The Best Player-Seasons in Pirate History: A Statistical Evaluation

Who are the best players in the history of the Pirates? Everyone responding to that question will, of course, approach it differently. Some may rely on their personal recollections of thrilling plays and great moments, others might look at old newspapers and read about World Series exploits, and still others will review the record books and encyclopedias.

In this article, I’ll try to do the job “scientifically,” removing as many biases and preconceptions as possible, and specifying the methods of analysis and the assumptions used in the evaluation. I’ll be using “state of the art” sabermetic tools, some of which have been developed, discussed, and honed on the Internet but haven’t been seen much in print.

Before starting, let’s consider for a moment what “best” really means. Should we look at a player’s career accomplishments with the Bucs or his value at his peak? Neither viewpoint is superior to the other, and both can offer interesting results. I’ve decided to present peak performances in this article — the best single seasons generated by players for the Pirates. Similar methodology will be applied to Pirate careers in a research presentation at the Pittsburgh SABR convention.

Ground rules

Pete Palmer kindly provide season-by-season data for the 1420 men who played National League baseball in Pittsburgh in the 108 years between 1887 and 1994. I’m looking solely at their records with the Pirates, a total of 3712 season records.

The primary tool employed in this analysis is Clay Davenport’s Equivalent Runs (EqR) approach. A hitter’s EqR is the estimated number of runs that he added to the league’s total for the year. Sabermetricians are interested in estimating runs because that is the “currency” of baseball — games are won by scoring more runs than the opponents do. The team’s hitters try to maximize runs scored, and simultaneously the pitchers try to minimize runs allowed.

The methodology adjusts EqR for year effects and park effects. Using the currency metaphor, a year effect is like monetary inflation (in high-offense years, each additional run is less “valuable” than in normal years.) Year effects include such broad concepts as the “deadball” and “lively ball” eras, the “pitchers’ year” of 1968 and last year’s offensive explosion. Park effects are like the disparity in cost-of-living between, say, New York City and a farm community. With Colorado’s entry into the big leagues, many people have noticed its extreme pro-offense park effect. The negative effect of, for example, Forbes Field on homers is much less noticed. Park effects vary from year to year, often due to moved fences — in the Greenberg Gardens/Kiner’s Korner years, Forbes Field was nearly neutral for home runs.

Adjusting for those effects places the estimate of a player’s runs (now called EPER) in a common context, a hypothetical neutral park in the 1992 American League. The specific league-season isn’t what’s important, though; the significance is that the adjusted stats are directly comparable — there’s no need to worry about whether it’s 1930 or 1968, Baker Bowl or Dodger Stadium. It’s analogous to the familiar procedure of adjusting for inflation by expressing all prices in, say, “1990 dollars.”

A similar approach is used to place pitching statistics in a common context. Because the structure of pitching staffs has changed so radically over the last century, individual innings pitched were adjusted to approximate the 1992 AL’s distribution, with the adjustment based on the league’s #2 and #4 pitchers in IP. For example. Pud Galvin’s 440.2 IP in 1887 were adjusted to 262.1 IP in this analysis. The measure used for ranking pitchers was “runs saved,” defined as the difference between the total runs allowed by the pitcher (using his adjusted innings and adjusted ERA) and the number of runs allowed by an “average” pitcher — one whose ERA was 4.00 — in the same number of innings. Thus, both quality (ERA) and quantity (innings) go into the measure.

For pitchers, individual batting and fielding were not used in the rankings. For the rest of the players, however, it was important to measure defense, even knowing that fielding measures are far less powerful than those for hitting and pitching.

The best way to assess fielding is to determine how many balls hit in a player’s direction he actually gets to. Only since Project Scoresheet, and its stepchildren STATS, Inc. and The Baseball Workshop, began to collect such data in 1988 has it been possible to compute such a measure. STATS produces its Zone Rating, and Sherri Nichols of SABR calculates Defensive Average (DA).

