Bruce, Lee, and the Goose
by Tangotiger
I recently introduced Crucial Situations. Based on the various game state (inning, score, men on base, and outs), we can determine the chances of winning at any point in the game. We can also determine how much swing (leverage) those situations present. The close and late situations are high-leverage.
Relief pitchers are a different breed because managers can pick and choose the situations they face. And, the better the pitcher, the more likely this pitcher will be pitching in crucial situations, and the more impact he will have to winning (or losing) the game. But how much more?
Bruce Sutter
Bruce Sutter faced 4,251 batters. I looked at each PA, one at a time, and determined how much leverage each PA had (based on my chart). For example, a typical situation has a leverage of "1". One of the most crucial situations of all (bottom of the 9th, ahead by 1, men on second and third, and 1 out) has a leverage over "10". That is, the result of that PA, because of the game state, is magnified 10 times (like having 10 typical PAs all rolled into one).
Bruce Sutter's Leverage Index (LI) was 1.90. That is, Sutter facing 500 batters is the equivalent of a typical pitcher facing 950 batters. Here are Bruce Sutter's career stats, along with the "leveraged" line, with all his numbers multiplied by 1.9.
Pitcher IP H ER HR BB SO Bruce Sutter 1042 879 328 77 309 861 Leveraged 1980 1670 623 146 587 1636
I have my own similarity scores for pitchers (adjusted for league run environment, but not for park), and I looked for the most similar pitchers born since 1940, to fit this leveraged profile.
Note: All rate stats are per 9 IP. "xH" is non-HR hits.
Pitcher IP ER xH HR BB SO Lvrg. Sutter 1980 2.8 6.9 0.7 2.7 7.4 Mike Scott 2069 3.5 7.3 0.8 2.7 6.4 Ron Guidry 2392 3.3 7.4 0.9 2.4 6.7 A. Messersmith 2230 2.9 6.2 0.7 3.4 6.6 Jose Rijo 1803 3.2 7.4 0.7 3.2 7.8
Bruce Sutter is one of those very good pitchers, with a few great years, but whose career was just not long enough.
Goose Gossage
Goose's LI is only 1.62. Remember, his games as a starter reduces his overall leverage index, but increases the total number of innings. Going through the same exercise, here are Goose's leveraged stats, and his most similar comps
Pitcher IP ER xH HR BB SO Lvrg. Gossage 2931 3.0 6.9 0.6 3.6 7.5 David Cone 2881 3.4 7.0 0.8 3.5 8.3 Mark Langston 2963 4.0 7.3 0.9 3.9 7.5 Dwight Gooden 2801 3.5 7.6 0.7 3.1 7.4 Fernando V. 2930 3.5 7.7 0.7 3.5 6.4
Well, Goose certainly looks a bit better that Bruce Sutter. The way you feel about Dwight Gooden, Fernando Valenzuela, and David Cone, will lead your decision as to whether Goose Gossage should go to the Hall Of Fame.
Lee Smith
Unfortunately, I only have access to Smith's play-by-play records until 1990. His LI until then was 1.73, and this will be the figure I will use.
Pitcher IP ER xH HR BB SO Lvrg. Smith 2230 3.0 7.3 0.6 3.4 8.7 Jose Rijo 1803 3.2 7.4 0.7 3.2 7.8 David Cone 2881 3.4 7.0 0.8 3.5 8.3 Bill Singer 2174 3.4 7.5 0.5 3.2 6.3 A. Messersmith 2230 2.9 6.2 0.7 3.4 6.6
Looks like a combination of players from Bruce Sutter and Goose Gossage! I don't see Lee Smith's credentials being that strong.
Other Considerations
While I have covered the aspect of leverage, there is also the aspect of timeliness of performance. That is, is it possible that Goose Gossage performed better in the high-leverage situations, and worse in the low-leverage situations? That would be the topic of another study, similar to the Mills Brothers' Player Win Averages (and their offshoots like Doug Drinen's Win Probability Added, etc). Of course, maybe Gossage's comps have also showed the same proclivities.
Conclusion
The save statistic is perhaps one of the most poorly used statistic in trying to determine the value of a player. We should look beyond the save total to determine a player's true value. Once we get the save numbers out of our mind, we are left with 3 very good pitchers, whose contributions were limited by the innings they threw. The impact of an 80-inning reliever is no more than that of a 160-inning starter. And that's how we should view them.
Acknowledgements
As always, this research would not have been possible without the event files that have been compiled by Retrosheet. The effort in producing this research was made far easier using Ray Kerby's A.S.S. software.
