The second greatest relief pitcher as of his retirement (though Gossage has comfortably passed him by '91; not sure about Quiz yet), will that be enough for him? Maybe.

Fingers probably has the best argument for strike credit in 1981 - he was phenomenal and lost 50 games to the strike. Peak voters especially should be boosting that season.

Peak voters especially should be boosting that season.

Should we, though (at least, a lot)? I mean, given the small sample size we're working with on relievers, isn't it plausible that a guy like this, with an insane half-season, would--or at least might--regress to the mean in the other half? (In a way it's almost a moot point, as in my PRAA system this is by far Fingers's best season even without extra credit.)

Nothing's certain either way, of course, but I'm not sure just assuming he'd have been equally lights-out for another 50 games makes sense with a relief pitcher.

>I'm not sure just assuming he'd have been equally lights-out for another 50 games makes sense with a relief pitcher.

Or with anybody. That was certainly not the customary method in adjusting NeLs to 162 games.

I'm deeply skeptical about Fingers. WS, while flawed has him among the top dozen relievers, but near the top. I'm going to do a little more research about LEV and IHRA and maybe WXRL before I make a decision either way. Goose seems better to me.

Which reminds me. I think someone out there has expressed reservations about the use of WXRL. Whoever you are, could you please restate those reservations? Is WXRL something we should be paying close attention to, or does it have difficulties that need to be taken into account?

Goose is better than Rollie for sure. And Jim Palmer is better than Fergie Jenkins, and Schmidt is better than Brett.

To me, the reliever list is Hoyt, Goose, Rollie and then some guy named Rivera some day. We have had relief specialists for 50 years now. You can argue whether to elect 4 of them, or any of them. But if you're only gonna elect 4, Rollie is one. (If 3, then probably not.)

I think you should regress the missed innings in 1981 to Rollie's mean performance, which would add PRAA since he was generally above average.

Goose is better. I like Eck's career better than Rollie's but his relief work alone wouldn't make the short list. If you're including Rivera you can add Trevor Hoffman to your short list also. Lee Smith and Fingers are closer than you might think.

But if you're only gonna elect 4, Rollie is one. (If 3, then probably not.)

I still haven't done research on Sutter and Quiz and Eck yet, and you may be right, but I'm still not convinced it's a slam-dunk. (I agree about Hoyt, Goose, and probably Rivera, from my quick looks at his numbers--hey, I'm a Yankee fan, so I'm more familiar with them.)

Anyway, I prefer to look at this the other way around: not with a quota of relievers, but simply how many meet the criteria necessary to get my vote. Right now, Fingers doesn't, quite--I have him slotted now right at the borderline where he may or may not eventually make my pHOM.

But I'm not saying that means I'll only take 3 relievers ever. I'll check out the other guys as they come up, and I think there's at least a chance that Sutter or Quiz have peaks large enough to get a vote from me. (There's also a decent chance they don't.)

I think you should regress the missed innings in 1981 to Rollie's mean performance, which would add PRAA since he was generally above average.

Absolutely. Which turns him into John Hiller plus a strong prime (Hiller's all peak). That (and some other things) get Fingers past Mike Marshall, but he still falls a bit short, IMO.

>Lee Smith and Fingers are closer than you might think.

Well, true, I overstated. I guess my point is, having eval Smith (and Sutter) I am pretty comfortable that I won't be voting for them. I haven't eval Eck or Rivera or Hoffmann. All are better than Smith, probably, not sure if they're better than Sutter, and not sure if they're above the line (Eck or Hoffmann, Rivera is above the line).

I am confused about Quiz, though. Peak looks incredible, Gossage-like. But the general word is "not good." I guess it's about inherited runners. But look at his OBA and OOB. Dunno. Right now he is well below the line, but more study needed.

Sutter strikes me as possibly more electable than Fingers. He's got a sweet peak/prime. That's all there is, but it's not short either.

Sunny, I agree that Quis merits more study because there's a disconnect between his performance and his reputation.

I guess it's about inherited runners.

I have a question about this, too: Has there been a study showing that preventing inherited runners from scoring is a skill, beyond just how good a pitcher you are in general? And even if so, is this along the lines of clutch hitting--it's a skill, but not a particularly huge one (or as huge as generally thought, anyway)?

I've long wondered whether the difference between two otherwise equal pitchers who have wildly different records with inherited runners comes down to luck, or largely so. I can't imagine this hasn't been studied, but an admittedly quick search isn't turning anything up that seems definitive (or even terribly analytical)....

>Sunny, I agree that Quis merits more study because there's a disconnect between his performance and his reputation.

Do you mean good performance, bad rep or good rep, bad performance?

Maybe the disconnect is between his old reputation and his new one.

