With all due respect, VAL, it's not the end of the story. GIDPs are also related to swinging on top of pitches, swinging at poor pitches, and a lack of speed down he line. And while I certainly remember that some of Rice's GIDPs were rockets off his bat, there were plenty that weren't especially hard hit.

I really don't want the GIDP issue to be perceived as a focal point of my case against Rice -- they are actually a tangential point. But added to his poor walk rates, his dropoff in production post-79, his mediocre defense and his lack of speed, I think the collective evidence leaves him short of the bar.

As I wrote last year, this is not a conclusion I planned (or wanted) to reach when I began researching that article. I'd love to be able to raise his flag on merits myself, if for nothing more than to spite the media d-bags who conspired to tarnish his image. But the facts as I see them compel me to do otherwise.

If Ozzie Smith makes it (I don't think he should have), then Jim Rice makes it. I look at it this way. Would anyone trade Ozzie Smith in his prime straight up for Jim Rice in his prime ? No. It would take 2 Ozzies to get 1 Rice, at least. Maybe Jim Rice should have retired 2 years earlier but he shouldn't be punished for it.

First, this isn't an exact science, and even if Smith had been a mistake, it wouldn't make sense to compound the error by making another mistake. If Player X sneaks in and is the worst player in any Hall of Fame, you can't let in everyone who was as good as X, or you'd have to increase the size threefold or more.

Second, anyone who traded Rice for Ozzie at any point after 1979, when Rice looked like a Hall of Fame lock, would have made a brilliant trade. After that point, Ozzie picked up 298 win shares to 154 for Rice. Ozzie's best year (his prime?) was probably 1987, and if you had traded him for Rice straight up after that season, you'd be giving up 136 future win shares and acquiring 11.

Ozzie's major strength, obviously, was that he was a surreal defensive shortstop for a very long time. He started out as a poor offensive player and eventually became decent: no slugging but a better-than-average OBP and excellent baserunning skills. He didn't do all that much before he reached age 27, but then he played at a high level until he was past 40. Rice, on the other hand, had his best seasons in the rear view mirror by the time he reached 27.

Sure, I know it's not the focal point, Dog, and I understand what you are saying. He did have about a .01 GIDP/PA higher-than-average GIDP/PA rate over his entire career, but that amounted to only about 2 to 3 runs a year, which is very small. As you say, there were times when his GIDP rate wasn't all about hitting it hard, and he really did most of the GIDP damage during the mid 1980's (1982-85, 87) and the rest of the time he was pretty much an average GIDP hitter, including the peak seasons.

His walk rates weren't poor, btw, they were in the average range. Since his walk rate is factored into all the key metrics (OPS, BRC, WS, VORP, etc) we can't really add that on to the discussion anyway, other than to say that his numbers would have been higher if he walked more.

Looking back at the last list on the page, I omitted Rice's career numbers. He had 1287 batting runs created, compiled at a rate of .209 BRC/PA. I think the cutoff for hitters in isolation for HOF selection is about .200 BRC/PA, so I think both Rice and Evans have a legitimate claim to HOF selection. Rice was a better hitter than Evans and Dawson and several others mentioned, and you just can't take that away from him by cherry picking tangential stats like GIDP, or implying that the Fenway park factor aided him more than it did Evans, or Fisk, or Yaz or Boggs, or by saying that he could have been better if he walked more. He didn't have three great years and 11 bad ones - he had three great years, six very good years, and a five that were average.

I guess that I had an advantage of having been in Europe for most of the 1980's, so I didn't really see the Jim Rice in decline that you guys did - I only remember the one from 1975-1981 and he was as good a hitter as there was in the big leagues during that time, bar none. Maybe this gives me rose-colored glasses to view him through, or perhaps the converse is true. He was who he was, a strange perplexing man who titillated us with his batting prowess for half a career, then let us down in comparison over the second half of his career.

Let him in.

A couple of observations and a question...

According to Retrosheet event files, Jim Rice grounded into nineteen double plays in which a runner scored (8 x 643; 5 x 543; 5x 463). That's about 6% of his GIDPs and they cost him 19 RBI.

If you simply take away Rice's last season (1989), I think his numbers would look better to the average HoF voter. He went into that season with an even .300 batting average and his lifetime OBP, SLG, and OPS would all be sligthly higher...at the cost of three home runs and 28 RBI. I don't have much doubt that if he could have had one season at the end that allowed him to keep his AVG at the .300 mark and boost his HR total to 400, he would be in the Hall already.

As for saying he had three great years, six better than average, and five average ones, what are you basing that on? Just against his peers or in the historical hierarchy? I'm not trying to quibble, but just understand your metric.

Well, using the BRC metric, which is the cumulative Batting Runs Created (BRC - formula is on the Wiki) over the course of a season, the general cutoffs I would use would be:

120+ - MVP quality season
100+ - All-star quality season
80+ - Excellent season

From a rate perspective, I use BRC/PA as a relative value of how well a player hit independent of his PA's. I consider BRC/PA values of +.200 to be all-star quality production and this helps us to gauge partial seasons.

