Win-Win Trades

If you are anything like me, you probably think of trades, especially those that have occurred in the past, as zero-sum games. We ask who won and who lost, who got the better of the deal and who got the worse. We imagine that two participants (general managers) enter into a contest, battle over a fixed payoff (surplus WAR), and one person’s loss is the other’s gain.

But as students of game theory know, not all games are zero-sum. Some require—and reward—cooperation. When it comes to trades, nothing illustrates this principle better than win-win trades, in which, theoretically anyway, general managers do not beggar thy neighbors but, rather, both profit from the exchange. Of course, at least when they happen, all trades are win-win, otherwise one general manager or the other would back out. Like the tango, it takes two to trade, and both dancers must like the song. But win-win trades in particular underscore this point about how every trade must be mutually beneficial for it to occur.

Below, I describe three such win-win trades from across baseball history. As these trades show, the magic of win-win trades happens when teams trade from strengths to fill weaknesses. Both teams surrender valuable players, but by redistributing their assets, both teams end up stronger than before the trade.

Jan. 22, 1918: New York Yankees trade Les Nunamaker, Fritz Maisel, Nick Cullop, Urban Shocker, Joe Gedeon and $15,000 to St. Louis Browns in exchange for Eddie Plank and Del Pratt.

In 1917, the New York Yankees and the St. Louis Browns both finished in the second division, as they had virtually every season since they began playing professional baseball (first division meaning top four teams in the then-eight team American League, and second division meaning the bottom four). True, the Yankees had finished fourth in 1916, and second a couple of times in the previous decade, but they had not sniffed a pennant in a long time. (From 1921 to 1964, New York would win 29 pennants; the 1917 Yankees were not—yet—those Yankees.) The Browns, meanwhile, had an even less inspiring record. Since the franchise started play in 1901, they had finished in the first division just twice and had also never won a pennant.

Both teams, however, had reason to believe that better days lay ahead. The Browns had a new owner and a young, slugging first-baseman named George Sisler. For their part, the Yankees had recently purchased third baseman Home Run Baker, could play defensive wizard Roger Peckinpaugh at shortstop every day, and sported a deep rotation led by a young Bob Shawkey.

By the end of the 1917 season, though, the Yankees had something more than a hole—they had an abyss—at second base. In August of 1913, they had acquired Fritz Maisel from the then-International League Baltimore Orioles, and Maisel initially played well, hitting a bit, playing passable defense at third base, and tearing up the base paths. (He stole 74 bases in 1914.) In 1916, though, he moved to second base to make way for Baker, and things fell apart. In May, he broke his collarbone, and when he returned in August he struggled. In 1917, his struggles worsened. He hit .198 in 466 plate appearances and, perhaps needless to say when one hits like that, was worse than a replacement player.

On the bright side, the Yankees had more pitchers than they knew what to do with. In 1917, they gave nine or more starts to eight pitchers, each of whom was under 30 years old and within spitting distance of a league-average FIP or better.

Cue the Browns, who had a premier second basemen in Del Pratt, even if he had slipped a bit in 1917, his age-29 season, because of wrist and knee injuries. Pratt had also made himself expendable by bringing a libel suit against Browns owner Phil Ball, who had charged his players, including Pratt, with “laying down.” In 1917, the Browns also had a rotation led by Dave Davenport and Allen Sothoron, both of whom were worse than replacement pitchers that year. Still worse, one of their best pitchers, 42-year-old future Hall-of-Famer Eddie Plank, had announced in August that he would retire.

In short, the Browns had a second baseman they needed to trade and a rotation in name only; the Yankees needed a second baseman and had pitching to spare. The trade stars aligned. On Jan. 22, the Yankees sent a passel of players to the Browns for Pratt and Plank. Although the Yankees gave up a lot, the trade looked fair for both teams. Pratt could be expected to bounce back, even if he would soon turn 30, and Plank could still get batters out, assuming, that is, the Yankees could lure him off his farm in Gettysburg, Pa., and into the big city.