The two are similar in their focus. They’re vastly more useful than any previous fielding assessments, all of which are flawed because they consider only chances accepted, so a ball that might have been handled by a better fielder is called a hit, and lot charged in any way against the immobile fielder. Perhaps the least problematical of previous rankings is Palmer’s linear weights-based Fielding Runs, which is used here for the years 1887-1987. For 1988-1994, Dale Stephenson’s Fielding Runs statistic (based on the Nichols DA data) is the measure of defense. It must be noted that catcher defense remains all but completely inscrutable. Palmer uses team ERA as one of the components of his rating, and DA simply ignores catchers.

Players were assessed only at their primary positions for the season, and only sea sons where the player spent at least half of his defensive games at the primary position were considered (this condition affected the composition of the roster at only one position). While the complete batting record was used, fielding only at the primary position was included in the overall rating of a player. The final ranking for non-pitchers was based on the number of runs the player generated, compared to what an “average” player would have produced while consuming the same number of outs. Two players made the roster at each position.

To demonstrate the process, consider Lloyd Waner’s rookie year of 1927. Little Poison set the all-time record for hits by a rookie with 223 and compiled a .355/.396/.410 (BA/OBP/SP) mark. He added 102 EqR to the NL that year; adjusting for Forbes Field (a hitters’ park but not a home run park) and the high offense of the season decreases him to an EPER of 94, or 97 in a 162-game season. Lloyd’s glove was poor that year, -11 FR. Adjusted to a neutral 162-game season, that’s -10 FR, so his total for the season is 87 runs produced . An average 1927 player making 406 outs would have accumulated 75 EqR, which adjusts to 72 EPER, plus zero Fielding Runs.

Therefore, after all adjustments Lloyd Warner produced only about 15 (87-72) runs more than an average player. Which is about the same as say, Billy Sunday in 1888 or Mike Easler in 1982.

The players

Leading off for the Pirates’ seasonal all-time team is center fielder Max Carey, 1917. Though he hit just .296/.369/.378 for the cellar-dwelling Bucs, it came in one of the deadest of the deadball years. Carey’s superb fielding also contributed mightily. Just a bit behind Carey is Andy Van Slyke, 1988; though Slick hit a bit better in 1992, his once-fine fielding had deteriorated significantly by then. Ginger Beaumont, 1902, is next in line among Pirate CFs.

Our number two hitter is surely the least-known player on the roster, third baseman Jimmy Williams, 1899. Has anyone, anywhere, had a better rookie year than his spectacular .355/.416/.532 numbers? Williams, who had never before had even a cup of coffee in the bigs, led the league in triples and finished in the top five in seven other batting categories that year. He played just two years for the Bucs, jumping to Baltimore in 1901 and spending the rest of his career in the American League. Williams even fielded deftly enough to move to second base, which he played quite well, for most of his career. His backup is Tommy Leach, 1902, who combined fine fielding and solid power (he led the league in both triples and homers). The 1923 season of Hall of Famer Pie Traynor, who most readers would have expected to man this position, came in a year of high offense. His good but not great fielding left him behind the underappreciated Leach. Indifferent fielding put Bill Madlock, 1981, out of the running.

The only unknown at shortstop on this team is which Honus Wagner season ranks highest. As it turns out, 1908 was a poor glove year for the Dutchman, but his .354/.410/.542 absolutely towered over the worst year for offense in National League history Fred Tenney of the Giants scored one more run, Brooklyn’s Tim Jordan hit a couple more homers, and Wagner didn’t walk very much; otherwise, Honus led the league in everything, often by a hefty margin (an astonishing 90 points in slugging percentage). Arky Vaughan, 1935, took the runner-up spot at shortstop (well, runner-up behind seven more Wagner seasons), and his 1938 season, when he fielded extremely well, was almost as good. The two — unquestionably the best shortstop of all time, and a plausible choice for second-best — dominate the position, of course. Jay Bell, 1993, shows up on the list behind ten Wagner and four Vaughan seasons, with another Wagner and four more Vaughans between Bell and Dick Groat, 1960.