Here are Rice's career numbers, sorted by his highest BRC seasons:



This is the basic source for my statement about the "three great seasons plus six." I consider the seasons 1921-1992 to be more/less equal for comparison purposes. There were a handful of outlier seasons, which I have listed in earlier posts. I'll post a table of different players for comparison tomorow.

Thanks for the follow-up, but the Wiki definition does BRC does raise the question of how well players from early eras are valuated. Also, the coefficients inside the parens in this phrase, -.333*(.29*K/PA + .31*GO/PA + .28 *AO/PA), only add to 0.88. Is this to account for outs on base? Finally, do you ignore hit by pitch or lump them with walks?

BRC just represents the isolated batting events of a larger formula, so does not include runs created through running (outs on base, extra bases taken, stolen bases, caught stealing), pitching (HBP's, balks, WPs), or defense (errors, passed balls, double plays). Those calculations (where possible) are included in other variations of the run formulas for team, league and pitching.

The out values are based in large part on Tango's work on probabalistic run scoring. The problem is that with probabalistic thinking, run-scoring is calculated on one out, but in reality runs are scored in units of three outs, so the actual run value of an out is only 1/3 the probabilistic run value. It's also nearly impossible to separate GOs from FOs in historical analysis, so in historical analysis I use a single out value (-.105) in my calculations.

I'm open to suggestions about out valuation, since this aspect of the formula is still evolving. I'd also be interested in hearing more about how you think these metrics might undervalue or overvalue players from earlier eras. I find that the values pretty much work consistently across all of modern history. Hitters before 1920 had an obvious disadvantage, and hitters after 1992 had an obvious advantage, but I think that run-scoring from a hitting aspect was pretty constant from 1920-1992, except for about 10 outlier years where hitting (or pitching in 1942-1945) had brief one or two-year spikes due to rule changes or expansion).

As an interesting aside, the one-out vs three-out out value observation is in large part my basis for the argument that advanced defensive metrics seriously overvalue defensive miscues, but that argument doesn't really belong in the HoF thread

Just to clarify this one point, I should say that run-scoring through batting was consistent throughout all of history, i.e. the formulae I use for run-scoring works regardless of era. What was different before 1920 and after 1992 was the number of runs scored through contact events, or contact value (CV - formula also in the wiki). The league CV was pretty uniform from 1921-1992, but was much lower earlier, and much higher later, so from a hitting perspective, there were really only three distinct eras in baseball - Deadball Era, Classic Era (for lack of a better term), and Weight-training Era. I don't think you can compare players using raw number across those eras very well, especially the hitters from the past 15 years.

Deadball is a different story in some ways. The greatest hitters of Deadball found a way to create batting runs at levels that compare favorably to Classic Era hitters, including Cobb, Wagner, Crawford, Collins, Speaker, Jackson, etc, but there were just fewer of them. There were just fewer guys reaching this elite status than in later seasons. This has interesting connotations, really, and one might argue that Deadball was really more of a reflection of a thinner talent pool and a general hitting paradigm rather than an across-the-board disadvantage to hitters. The ball wasn't really "dead" afterall, just dirty (and punky in the later innings, obviously).

Another observation in baseball historical evolution is that teams scored about three times as many runs through shoddy defense in 1910 as they did in 1980. This obviously should mean that excellent defense was much more important in 1910 than it was in 1980, yet we elect guys like Ozzie Smith to the HoF, but ignore guys who probably had a bigger impact on run-scoring through their defensive excellence, like Dave Bancroft. And Bancroft could hit a little too, lol.

So there's lots of cool stuff in here, and lots more that can be unearthed. I haven't done as much research on pitching as I would like to yet, but I think that there is a more accurate and telling metric that can and should be used to evaluate historical pitching than the current ERA-based standards. My feeling is that pitching evolution is more related to defensive evolution than offensive evolution, so there are probably different distinct pitching eras than hitting eras, if that makes sense. The biggest single difference in defensive evolution has been glove technology, but there have also been philophical changes in baseball strategy regarding the importance of defense. The pitching resurgency in the late 1950's and early 1960's was a reflection of both of these factors, along with the insertion of black and latin players into the lineups (and eventually rotations), but like Deadball, the great hitters stayed pretty much the same, there were just fewer of them which supressed run-scoring quite a bit.

The run value of the out is around -.25 to -.30 runs, and it does not need to be divided by 3.

Tango, I'm very familiar with your work and I think it's absolutely brilliant, but I can't get it to work in a cumulative model. If I am doing things wrong (and I probably am), I would most surely appreciate you pointing it out.

In the cumulative run model, I subcategorize runs as follows:

I. Batting Runs Created - the linear weighted values of walks, singles, doubles, triples and home runs:
(.47*1B + .78*2B + 1.09*3B + 1.40*HR + .33*BB)

II. Speed Runs Created - the linear weighted value of baserunning events. Since in historical analysis, only SB's and CS's are available, I calculate those according to lwts (.193 SB - .437 CS). On a team basis, the SRC from SB/CS is very small - normally +/- 10 runs per team per season. Runs from outs on base or from extra bases stretched are entropic, but probably within the same +/- 10 run range.