Did some more searching, found this from mgl in 2005:

Given enough innings, ERA (adjusted for park and league) tells us pretty much everything about a reliever we need to know. Their strand rate, even adjusted for bases/outs, over and above ERA, again, in the long run, is simpy random variation (i.e., luck). In the short run, of course, strand rate helps us to determine how "effective" a releiver actually was (in terms of helping his team actually win games or not). It doesn't really tell us how effective a pitcher he was (in a context neutral sense). For example, if a pitcher comes into a game with runners on first and second and two outs and allows a bases clearing doublem, he was not very "effective." If he comes in with 2 outs and no one on and allows a double, no problem. Both pitchers pitches the "same" but in terms of value in retrospect, one was way more valuable than the other.

The problem with articles and stats (strand rate) like this is that they serve to further blur the line between past value and true talent/future value. People too often equate the two. For example, if pitcher A has a 3.00 ERA and a great strand rate and pitcher B has a 2.50 and a terrible strand rate, while pitcher A was probably much more valuable than pitcher B in retrospect, the pitcher you would want pitching for you in the present or future is clearly pitcher B. How much confidence you have that pitcher B will perform better than pitcher A in the future depends on how many innings those ERA's are based on, but in no way shape or form, does it depend on either pitcher's strand rate.

It is disheartening (and inconsistent) to me to see sabermetric oriented people and knowledgeable fans clamoring for "inhereted runner" (strand rate) type stats for relievers and VORP-like stats for batters (for example to help determine the MVP). They are completely different types of stats. One is 100% context neutral (VORP) and does not necessarily have anything to do with how much value a batter actually produced for his team (how much he actually helped them win), and the other (an inhereted runner stat for pitchers) is just the opposite in that it gives you an idea as to how much value a reliever created for his team, at least in terms of runs allowed (although not necessarily in terms of wins). The equivalent stat to VORP for pitchers (relievers and starters) is an ERC (component ERA) turned into a runs above average or replacement saved. Again, VORP for batters or for pitchers or an ERC turned into a VOPR-like stat is strictly a context neutral "theoretical value" type of stat and nothing necessarily to do with actual value produced (although it will obviously correlate well with actual value produced, especially as you increase sample size). This is not the right thread for this, but using a theoretical, context-neutral stat like VORP or linear weights (absolute or rate) is ridiculous for an MVP award, for obvious (to me at least)reasons, yet, again, so many saber-types try and do this. Enough of my late-night rant...

So, as I was starting to realize on my own, we're dealing with the difference between value and skill here--if preventing inherited runners from scoring is largely a matter of luck (beyond how good a pitcher you are in the first place, anyway), should that matter for our purposes, since we're not trying to figure out how well any of these guys will pitch for us next year?

Yet somehow, to me, the luck vs. skill factor does matter somewhat. If the difference in a given system between Fingers and Quiz comes down to inherited runners--and we feel that preventing them from scoring does indeed come down largely to luck, which I realize is far from established thus far, at least by me--I'm inclined to adjust things so that the difference doesn't count for quite so much. Results matter most, of course, but someone can be unlucky and still remain a dominant (and, to me, worthy) player.

Obviously, if BABIP is random, then we have to factor random events into the evaluation, which is just giving due consideration of all things that create value.

we're dealing with the difference between value and skill here

Lot of such discussions here lately. For example:

I'm not sure just assuming he'd have been equally lights-out for another 50 games makes sense with a relief pitcher.

Doesn't matter. The 1981 season was 109 games long, not 162. The VALUE of every single thing that happened in those 109 games was magnified by 50%, because a pennant was still a pennant. Could Rollie have sustained that level of excellence over 116 innings instead of 78 innings had there been no strike? Irrelevant, because there WAS a strike. We're trying to measure the value of what he actually did, not try to figure out what he might have done. He did what he did, and it had an IMPACT on a 109 game season that was EQUIVALENT to the impact a similar rate over 116 innings would have had in a 162 game season.

The question of whether he could do it for 116 innings only becomes relevant if you're curious if he could ever have that much impact in a regular 162 game season. Probably not, but that doesn't take away from what he actually did. Giving everybody two months off in the middle of the season may very well help the relievers more than anybody else because they could possibly be paced differently in the second half, taking on more innings and/or appearances due to the time off. It doesn't make him more skillful though. The fact is that Fingers took advantage of a special circumstance to have a uniquely valuable season. Good timing.

The fact is that Fingers took advantage of a special circumstance to have a uniquely valuable season.

Good point. I've adjusted accordingly, but it still doesn't get him all that close to my ballot.

OTOH if that was the final word on schedule adjustments, the HoM would be full of pre-1893 pitchers.

Can I interest anybody in Tommy Bond? Jim McCormick?

OTOH if that was the final word on schedule adjustments, the HoM would be full of pre-1893 pitchers.

Can I interest anybody in Tommy Bond? Jim McCormick?

Only if you're a peak voter using out-of-the-book Win Shares ;-)

if BABIP is random, then we have to factor random events into the evaluation

And if it isn't? (hint, hint...)

-- MWE

To sum up what I originally posted about Rollie and his post-season heroics on a discussion thread:

I will add about 10 career "wins" (or one win per prime year) to Finger's career totals, based on what he accomplished in Octobers between from 1972-74. He basically added almost one whole World Seires trophy to the A's collection in those 3 years, compared to what likely would have hapened had a 'typical' good reliever been in his place; or even if Rollie had pitched no better than he did throughot his career.

