ANALYZING TRADES

PAUL ELSTEIN

EVERY YEAR dozens of players are traded or sold. While the success or failure of trades has been analyzed from the long-run perspective of the teams involved, most notably by Bill James, the immediate impact of a change of scenery on a player's performance has not been evaluated.

  This article will try to answer the question of whether a trade to a much better or much poorer team affects a player's statistics. There are two schools of thought on the subject. By far the most prevalent view is that surrounding a player with high-caliber teammates will improve his play considerably. Conversely, if a player is sent to an inferior team, many persons feel his play will decline. A less widely-held opinion is that some players do better with a losing team. Whether the pressure of a pennant race impacts adversely on their performance or whether they prefer the more relaxed atmosphere of a non-contender is not really known.

  A sample of 178 players, including 52 pitchers, was selected in the following manner:

    All were traded or sold during the off-season and remained in the same league.

  - All played with only one team in each of the two successive seasons.

   - All pitchers had at least 15 decisions in each of the two years. Non-pitchers played in at least 100 games each year.

  - Each was traded to a team whose record the season preceding the deal was at least ten games better or ten games worse than that of the team which he left. For example, Tris Speaker played for the Boston Red Sox in 1915. Shortly before the start of the next season he was sent to Cleveland, which had finished 44 games behind the Red Sox. Thus Speaker's performance in 1916 was included in the study.

  We then looked at four criteria of performance among non-pitchers. These included batting average, slugging percentage, runs batted in and runs scored. We compared each of these over the two-year period for each player. Similarly, pitchers were evaluated using won-lost percentage and earned-run average.

  Of these factors, runs batted in, runs scored and earned-run average are dependent directly on the caliber of a player's teammates (except, of course, when the player hits a home run, which gives him both an additional run scored and run batted in). In addition, a pitcher's ERA might be affected by the fielding skills of his teammates because some good fielders handle balls that otherwise would be hits that result in earned runs.

  Slugging percentage and batting average should not vary according to the degree of success of a player's team unless the hypothesis is accepted that surrounding a batter with hard-hitting teammates will improve his average. The theory has been that in situations of this type pitchers will not be able to pitch around such players and they will have more good pitches to hit.

  Of the 52 pitchers who were being studied, 21 improved both their ERA and won-lost percentage after being traded. Of these 21, 17 had been sent to better teams. For example, Jack Kramer's trade from the St. Louis Browns in 1947 to the Red Sox helped him considerably the following season. His ERA dropped from 4.97 to 4.35, and his winning percentage rose from .407 to a league-leading .783 as he posted an 18-5 record. Only four pitchers traded to poorer teams improved their won-lost record and ERA. Conversely, of the 18 hurlers whose ERA and percentage both declined, only four were sent to better teams and 14 to teams with poorer records.

  This outcome should be expected. Because won-lost percentage depends to a great extent on the play of a pitcher's team, the fact that so many of the pitchers whose percentage improved were sent to better teams was not surprising. The improvement in ERA might be less obvious, although the point made earlier about stronger teams having better fielders could apply. However, of the six pitchers traded to better teams who had a decline in one of the two categories (ERA and won-lost percentage), all six wound up with higher ERAs. Thus Dick Donovan with the expansion 1961 Washington Senators went from a 10-10 record to 20-10 with Cleveland in 1962. However, his ERA rose from a league-leading 2.40 to 3.59 despite his improved percentage. Similarly of the seven pitchers sent to poorer teams who improved in one of the two categories, five had a lower ERA and only two had a higher percentage.

  Perhaps the most interesting case was Milt Gaston. His record in 1927 with the seventh-place Browns was 13-17 with a 5.00 ERA. Not very impressive, but when traded to the Senators, who finished 26 games above the Browns in 1927, Gaston was even worse. His record fell to 6-12 and his ERA rose to 5.51. This performance got him sent to the Red Sox, who finished 28 games behind Washington in 1928. Pitching for the last-place Sox in 1929, Gaston saw his won-lost percentage rise to .387 and his ERA drop to 3.73. Therefore, Gaston defied the expectations twice in successive years.

  Of 28 pitchers sent to better teams, ten of them - or a little more than one-third - finished the next season with a higher ERA. Of the 25 pitchers shipped to poorer teams, nine or about the same percentage had lower earned-run averages. It would seem to indicate that a pitcher's statistics, outside of his winning percentage, do depend to some extent on his team.

  When we look at players other than moundsmen, a different picture emerges. While ten batters improved in the four offensive categories, 11 declined in all four areas. Furthermore, three players went down in three categories and remained the same in one, while only one went up in three areas and stayed the same in the other.

  The real shocker was in the performance of players who were traded to teams with poorer records. Twenty-one of these players improved in all four categories, while only nine declined in all four! Two others went down in three categories and had no change in the other.

   A couple of examples are indicative of this trend. Chuck Klein was expected to give the Cubs a big lift after his 1933 acquisition from the lowly Phillies. Alas, it did not happen. Klein's run production went down by 23, his runs batted in declined by 40, he lost 67 points on his batting average and his slugging percentage slipped 92 points.

   A change of scenery twice was good for Rabbit Maranville even though his new team was inferior to his old one. When he was traded from the Braves to the Pirates after the 1920 season, his run production increased by 42, his RBIs climbed by 27, his batting average rose by 28 points and he added eight points to his slugging percentage. After his 1928 trade from the champion Cardinals to the Braves, a similar result took place. The Rabbit more than doubled his runs scored, added 21 runs batted in, increased his batting average by 44 points and his slugging percentage by 24 points.

   What about Speaker, whom we mentioned earlier? Following his trade from the champion Red Sox to the lowly Indians in 1916, the Grey Eagle's runs total declined from 108 to 102. However, he added 14 RBIs. More important, his batting average rose 64 points and his slugging percentage went up 91 points. Even without Speaker, Boston still won the world championship. However, Cleveland did move 30½ games closer, although still finishing 14 games behind.

   If we look at those hitters who went up in three categories and down in one, the pattern continues. Of the players traded to better teams, nine improved in three categories and 11 went down in three areas. For those traded to poorer clubs, 11 improved and seven declined.

   How can this be explained? Possibly the team that was better one year slipped the next. Nap Lajoie went from the last-place Indians of 1914 to the defending American League champion Athletics. We all know that Connie Mack broke up the Athletics after the 1914 season, and we might expect that Lajoie's record would be worse even though he was traded to a team that had been much better than his old club the previous year. Although Philadelphia finished in the cellar and 14 games behind Cleveland in 1915, the 40-year-old Lajoie improved in all four categories.

   It also is possible that a poorer team may, nevertheless, be better offensively than one with a superior record. However, while there have been teams like the 1930 Phillies or the Hitless Wonder White Sox, the statistical probability has been very low.

   Finally, the numbers may be small enough to be misleading. Although this is possible, the sample would seem to be sufficiently representative to indicate that playing for a. better team does not necessarily result in improvement in a player's performance. Although pitchers traded to better clubs did improve, one-third still had higher earned-run averages with their superior teams, while one-third of those traded to poorer teams wound up with better earned-run averages. For hitters, a trade to a weaker team more often was a shot-in-the-arm than a trade to a contender. While the data are relatively straightforward, the explanation is indeed far from simple.