III. Pitching Residual Runs Created (RRC) - the net result of things a pitcher does to hurt himself, including wild pitches, balks, and hit batsmen. Depending on whether you consider to be a positive batting trait (incorporated into BRC) or a negative pitching trait (Incorporated into RRC), you could also include walks here. Perhaps the truest sense would be to add 50% to the batter and 50% to the pitcher, but that doesn't really matter when calculating team runs. I use the linear weighted values for these events, and they are a relatively constant +15-20 runs per team per season throughout all of history.

IV. Defensive Runs Created (DRC) - the value of errors, using the linear-weighted value for errors (.478). There is going to be some seasonal variation because this is an average value of all errors, but there is obviously quite a difference between a one-base and a three-base error in terms of actual run value. This value has decreased notably thoughout history, from about 150 team runs per year in the 1900s to about 50 team runs per season today.

I normally incorporate out values into BRC, but for the purpose of this exercise, we will leave it as a separate entity, and compare it to the remainder of runs left over from the other run calculations at the team level.

Here is a listing of league run calculations using the formula I outlined above: BRC + SRC + RRC + DRC = Theoretical Runs Created (TRC):



BRC in the above chart is calculated without the out valuation. This leaves a run Delta between TRC and actual Runs Scored (RS) which I divided by the batting outs (BOuts) and came up with Runs per batting out (R/BO) to illustrate what I think the cumulative value of batting should be to approximate total runs (RS).

The .105 standard out value I use is just an estimate, but it could be higher, possibly around .110, depending on how much entropy there is in a year-to-year model. I also believe that batting out rates are probably more/less constant. The fluctuation we see in the chart above is, I believe, related both to contact value (CV) and to the GB/FB ratio. As you have pointed out in your work, Ground Outs have a higher negative run value than fly outs, so in seasons where the GO/AO ratios were low (such as the 1930's and post 1992, outvalues were probably lower than average. I am also not completely convinced that SRC calculations are quite accurate, and there may indeed be more runs created through baserunning than we currently believe. Defense also probably has some influence over consistency in this model, as does just plain old random variation. Hopefully, further study of individual and team run scoring will help further clarify the baserunning and defensive influence on run-scoring especially through your work in the present tense.

So if there are errors in my calculations or assumptions, please point them out.

So

I’m going to take up the case for Rollie Fingers one last time. I was certain that the people voting in this thread would never elect a pitcher who played part of his career as a starter, and part as a reliever. Then Dennis Eckersley was elected in his first round with more than 90% of the votes. Where were all of the arguments against relief pitchers then?

Everyone in this forum establishes their own standard for which players should be elected. I would hope that the standard would be consistent. Of all the relief pitchers considered so for, Eckersley has a career most similar to Fingers. Was Eckersley better? If Fingers is better, doesn't he deserve to be elected?

Tudor Fever points out that Fingers ERA+ trails many other pitchers. This is true, but Eckersley's career ERA+ (116) trails Fingers (119). A look at career ERA reveals that Fingers' 2.90 is significantly better than Eckersley's 3.50, despite Eck’s ranking in the top 10 for ERA in 1975 and 1977 – 1979. How is this possible? Because Fingers was more consistent. His ERA for a season was more than 3.00 five times in seventeen seasons, and one of those is 1968 where he only pitched 1.3 innings. Eckersley’s ERA was over 3.00 for 16 seasons, or two thirds of his career.

The assertion that Fingers “wasn’t that elite” is nebulous at best. It is true that James doesn’t include Fingers in his top 100 pitchers. It’s also true that with James contributing to the Red Sox in 2006, the Sox ended up third in their division. Where do others rank these two pitchers? One source, Ranking Baseball's Elite: An Analysis Derived from Player Statistics, 1893-1987, by A.W. Laird, McFarland & Company, Inc., 1990 places Fingers at #3, behind Gossage and Wilhelm. Another source, The Sporting News Selects Baseball's 100 Greatest Players, by The Sporting News, The Sporting News Publishing Co., 1998 ranks Fingers as #96, and Eckersley at #98. One online listing, thebaseballpage.com ranks Eckersley #5 among relief pitchers, and Fingers #6.

How did each rank against his peers? Eckersley was among the top 10 in saves eleven times. He won the Cy Young and MVP in 1992, was the ALCS MVP in 1988, and received the Rolaids Relief Award in 1988 and 1992 . Eck was selected as an All Star 6 times. By comparison, Fingers was among the top 10 in saves fourteen times. He won the Cy Young and MVP in 1981, was the World Series MVP in 1974, and received the Rolaids Relief Award in 1977, 1978, 1980 and 1981. Fingers was selected as an All Star 7 times. Neither pitcher had a remarkable career as a starter. Eckersley did have one 20 win season, but he didn’t pitch consistently at this level.

The bottom line is that Fingers accomplished what Eckersley did, only Fingers did it first, and he did it better. Looking objectively at his career, Fingers belongs in the SoSH Hall of Fame. All I’m asking is that you take whatever analysis you used for Eckersley, and apply it to Fingers.

My vote for 2005:

Rollie Fingers
Tom Henke
Bruce Sutter

Majordad,

1. Eckersley would not have received my vote as a reliever alone; his fairly long career as a decent starter put him over the top. (Fingers was pretty bad in his 37 starts.) Eckersley pitched almost twice as many innings as Fingers, 3285 2/3 to 1701 1/3.