December 17, 2002 - Dan McLaughlin _The Baseball Crank
(www)
(e-mail)
One little useful fact in the same vein, although I'm not 100% sure it means anything: from 1977 to 1984, an 8-year span, the Goose's teams exceeded their Pythagorean projections by 21 games, almost 3 full games a year. The biggest effects came, generally, in some of the seasons when the Goose pitched the most - 1977, 1980, 1984. (Dan Quisenberry has a similar, even more impressive record: for the six seasons of his prime, from 1980 to 1985, the Royals exceeded their Pythagorean record by 20 games.) Bruce Sutter's teams exceeded their Pythagorean records by 19 games over 9 years (1976-84), although the biggest damage (+7) was done when he was a rookie setup man; the numbers break down to +16 for his first three seasons and +3 for the next 6 years when he was mostly used in save situations, albeit with a much heavier workload than the modern closer.
Lee Smith? -8 games from 1982 to 1995. I don't hold Smith responsible for that, but it's another fact suggesting that his impact on his teams was less significant than guys like Gossage and the Quiz who threw 100-135 IP a year at their peak, with ERAs in the low 2s and the 1s.
Does this prove anything? Logically, you expect teams with great bullpens to win the close ones. It's noteworthy in Gossage's case that the biggest seasons were the ones when he was paired with other good relievers (Kent Tekulve, Ron Davis). I think if memory serves, I've seen some studies have shown a slight overall effect for teams with good bullpens (witness the Braves this year), but I could be wrong about this, and I obviously haven't studied the question. You would think that it should be an extra feather in a guy's cap if his team won an unusual number of close games when he owned the 8th and 9th innings.
December 17, 2002 - RN
Tom, Great stuff.. If it's not too much trouble, I'd be interested in seeing Hoyt Wilhelm, Rollie Fingers and Dennis Eckersley's numbers crunch in similar fashion. I recognize that Hoyt and Eck, like Goose also started games early on in their careers..
Rod
December 17, 2002 - Vinay Kumar
Wow, I was just hoping to see something just like this. Great work, Tangotiger, it's very interesting work. And, at first glance, it seems to prove me wrong (I guessed that Goose's innings were higher leverage than Smith's). A couple questions and thoughts, though:
What is the LI of a typical starting pitcher? Tangotiger multiplied each reliever's stats by his LI and then found comparable pitchers; but that assumes that those other pitchers have an LI of 1. If those starters have LIs around .9, then the relievers are beind underrated by 10%.
I think the adjustment really needs to be applied on a year-to-year basis. As you point out, Goose pitched more innings as a starter, but that lowered his LI. But those starting innings are still multiplied by his career LI. This is especially painful for Goose, as his year as a starter was much worse than the rest of his career. This is what you're getting at when you talk about timeliness of performance, although it seems that you're talking more about clutch performance, while I'm referring to a pitcher's role and performance both changing over the course of a career.
December 17, 2002 - Vinay Kumar
I was curious how big a difference adjusting for LI year-by-year would make (as opposed to doing it for the career totals, like Tangotiger did). I assumed that the LI for a start is .9 (that's a guess, it could certainly be closer to 1), and worked backwards to see that the LI for a Gossage relief appearance was 1.73 (interestingly, this is exactly Smith's career LI). Then I estimated his LI for each season, prorated his stats by that LI, and summed his career: 2940 IP, 960 ER, 2.94 ERA and 129 ERA+. That's nice, a difference of about 3 wins over his career, but not really enough to alter a HOF vote.
Of course, if we took it one step further and adjusted each season by the actual LI (as opposed to my estimate), that would change things a tad bit more, too.
December 17, 2002 - tangotiger
(www)
(e-mail)
Vinay, the starter is almost exactly 1.00. I looked at two starters, one who was good and went long (Blyleven), and one who went short and was not so good (Knepper). Bert was .99 and Bob was .98.
As for historical, I have the LI for the 20 pitchers with the most relief games from 1974-1990:
pitcherid Leverage Index suttb001 1.90 smitl001 1.76 fingr001 1.75 gossr001 1.72 rearj001 1.69 smitd001 1.52 orosj001 1.48 laveg001 1.46 quisd001 1.45 mintg001 1.45 tekuk001 1.41 garbg001 1.38 lyles101 1.35 campb001 1.31 stanb001 1.30 martt001 1.24 leffc001 1.20 hernw001 1.18 baird001 1.10 andel001 1.03
This is the LI only while as a reliever.