Without this credit, he would not get on my ballot; he just did not save as many runs as many other backlogged pitchers. But with this, he edges on the bottom, and edges about even with the Goose thru 1990 as the 2nd best reliever to date.

OTOH if that was the final word on schedule adjustments

Yes, I can be a little dogmatic about the schedule adjustments. ;-)

My point is that when it comes to the value numbers, we're scaling, not extrapolating. The impact of 12 WS in 81 games is the same as the impact of 24 WS in 162 games; scaling is simple arithmetic.

If we're trying to figure out what the guy would have done if he had actually PLAYED 162 games, all of these extraneous considerations come in when we extrapolate. Was it even feasible to do that? (For 19th century catchers and pitchers the answer was often no; they would be too banged up. For the other guys, why not, later generations did it without much problem.) What was the guy's individual likelihood of getting injured during the extrapolated playing time? Was he having a good season or a bad season? (If it wasn't typical of him, then regression is called for.)

But all of these questions when we extrapolate have different answers for different players, and they can change the ordering of certain players when compared to other players who played in the same season. Bennett or McCormick may have been more valuable than Anson over the sprint of an 84 game season, while Anson might have been more valuable if they'd gone out and played the 162 game marathon. But they only actually played 84 games, so extrapolating and then concluding that Anson was the more valuable player is not a true result.

Similarly, the 1981 season was only about 108 games long. Fingers probably couldn't have sustained that pace over 162 games. But to downgrade his season relative to the position players that could have sustained their pace? That does not yield a true picture of his impact during that season.

OTOH if that was the final word on schedule adjustments, the HoM would be full of pre-1893 pitchers.

Can I interest anybody in Tommy Bond? Jim McCormick?

Only if you believe the way WS splits credit between 19th Century pitching and fielding is correct.

Didn't we all have this exact discussion a couple "years" ago?

What happens if I just run the straight RA+ Pythpat system, with no leverage adjustments or anything. Fingers comes out at 111-78, compared to 88-50 for John Hiller, 108-80 for Stu Miller, and 89-65 for Mike Marshall. His 1981 checks in at an equivalent 8-1, but constast that to 1979 when he was an equivalent 4-6. Actual conversation I had in 1981: "Is Fingers really that good?" "No, because no one is that good."

Fingers had a lot of decisions, considering how much he pitched. It works out to 7.33 IP/decision. But the same can also be said for Miller (8.14 IP/decision) and Marshall (6.63 IP/decision). And many of the "extra" decisions were losses, which can also be said about Hiller, Miller, and Marshall - the bad W-L record with the "extra" losses just comes with the role.

Is what Fingers did good enough to clearly separate himself from the likes of Hiller and Miller? I don't see much separation.

Joe has argued, and I agree, that just looking at RA+ or ERA+ as a measure of effectiveness will underrate Fingers relative to his contemporaries because

1) he was more successful in preventing inherited runners from scoring
2) he was more highly leveraged
3) he was, in a career context, much more durable

There are a variety of ways of looking at Fingers that don't fully take these factors into account, and they all show Fingers to be excellent, but maybe not excellent enough. The big question for ranking Fingers, it seems to me, is whether one should take these factors into account, or not. I think they are all important to a reliever's merit, so I am convinced that Fingers should be elected.

Chris, I think mark's/mgl's point above, however, is that maybe all of those things aren't equal. I'm talking somewhat theoretically here, not about Fingers specifically.

If pitchers have either no ability to prevent BIP hits or a limited ability to do so, then it seems speculative to weight IHRA too much more than we would general effectiveness. This may be particularly true for certain types of pitchers. Quis for instance was a pretty extreme sinkerballer (I think). Would an extreme sinkerballer have a propensity toward BIP hits? And more over, would an extreme sinkerballer like Quis (without too many Ks) have a propensity to give up more IHR because he allows more balls in play even without hits: so giving up groundballs that score runners from third (or second on conseutive ground outs in the appropriate base-out situations).

But on the other hand, doesn't leverage, in essence, take IHR into account? If a guy enters the game with no one on and no outs and a three run lead in the ninth, his leverage is X. If he enters with a three run lead in the ninth with three on and no outs, his leverage is much higher. Those inherited runners are already part of the leverage function. So part of my concern is the possibility of double-counting. If we use leverage as Karl has suggested (as a simple multiplier) we're accounting for performance and for inherited runner situations. If we do that AND we figure out effectiveness with IHR, I wonder if we're overrepresenting IHR in our figuring.

Finally is career durability all that important for a reliever? In a way I could see arguing it's not terribly important. High-peak relief seasons, it seems to me, should be more valuable for relievers than perhaps at other positions because they do pitch in specifically leveraged situations. Because of the leverage, dominant performances are of magnified importance. Now if you wanted to argue that X pitcher displayed a lot of in-season durability, well that's something else. Being available for multiple inning stints mulitple times a week, particularly when you're pitching at peak efficiency, seems more valuable to me than being able to string together lots of one-inning appearance seasons.