2. Unadjusted ERA is relatively meaningless here; Fingers spent his entire career with teams with pitchers' home parks (A's, Padres, Brewers), as opposed to Fenway and Wrigley, and runs were a lot scarcer in Fingers's era than in Eckersley's.

3. Are you saying that the Sox' record in 2006 destroys James's credibility as an analyst? By this logic, the 2004 championship proves he's the greatest analyst ever. Obviously, the logic of both assertions is absurd.

4. Neither Fingers nor Eck deserved to win their Cy Youngs/MVPs. Not even close.

5. Save totals are more a function of usage and roles than of talent. Jose Mesa (320 saves, 102 ERA+) is Exhibit A for this.

6. I'll say one nice thing about Fingers: he threw a lot more innings in his best seasons than any closer does today (unless you count Rivera's postseason innings).

7. No vote for Wade Boggs?

Val -

Your calculations are, obviously, a form of Linear Weights and, as such, you are dealing with average values for offensive, defensive, and pitching events. When you include average out values with these events the total should equal zero, not the difference between the batting runs created and the actual runs. That is why your out value is skewed.

edit- meant to say out, not run value

I see what you are saying, and zero-basing makes sense in the probablistic model, but in the cumulative model - looking back at seasons and careers, it seems to me that we should be looking at the effect of outs on total runs scored, and that the value of positive and negative influences on run-scoring should equal the actual runs scored, not zero.

I'm not looking at predictive value with this system, I am looking at historical influences and using the lwts system to study team dynamics and player performances (and their contribution to team run scoring) in the past tense. The basic formula works on the league level, the team level and the individual level, on both run scoring, and runs allowed.

Does this make sense in that context?

You set the run value to -.25 to .30 (whatever it takes to get the overall total to zero). Technically, you should use something like this:
http://www.tangotiger.net/customlwts.html

Anyway, from the zero-based approach, you add:
(lgRuns/lgOuts * tmOuts) * (playerPA/tmPA)

So, a player who is +100 runs, and has say 200 outs and 500 PA (Bonds-like), he gets bumped up based on his high PA, not low outs.

In the above equation, the you are portioning out what an average team would have done, giving the IP of the team in question. The second part is the percentage of the PA that the player has.

HOWEVER, using .10ish runs is a fine alternative for most cases. That figure is simply the .25 to .30 run value minus the lgRuns/lgOuts (usually around .17 or so).

I don't want to hijack this much longer, but Tango is correct, Val. Your "out" value is effectively the actual out value less the average number of runs per out, not "run-scoring is calculated on one out, but in reality runs are scored in units of three outs, so the actual run value of an out is only 1/3 the probabilistic run value." as you previously stated. Your work still, for the most part, illustrates what you intend it to, "looking at historical influences and using the lwts system to study team dynamics and player performances (and their contribution to team run scoring) in the past tense." However, in your historical work you may want to use Tango's custom linear weights numbers for different run environments, as the value of offensive events can vary drastically depending on the league average runs scored.

If any measure is a function of usage and roles and not talent, it’s innings pitched. The manager controls who pitches when. The pitcher has to deal with the situations that are presented. The fact that Eckersley started more games than Fingers skews this number without any regard to the quality of the pitching. Number of inning pitched alone means that the pitcher had a steady job, and not much more.

The quality of the pitching needs to be quantified. Whether that’s done with wins and/or saves, ERA, or ERA+ depends on the individual making the evaluation. Unadjusted ERA isn’t meaningless. It’s the cold hard reality of how a pitcher performed in the circumstances presented at the time. It stands on its own merit as a measure of quality for that pitcher. The difference between Fingers and Eckersley with regard to ERA is striking, and it can’t all be attributed to venue. Even adjusted, Fingers ERA+ is better than Eckersley’s.

Eckersley started for 12 seasons, averaging just under 12.6 wins per season. His ERA was under 3.00 for only 3 seasons. It was over 4.00 for 4 seasons. You call it decent, while I would call it mediocre. It certainly isn’t a performance that would put a player “over the top” in my judgment.

You may think that these players didn’t deserve the awards they received. That’s your opinion. The fact is that they did receive them, and they are also a measure of how elite they were when they played.

I should have been more specific in my commets on James. My point is that he is not the be-all and end-all when it comes to statistical analysis. Other lists compiled by other sources rate Fingers as better than or close to Eckersley. Whether numbers or lists are used, Fingers is at least as elite as Eckersley.

I am not arguing that Eckersley should not be in the Hall of Fame. I’ve stated before that I would support him, and would have cast a vote for him if I had gotten the chance. I am saying that if you look at the players objectively, both deserve your vote.

I'm not following this thread very closely, so if this has been discussed, you may ignore.

There is a mountain of difference in reliever/starter performance for the same pitcher these days. We are talking about 1 run per 9 IP, or about .090 wins per game. That is, a pitcher who is a ".450" pitcher as a starter would be a ".540" pitcher as a reliever.

This is true for 1999-2002, and likely true for post 2002 as well, and probably true for post 1986.

However, I do not know what that figure is for the 60s-80s.

Just picking at some random players:
Bob Stanley
http://www.retrosheet.org/boxesetc/Lstanb0010.htm
Struck out 3 times as many batters as a reliever, even though he only faced twice the number of batters. ERA difference of 1.12.