For their part, the Browns received three position players. Among them was Maisel, who, if healthy, could possibly fill out the Browns infield; Joe Gedeon, a 24-year-old second baseman who had not quite lived up to his early promise but who might benefit from a change of scenery; and Les Nunamaker, the Yankees’ passable though aging starting catcher. But the real prizes for the Browns were the Yankee pitchers, Nick Cullop and Urban Shocker. Cullop, like Nunamaker about to turn 30, had struggled in 1917 but was just one year removed from a solid 1916 season that saw him post nearly 3 WAR. Shocker, meanwhile, looked like he might be an ace. In 1916, as a 25-year-old, he had mowed down the International League while pitching for the Toronto Maple Leafs — then a Yankees farm team — and he continued to pitch well when the Yankees summoned him later in the season. In limited duty in 1917, he had also impressed.

In retrospect, as every model win-win trade should, the trade worked out for both teams. Disappointing the Yankees, Plank could not be talked out of retirement. However, Pratt returned to form in 1918, posting 3 WAR that year and 10 more over the next two years. After that, the Yankees traded him to Boston for, among other players, a young Waite Hoyt. (Overall, Pratt produced nearly 23 WAR until he retired in 1924.)

As expected, with the enormous exception of Shocker, none of the players the Browns received from the Yankees turned into much. Maisel had a decent year in 1918, but he soon left the Browns for a long career with the Baltimore Orioles. Gedeon started at second base for the Browns for three years, and just as it seemed that he had figured things out, he was banned from baseball for his role in the Black Sox scandal. (Privy to the plot, he had bet against the White Sox.) Nunamaker had a good year for the Browns in 1918, but since the Browns already had a catcher in Hank Severeid, he would soon leave town in a trade for (weirdly) yet another catcher. Cullop did not play in 1918—I assume he was drafted into the military—and though he would have a nice career with the Louisville Colonels of the American Association once he returned to professional baseball, he would pitch only a few innings for the Browns, and those came in 1921.

Shocker, however, more than made up for these disappointments. Over the remainder of his career, he posted nearly 52 WAR. If you think of the trade as Pratt for Shocker, clearly the Browns got the better of the deal. Balancing the scales somewhat, however, is that the Yankees reacquired Shocker in 1924 for three mediocre pitchers after Shocker and Browns management soured on each other.

Both Pratt and Shocker made their teams better, too. Beginning in 1918, the Yankees finished in the first division in each of the next three years. The Browns also improved, though they had further to climb. After finishing seventh in 1917, they finished fifth in 1918 and 1919, fourth and third in 1920 and 1921, and, in 1922, with career years from Sisler and Shocker, finished just a game behind the Yankees in the pennant race, which is about as good as things got for the Browns in their half century of mediocrity.

Oct. 11, 1946: New York Yankees traded Joe Gordon to Cleveland Indians in exchange for Allie Reynolds.

Unlike the 1918 trade, which saw the Yankees trade pitchers for an infielder, another classic win-win trade saw them trade an infielder for a pitcher. In 1946, New York had slipped from its perennial pre-war place atop the American League to third place, 17 games behind a dominant Boston Red Sox team that won 104 games thanks largely to perhaps the greatest season Ted Williams ever had. Lower down in the standings, the Cleveland Indians finished a distant sixth, nine games below .500 and 36 games behind the Red Sox.

Heading into the 1947 season, the Yankees rotation was top-heavy and aging. Their ace, Spud Chandler, would be 39, and their next best pitcher, Bill Bevens, seemed unlikely to repeat the career year he had in 1946 at age 29. At the bottom of the rotation, the Yankees gave 10 or more starts to the likes of Joe Page, Tiny Bonham, Al Gettel, Randy Gumpert and Cuddles Marshall. Over 65 starts, this crew produced just 2 or 3 WAR. If the Yankees had a hollow rotation, their infield was bursting at the seams. They had at least five viable players—Joe Gordon, Phil Rizzuto, Snuffy Stirnweiss, Billy Johnson and veteran Frankie Crosetti—for just three positions.

By contrast, the Indians could not find a second baseman for love or money. In 1946, they gave starts at second base to Dutch Meyer, Ray Mack and Jack Conway, who together were ever so slightly worse than a replacement player. And while not exactly swimming in quality pitchers, the Indians had more options there than they did at second base.

So on Oct. 11, 1946, the New York Yankees traded their all-star second baseman, Joe Gordon, to the Cleveland Indians for starting pitcher Allie Reynolds.