Batting cleanup, with undoubtedly the finest season ever compiled by a Pirate not named Wagner, is left fielder Barry Bonds, 1992. He did it all with the bat (.311/.457/.624 in a year when the National League was weak on offense), more than compensating for a merely good season afield. Barry’s 1990 season was nearly as good — his hitting wasn’t as great but he had an exceptional fielding year. He is in some very strong company, however. The huge difference between Bonds’ glove and his own subpar fielding kept Ralph Kiner, 1949, out of the top spot in left; at the plate, Kiner put up superior numbers. And Willie Stargell, 1973, was close on Kiner’s heels.

Right fielder Roberto Clemente, 1967, bats fifth in this lineup. The Great One combined the National League batting title, in a low-offense year, with splendid defense. The difference between the pitching-dominated `60s and the lively ball `20s is evident when you look at his backup — Paul Waner, 1927, had much better raw numbers and was also a fine fielder, but his adjusted results didn’t quite match Clemente’s. Dave Parker, 1977 (an extraordinary year with the glove) and Kiki Cuyler, 1925 (in his only right field season as a Buc), also fine fielders, couldn’t quite match Clemente and Waner.

Batting sixth is first baseman Elbie Fletcher, 1941. With all the superb outfielders the Pirates have showcased over the years, it’s surprising that first base has been such a weak position for the club. Fletcher had good gap power, drew tons of walks, and played fine defense, but he was nowhere near the level of the Bucs manning the other “offensive” positions on this roster. Backing him up is another big surprise, Bob Robertson, 1971, whose talents were muted by the low-offense era he played in.

How weak has first base been in Bucco history? After Robertson come Jason Thompson, 1982, and slick-fielding Sid Bream, 1986. Hall of Famer Jake Beckley, playing in the high-offense 1890s, was well down on the list.

Catcher Tony Pena, 1984, hits in the seventh spot for the club. In a low-offense year, catching the league’s best pitching staff (this was the season in which the Pirates led in ERA while finishing dead last), Pena takes the starting position ahead of Manny Sanguillen, 1975. It should be pointed out that two long-ago Pirate backstops, Fred Carroll, 1889. and Doggie Miller, 1891. who would have ranked first and fourth, respectively, were eliminated from consideration due to our requirement of playing at least 50% of the season’s defensive games at his primary position. Carroll played 43 games at C, 41 in the OF, seven at 1B, and one at 3B (and was actually the 1889 backup to Miller). In 1891. Miller played 37 games at SS, 34 games at 3B, 24 games in the OF, and one at 1B along with his 41 games behind the plate. Those guys could hit, but they don’t quite seem like full-time catchers.

By far the greatest impact of adding a measure of defense to the evaluation process appears at second base. Looking only at hitting, unknown Jim Viox, 1913, who seems to have known very little about what to do with that leather thing he had to wear between plate appearances, topped the Bucco list. He was followed by George Grantham, 1930, Dots Miller, 1909, and Johnny Ray, 1984, all known for fielding ineptitude. Meanwhile, arguably the finest fielder in the history of the game, at any position, played many years of 2B for the Pirates. Our number eight hitter, then, is Bill Mazeroski, 1963, whose fielding brilliance (in that and several other seasons) outshines every other Bucco second sacker. Another good field/no hit player, Pep Young, 1938, is Maz’s backup.

The pitchers

With all that firepower among the hitters (Traynor, Lloyd Waner, Stargell, Cuyler, Clarke, Beckley, Lindstrom — Hall of Famers all — didn’t make the club), you probably don’t need to depend on good pitching. That’s fortunate, because, to be frank about it, the Pirates have been weak in that department throughout their history. No pitcher has ever been elected to the Hall of Fame because of his feats with the Bucs; newly-chosen Vic Willis probably comes the closest, though most of his prime years were spent with the Braves.