Woodie Fryman:
http://www.retrosheet.org/boxesetc/Lfrymw1010.htm
Struck out 14% more batters in relief than as a starter, even though over half his relief games were as a 39+ year old. ERA difference of 1.08.

These two prove nothing of course. But, it should certainly mean that someone should expand the study I did in The Book.

For lack of a better metric, I have decided to let the All-Star voters of their times decide which relievers I vote for. Here is a fairly comprehensive list of multiple all-star relievers (feel free to add some if you can think of any I have missed):



* denotes active player

Seriously, all the stats are flawed, and for the most part, I think AS selections are a fairly true representation of the general consensus of the time (although some managers did make a few token selections, most were marginal guys with one or two AS appearances).

Using this criteria, I don't see how we can ignore Fingers and Smith and try to select pet guys like Henke and Quisenberry. Remember also that Fingers was getting selected to the AS team before there were designated "relief slots". I don't care what the numbers look like 25 years after-the-fact, Fingers deserves to be selected.

My Ballot:

Boggs
Fingers
Rice
Smith
Sutter

Wade Boggs
Andre Dawson
Dwight Evans
Fred Lynn
Jim Rice

I agree with the a lot of the points VAL and MajorDad are making about Fingers. I think Rollie had a better career than guys like Henke and Quisenberry. There are only a handful of relief pitchers I would put above Fingers: Wilhelm, Gossage, Rivera, perhaps Hoffman and Mike Marshall. As of this moment, I would only vote for the first three. Relief pitchers are simply not as important as most people think they are.

Lahoud et al - I'm leaving for New Hampshire, and won't be back until the new year. I've made my arguments. Fingers is on his final round of voting. If you are a part of the 90% that voted for Eckersley, it's time to sack up and admit that Fingers also belongs in the SOSH HOF.

For what it's worth, I would place Fingers second behind Wilhelm, ahead of Gossage. I'll rank Rivera when his career is complete.

I wanted to look at Jim Rice's career from some other perspectives so I chose his near-contemporaries, Hall of Famers Dave Winfield and Eddie Murray as comparisons.

Table 1 shows Rice's career numbers for the first thirteen years of his career (omitting his last three) along with the numbers compiled by Winfield and Murray from the starts of their careers until they had similar amounts of plate appearances.

Rice has the advantage. It's not large over Murray but Winfiled does not look like a potential HoF'er at this point.

Table 2 compares those two players to Rice when they had approximately the same number of PAs as Rice did at the end of his career. As can be seen here, Murray overtakes Rice while Winfield's stats remain flat.

Table 3 shows the respective careers with the three highest consecutive years (based on OPS+) removed from each player's career. Their OPS are identical, but what is obvious is that Rice lags in the counting stats (H,2B,HR,RBI) that HoF electors like.

Table 4 shows the stats at career-end for the three players and, again, Rice suffers from a shortfall in counting (longevity) stats.

As much as I liked Rice as a player, I don't think his three peak years are enough to overcome his lack of counting numbers. And, if I don't think that, then I'm willing to bet that there are many sportswriters who will think the same way.

Otto, if your goal is to evaluate the players, as opposed to predicting how sportswriters will actually vote, then your numbers need to be park-adjusted. Here are the career home/road OPS splits for these guys, courtesy of BR P-I:

Rice: 920/789
Murray: 834/838
Winfield: 812/841

Another point in favor of Murray and Winfield is that Rice (9,058) had far fewer career PAs than either Murray (12,817) or Winfield (12,358). Rice was toast at age 36, while the other 2 played into their 40s, with multiple good seasons after age 36.

2005
Wade Boggs
Andre Dawson
Dwight Evans
Fred Lynn
----------------

Fingers definitely is closer to consideration than Lee Smith or Tom Henke (wha?), but I agree with Lahoud that relief pitchers, individually, are just not as important as they're made out to be.

First, there's Fingers himself. He was durable and consistent, but there were plenty of years where it's not clear he was the best reliever on his own teams. He had some dominant years, and then he had some years when he was outpitched by guys like Mudcat Grant, Jim Roland, Paul Lindblad and John D'Acquisto.

But maybe some people are willing to overlook that because they're bothered by a "shortage" of relief pitchers in the SoSH Hall of Semi-Fame. They shouldn't be, IMO, because that's not the same thing as having a shortage of catchers or third basemen. "Relief pitcher" is not a position. "Pitcher" is a position, and we have no shortage of those.

There's nothing wrong with being a very good pitcher who is not in the Hall of Fame. Rollie was nowhere near as valuable to his teams as Jack Morris or Dennis Martinez, who I also wouldn't vote for.

Following up on Otto's Rice/Murray/Winfield comparison, I thought I'd add one more chart, just for kicks. Among his comparisons, Otto put JR's first 12 years next to Winfield and Murray's. One "advantage" Rice has in that comparison is that his first 12 years (not including the '74 cup of coffee) were his 12 best years.

So I did a comp of Rice's 12 best seasons ('75-'86) vs. Winfield and Murray's 12-best non-sequential seasons (the years were chosen by straight runs created):




Winfield is closer in this comparison. It's easy to see how borderline his and Murray's candidacies would be if they, like Rice, had just the 12 good years. I don't know what that has to do with Rice's candidacy, but it was a nice excuse to try out that chart function.