I don't really know the answer here, but my own thinking suggests that leverage is the most important factor, then maybe IHR, then durability, but that the difference in importance between LEV and the others is pretty big. But that's just a recently formed opinion, and I hope others will be able to shed more light on this.

Rollie Fingers – I have been thinking about what the minimum amount of career win shares it would take to make my ballot. Koufax has the fewest of any player I have voted for, 194 and he had 3 MVP type seasons. I know Fingers won an MVP, but I don’t think of that as an MVP type season.

So what is Rollie's arguement? Best at his position? Because I don't really see that.

Gossage had 3 20+ win share seasons, Quiz had 4, Fingers none. I didn't really have time to has this out last week, so I hope to this week.

Mike, I think almost all of us know how impressive Gossage was, so Fingers doesn't really need that to get him there.
What he had that others didn't is incredible inherited-runner stops and also stellar WS results - if a pennant is a pennant, what's being a key to multi-WS titles?

Everyone should at least mull over Dimino's stats on this.

Like most of us, I'm not looking to elect a platoon of relievers either. We all liked Wilhelm, most will like Gossage and Rivera and quite possibly Hoffman.
Fingers is the challenging case, I think.

What he had that others didn't is incredible inherited-runner stops and also stellar WS results - if a pennant is a pennant, what's being a key to multi-WS titles?

I just don't know that this is reasonable grounds for a HOM case, in fact, I think it's pretty shaky.

Why do I say that?
1) No one has shown us that IHR prevention is a skill/talent, but there is some evidence that relievers have less control over it than we might think based on the divergence of IHR rates among the top relievers we've looked at. IHR stops DO happen, just as clutch hits DO happen. But without compelling evidence that IHR stops are a skill/talent, I'm concerned that we will use it as a measuring stick for merit where it may only be appropriate for accounting and may be inconsistent with the use of the many context-neutral stats upon which we all base our evaluations (or nearly all of us).

2) Giving post-season credit is truly a matter of taste among the members of the electorate. I don't think I'm likely alone in believing that it's a potentially unfair practice, particularly if it is selectively applied (I don't know if that's the case, but I don't see lots of October credit citings in the comments, but I could have missed them, of course). There's many reasons for believing that October credit is a dubious practice, most important among them that it rewards circumstances and opportunities which are not distributed anywhere close to evenly throughout the league and which clearly benefit certain groups of players, for instance: Yankees, Early 1970s A's, 1970s Reds, 1990s Braves, 1990s Indians, 1910-1914 A's or Giants. If October is a very, very small portion of a system, then maybe it's fine, but if we're considering hanging a guy's plaque up in some large part on the basis of post-season accomplishment, particularly a second-best reliever, then I think we really need to talk out the implications of post-season credit.

Anyway, if, indeed, Fingers can be summarized as having IHR and October as his leading arguments, I'm wondering if we've overbuffed the shiny new toy.

However, despite my somewhat combative tone, I agree with Howie on this point for sure:

Fingers is the challenging case, I think.

Eric . . . what inherited runner stops tell is how many runs a pitcher's performance prevented. That's incredibly important, and for guys that pitch a lot of partial innings, it can be much more important than ERA. If a guy comes in throws 1/3 of an inning, gives up 2 hits and a walk and no runs, his ERA is 0.

But if there were two runners on and he allowed them to score, that's much more representative of his performance than his ERA, isn't it? I mean what's the run expectancy for bases loaded one out? It's 1.65 runs (based on 1999-2002).

Inherited runners isn't a 'clutch' stat. It's one more way to look at a pitcher's performance, besides the all too often misleading ERA.

And inherited runs prevented is compared to an average pitcher, it's not a counting stat like RBI. I think it's a mandatory modifier on any reliever's record.

If October is a very, very small portion of a system, then maybe it's fine, but if we're considering hanging a guy's plaque up in some large part on the basis of post-season accomplishment, particularly a second-best reliever, then I think we really need to talk out the implications of post-season credit.

2nd best reliever? What do you mean? Please elaborate . . .

2nd best reliever? What do you mean? Please elaborate . . .

I see gossage as the best reliever of the era. Fingers probably second best. Based on commentary in various threads, this opinion seems to be held by a pretty large percentage of the electorate as well.

But if there were two runners on and he allowed them to score, that's much more representative of his performance than his ERA, isn't it?

I don't disagree with you here. This is truistic. But I'll fire back with my own situation: Bases loaded, no outs, ninth inning. Fingers comes on, gets two easy grounders then the third out, but two runs score before the final out on those grounders. Three-up, three-down, as Fingers trades outs for runs and finishes off the win. Sure he lowered the WinX, but on the ground, he's also just had a 1-2-3 and salted away a win by inducing weak contact. All IHR may not be created equal, nor all IHR that are allowed to score.

