Triples and Defensive Position: How the Distribution Has Changed Through the Years
I have a questionable habit of introducing subjects to you no matter how I came to them, whether it was interesting or not. At least this approach is consistent with the general form of my presentation. I not only take you through the development of the research, but how it came to be in the first place. To start off these screwball topics out of thin air, to me, would be just like bringing up a topic out of thin air in conversation. The alternative to presenting the embarrassing truth would be contrivance, and that is no real alternative.
So the genesis of this one is, I thought the ‘54 Baltimore Orioles would be an interesting team to study because they were really bad, particularly offensively. But randomly, I ended up noticing that they hit 49 triples, 3rd in the American League. This was more than the ‘54 White Sox, who had Minnie Minoso, and Minoso hit 18 to lead the American League. As a ratio, that would be like a 200-home run team having a hitter with 77 home runs. You’d have thought Minoso would have put the White Sox over the top in triples. The Orioles’ top triples hitter was Cal Abrams, who only had 7. Then, not only did the White Sox have Minoso, they had Nellie Fox and Jim Rivera, both of whom had 8 triples, more than Abrams had. How on earth then did the Orioles hit more triples than the White Sox?
Trying to break this down, I obtained statistics for each position, and noticed something interesting.
‘54 White Sox triples by position: LF 12, CF 3, RF 10
‘54 Orioles triples by position: LF 8, CF 2, RF 8
For both teams, the total for center field was much lower than the total for the other two outfield positions. We know that speed is central to triples, and today, that center fielders are on average the fastest of the three positions, so this division was curious. I had done a massive study of triples in what I called “the DiMaggio era,” and believed I had observed the same thing then I saw now with the White Sox and Orioles, or at least that left fielders and right fielders were sometimes the leading triples hitters of the day. So I got it in my mind to use Stathead (Thank you thank you thank you Stathead, I never stop marveling at what you allow us to do) to get the whole positional breakdown by year, or at least back to 1924, the first year where play-by-play coverage (and thus positions) exceeded 90%.
Analyzing total triples over time was worthless without accounting for the number of teams, so I pivoted to the percentage of triples that each position contributed. There have been many more games since 1924 without Designated Hitters than with them, so I excluded them. Pitchers have sometimes been supplanted by DHes, of course, and even when they haven’t been, the number of at-bats they take and the number of at-bats that pinch hitters take has varied, so I threw out pitchers, too, not to mention pinch hitters. So I consider eight positions. An average position, then, contributes .125 of triples.
Over the 99 years, average triple representation by position has been
CF .175
LF .144
RF .142
2B .132
SS .130
3B .108
1B .101
C .068
From 2010-2022, the data has been thus
CF .210
SS .148
LF .146
2B .145
RF .133
3B .099
1B .069
C .050
Another telling snapshot is 1924-1949
RF .153
CF .149
LF .148
1B .137
3B .115
2B .114
SS .105
C .079
The two above lists, by the way, are weighted by total number of triples for each year, rather than being an average of the individual-year averages.
The next list is of each position’s standard deviation for triple percentage over the 1924-2022 period. Normally, a high standard deviation indicates inconsistency, but here it works well as a proxy for change. In other words, if a position used to be responsible for a good number of triples but now isn’t, this shows up as a higher standard deviation. The standard deviation in this analysis doesn’t seem to have much to do with year-to-year deviations, as I imagine each position has an equal amount of randomness to its numbers. Nor is it the case that triple percentages for positions are marked by multiple peaks and valleys. For every position, it seems there has been a dominant trend of rise or decline or general uniformity. For some positions, the rise or decline has settled, but the effect on the standard deviation is felt. So you can see that the greatest change over the years has been for 1B, CF, and SS. Third base, LF, and C haven’t changed much. If you want to combine the mean and the standard deviation, approximately 68% of the first basemen triple percentages have been between .073 and .129, and you can apply the same technique to the other positions.
Standard deviations (when same to three digits, they are listed in order of the exact number)
1B .028
CF .028
SS .024
2B .020
RF .017
3B .015
LF .015
C .015
Then I would be remiss if I did not also list the standard deviation by decade (or, in the case of this decade, 2020 - 2022, and in the case of the ‘20s, 1924 - 1929). So what this is — I figured the share of triples of each position over a decade, and then figured the standard deviation of those percentages. In the 1950s, Baseball Reference has positional data for 6,363 triples. Each position has a percentage of these, and then I figured the standard deviation for those.
Standard Deviation of Position Triple Percentage, by Period
1924 - 1929 .023
1930 - 1939 .026
1940 - 1949 .031
1950 - 1959 .026
1960 - 1969 .027
1970 - 1979 .029
1980 - 1989 .039
1990 - 1999 .046
2000 - 2009 .048
2010 - 2019 .052
2020 - 2022 .042
The general takeaway is certainly that is has gone up. But after there was little change really from 1930 - 1979, standard deviation increased 59% over the next two decades, and continued to rise in the two decades after that. Whether the dip in the first three years of this decade means anything is unknowable.