Sorry for the redundancy, but any comparison of Rice to Murray and Winfield that ignores park effects has no credibility.

I agree with this, but there are problems with standard park effect adjustments so it's unsure to me how much we should penalize or credit players for the parks they played in, especially in assymetrical parks like Fenway. On a whole, Fenway in this period increased doubles and HRs fairly significantly, while surpressing triples and singles. Almost all of this effect was created by the presence of the Green Monster, of course. The main problem with individual players is that the parks affected them differently and this is not currently reflected in standard metrics such as Plus ratings.

Evans' home/road numbers for instance display this effect almost perfectly, but Rice's do not. Rice hit more singles in Fenway than on the road and this was almost certainly a reflection of wall-ball singles. What exactly those singles might have been had he played in a different home park will never be known, so it's very difficult to say. We can look at their home road numbers as an indicator, but again this is problematic because hitters adapt to their home parks, change their swings, etc, and those numbers get skewed too.

Frank Robinson is a good example. Had old Robby not been traded to the Orioles for Milt Pappas and played most of his effective career at Crosley Field, people would have questioned his numbers too. As it turns out, he hit just as well at Memorial Stadium as he did in Cincy.

So we don't really know how Rice would have turned out had he played his entire career somewhere else, and I somehow don't believe that he would have been a mediocre player. I think to some extent the Monster ruined Rice and that he would have put up Winfield-type numbers at least (I think they were very comparable in skill).

This is all very speculative, of course, and we don't have the tools to analyze this much further than we already have.

I do think there is some serious anti-Rice bias in this forum, though. I think we are perhaps too close to Rice as a player, and our general love for him has turned into a general loathing perhaps based on our impressions of what he should have been and how he seemed to decline in the 1980's. What especially irks me is how we seem to be using the park effect argument against him, but not against Evans and Lynn (or even Fisk, Boggs or Yaz). If Rice's claim to greatness is curtailed by his Fenway numbers, than Evans' must be curtailed at least equally and his claim to greatness (as opposed to accumulated numbers) is weaker to begin with.

So I can see the argument against Rice, and I go back and forth on it myself, but if you are not going to elect Rice, than for god's sake take Evans and Lynn off your ballots.

Just to put that GIDP debate to bed, Rice hit into double plays over the course of his career in at a rate of .035 GIDP per plate appearance. The league average GIDP rate is.025 GIDP/PA. This means that an average ballplayer will cost his team 3.58 runs over the course of a 650 PA season by hitting into double plays, while Rice cost his team 5.01 runs per 650 PA, a whopping total of 1.43 runs per year.

I hear ya, VAL; if we count park effects against Rice (which we should), then we need to count them against the other Red Sox as well. However, for me Evans makes the cut and Rice doesn't for two reasons: defense and career length. Check out URI's aptly numbered Post 666 from earlier in this thread.

Meaningless. Look at it in the context of PA's with runners on base, as I did earlier in the thread, and in my MSP article. The gap between Rice's GIDP rate and that of those of his peers is far greater. Who cares what his GIDP rate was in PAs where the bases were empty?

I was primarily thinking about how the writers vote. While the level of sophistication in stats is increasing among sportswriters, they still are swayed by the "magic" numbers--500 HR, 3000 hits, 1500 RBI or RS and rate stats--.300 BA, .500 SLG to name a few. I think longevity is what propelled Winfield and Murray into the Hall. They certainly did not outshine Rice at the 13-year mark (equal PA). There are other players who are in a similar state as Rice--Dave Parker, Dale Murphy, Andre Dawson come to mind. The peak is not enough (especially when it was years ago) and tehy don't have the magic numbers (or enough of them).

Wow .. I voted way back in the first few years of the SOSH HOF back on the old board .. finally I'm back and am ready to cast my ballot for 2005..

Wade Boggs
Rollie Fingers
Andre Dawson
Bruce Sutter

Without meaning to be snarky (and giving you time to correct it), how often did Rice hit into a double play without men on base?

I hear what you are saying, and he did ground into a lot of double plays, but as you point out yourself, not significantly more than a lot of guys with his same talent level, including Parker, Winfield and Puckett. All I was saying was that in terms of runs, the GIDPs were insignificant in terms of his career run production.

What we know is (1) Rice played in an extremely advantageous place for offense, and (2) he displayed an extreme park advantage himself, hitting just .277/.330/.459 on the road in his career. There are very few good hitters in baseball history with such a dramatic home/road split.

Personally, I don't care that much about (2), since this could in fact be a skill. Perhaps in another park he would have adjusted and hit well in that park, too? I don't know, and I don't care.

All we need to concern ourselves with is (1). It took an awful lot of runs to win a game at Fenway Park in Rice's prime, so the statistics need to be normalized. We know how to do this, everyone does it the same way, there is no need to reinvent the wheel here.


