david@fisher.UUCP (David Rubin) (08/06/85)
PREFACE With this article I begin my case for Gary Carter being a greater asset than Tony Pena. Due to the extreme length (and heavy reading) of my essay, I've broken it down into three parts. Part I, this article you are now reading, begins with a bibliography and a point by point answer to Paul Benjamin's last article. As it has been some time since that article appeared, I have reproduced it fully to refresh the memories of those of you who can't remember its details, and have outlined the nature of the response that will appear in Parts II (Offense) and III (Defense) to his particular points. Part II is already written, and will be posted one day after this introduction. Part III is nearly complete, and should be posted by the end of the week. SOURCES The Baseball Encyclopedia (sixth edition). The tome. It is limited by its "official" character. Alternate statistics (such as OBA) must be calculated on your own, and some information (such as CS) is not listed. But any book that contains 2500+ pages of baseball statistics definitely has something going for it... The Hidden Game of Baseball, by Thorn and Palmer. The most cogent explanation of the philosophy of the "new" baseball statisticians and the discoveries it has led to. At times, the authors appear to undervalue simplicity. The Baseball Abstract, by James. I have the 1984 and 1985 editions (it is an annual yearbook). Not as sophisticated as Thorn and Palmer, much more intuitive, and always common sensical. POINT BY POINT [>> = my last article, > = Paul Benjamin's last article] >> The most important measured defensive contribution a catcher makes >> is the slowing of the opposition's running game. Two teams have >> consistently been the toughest to run on the past few years. They >> featured veteran pitching staffs of high quality and rifle-armed >> catchers behind the plate. They were, of course, Montreal and >> Pittsburgh. No objective difference between the two club's ability >> to shut down an opposition running game are discernible (Montreal was >> a wee bit tougher); thus, no difference between the two catcher's >> abilities in this respect are demonstrable. >Some statistics, please, for these assertions? >Even if the stats are true, the fact that Montreal has a >much faster team than Pgh (check the stats for last year and >this one) means that Pgh faces a tougher average opponent >than Montreal, as far as speed is concerned). Thus, if there >is "no objective difference" in how they stop thefts, then >Pgh is clearly better at stopping the running game. (If you >want me to add the thefts for the league without Pgh vs. the >league vs. Montreal, figure the percentage difference, and >post it, just yell.) I presumed the statistics you are requesting are the ones on the ability of Montreal and Pittsburgh to shut down running games; what I could find are in Part III, and, though skimpy, are at least consistent with my claims. As for the difference between Pittsburgh's and Montreal's running games in 1983-4, this can be summed up in two words: Tim Raines. How many bases could have Raines "inflated" the SB-against total of Pena in 18 games a year? Something on the order of about 7?? As for your offer, I am not clear on the relevance of the suggested statistic. If, however, you have the SB%-against for Montreal and Pittsburgh, excluding all Montreal-Pittsburgh games, by all means post the two figures. >> Somewhat less critical is the number of bad pitches stopped by the >> catcher. I have watched both Pena and Carter with equal frequency (up >> until this year). I can see no difference. I doubt anyone really >> can. >I hate to point this out again, but the Gold Glove voters >obviously can see a difference. They probably know more than >you or I do about the relative strengths of their arms, and >their ability to stop bad pitches. What I have read is that >Pena's arm is much better. It looks that way on the tube, >too. Ahem. If the sum of your case for Pena is the Gold Glove was given to him, you have not refuted my basic contention: if Pena's defensive play is to make up for Carter's obvious offensive superiority, then it would certainly reflect itself in some quantitative way. For all I know, the Gold Glove award may be arbitrarily awarded if the two contenders are approximately equal; the Gold Glove itself is not sufficient evidence of a significant difference. Excuse me if I do not accept the verdict of the electors as divinely inspired... >By the way, where is evidence for your statement that this >is less critical than throwing out runners? I might agree on >this, but without evidence, it is just unsupported opinion. It is an informed opinion. The statistics that might prove or disprove it are not kept. I am hypothesizing that the borderline stolen base, where an effective catcher will produce an out and the mediocre will not, is a far more common occurence than the borderline squibbler in front of the plate. >> If by higher percentage, you mean a higher batting average, some years >> yes, some years no. If you mean a higher on base percentage, you are >> dead wrong: Tony Pena doesn't know the meaning of the words "base on >> balls" (odd, for a catcher, isn't it?). Finally, Carter's power and >> Pena's power are of different orders of magnitude. >Pena's batting averages for the four years he has played are >.300, .296, 25 points higher than Carter. Do you have fig- >ures to support your statement that Carter more than makes >up the difference in on-base pct? You are correct about Pena's batting average. I am correct about on-base percentages -- the difference in walks drawn is far greater than the difference in hits. >Pena hit 15 HRs each of the last two years. Carter has aver- >aged about 23 HRs a year, including seasons of 16, 17 and >19. You call the difference between 15 and 23 an order of >magnitude? Quite a feat of ledgerdomain, comparing