In retrospect, this trade would look win-win, but it could have just as easily turned out lose-lose. Neither player was a sure bet to fill the hole for his new team. Indeed, both players looked like they might be past their primes. After dominating American League pitching for his first six years in the majors, Gordon went off to serve in World War II. His first season back, 1946, did not go well. He made the All-Star team, but more on past than present glory. In 1946, suffering from injuries and the layoff, he hit just .210 with 11 home runs; in the six years before the war he had averaged .278 with 24 home runs. Now Gordon would be 32. It was not at all clear what he would do in 1947.

Reynolds’ future looked no more promising. Unlike Gordon, Reynolds did not serve in the war. He stayed back and pitched well—but not as well as one would think—against the 4-F leftovers who comprised American League batting lineups during the war years. In 1946, when players drafted into the armed forces returned to the major leagues, and Reynolds suddenly had to face real hitters, he did not fare nearly so well. During the war he had been a solid starter, but in 1946 he was not even replacement level. Like Gordon, Reynolds would probably bounce back, but here too you would not necessarily want to take that bet, not least because he would turn 30 in 1947.

As it turned out, both Gordon and Reynolds flourished. In 1947, Gordon returned to form, hitting .272 with 29 home runs, good for seven wins above a replacement player. He played three more years with Cleveland before retiring in 1950, accumulating about 20 WAR.

Like Gordon, the change of scenery did Reynolds good too. In 1947, he won 19 games for the Yankees with a 3.20 ERA, contributing about 2 or 3 WAR. He was good if not great for the next two years, and then from 1950 to 1952 put together three very good years, including a stellar 1952 that saw him win 20 games and post a 2.06 ERA. From his first year with the Yankees until he retired in 1954 at the age of 37, Reynolds would contribute close to 20 WAR, virtually identical to Gordon’s figure.

The trade worked out for both players, which meant it worked out for both teams, too. In 1947, with Reynolds in the rotation, the Yankees won 97 games and beat the Brooklyn Dodgers in a classic seven game World Series. After missing the playoffs in 1948, the Yankees went on to win five straight pennants and five straight World Series. Things did not go quite as well for the Indians in 1947; they finished in third place, 14 games back of the Yankees, but they went considerably better than they had in 1946. In 1948, however, Cleveland ended the season tied with Boston, defeated the Red Sox in a one-game playoff, and went on to win the World Series over the Boston Braves. They needed every one of the six or seven wins above replacement Joe Gordon contributed that year, in what is still their most recent World Series triumph.

Who won the trade? It depends on whether you prefer peaks or plateaus. If you like peaks, Gordon contributed his 20 WAR for the Indians over just four seasons. And when he retired in 1950, he may have had one or two good years left in him. If you prefer plateaus, you might think the Yankees got the better of the deal. Reynolds provided the Yankees with six seasons of above-average pitching. Regardless, I imagine that with perfect hindsight, each team would do the trade again.

May 25, 1984: Boston Red Sox traded Dennis Eckersley and Mike Brumley to Chicago Cubs in exchange for Bill Buckner.

The same may or may not be said for another pair of teams involved in a classic win-win trade from more recent baseball history. Nearly two months into the 1984 season, the Chicago Cubs had played their way into first place. After a string of five straight last or next-to-last finishes in the National League East, the team decided to take advantage of its early success. Most pressing, the Cubs did not believe they had a championship rotation. The starting staff was led by the recently acquired Steve Trout and also featured a solid Scott Sanderson, but after that it took a turn for the worse. Rick Reuschel had been good, and would be good again, but in 1984 he was still recovering from rotator cuff surgery and did not belong in a major-league rotation. Nor did Dick Ruthven or Chuck Rainey. The sun had set on Ruthven, and it had never truly risen for Rainey. In short, the Cubs needed pitching. Fast.

They also had one too many slow, slugging corner players. Before the season started, they traded for outfielder Gary Matthews, who joined Leon Durham and Keith Moreland in the outfield and Bill Buckner at first base. For 1984, Durham moved to first base—which helped defensively—but the shift left the veteran but still useful Buckner without a position.