I think there is a weird pro-Rice bias in this forum. I have been voting for Rice, ironically, but I have often find myself disagreeing so much with the pro-Rice arguments that it is getting harder and harder to justify my vote. We have otherwise smart people who don't want to adjust for Fenway Park. We have smart people who apparently believe that offensive rate stats are all that matter, when players like Fred Lynn and Dwight Evans did everything on a baseball field better than Rice except hit, and they were both excellent hitters. We have smart people making comparisons based only on the first 12 years of career. All of these arguments are ones you would pick out not to shed light on a situation, but to rig the argument for Jim Rice.

If you compare Rice and Evans purely as hitters, and you don't rig the data, it is too close to call on the rate stats. When you add in the career length, and the gigantic defensive differences, it starts to get silly.

2005 Ballot:
Wade Boggs
Dwight Evans
Jimmy Key
Dave Parker
Dave Stieb

Boggs
Rice
Dawson
and a trio of relievers:
Fingers
Quiz
Sutter

Quiz
John
Evans (Dwight)
Boggs
Darryl - I felt so bad for that guy. My aunt was a big Mets fan from the 60s onward and Doc and Darryl's sad careers more or less killed her as a baseball fan. I don't think she even watched any of their playoff games in the late 90s. She's just starting to make a few small noises about going to Shea with me when I'm in Connecticut again but it's been 15 years.

Looks like I made it under the wire....

2005 ballot
Wade Boggs
Don Mattingly
Dave Stieb

Boggs is in, Evens is one vote shy for the 3rd straight ballot, and we say good by to Rollie Fingers...

2006 Ballot

Will Clark
Andre Dawson
Dwight Evans
Fred Lynn
Dan Quisenberry

2006 ballot
Albert Belle
Will Clark
Andre Dawson
Dwight Evans
Don Mattingly
Graig Nettles
Willie Randolph

I didn't vote for either, but there seems to be zero justification for voting for Sutter and not Quisenberry. They finished their careers within 1 IP of each other, and Quisenberry had a big edge over Sutter in both ERA+ (146 to 136) and DERA (3.53 to 3.31).

Belle had 5 seasons (1993-96 and 1998) in which his average WARP3 was 11.84. Rice broke 10 once, with 10.4 in 1978.

I think that as a group we have set the bar for relievers very high (if there is a bar at all) and am withdrawing all of my previous support for relievers based on my All-Star appearance criteria from last week. In other words, I won't vote for anyone with fewer AS appearances than Fingers, and that includes Sutter and even moreso Quisenberry (whose ERA+ numbers seem way out of proportion to his general value).

I'm also not voting for anyone whose primary career peak seasons occurred after 1992 (Belle is the first to come up in our polls) because there was about a 20% spike in contact value on a league level that occurred in 1993 and persisted to the present. I want to study that spike further, and I don't know if WARP3 adequately compensates for that spike.

Reading Lahoud's excellent synopsis of the whole Rice-Lynn-Evans thing, I think that he is absolutely right. I think what I was reading as an anti-Rice bias was probably a pro-Evan bias that still confounds me. I'm now leaning heavily toward none of them meeting HoF criteria.

My 2006 ballot is, therefore, empty at this point. I may add a couple of names later.

Wouldn't it be something if Albert Belle got into the HoF and Mac didn't? I mean, that's just crazy enough for baseball.

If look back in this thread to post #666, I did a quick and dirty look at Evans/Lynn/Rice/Singleton using WARP3.

Now, I'm doing the same thing with Evans, Rice, Dave Parker, Andre Dawson, and Willie Mays. I added Mays just to put a 'dominant' outlier in there for perspective. I am not saying that because these guys aren't Willie Mays, they aren't HOFers. More like, "Here is the Gold Standard...where does your borderline fall."

I have Trending by Age, Cumulative Career Totals, Best 5 seasons, Best 3 overall seasons, and finally something new...a list of all the combined seasons above 5 WARP3 (a good demarcation line I just made up between a great and good season), complete with frequency of 8 and 5 WARP3 seasons. I defined a season as any year with 100+ games played. Don't like it, make your own charts.

The charts are in *.jpg format, so if you can't see them due to scrolling issues, or have images turned off, right click and save, or bugger off.

Here is some further discussion on run-scoring over history, with particular emphasis on how the hitting game changed in the 1990's and why I think current metrics do not adequately reflect the changes in the game.

I did a study on Contact Value a few weeks ago and found some interesting things out. CV is the linear-weighted run value of all contact events (balls in play - hits and batting outs) expressed in runs per plate appearance. Of particular interest to this discussion, I found that league average contact values didn't really change from 1920-1992. Here are the average values for different eras:



Not only did the average values from 1920-1992, not change, but seasonal variation was much smaller than expected, with excursion from the average in all but about a dozen outlier seasons with +/- 5%. In other words, there appeared to be a fairly constant mean league CV in baseball for 73 years.

The outlier seasons are interesting and were caused mainly by changes to the overall player pool rather than any rules changes, except for the live ball season of 1930 and for the addition of the DH rule in the A.L. that added about 10% to the A.L. mean CV from 1973-1992. The DH rule didn't really add to the player CV, however, it just added an extra hitter to the lineup. During the WW2 seasons of 1942-1945, league CV dropped about 15% and in seasons following league expansion, jumped about 15%, but in all cases, regressed to the mean with the next couple of seasons as the player pool normalized. On interesting spike was a 10% increase in N.L. CV during the mid and late 1950's. This 10% was directly linked to the addition of black and latin hitters such as Mays, Aaron, Clemente and Robinson to the N.L. player pool. This creates a unique challenge to seasonal adjustments, since you can attribute a 10% change in run production to a half-dozen or so individual players.