While I agree that the accounting of a relievers' effectiveness is changed by his performance with runners on board, I still don't know whether or not his actual performance with runners on board is a function of his skill/talent or whether it is a random thing. For example, let's say for discussion's sake that Fingers allowed the exact same WHIP and opp SLG (relative to league) as Quisenberry. Yet Fingers is miles ahead of Quis in IHR prevention while Quis is fairly well below average in that department. What does this tell us about them as pitchers if anything? Does it tell us that Quis couldn't bear down and get the big outs? Or that Fingers could? If that is what it tells us, that's important information to have about their merit, and I'd be excited to know that!

Or does it say nothing about them?

And if it says nothing about them, then is IHR performance an argument for Fingers that makes him meritorious? Or does it mean that he had a verifiably amazing string of fortunate outcomes in IHR situations that are now accounting for a moderate boost to his overall run prevention? And if so, then what?

The problem I have with IHR is that the modern closer almost never has any IHR to prevent or allow, whereas the fireman model frequently faced batter with IHR. It's trickier than it looks, if you ask me.

If October is a very, very small portion of a system, then maybe it's fine, but if we're considering hanging a guy's plaque up in some large part on the basis of post-season accomplishment, particularly a second-best reliever, then I think we really need to talk out the implications of post-season credit.

I mentioned something similar a few weeks back to OCF via e-mail. Something just seems wrong to me that a reliever has a disproportionate advantage over position players and starting pitchers when combining regular season and postseason play because relievers don't pitch that much during the regular season. I think some type of postseason relief adjustment may be in order in our analysis.

I have to say the more I look at relievers the fewer I want of them. Papelbon is clearly, from the team's actions, more valuable to the Red Sox as a #2-3 starter than as an elite closer. I'm glad we elected Wilhelm but even Gossage looks very iffy. Their ERA+ just isn't that great, and if you adjust IP to reflect an equivalent starter you have to adjust ERA+ as well, at which point they fall below the borderline (except for Wilhelm and Mariano.) Going into the analysis, I would have expected the elite relievers to have unadjusted ERA+ of 140-150; if they don't it suggests they were mostly overhyped midlevel starters.

I wasn't saying that Fingers' case is mainly inherited runners and postseason - I was saying that he already was a close call, and those other things help a bit. I'm not a huge postseason credit guy at all. It's just part of a pattern - Fingers is very good all the time, and the bigger the spot, the greater he got.

The Yankees' Wang fairly clearly right now is deliberately inducing ground balls all the time, except when he needs a K - then he dials it up. I guess some SABRmeticians just think he's lucky, but this is a skill.
I don't know the exact Fingers analog, other than that his inherited runners numbers seem too far ahead of the norm to ASSUME that it is luck. At a certain tipping point, the burden falls to those who want to prove something is NOT a skill. I think this may be one such example. If not, he should take Lou Gehrig's "luckiest man on the face of the earth" comment away and claim it for himself.

Re IHR:

Something Joe hasn't said in defense of _his use_ of IHR in his _comprehensive, quantitative evaluation_ that perhaps should be said. Joe is looking at all the IHR data and including it in his total assessment of runs prevented. He's not looking at IHR independently as a rate stat, like we would do when we compare hitters by OPS+. If a reliever doesn't ofen enter a game with runners on, then IHR will have little impact on Joe's assessment. If he often enters with runners on base, then his IHR will have an important effect. So, saying IHR is problematic because it doesn't apply to modern closers is beside the point, at least in Joe's system, which does not penalize pitchers for dealing with few inherited runners, but rewards pitchers with many inherited runners for preventing them from scoring, which his system should do, because that is what, generally, relievers who have many inherited runners are being asked to do. If Quiz, for example, has a low IHR but actually had few inherited runners to deal with, then that low rate shouldn't affect his value much.

If it is possible to post _aggregate data_ showing total numbers of inherited runners and number who scored, per season and per career, as well as rates, that might help voters judge the importance of this statistic more effectively.

In the case for Fingers, Joe is not arguing (and I am not arguing), that because Fingers had a great IHR, that makes him a great reliever.

The case, instead, is this. By straight ERA+ or DERA, Fingers' effectiveness seems to make him a borderline candidate. His effectiveness over his career in preventing inherited runners from scoring, which was exceptional, raises his effectiveness a significant amount that is not reflected in his other statistics.

The thought-experiment case where a relief pitcher enters with the bases loaded, no outs, and three run lead and allows two runs to score on groundouts in an interesting case, but it is not really meaningful. In this case, wouldn't we say the reliever who comes on and strikes out the side, preventing any runs from scoring, has performed better than the pitcher who has allowed two runs on ground outs? In the context of this particular win, the two pitchers will both have done their jobs in getting their teams the win, but the one who has allowed no runs has clearly been more effective as a pitcher. I think the cases where allowing inherited runners to score is good strategy are fairly infrequent, and even when it _is_ good strategy, not allowing inherited runners to score is _better_, if it can be managed. Therefore, in the aggregate, IHR ought to be a meaningful statistic about the pitcher's performance.