Sorry, but my data is limited to the pbp provided by Retrosheet.
I agree that doing the LI by year, and then doing the multiplication by year would add the "timeliness" aspect as well. I might do it for one of these guys, maybe Quiz.
December 17, 2002 - Oliver
Fascinating, innovative stuff. Thanks. How much impact does the team have on a reliever's LI? One might expect very good and very bad teams to offer fewer crucial situations to their relief pitchers. Have you calculated any team LIs, and how close is the correlation between a team's LI and its relievers' LIs?
December 17, 2002 - tangotiger
(www)
(e-mail)
Thanks... good question. I haven't calculated it yet, but I have another tool (which for a lack of a better name I call the Tango Distribution), which shows the expected runs per game distribution, given the runs per game of a team. Using this, I can figure out the win% for any two teams, broken down by run differential. Surprisingly (to me), there is little difference between a .400,.500, and .600 team in terms of "number of close games". I would suspect that I could extend this to "number of crucial situations" as well, and therefore, expect that most team face a similar number of crucial situation. The more you get away from being a .500 team, the less the number of crucial situations. I'm not sure what the relationship is between team win% and crucial situations (yet). I'll keep this in mind next time I'm working on this. Great question!
December 17, 2002 - Anonymous
December 17, 2002 - Craig B
Tangotiger:
Since a typical starter has a Leverage Index around 1.00 or so, and a closer can be up around 1.90, what is the figure for a typical long reliever? I would imagine it must be quite a bit lower than 1.00...
Maybe a season like Mark Eichhorn, 1986, where he threw 157 innings in relief and would have won the ERA title if he had five more. I'm betting his leverage index was well below 1.00.
December 17, 2002 - Devin McCullen
Good analysis, but it doesn't really get at the major issue in dealing with relief pitchers, which is "What is the Hall of Fame standard for relief pitchers?" I realize tango tries to translate the reliever's record into something comparable to a starter's record, but I don't know if that's the right answer. Looking at the similarities, I wouldn't put Gossage's comparables in the Hall, but it doesn't mean he should be out. If you look at the numbers, the best relievers are not going to have as much impact as the best starters, so you're left with 2 choices: 1)Keep relievers out of the Hall of Fame or 2)Accept that less valuable players will recieve the honor. I would go with the second choice, because relievers certainly play an important role in modern baseball, and this should be recognized in the Hall of Fame, and by far the most important part of the Hall is the inductees.
But the stickier question is, if you're going to honor relievers, where do you draw the line? For starters and position players, although it's far from clear, there's a pretty good idea of the qualifications that takes care of 90% of the candidates (and we have fun arguing about that last 10%). But for relievers, it really hasn't been defined, even among the sabermetricans. I don't have a good answer, other than my own instincts. I think Gossage is a better candidate than Sutter or Smith, but good enough for the Hall? Until there's some kind of standard for that, we're not even sure what we're discussing. To sum up, before we can decide if these men are Hall of Fame relievers, I think we need a better definition of what a Hall of Fame reliever is.
One other question for tango. I'm a little bit concerned about the same pitchers showing up as similar players for these guys. Since I don't know what your similarity criteria are, I'm wondering if there's something in certain starting pitchers' records that tends to make them 'similar to relievers'. If there's some kind of bias, your system might not be giving you the truest equivalencies. I don't know, it just seems like correspondances between Sutter and Smith, and between Smith and Gossage, is too coincidental.
December 17, 2002 - Charles Saeger
Tom -- nothing to add, but I must say that this was a great article.
December 17, 2002 - MattB
Without knowing the formula used, it sounded to me like the lowest possible number assigned to a situation was a "1". That would mean that, since every pitcher faces SOME crucial situations sometime during their career, the baseline should be somewhere above 1. If that is so, then aren't all of the relievers discussed actually overvalued by this method?
December 17, 2002 - tangotiger
(www)
(e-mail)
Craig: Mark Eichorn, in 1986, was 1.32. If you look at the list of 20 players I listed in my followups, you will note that Bob Stanley is 1.30. I think this is probably what you'd find with your multi-inning non-closing firemen. Eichorn did have 10 saves, and finished half his games, so you might be careful about how you extrapolate his usage to other relievers. In any case, his 157 relief innings is equivalent to 207 typical innings. The guys below 1.00 in LI are the true mop-up guys.
Devin: Your point is valid regarding what it would take. Since the HOF is a self-defining institution, I don't see how I can answer that question with any basis. Writers, like fans, are flying by the seat of their pants in trying to establish the potential impact a reliever has.