All of this changed in 1993 and thereafter, when CV's in both leagues jumped about 20% and never regressed to the mean. In other words, the league suddenly began to score 20-25% more runs per season through contact than they had in the 72 previous years. This is essentially the same leap in run-scoring that occurred between 1918-1921 (see Note 1), which should put it in context.

There are several factors that probably contributed to this jump, including expansion in 1993 (including the addition of Coors Field, which represented about a 4% increase in N.L. CV all by itself), smaller ballparks in general, and a greater emphasis on power pitching, but the single most significant reason for this spike in contact offense was almost certainly weight training. By weight-training I mean exactly that - the widespread shift in training for all players to increase muscle. Although PEDs were certainly a part of this, it was by no means the only part - just a subset of the whole weight training regimen.

At any rate, weight training seems to have changed the game forever and it wasn't just a handful of players that changed, it was nearly the entire set of major league players hitting the ball 15-20% harder than they ever had before.

This presents a unique challenge to normalizing statistics from players before and after 1993. CV only reflects about 65% of total runs scored (the other runs coming from non-contact batting events (hits and walks), defense and baserunning), so a 20% increase in CV doesn't necessarily translate into a 20% increase in run scoring across the board, so league average run normalization isn't going to be terribly accurate. OPS+ and other adjusted metrics are probably out-of-whack, since nearly all of the run increase in this era was related to SLG, not OBP events. More work needs to be done on this era before accurate comparisons can be made.

One way you might be able to get a quick snapshot of player changes, would be to multiply player's SLG by .8 for seasons after 1992. Does this work for Albert Belle? Well, his SLG from 1989-1992 was .483. His SLG from 1993-2000 was .587, so the difference the after/before in his SLG was about 82%. Now Albert also gained some OBP skills after 1993 for which he should not be penalized and even after reducing his SLG by 18% he still put up some very good .800 and .900 OPS seasons, but those four 1.000 OPS seasons were fairly bogus in comparitive terms. So Belle must surely suffer from this era shift, as must many others whose numbers are, in relative terms, too good to be true.



Note 1. Although the averages for the Deadball Era in my table above are given in the high 70's, there were actually three distinct periods within the Deadball Era - and league average CV's during the second decade of the Deadball Era were actually about 10% higher than the overall Era average.

VAL, forgive me if I'd oversimplifying here, but if I understand it, your basic point is that all offensive stats need to be adjusted post 1992. This is obvious, but if a player excels relative to his peers, as Belle did, he should get as much credit for that as would be the case in any era. In the 5 years in question (1993-96 and 1998), Belle finished 8th, 2nd, 2nd, 7th, and 1st in the league in OPS+.

You also make the point that Belle improved a lot after 1992. True, but the most plausible explanation for this is that he was simply entering his peak years at age 26.

One can legitimately argue that Belle's Hall-worthiness was hurt by the short length of his career, but his peak was better than that of at least half of the Hall of Famers.

2006 Ballot:

Albert Belle
Andre Dawson
Lee Smith
Bruce Sutter

You can certainly compare Belle to McGwire, Bonds, Griffey, Thome, Ramirez, etc, but not to Jim Rice, Dave Winfield or Lou Gehrig any more than you can compare Tris Speaker to Mickey Mantle.

Something happened to the game in 1993 that changed hitting as much as it changed at the end of the Deadball Era. I don't think that conventional seasonal corrections capture it, because the effect was all in slugging - players began hitting the ball harder in 1993, pure and simple. The fact that the change was almost all in slugging means that conventional correctional metrics such as OPS+ are probably skewed as well, so if you are doing peer comparisons with post -92 performances, you might be better off using raw numbers, perhaps park-corrected only, and perhaps using the .82*SLG correction I applied to Belle.

Something also happened to Albert Belle in the mid 1990's besides the natural maturation of a 26-year old. Maybe he just really started working out hard in the weight room, maybe perhaps that regimen included juicing, or maybe the league entered a livelier ball into the game - I have no way of knowing. All I know is that the league changed and so did Albert. His peak was, frankly, an illusion.

You can compare how they did relative to their peers. You can compare how many runs they created, as adjusted for how many runs were scored in the league that particular year.

By the way, nice try grouping Rice with Winfield and Gehrig. I'm wise to your subliminal games!

Sorry, this makes no sense. If Belle was one of the very top hitters in the game in the mid-1990s, why does that differ from being one of the very top hitters in 1915, 1035, 1955, or 1975? It all comes down to how many runs a player produces, relative to how many runs are needed, on average, to win a game in that particular environment.

And your assertion of how the game changed is irrelevant to Belle, because his OBP in 1992 was .320, and for the next 4 years it was .370, .438, .410, and .410.

Juicing is a different issue, and one we'll discuss for sure when McGwire comes up. But saying his peak was an illusion is absurd. He created 12.30 runs per 27 outs in 1994, and 10.05 in 1995. Were these illusory runs that led to illusory wins for the Tribe?