As to the argument that IHR might be random and have no link to a pitcher's skill, an obvious approach to assessing that would be to look at the data for a lot of relief pitchers and see if there is any correlation between success at preventing inherited runners from scoring and other kinds of success. Maybe I'm incurious, but I have to suppose that the good stat heads at BP wouldn't be counting the things if they didn't think the stat was meaningful. I'm certainly open to considering data that shows that it _is_ random. Nevertheless, since it is generally considered to be a relief pitcher's job to prevent inherited runners from scoring, the presumption should be that we should give them credit for doing so, unless there is strong evidence that pitchers have little ability to affect the rate at which runners on base score. And phrasing it that way, I have to say that I am highly skeptical of the idea that they don't, as we generally believe that the quality of the pitcher has a significant effect on runners' scoring in general, and that most runs that a starting pitcher gives up begin as runners on base, which he either strands or allows to score. He doesn't have _sole_ responsibility for these runs, as his fielders will play a role, but that's a division of responsibility we are pretty comfortable with quantifying. I can't see why the IHR case should be significantly different from the general case.

Since I introduced the IHR topic, particularly with regard to rates of IHRS, let me put forth some stuff.

These are the MLB IHR/IHRS/IHR% figures I calculated a couple years ago. At the time, what I did was to simply copy all of the IHR and IHRS data for all relievers in BP's IHR/BQR tables for the years 1972-2005. Those tables have since been reformatted and no longer include this data.

Here's the yearly IHR data:



Sunny may be generally correct that the incidence of IHR per relief appearances is trending downward from in the high 90% range in the 1970s to now about 50%. But this is likely a function of the fact that there's simpley more games in relief nowadays and only so many IHR to go around, so take it with some salt. In general, however, it does appear that slightly fewer IHR are scoring today than did in the 1970s and even the 1980s, with the numbers generally trending down from the mid 30%s to the low 30%s.

OK, that's your overview. Back with cases in a little while.

If only MLB had adopted different rules w.r.t who gets charged with the run when more than one pitcher appeared in a certain inning, this issue would be much smaller.

I'd personally prefer something like

reliever enters game with 0 outs in the inning - starter gets charged with all inherited runners
1 out - reliever gets charged with runner on first if he scores, but starter with runners he bequeathed on 2nd and/or 3rd
2 outs - reliever gets charged with all inherited runners who score

outs runners on
.. 1st 2nd 3rd
0 .St .St .St
1 .Re .St .St
2 .Re .Re .Re

But they didn't ask me, and it ain't so; ergo, the IHR analysis is one salient point for evaluating relievers. Their saves/blown saves record will reflect their IHR success, which is probably one reason Saves have at least 10% of the relevance most writers and fans give them.

Of course, BS almost always also result in an L and a modern closer almost never gets an L without a BS.

Who are guys who have faced a lot of IHR? Who are the guys with the best career IHRS%? The following is an overview.

Here's the top 25 guys in the raw total runners inherited and in the unindexed rate of IHRS.



1970s Relievers: We Eliminate the Middle Man.



If we narrow it down to guys with 200+ career IHR…



On the back end of it, the worst guys are a little over 40%. Interesting to note the presence of Hoffman, Myers, Percival, and Eck among the closers. The 1970s aren’t well represented here at all, if at all.

One group that’s extremely well represented are lefty set-up men. Were I to guess this might be because
1) They don’t face a lot of hitters
2) They have a very strong platoon advantage to begin with
3) They may have an advantage over righties at holding runners on first base.

So who are the top 1970s guys? I eyeballed the list and pulled these guys off who seemed like they were 1970s guys to me. BTW, the average MLB IHR% in the 1972-1980 period was 34.8% to give some context. As you move closer to now, it gets lower, but by 1985 the average hadn't really begun to drop yet and remained around 34.7%-34.8% (depending on whether you use the average of the averages or weight the average by the number of IHR).



And among 200+ IHR guys



The lefty issue is pretty much eradicated since these guys are mostly right handed and if not tended to face more batters anyway. As noted above, this was a pretty loose definition of 1970s reliever, so it’s possible I’ve left some guys out.

Back in a little while with some career IHR season-by-season breakdowns.

In this case, wouldn't we say the reliever who comes on and strikes out the side, preventing any runs from scoring, has performed better than the pitcher who has allowed two runs on ground outs?

This is a pretty specialized case, but I would say no. In the ninth inning, baserunners who, if they score, won't tie the score simply don't matter at all; this is why no one holds runners on then, and fielders (if they're smart) pay no attention to them except as potential outs on the bases (and even then only if they're damn sure they can get them).

So to me, a reliever coming in that situation who gets out of it without losing the lead--however he achieves this--has done the job. I honestly don't think he's done a better job if he did it with three straight strikeouts; that will just be more exciting, and look better. The result is what matters, and that's the same regardless.

I think the cases where allowing inherited runners to score is good strategy are fairly infrequent, and even when it _is_ good strategy, not allowing inherited runners to score is _better_, if it can be managed.