Pete Palmer and Bill James try to answer this question by using a combination of SV, GF, and G to come up with a reasonable estimate. I'm offering the same type of solution from a different angle. Therefore, I think it is irrelevant how we think they impact, and how they've changed the way the game is played and managed. The fact is that the impact of the best relievers, while real, is not substantial enough to catapult them to the levels of the superstars. And the best of the lot is good enough to put them in line with star pitchers who lack longevity. This is why relievers are paid they way they are. GMs may have figured out their true value already.
However, your point is just as valid, and that the HOF may not simply be about "overall value". And perhaps relievers do deserve a special spot. I don't know, and I think that the writers also don't know.
As for my sims, I'll run a couple more, like for Quiz and Reardon and Bob Stanley, using their LI. I'll let you know what shows up.
Charles: thank you! I had alot of fun doing this piece! I just wish I could devote more time to this.
December 17, 2002 - Marc Stone
Perhaps I missed this, but how do we know the Leverage numbers are scaled correctly? I accept that a "10" is a more crucial situation than a "5" but is it really twice as much? Or is it 10 times or 1.2 times as much?
December 17, 2002 - Devin McCullen
I agree completely about the writers. There's no rational reason for Rollie Fingers to be a HoFer and Goose Gossage not to be. This is going to be an interesting subject for the Hall of Merit to consider as well (although with the rules there, "No Relievers" may be a perfectly justifiable demarcation).
December 17, 2002 - tangotiger
(www)
(e-mail)
No, you didn't miss it. I explained it in another article.
If you go to the first line of the article, and click on the link, you get a general explanation of how I determined the leverage of the situations. If you then go into the comments section, in one of the December 4 comments, I elaborate on how I derived the leverage values. Hope that's good enough?
I apologize for making each of these win expectancy articles links to links to links. They're all related, and it's very tough (for me) to write it adequately, without making it a mathfest.
December 17, 2002 - tangotiger
(www)
(e-mail)
If ever I get pbp prior to 1999, and I get the World Series pbp, I'd *love* to look at Mariano Rivera. He may turn out to be a borderline candidate like Goose or Lee Smith, based on regular season numbers. But when you add in his tremendous playoff performance, that may be enough to get him over.
In fact, I am surprised how little play playoff heroics get. In the NHL, they have the same problem. The NHL and NBA is *all* about the playoffs. But the awards and HOF, etc, is mostly about the regular season. Rarely do you see the two combined. I believe in soccer, they combine all games, regardless of "league". Pele has 1,241 (or whatever) goals, with no split.
December 17, 2002 - ColinM
Really fantastic work. I have a question regarding peak value for relievers. It would seem to me that the years in which a reliever would be used with maximum leverage would be those seasons in which he was at his best. So, for example, Goose had a LI of 1.62 for his career. But Goose also had a number of mediocre years at the end of his career where he probably wasn't used in high leverage situations. So isn't it likely that his LI for his peak seasons may be quite a bit higher, maybe more like 2.0 or so. If this is true that would make his best seasons, 75, 77, 78, extremely valuable, HOF caliber.
I guess my theory is that while we can determine which starting pitchers had a similar career value using a relievers career LI, a top reliever such as the Goose is more likely to have a greater peak value than those starters because his LI will be greatest during his best years. So it may be that while it is almost impossible for a relief ace to compile HOF level career numbers, his peak value might be high enough to justify election along the lines of Koufax or Dean. Do the numbers support this, or am I off base here?
December 17, 2002 - tangotiger
(www)
(e-mail)
Thanks Colin.
Yes, your point is very valid, and Vinay also brought up the issue in his comment. Essentially, the next breakdown is to look at each PA within the context of the leverage of the situation (which is what the Mills Brothers and Doug Drinen did). So, while Bruce Sutter may be 1.90 overall, he'd have say 30% at a 4.0 leverage, and 40% at a 1.50 leverage, and 30% at a 0.3 leverage. And maybe during the 4.0 leverage situations, that's when he was his best (whether because his manager used him during his peak years, or he rose to the occasion, or by luck), and therefore, he was even more valuable.
It's possible that there is some impact here, especially with the relief wild card.
It requires some upfront work on my part to get the whole thing set up. I'll see if I can devote some time to this. Perhaps after XMAS.