I couldn’t decide whether to carry five starters and four relievers, or vice versa, so let’s just expand the team’s roster by one, presenting the live-man starting rotation and five-man relief corps for the Bucs’ seasonal all-time team. As described earlier, pitchers are ranked by the total number of runs they saved the team in comparison with an “average” pitcher (defined as 4.00 adjusted ERA) in the same number of innings. This measure accounts for both quality of pitching (low ERA) and quantity of pitching (maintaining that ERA over more innings will save the club more runs). Note, however, that innings pitched for the early years of the game are adjusted to current-day practices — no 440-inning seasons for Pud Galvin and friends.

The very best season by a Pirate starting pitcher, Preacher Roe, 1945, seems worthy of extended study. Even if we discount it as a war-year anomaly, an analysis of the run support differences between Roe and the other starters on the club would be instructive. How else to explain how a pitcher who saved his team nearly 40 runs over the course of the season, on a club finishing 10 games over .500, who led the league in strikeouts and had pinpoint control, whose estimated record with an average offense in a neutral park in a neutral year would be 19-8, ended up as a 14-13 pitcher? Whatever happened to Roe certainly didn’t affect teammate Nick Strincevich, who went 16-10 with an ERA half run higher.

Following Roe are pitchers much more commonly associated with the Bucs. Babe Adams, 1913, was the best of many solid years for the leading hurler of the deadball era Bucs. Similarly, Vernon Law, 1959, and Bob Friend, 1955, were prime performances by the pitching stalwarts of the legendary 1960 World Series champs, and Rick Rhoden, 1986, capped the Pirate career of a pitcher who re-established himself during the gloomy mid-1980s.

Close behind this quintet are Rick Reuschel, 1985, a brilliant season with too few innings to join the rotation, Bob Veale, 1964, and Wilbur Cooper, 1922.

The indisputable ace of the bullpen spent just one season as a member of the Pirates, but what a year it was! Only eleven seasons by starting pitchers have ever saved the Bucs more runs than Goose Gossage, 1977, and most of them needed about twice as many innings to do it. The rest of the bullpen on the club includes Roy Face, 1962 (not the year he went 18-1), Kent Tekulve, 1983, Al McBean, 1964, and Bill Landrum, 1989. All five relievers are righthanded; Rod Scurry, 1982, and Ramon Hernandez, 1973, are the best of a rather weak collection of southpaw relievers.

Conclusions

Undoubtedly, most of you find some of my selections to be ridiculous or silly. I tried to carry out these analyses with as few preconceptions and as few biases as possible, letting the numbers speak for themselves. None of the people whose analyses I employed — Davenport, Palmer, Nichols, Stephenson — try to steer their results in any way; we’re all interested only in trying to assess what the documentary record of baseball tells us about the game.

Still, there are assuredly flaws in the models or the reasoning involved in the systems developed for this analysis. There is always room for improvement. and everyone involved in this project would be delighted to hear constructive criticisms or to discover additional raw data that would enhance the methodology. The deficiencies in our knowledge base are far greater when it comes to fielding than any other aspect, but evaluation of hitting and pitching data can always stand improvement as well.

Along with the researchers whose work I’ve used in these analyses. I thank David Tate, David Orabiner, and many others on the Internet and elsewhere for their assistance and guidance in this project.

 

The Best Buc Single Seasons: Hitters

 

The Best Buc Single Seasons: Pitchers

(Click images to enlarge)

---

Key

• BA = Batting average
• OB = On-base percentage
• SLG = Slugging percentage
• EqR = Equivalent Runs (unadjusted)
• FR = Fielding Runs (linear weights 1887-1987, DA-based 1988-1994)
• EPER = EqR adjusted for park, year, and length of season
• Above Avg. = number of runs above what an ‘average’ player would do while consuming the same number of outs
• Rs (Runs) Saved = difference between the pitcher’s own (adjusted) runs allowed and the number allowed by an average pitcher in the same number of (adjusted) innings