Meanwhile, the Red Sox had not started 1984 nearly as strongly as the Cubs, but they had reason for hope. They had won 89 games in 1982—before winning just 78 in 1983—and would eventually go on to win 86 in 1984, good for fourth place in a very competitive American League East. They also had arguably the best young rotation in the league, led by Bruce Hurst, Bob Ojeda, Oil Can Boyd—all 26 years old or younger—and 21-year-old Roger Clemens.

They also had 29-year-old Dennis Eckersley, who had not yet transitioned to the bullpen but who had, because of injuries and ineffectiveness, suffered through a dismal 1983 season. He had not begun 1984 much better, and he would be a free agent at the end of the year. At first base, meanwhile, the Red Sox still had Dave Stapleton, the utility man whom they had installed there in 1983 and who had had an epically bad year, to the tune of -1.6 WAR. He started the 1984 season even worse.

So, naturally, in late May the Cubs sent Buckner and prospect Mike Brumley to the Red Sox for Eckersley. Brumley never amounted to much, but he had been drafted in the second round of the 1983 draft, so he must have represented the difference, in the minds of the Cubs and Red Sox general managers, between Eckersley and Buckner, even though Buckner had of late been the more consistent player.

The trade turned out well enough for both teams, though the Cubs definitely got the better of it. (Without a doubt, though, it was win-win in terms of MAR—mustaches above replacement. Both Buckner and Eckersley had/have beauties.) Buckner could hit for average, and would as the starting first baseman for the Red Sox for the next three years, but his defense at first and his stubborn refusal to take a walk dragged down his value, especially in the years after the trade. He had a passable year in 1985, but for the remainder of his career, including 1984, he was usually below replacement level. True, in 1984 he was probably better than any one else the Red Sox could play at first base, but that is damning by very faint praise indeed.

Eckersley rediscovered his former self, and not by switching to the bullpen, which would come later. After joining the Cubs in May, he started 24 games and contributed 4 WAR. A few weeks after acquiring Eckersley, the Cubs sent three prospects, including Joe Carter and Mel Hall, to the Cleveland Indians for starter Rick Sutcliffe, and with Sutcliffe and Eckersley Chicago made the playoffs for the first time in four decades. The Cubs re-signed Eckersley the following year, and he gave them a fantastic 1985 and an average 1986 before departing for Oakland and the bullpen.

…

What can we learn from these trades? To risk stating the obvious, all trades, even win-win trades, involve risks. No one can see the future, especially when it comes to something as overflowing with variables as the game of baseball and the athletes who play it. Even so, more than most trades, what strikes me about these three is just how arbitrarily they turned out. Of the three trades, at the time they were made, only Shocker for Pratt seemed like a definitive overpay. Even so, the Yankees wanted to win now, and they could afford to give up Shocker, and as everyone knows, There Is No Such Thing As a Pitching Prospect, so who knows.

As for the 1946 trade, Gordon had a better track record than Reynolds, but Gordon was coming off injuries and Reynolds was younger. Call it a toss-up. In 1984, I might have preferred Buckner to Eckersley, not because Buckner was very good—even in his best years he was no more than a solid starter—but mostly because the projections for Eckersley did not look good at all. He had intermittent but serious arm trouble, and in two out of his last three years he had not pitched well at all.

All of which is to say that at the time, each trade looked fair, which makes sense since they were designed to help both teams win now. But going strictly from past performance, none of these players looked like particularly good bets to produce much. Some of them did, and some of them did not, but none strikes me as having been foreordained to succeed or fail. Just the opposite. It would not have surprised me at all if Pratt never rediscovered his swing; or Shocker became a good but not great pitcher; or Gordon never recovered from his layoff; or Reynolds never developed into a dominant pitcher; or Eckersley never got his velocity back; or Buckner did not lose his value completely. Indeed, all these possibilities seem at least likely to have happened as what in reality did happen.

So beware those who gloat over the egregiously lopsided trades of the past. Those trades may not have looked so lopsided at the time, and in another universe only a little different from ours, they may have tipped in the other direction.

Some general managers make very obviously bad trades, and everyone at the time knows it. But these are fewer than you might think. More often, a trade reveals itself only in time. It goes from a state of equilibrium to disequilibrium. However, predicting the outcome of a trade is like predicting the weather: there is some science involved, but equal parts chance.