December 17, 2002 - Marc
Congratulations, leverage is certainly a valiant attempt to understand why top relief pitchers are "stars" when they're active and lousy HoF candidates when they're not. We understand leverage (ie. clutch) situations intuitively and qualitatively, but we have no system for considering them quantitatively. Until now, maybe.
But it seems that leverage does not really treat top relievers quantitatively in the way we relate to them qualitatively. Sutter, Smith and Gossage, most would agree, I think, were really better than their "comps."
BTW, it seems to me that because of the high strikeout numbers, the comps for modern closers are going to be limited to recent starters. This limits the perspective that the comps provide, unfortunately.
As to how the HoF should or might proceed to evaluate relievers, right now you've got a range from Wilhelm to Fingers. If you really check out Hoyt Wilhelm, you'll agree that nobody has pitched in relief at his level, ever, he da man. And Fingers, as has been pointed out, we hope the Hall will not fall below this level. But even so, only Bruce, Lee and the Goose, so far, seem to be competing at or near the Fingers line. Eck will also do so in ought-four.
I think the best that we can do is to make sure we know, at least, who the best HoF candidate among the relievers is, and focus attention there. If the vote is split (and I don't think anybody knows if it is, but if the vote is split) between two or three of the relievers instead of focusing in on the best one, then none will get elected.
So this year it seems to be the Goose, but next year Eck.
December 17, 2002 - Joe Dimino
(e-mail)
Good stuff Tango, but I agree with Vinay and Colin, that you have to go year by year for the numbers to be meaningful. For example, Goose's starter year is weighed way too much. Not only did he pitch a disproportionate share of his innings that year, but those innings are then being weighed 1.62, when the leverage wasn't nearly that.
I'd say the 1.9 for Sutter is interesting, because that is about what Bill James guessed a typical closer's value was, in his relief study in the NHBA. Maybe we're getting consensus on that at least?
December 17, 2002 - ColinM
That would be great Tango, to see the breakdowns by situation. I look forward to seeing the results. After this step, it would be nice to see if we could come up with a better way to estimate leverage for relievers for those seasons where PBP data isn't available. For the last 5 years or so I have used my own little method for relief aces, to estimate historical leverage-equivilent innings. Simply IP + (G-GS)/2 + SV. It seems to give results that make intuitive sense, and looks reasonably consistent with your leverage numbers. I still like it better than Bill James' save eqivilent innings which just has a hard cap. Of course it only "works" for top relivers as the assumption is that relief games are more valuable than GS. It would be nice to use the PBP data to come up with a formula that isn't quite as rough an estimate as this.
December 17, 2002 - tangotiger
(www)
(e-mail)
I think this is the exercise that Vinay did, but in response to Joe's point, let's go through it step-by-step.
Let's assume that Gossage has an ERA+ of 126. Let's assume that he had 251 IP as a starter, and 1558 as a reliever. Let's assume that as a starter, his ERA+ was 100, and as a reliever it was 130. Fair enough?
Now, his LI as a starter was 0.99. His LI as a reliever was 1.72. As we see in the above paragraph, he pitched his best as a reliever. So, if we take his 1558 innings and multiply by 1.72, that gives us his "adjusted typical" innings. Do the same for the 251 x .99. Good so far?
Now, weight the 130 ERA+ by 1558x1.72 and the 100 ERA+ by 251x.99. You end up getting an ERA+ of 127. That's compared to the initial value of 126.
The point is that because very little of Goose's innings came as a starter, the change won't affect much. If this was Eckersley, then that's a different story.
That said, while the impact is small in this case, we should still do the breakdown as I mentioned in the previous post, so that we are leveraging each particular PA, and not applying an overall leverage on the sums of the PA.
December 17, 2002 - Vinay Kumar
Yes, Tango, that's essentially what I did. The final ERA+ figures that we came up with differ a bit, and that's because the numbers I estimated were slightly different than yours. Oh, and when you averaged ERA+s, you used (LI-adjusted) IP as the weight, when you have to use (LI-adjusted) ER as the weight.
I'd say the 1.9 for Sutter is interesting, because that is about what Bill James guessed a typical closer's value was, in his relief study in the NHBA. Maybe we're getting consensus on that at least?
Joe, that's very close to what what Doug Drinen's data suggested as well. Although in any given year, that value could be much higher or lower for an individual closer (for instance, win probabilities said that Trevor Hoffman was much more valuable than any starting pitcher in the NL in '98, and it would take a multiplier >2.0 to reach that conclusion; of course, Hoffman went 53/54 on saves that year, so that helps).