Pena's BEST year with Carter's AVERAGE year. Moreover, your citation of Carter's bad years ought to have carried an explanation: the year he hit 17 (1983), Carter was severely restricted by a knee injury all year, and was playing in pain; the year he hit 16 (1981) was the strike year, and had the season gone the full course (say 550 AB), the 16 would have extrapolated to 24-25 home runs; there is no year that Carter hit 19, but you may mean 1978, when he hit 20 in his second year as Montreal's starting catcher -- perhaps this was the proverbial sophomore jinx. Finally, it is well to note that the Carter off years are still plateaus to which Pena has yet to do more than aspire. More in Part II. >...........Perhaps you should study a bit more math :-) By >the way, Carter hit 31 HR's in '77, promising a great power >career, but never hit 30 again! He did hit 29 twice, though (1980,1982). Your implied standard of power (there are two kinds of power: greater than 30 HR's, and less than 30 HR's) is quite rough. >.............................If you want a power-hitting >catcher, try Parrish - he averages about 30 a year, and is >improving. (That means that Parrish outpowers Carter by >about the same margin that Carter outpowers Pena.) Untrue. Parrish has hit HR's in 4.8% of his AB's, Carter 4.3%, and Pena 2.4%. Parrish plays in a park that favors hitters substantially, Pena plays in a park that favors hitters moderately, and Carter played in a park that favors pitchers substantially. As suspect as cross-league comparisons are, and despite the fact that Parrish is irrelevant to the issue at hand, to say that Parrish outpowers Carter by the same degree Carter outpowers Pena is refuted by the frequency that the three connect for dingers. Parrish, by the way, has NOT averaged about 30 homers a year, and the issue here is NOT Carter vs. Parrish. >> Remember also that >> Pena's stats were generated in a park moderately favoring hitters, while >> Carter's were generated in a park second only to the Astrodome (before >> dimension changes of this year) in depressing hitting statistics. >I see. Park matters more than surrounding lineup. Where is >your evidence for this? Evidence will be summarized in Part II. There is irresistable evidence that park effects insinuate themselves in all statistics, while surrounding lineup only affects such statistics as runs scored and rbi's that measure the actions of the players teammates. >.............................I disagree completely. I think it >varies on a situational basis. In the Pirate lineup, the >only two hitters that represent threats (Jason Thompson and >Pena) are constantly pitched around. Hence, try to put them >together in the lineup. But, as you point out, Pena is not a >good #3 hitter - he doesn't make enough contact to bat third >(neither does Carter), so this was abandoned. First, the hypothesis of lineup effects on batting average, slugging average, on-base percentage, etc. has not been supported by the accumulated evidence (park effects have been so demonstrated). Second, it is not critical the the #3 hitter be a contact hitter. It is critical that the #3 combine spare consumption of outs with the ability to move runners around; in short, the number three hitter should have both a high OBA and a high SA. Most number three hitters do strike out frequently (power hitters must pay the price for their big swings); most of them also walk frequently. >.................................................However, Pena >is still hitting as well as he did at this point two years >ago, when he finished at .300 and 15 HRs. Like many latin >players, he heats up when the weather does. Irrelevant. I have only considered records for entire seasons in what follows. >At any rate, Pgh's park is hardly a great hitter's park. ERA >championships are not won by teams in hitter's parks. Whether a park is a hitter's or pitcher's park for a given season depends on many non-constants, and may actually change from season to season, depending on things such as the percentage of day/night games (Shea, for example, is a hitter's park by day and a pitcher's park at night), weather (days with wind blowing in, for example, or even average temperature that season), dimensional changes (e.g. moving the fences around or filling up foul areas with seats), etc. It is complex enough that one cannot be certain that a pitcher's park will be again a pitcher's park, etc., but one can, AFTER the season is over, get a pretty good idea about whether that field favored offense or defense, and to what extent, the past season. Indeed, the figures for parks which are not tinkered with are so consistent from season to season that it would be incredible to suggest that the observed effect was random. But see Part II for an explanation of how park effects may be calculated. >> He does have better speed. Excellent for a catcher, unremarkable for >> anyone else. Again, this does not win many ballgames. >Wrong again! As just one example, do you remember the 1982 >St. Louis team? They were a speed team, crafted for their >park, but they won anywhere. Speed is a crucial factor in >baseball. By this, I do not mean just stolen bases, but >speed, e.g., going first to third on a single, causing an >infielder to rush a throw, avoiding a double play, etc. But all the things you mention are reflected in offensive statistics (the extra single, the stolen base) with the exception of going from first to third on a single, which is not recorded as a statistic and remains of marginal importance. St. Louis's speed may have won them a pennant in 1982, but it was the speed in getting from home to first that was the most valuable, and one should remember that the infield single that speed makes possible is slightly less valuable then the typical line drive single to the outfield. >.................................................................The >opinions of many baseball