Sure, if you're confident the guy you have out there can manage it. I admit that I'd rather have vintage Koufax pitching my ninth innings than CM Wang. But when you become less confident your guy can strike out the side, the idea of having a guy you know (or at least are reasonablly confident) will induce weak groundballs to the right side becomes more appealing.

And unless you're talking about extremely high-K pitchers, I think the question really comes down to what you go on to discuss: is preventing inherited runners from scoring a skill, or luck (or some combination, most likely, as with most things)? I'm less sure than you are about this--just looking at Eric's '70s list, the names aren't breaking down as you'd expect, exactly. You do have some of the stars of the era at the top (Fingers, Tekulve, plus guys like Burgmeier and Grant Jackson who had some great years), but there are also some guys sneaking down to the league average for the era (Gossage, Lyle, Marshall).

I'm looking mainly at that last list, with the over-200-IHR guys, so we're unlikely to see any really bad pitchers here. But the distribution of these names at least seems like it could be random.

Maybe, in fact, all those guys at the bottom are indeed a little overrated because they let all those inherited runners in. And maybe Dave LaRoche was a lot better than I realized at the time. But that list just makes me want to see more evidence that IHR isn't similar to clutch hitting (a skill, but not nearly as significant a skill as it would appear).

Thanks Chris Cobb for your post #36, sums up my system better than I have . . .

As for how I count inherited run prevention - I'm not using any percentages. I'm using the BPro stat, Inherited Runners Prevented, which gives the total number of runs prevented (or allowed - the stat goes negative) compared to the runs an average pitcher would have allowed.

The two biggest negative numbers are lefties (Tom Burgmeier and Darold Knowles); this makes me think they are looking at something along the lines of base-out state as opposed to just raw inherited runners, but I'm not positive.

To elaborate, I use this number and add it or subtract it as appropriate from runs allowed. So as Chris says if the pitcher never inherits any runners, it's a big zero.

I also use their Bequeathed Runs Prevented stat, which adjusts for whether the relievers that inherited the pitchers runners were good or bad. And for starters I use their bullpen support stat, which works the same way.

Nearly all relievers have a positive number for BRP (which means they are helped by their relievers), which again shows me evidence they are using base-out state, as opposed to raw numbers. Also, I think this is where much of the "reliever ERA advantage" comes from, when you adjust relievers for this, it brings their ERAs back closer to the starters.

For Fingers specifically:

*As a starter, his relievers cost him 7.6 runs.
*As a reliever, he prevented 33.1 more inherited runs from scoring than an average reliever.
*His relievers prevented 22.4 of his inherited runs from scoring (which is typical for a reliever with as many innings as Fingers - an average reliever that I've studied would have 22.9)

So the net total is that I subtract 18.3 runs allowed from Fingers for his career.

His defenses were slightly below average, so that basically knocks .01 off his RA in a 4.50 R/G environment. His leagues were weak (as were all expansion era pitchers, especially those with significant time in the AL) so that adds .08 back on.

So how do I get his DRA+ up to 124 from 119 - it's mostly the 18.3 runs I subtract and a little bit that he didn't give up a whole lot of unearned runs either, despite playing for slightly below average offenses. 10.7% of Fingers' runs were unearned, Lindy McDaniel was at 12%, Wilhelm 18.2%, Gossage 9.7%, Stu Miller 12.5%, McGraw 11.6%.

There's no doubt in my mind that Fingers was a better pitcher than his ERA+ suggests.

Oh and I don't see why people discount Fingers' 1981 season.

Maybe (and that's a real maybe, it's not impossible that he did), Fingers didn't deserve the MVP, but he sure as hell deserved the Cy Young Award. It was an amazing season. He gave up NINE runs in 78 innings. He has 28 saves in his 47 appearances (note the IP/G ratio, these were not 3 out/3 run lead saves). His IRP and BRP move his RA total to 9.4, so that 9 is legit, it's not like he gave up a ton of inherited runners or left behind a bunch of messes.

His innings were leveraged at 1.78 that year - so it was, over a full season the equivalent of a starter throwing 198 innings with a 1.43 RA (not ERA, RA) in a 4.50 environment. That's one helluva valuable season. The 3rd most valuable reliever season I've found so far, behind only Hiller's 1973 and Gossage's 1977. And his team won the division and made the playoffs for the first time in their history.

So Fingers' reliever peak is just fine by me. He's got the 3rd best 'best season'. He's got the 8th best '2nd-best season'. He's got the 5th best '3rd-best season'. And he's second only to Gossage on 4th and 5th best season.

So I'd describe his peak as 'damn-good' and his prime as 2nd best, only to Gossage.

Maybe the LHP tend to come out when still in mid-inning, minimizing the opportunities for inherited runners to score while they are actualy in the ballgame?