December 17, 2002 - Gilbert
Even as a non-St.Louis fan, I still think of Bruce Sutter as the split-finger fastball guy before Roger Craig tried making it a staple. He should get some credit under what has been called the "Candy Cummings" Keltner test criterion. Even if he wasn't a pioneer of it, he inspired more successful imitators than Christy Mathewson inspired "fadeaway" imitators.
December 18, 2002 - Steve Cameron
I'm not sure what to make of this analysis. I don't like the strict comparison of relievers to starters as comps, I'd there should be some other relievers who'd pop up.
The point of this analysis seems to be that none of the relievers - and that relievers in general should not be inducted. I base that on the following quote:
"The way you feel about Dwight Gooden, Fernando Valenzuela, and David Cone, will lead your decision as to whether Goose Gossage should go to the Hall Of Fame."
Like it or not, "relief pitcher" (or more specifically "relief ace" or "closer") is now being generally considered a separate position. Therefore comparing these guys to starters is not worth very much when assessing their career value in retrospect regarding HoF credentials. It's worthwhile when dealing with current events, lkike trades or pitching roles, but we need to look at what they did & how they were used. It'd be like comparing Catchers to First Basemen - we can do it, but it's not really relevant to how good they are/were for their team.
Not sure what else to say, other than I wish there was a Keltner or Gray/Black ink analysis.
December 18, 2002 - Tom N
How much many high leverage plate appearances do relievers cause for themselves? If a relief pitcher comes in with the bases empty, none out and proceeds to load the bases and then work his way out of the jam, then he's turned a moderately leveraged situation into an extremely high leverage situation. Wouldn't his Li be higher for that appearance even though he's endandered the teams chances of winning? You may have addressed this some place or another, if so, I'd be interested in seeing a link.
December 18, 2002 - tangotiger
(www)
(e-mail)
Good point.
There are two things to consider with "leverage". You can take the position of what was the leverage of the situation, assuming that the pitcher will pitch to the end of the inning. So, if it's top of 9th, 1 out, man on 1B, up by 1, the leverage is not that particular situation, but rather that particular situation as the starting point, until the end of the inning. It could be that that particular PA may have a leverage of "4", but the "starting from that PA to end of inning" may have a leverage of "2".
Furthermore, you can also take the point of view that if a reliever gets himself into a jam that the manager is "bringing him in" to get himself out. That is, after every PA, the manager is deciding whether to bring in his existing pitcher, or bring in a new pitcher.
Remember, my point of view is crucial PAs. So, PA by PA, what is the leverage. I don't know if it's the pitcher or the fielders that caused the change in leverage. And really, I don't care. What I care about is how often did he face a high-leverage situation.
It is important that you don't make a stat do what it wasn't designed to do.
If I were designing a model to decide when is the optimal point in the game to bring in a reliever, such that he will pitch to end of inning, I would have different leverage numbers. And if I design a model, such that my pitcher will pitch to end of inning, plus one more full inning, I'd have again, different leverage numbers.
All these methods are good, within the context of their design assumptions.
December 19, 2002 - tangotiger
(www)
(e-mail)
To add to the point about "how often do reliever cause high-leverage PAs for themselves": Bruce Sutter comes into a high-leverage situation, and he keeps it high-leverage. Fat Rojas comes into a high-leverage situation, and turns it into a low-leverage situation (by giving up 3 run HRs).
So, there are various reasons as to turning a type of leverage situation into another type of leverage situation. It's not just a "if he's bad, then..." kind of deal. It's alot more intricate than that.
December 19, 2002 - Scott
"The way you feel about Dwight Gooden, Fernando Valenzuela, and David Cone, will lead your decision as to whether Goose Gossage should go to the Hall Of Fame."
I disagree. Comps are useful, but only go so far. Goose's comps had substantially higher ERAs (and gave up more hits and HRs).
Goose had about a 10-year stretch there where his ERAs were unbelievably low -- under 2.00, I think. It's unfortunate he's being rolled together with Smith & Sutter. In a way, it's the opposite of what haqppened to Duke Snider: in a group of 3 (Mantle, Mays, Snider), two deserved it, and the third only marginally did, so the strong two carried the third in. Here, there are three relievers, one of whom deserves it (Goose), but the other two probably don't, and they're dragging him down into a pool of "three top relievers who don't deserve it."
December 19, 2002 - tangotiger
(www)
(e-mail)
Gossage gave up .5 more walks than Gooden, but .7 less non-HR hits. Gooden's run environment was slightly higher than Gossage. The actual runs allowed by relievers are a little suspect because of the "accountability" issue. Not that it should be ignored, but just you have to account for it.