people that I have read recently >seem to agree that speed is becoming more and more impor- >tant, especially with the advent of artificial turf. (Also >check out this year's St. Louis team.) Again, the speed that is really important, the speed that creates baserunners, is already accounted for in the offensive statistics I have considered; I see no reason to weight it doubly. >> He is not a .300 hitter, and he doesn't hit 15 or so homers a year. >> If my memory serves me well, he's more a .280 hitter who hits 12 or so >> homers a year. A thin .280, too -- few walks. If you wish to contest >> me on this point, I will actually look it up. >You'd better look it up. Perhaps your disks need rebooting! You are right about Pena's average, I am right about its thinness. >To put Pena's offensive ability in focus, perhaps we might >contrast it in two ways: >1) with Dave Parker last year (his stats are on the same page): > AVG R HR RBI SB > Parker .285 73 16 94 11 > Pena .286 77 15 78 12 >despite Parker's getting 10% more at bats, and hitting in a stronger >lineup. (Everybody who didn't play for Pgh. hit in a stronger lineup!) >If Dave Parker had been a Gold Glove catcher last year, you'd have >heard a LOT about it! But Pena plays in Pittsville for a lousy team, >so you hear relatively little. (1) Parker also walks umpteen times more than Pena, (2) Parker creams Pena in OBA, (3) Parker had a sub par year, (4) You consider the wrong statistics. R and RBI depend on what teammates accomplish rather than the individual, and SB's are of relatively little importance. >2) with Carter last year (Carter's best all-around offensive year) > Carter .294 75 27 106 2 > Pena .286 77 15 78 12 >Carter does have the HR's, but in total runs (R + RBI - HR) we see: > Carter 154 > Pena 140 >This reflects the contribution to the team's scoring, which >is the ONLY important thing for winners. Indeed. Now if you'd only measure what Carter and Pena did instead of what their teammates did...By the way, do you have a good reason to subtract HR's? As flawed as such a statistic is, you make it quite worse by awarding two "total runs" on every run scored (to the scorer and the hitter) EXCEPT on a home run... >............................................If you look at >Montreal's scoring versus Pgh last year, you see (once >again) that Montreal outscored Pgh by much more than 10%, so >that Pena contributed a great deal to Pgh's offense. It >requires great obstinacy to say that if Pena had batted in a >good lineup, that he wouldn't have gotten 14 more total >runs. Thus I reiterate, "If Pena played on a good team, he >would be a star." We need not move Pena to a different team. The evidence is already there... if you could only free yourself from staring at R and RBI. I thought Pena was already a star. That is not what I am contesting. The issue is Pena vs. Carter, not Pena vs. the Press. >> The Pirate lineup hasn't been as bad as you make it out to be. >> Pittsburgh lost it's oomph in the last few years. >But Pena has played only the last few years. And the Pirates >have more than lost oomph, the offense has disintegrated. >Only the Giants are worst in scoring runs, and by a decimal >point. The good offenses in the NL are outscoring the >Pirates by well over a run a game. You seem to underplay the >importance that the surrounding lineup means to a player's >stats. Again, only if you look at the wrong stats. >> Before this season (and last), Pena had >> played on generally good offensive teams. Certainly better than the >> Expos of the first half dozen years of Carter's career. >So, you admit the importance of the surrounding lineup! Only with regard to R and RBI (the wrong stats). >Again, Pena has played fulltime for only the last 3 years, >and parttime for one year before that. Pena has been Pittsburgh's number one catcher for four years. >.........................................Also, you seem to be >blithely ignoring the fact that Carter was part of those >Expo teams that were supposed to win it all for a few years >in a row, but failed every time. Those were very talented >teams. So? Actually, you made me look it up; I can't stand the continued whining (albeit irrelevant) of poor Pittsburgh. The Pittsburgh teams of 1981-1984 were quite competitive with Montreal teams. In 1981, Montreal beat out Pittsburgh by 11 games (this was Montreal's title year); in 1982, by 2 games; in 1983, Pittsburgh did 2 games better; and in 1984, Montreal by 3. As irrelevant as the team dominance issue you try to scare up is, it does not even exist: since Pena became the starting Pirate catcher, Pittsburgh and Montreal have performed about as well. >> Game winning rbi's is the silliest statistic yet invented. It's a >> random crap shoot among people batting in the middle of good line ups. >Where are your numbers to support such a statement? When a >player makes a relatively heavy contribution to a team's >scoring, and drives in so many game-winners, it CAN'T be >ignored, (unless you really are trying hard :-) ). (1) Because game winning rbi's don't necessarily win games. The idea of picking out ONE run and declaring it THE critical one (unless the score is 1-0) is lunacy; moreover, as it's defined, most game winning rbi's are not awarded for "clutch" pressure performance, but because the batter in question got an rbi in the third inning... I've just seen too many GWRBI's go to guys who help kill off a first inning rally by hitting a "sacrifice" fly that scores the first run of the game to have much respect for its formulation. (2) Because game winning rbi's are not available to players on teams that don't win games. If your team wins 30 fewer games than some other team, it is unlikely that you can hope to compete for the dilute honor of leading the league in GWRBI's. (3) Because game winning rbi's share all the grievious faults of garden-variety rbi's. David Rubin {allegra|astrovax|princeton}!fisher!david