To add some useless anecdotal observation to all this: I've watched just about all of Mariano Rivera's career closely, and I'm quite convinced that he, at least, ignores runners in the ninth that aren't the tying or go-ahead runs--he really appears not to care at all if they score. He does generally seem to bear down more to be sure the runners that can tie the game up or win it don't come around, Luis Gonzalez notwithstanding. Hey, he was tired that day. :)

Of course, most of these runners these days aren't inherited, since he doesn't come in with runners on all that often, so this may be a moot point as far as strict IHR numbers for him. Eric, where does Mariano fall on your charts--I notice he's not there, but is that because he has a mediocre rate, or because he just doesn't have a ton of inherited runners at all (or both)?

One thing to clarify - the big numbers for lefties are in the Bequeathed Runners Prevented, not Inherited Runners Prevented. In my mind I've kind of merged the two, so I just wanted to clarify what I wrote in #43.

Burgmeier and Knowles have fairly normal IRP, but their BRP are through the roof. It's probably because of what you say Tom, that they leave mid-inning, and therefore leave more messes behind.

BTW, the two worst IRP by far that I've seen are Jack Baldschun at -29.2 and Quisenberry at -27.5. I'd think this is probably because of Quisenberry's low K and high H/IP rates. Quisenberry's WHIP was very good because he never walked anyone, but he did give up a decent number of hits, a little more than a hit per inning.

Baldschun had a good K rate, but he threw a ton of WPs, 49 in 704 IP, and he wasn't all that good anyway.

Oh, and Quisenberry's DRA+ is 132, compared to his ERA+ of 146, so that's the kind of impact extreme Inherited Runner stats can have. Quis also had pretty good D behind him and pitched in fairly weak leagues (similar to, but slightly worse than Fingers).

Quiz will not be on my ballot, partly because of what you said here, Joe; the H/9 are high, which meant he was not as effective coming in with runners on. I played a simulation league when Quiz was a star reliever, but despite the stellar ERAs and surreal control, the high H/9 were a red flag and I thin Quiz' "owner" was disappointed with those years.

My thought about Fingers - and relievers in general - is the last guy on my ballot has 302 career win shares, 3 30+ Win Share Seasons, and 7 seasons over 20. Fingers doesn't seem to provide A) the total career value B) the best single season values - or strung together peak values.

I understand the concept of leverage, but suppose the A's didn't have Fingers. Well, then Knowles or Bob Locker or Diego Segui, would have filled in. Suppose they didn't have Bando, then Ted Kubiak would have filled in.

I guess I am just saying that finding a closer is doable, finding a Bando, or a Murcer, or a Singleton or a Campaneris appears to be much harder.

OK, before I do season-to-season breakouts, I forgot to talk about indexing a reliever's IHR% versus his league. I did this for most of the important relievers of the 1970s-2000s. Not all of them, and now with the data gone it's too late to go back, but I can share what I've got. The method I used goes like this:

For season
1) Find RP's IHR% by dividing IHRS by IHR
2) Divide into league's IHR%

For career
1) For each season, take IHR% and figure how many runs the league average pitcher would give up given the same number of IHR
2) Sum the league's IHR
3) Divide player's career IHR%, divide into league's.

OK, so when we do that for my pool of 45 top relievers, here's what we get:


name IHR%+
---------------------
hoffman 1.65
wagner 1.45
myers 1.43
orosco 1.34
percival 1.33
smith, l 1.32
eckersley 1.27
fingers 1.23
tekulve 1.23
righetti 1.19
benitez 1.17
nen 1.14
beck 1.13
campbell 1.13
reardon 1.13
aguilera 1.12
sutter 1.12
henke 1.10
kern 1.10
rivera 1.10
hiller 1.09
hrabosky 1.08
mcgraw 1.07
worrell 1.07
giusti 1.05
smith d 1.05
hernandez w 1.04
marshall 1.04
mcdowell 1.04
wetteland 1.04
franco 1.03
gordon 1.03
gossage 1.03
hoerner 1.00
lyle 1.00
garber 0.99
eicchorn 0.98
carroll 0.96
hernandez r 0.96
stanley 0.96
granger 0.95
olson 0.94
jones 0.92
quisenberry 0.90
montgomery 0.87
==================
AVERAGE 1.11


Generally, the top relievers are indeed above the league average as a group. It is interesting to note that the ones we'd flag as possibly worth studying the merit of are dispersed throughout the rankings. For example, Quisenberry, Gossage, Marshall all fall below the group average. So do Wetteland and Lyle, even Mariano is a hair below average. On the other hand, Fingers, Hoffman, L. Smith, and Eck are all safely above. Sutter is just a hair above average for this group.

(to be continued...)

When you get down to cases what do you see? First to cut a wide swath, I just took a look at each guy and figured out in how many seasons his IHR%+ was within .50 of his career average. The average guy in this pool was within .50 of his career mark, under or over, 76% of the time. They were within .25 of their career marks 47% of the time. They were within .10 of their career marks 20% of the time.

Here's some of the most consistent relievers in this group in terms of their IHR%+. These are guys with the highest percentages of seasons within 10% of their career average. This is not the best way to measure this, I’m aware of that, but I’m also not sophisticated enough statistically to know the best way to measure consistency.