On a "rate" basis, of pitchers born since 1950, I have Guidry, Cone, Rijo, Sutter, Blylven, Gooden, Gossage all being "equivalent". In terms of IP or leveraged-IP, clearly Blyleven is the one that stands out here. By the way, John Smoltz is also in this group.
Gossage is borderline, in my view.
December 19, 2002 - Paul Wendt
(www)
Tango, I have wondered where you work ;-)
Two relievers "off the map" by TPR are John Hiller (single season 1973) and John Franco (career). Introduction of the decision factor (my term), derived from saves per inning and wins & losses per inning, vaulted Hiller and Franco to the top sometime between 1990s editions of Total Baseball.
It will be interesting to see how your method of Leveraged Innings will assess Hiller and Franco. Maybe a Tigers fan will volunteer time to help Retrosheet complete 1973.
Paul Wendt, Watertown MA
December 19, 2002 - Walt Davis
Just a couple quick comments....
One problem we're having here, I think both in terms of assessing them by traditional standards and by Tango's new-fangled standards, is that we've yet to see a closer (or fireman) hold the job for a truly extended period of time. Smith and Gossage both did it for about 13 seasons. Maybe this means comparing them to starters isn't the best way to go. With a few exceptions, HOF starters held that job for a good 18-20 years. Relievers, at least high-leverage ones, seem rarely able to do that (Wilhelm and ???).
Which brings us to the idea of making a correction for position. 3B, 2B and especially C tend to have shortened careers. Why? I don't think we know that (except for C's), but I think most of us make some "correction" for this when evaluating their HOF standards against those of a 1B or OF. So while we don't know why relievers should have shorter careers (in fact it seems counter-intuitive), maybe we need to compare relievers to relievers.
Of course maybe the reason is that they aren't that good and that's why they ended up in the pen to begin with. So maybe we shouldn't compare them to other relievers and maybe there shouldn't be any in the HOF any more than the best 4th OF in history should be in the HOF.
Now what I think would be neat is to compare relievers to the starters of yesteryear, but only in late-inning, high-leverage situations. If we think of the evolution of the closer/fireman/bullpen, the notion is that these guys are more effective late in the game than a starter who's pitched 7-8 innings. How does Gossage compare to the performance of HOF starters in high leverage 8th-9th inning situations?
I'm assuming we don't have enough PBP data to answer that question, but it's one I'd like to have before making an HOF decision on these guys.
December 19, 2002 - tangotiger
(www)
(e-mail)
Paul, welcome! I don't think I've seen you around here? I'd love for Retrosheet to get more PBP, and I'd love to run the 73 Hiller, and the Franco career through their paces.
Walt, my inclination is to say that they are in the pen because they are not Roger Clemens or Greg Maddux. However, they are David Cone or Ron Guidry, and those guys were pretty darned good. I don't have separate standard for catchers or anyone else. I look at it as how many wins did they contribute over some baseline. If you are a catcher, and you only play 120 games, and you are done by 34, then I don't have different standards. Not to say I'm right or anything.
You make a valid point that relievers can be considered similar to catchers (can't play long enough in a season or a career). So, you have to first resolve why they have shorter careers (because of the position, or the quality of the players there). Then you have to resolve if you want to have a different standard.
It's a tough call no matter what your perspective is.
December 20, 2002 - Marc
I've been adjusting my ratings of RPs upward for several years now...and catchers. I don't adjust anybody else. If you look at the records, shortstops do fine. Yes, there are no more Wagners but there are 25 guys who pop up to HoF candidate level by most numerical systems. If anything, it's the CFs who suffer, you just don't notice it because of the few who rate so very highly, but the mid-level guys fall off very rapidly.
But anyway, as to RPs, I don't agree that Sutter and Smith drag Gossage down. They are all dragged down by a lack of any useful standard. Wilhelm is the gold standard, the Wagner, the Ruth, but that standard is useless on mere mortals. And the Fingers line is suspect as not being high enough.
If anything, Sutter and Smith will help Gossage over the long term, as will Eck and any other really great RP that comes along just by helping to inch along toward a consensus of what the standard is. When we have ten guys at the Fingers-Sutter-Smith-Gossage level, then it will be easier to elect the best one. It will come too late, I am guessing, for the Goose to get elected by the BBWAA but will make it easier for the VC (though of course we don't know yet what the biases of the new VC will be).
