a run batted-in… I would make it so that you do not get an RBI for driving yourself in with a home run.

It's not simplifying RBI when your only 2 rules contradict each other.

even when a batter is clearly trying to get a hit via the bunt he's always credited with a sacrifice if a baserunner advances. That's bullbleep.

I disagree with the bullbleep-ness. Why "punish" the good bunters by taking away a sac?

The rule for sac flies as AB / not-AB has flipped back & forth a few times.
There's no good reason to keep sac flies from counting as ABs.

agree with those that think that sacrifice bunts shouldn't count against batting average, but that sacrifice flies should. One is clearly trying to make a "sacrifice" even if they are going for the base hit, while the fly ball is very hard to judge what was the batters intent.

any batted ball that results in a run scoring should get a run batted-in… I would make it so that you do not get an RBI for driving yourself in with a home run.

It's not simplifying RBI when your only 2 rules contradict each other.

In the first part, Poz is trying to say that he would give an RBI on a GIDP fielder's choice, or an error.

If you hit a three-run HR, he would give you two RBI. His justification for not counting the HR hitter as his own RBI:

This is not just double counting -- it's triple counting. The guy who hits a home run already gets credit for the home run. He already gets credit for the run scored. I don't think he should ALSO get credit for the RBI. Making this small change would clarify some things anyway -- it would bring back the rarity of the 100 RBI season for one. Anyway, my sense is that with RBIs we are really looking for how often they drive in OTHER runners. By "Good RBI Man" people tend to mean the guy who will drive in the runner from second with two outs or the guy who can get the runner home from third. Tacking on their home runs muddies up the concept, I think.

The rule for sac flies as AB / not-AB has flipped back & forth a few times.
There's no good reason to keep sac flies from counting as ABs.

Well, a sac fly has the positive result of driving in a run, so it isn't a wholly negative plate appearance. It is an out, though, so there could have been better outcomes for the plate appearance. Maybe the back and forth nature of the argument is the reason why the rule has flipped back and forth.

Maybe the back and forth nature of the argument is the reason why the rule has flipped back and forth.

Yep.
Though I've never seen a coherent argument for scoring the sac fly as no AB, but a run-scoring fielder's choice as an AB.

re:30, Oops, the excerpt made me think those were his only 2 rules about RBI.

Even so, his rationale is inconsistent and his rule isn't much simpler than the one in place. An RBI for a solo HR is triple-counting? That's like saying an RBI for a run-scoring double is triple-counting because you get a 2B, a H, and 2 TB's etc etc. Also he wants an RBI for a BB to score a guy, but not for a WP during an atbat.

ERA was invented in a time when defenses weren't as good. So for a time, ERA may have been able to really differentiate between pitchers who can pitch and pitchers who have bad defenses.

I disagree with the bullbleep-ness. Why "punish" the good bunters by taking away a sac?

Because he tried to get a hit and failed. No different than if he swung away.

Because he tried to get a hit and failed. No different than if he swung away.

He tried to do 2 things - get a hit and advance the runner. His failure at not getting a hit is the exact same failure as the inferior bunter who doesn't have the skill to even try to bunt for a hit.

He tried to do 2 things - get a hit and advance the runner. His failure at not getting a hit is the exact same failure as the inferior bunter who doesn't have the skill to even try to bunt for a hit.

Because in only the latter instance was the motivation of the batter truly unambiguous. We don't give sacrifices to guys who hit behind the runner and ground out to second because we don't really know what the intent was. We shouldn't give sacrifices to guys who hit long flyouts with a runner on third (but not on second). And we shouldn't give them when the batter is bunting for a base hit.

Now, as a baseball play, I'd love to see more teams teach their faster guys to learn how to square later and try to beat out the bunt/make the play tougher on the defense in these situations. But they shouldn't be sacrifices.

Errors should be dropped because they are arbitrary and capricious. Most BIPs could conceivably be fielded, some are and some aren't. "Ordinary effort" is a stupid distinction.

Ideally errors will be replaced with field/Fx "difficulty" ratings for all BIP.

SF/SH should be abolished and instead you can track baserunner advancement independently of hit/out.

Of course, there are already stats for these things, if you look hard enough.

Errors should be dropped because they are arbitrary and capricious. Most BIPs could conceivably be fielded, some are and some aren't. "Ordinary effort" is a stupid distinction.

Ideally errors will be replaced with field/Fx "difficulty" ratings for all BIP.

That will sound great on a linescore.

"So, the Angels had one run on two hits and had one batter reached safely on a BIP with .57 difficulty."

Errors help tell the story of a ballgame. You don't get rid of them so accounting can be a little easier for the tiny sliver of fans who are statheads.