I agree that Goose is the best RP on the '03 ballot, but Eck will be stronger in '04 and that will kill the Goose. If on the other hand the BBWAA does NOT elect Eck, then we've got a long long wait ahead of us.
Marc
December 20, 2002 - Scott
Interesting point Walt and others have made about how there are fewer 2B, 3B, C, & CF in the HOF. I'll ignore 3B for a moment, but C, 2B, and CF are three of the more grueling positions, physically. We all know about C. Note how far more CFs than LFs and RFs end up having surprisingly injury-shortened careers (Murphy, Lynn, etc.). And 2B end up being slid into far more than any other position (because more grounders go to shortstops who throw to 2B) -- Gantner, Vina, etc. Tthat's one reason the Yanks are thinking of moving Soriano, and Baseball Prospectus hypothesized that this is why more top 2B prospects under-achieve.
The point is that while the best 1Bs can be expected to have a 15-20 year career, the best Cs can be expected to have a 10-15 year career (of good years, at least), and that's why Carter and Murray deserve similar praise for their career achievements. The same is true to a lesser degree about CF and 2B, but much less recognized -- so they get hosed.
The relief pitcher situation is different. There's no reason to think the position is more grueling and therefore that the fewer inninge (even the fewer "leverage-adjusted innings") reflect that they did their duty in a physically grueling position. To the contrary, Smoltz shows that relieving can be a lot less grueling (setting aside oddities like Dale "The Horse" Mohorcic").
The point here is that there is a very good rationale for a "position adjustment" for not only C, but also CF and 2B. That rationale is not present for RP.
December 20, 2002 - tangotiger
(www)
(e-mail)
If I remember my post, Eichorn had 157 real innings, and 203 leveraged innings. It may be that that's as much mileage as you can get out of a reliever. That, because of all his warm up tosses, etc, etc, you won't get more than that. Then, he's got to do that for 15 years.
However, I don't know if this is a physical limitation, as it is for catchers (who play 130 games instead of 150, and who play 13 years instead of 18, e.g.). If the relievers are physically limited to 200 leveraged innings insead of 250 for starters, and 12 years instead of 16 for starters (just examples), then it may be fair to consider the relievers to have lower standards, like catchers.
However, this should be studied to the extent that catchers' careers have been, before we pronounce sentence.
Even after all this though, people can still choose to not lower the standards for the C/RP.
December 20, 2002 - Marc
>The point here is that there is a very good rationale for a "position >adjustment" for not only C, but also CF and 2B. That rationale is not >present for RP.
I'd suggest that what you've got is not so much a rationale as a hypothesis. I'm not going to do a respectable study, but if somebody did (tango?), here is what you might find.
Hypothesis: C and CF and 2B and RP have a difficult time having a normal-length career because of the rigors of the job. Test the hypothesis by determining whether fewer than expected (or, fewer than at other positions) last long enough for a HoF career.
Now check the evidence: Catchers--evidence confirms hypothesis. CF--evidence confirms hypothesis. 2B--evidence clearly does NOT confirm hypothesis, there are lots of deserving HoF 2Bs.
RP--evidence inconclusive. New position, not enough cases. Not proven but not disproven either. Maybe over the next 10 years there will be five better than Eck, then disproven. If none as good as Gossage or Fingers, then we're getting closer to a proof.
December 23, 2002 - Lenny
Why would make CF a grueling defensive position? There may have been a few excellent CFs (like Lynn) whose careers were shortened by injury, but I think every position has players like that (and I don't remember Murphy being injury prone).
Although I haven't studied it myself, CFs may well have shorter careers. But if that's true, my first hypothesis would be that its due to the fact that many CFs becone CFs due to their speed. When their speed starts to go, their defense declines and they need to justify their spot in the lineup more based on their offense - either by being able to move to LF, RF or 1B, or by being good enough offensively to carry their reduced glove in CF. This speed issue is not as significant for other defensive positions like SS and 2B (and especially C) so players at those positions can remain at a high level defensively even after their speed declines.
If that is the case, I cannot see using a lower standard for CFs than for other positions. While C and 2B (and maybe RP, although that hasn't been studied adequately) generaly have shorter careers due to the physical "gruelingness" of the positions, I think CFs have shorter careers (if they do) just because some of the players who get jobs as CFs early in their careers lack the skill set to remain effective once their speed starts to decline.
December 26, 2002 - tangotiger
(www)
(e-mail)
If there's anyone still out there, Eric Gagne's LI last year was 1.83, and Smoltzie was 1.79.