How do we make sense of this data? I do tend to think the very low number in the ‘20s speaks to an undeveloped, maybe not terribly high-level game. But we know baseball got better when the color barrier was broken. I don’t think that improvement was instantaneous, either; maybe it’s a bias against those no longer living, but I certainly think baseball was much better in 1979 than 1950. Yet nowhere in the 1930 - 1979 interval is there a sign of increased standard deviation in triple frequency across positions.
A way to further make sense of what happened is to look at the changes at each position and when they occurred. What I really notice in the ‘80s and the ‘90s is center field’s increasing in its percentage, and first base and catcher decreasing. Second base and shortstop showed some increase, too. None of that is too surprising, given the standard deviation increase. The high percentages got higher, and the low percentages got lower. If we take triple count as a proxy for athleticism, baseball became more efficient at putting the most athletic players at the positions most requiring athleticism. Another take would be that baseball became more specialized. Maybe the catchers and first basemen got bigger and bulkier, and thus able to do other parts of their jobs better, but not able to hit more triples. The “specialization” explanation again sounds like baseball got better, though, and more overall improvement from ‘79 to ‘99 than ‘30 to ‘79 would surprise me. I therefore favor the “efficiency” explanation (efficient placement of talent).
I’m now going to go position by position, discussing trends. A self-indulgent ploy is to look at a year where a position led in triples and then look at the players responsible. I essentially created a fantasy world with eight teams and observed the pennant race through the years. Supremacy, however, was determined by the mostly inane achievement of hitting triples! Something even the ‘54 Orioles did to the tune of 3rd-best in their league.
Catcher
We can derive comfort from the fact that catchers not only do not hit triples today, they never have. Then again, they’ve never stolen bases, either. It is part of what makes baseball, baseball. That they have the lowest standard deviation of any position speaks to this. Part of the low standard deviation, however, is just because their triple totals are so low to begin with. Catchers do comprise a lower share of triples than they used to, and one can’t help notice that their rate dove right when the standard deviation across positions rose, which is to say, in the 1980s. Their whole trend by period resembles the inverse of the standard deviation table (the correlation is -.95). Catchers had a triple percentage of 7.8 in the ‘70s, which fell to 5.9% in the ‘80s. It reached a nadir of 5.0% from 2010 - 2019.
There have been years when catchers beat other positions in triples, but there has never been a year when they’ve beaten more than one position. In ‘44, ‘46, and ‘47, they hit more triples than shortstops. Starting in 1972, they occasionally beat first base. The last time that happened was in 2015. Tim McCarver’s best efforts in ‘66, when he led MLB with 13 triples, caused nary a ripple in the overall standings. Catchers trailed 7th-place 1B by 16 triples. They comprised a lower percentage of triples that year than they did in 1962 or 1970.
1B
The story of first basemen and triples over history has been down, down, down. Down in every period, in fact, from 1924 - 1929 through 2010 - 2019, and significantly down. 2020 - 2022 has broken the trend, showing a 7.7% triple share for first basemen, compared to 6.7% from 2010 - 2019.
To have come down so far means you used to be a lot higher, and the 1924 - 1949 breakdown bears this out. Triples by first basemen were no threat to surpass triples by outfielders, but a respectable 4th right behind. Take just 1924 - 1929, however, and we almost have news. Right fielders hit 889 triples, center fielders 883, and first basemen 877. It’s incredible to me that, if you talked to a baseball fan in the ‘20s and told him first basemen didn’t hit triples, he might have pushed back, if he didn’t outright push you. And what’s more interesting is the possibility that first basemen may have been able to run, and the possibility that the contemporary stereotype didn’t apply.
But the bigger question here is, what has changed? Is it what goes into a triple, the skills that are needed to produce them, maybe because of smaller ballparks? Or is it which types of players man which positions?
There have been four seasons when first basemen topped the triples charts: 1926, 1928, 1936, and 1946. Given the data for the ‘40s (first basemen trailed center fielders by 21%), the ‘46 inclusion is the most surprising. It couldn’t have been a closer race: first basemen hit 96 triples, right fielders 95, and center fielders 94. Perhaps only the most diehard Cardinal fan can recite Stan Musial’s position for any given season, and for some reason, at just age 25, he was playing a lot of first base in 1946, 114 games. He banged out 20 three-baggers overall, four more than anyone in the majors, and twice as many as anyone in the NL. As for other first basemen in 1946, you don’t find anyone with more than 8 triples, which was done by Elbie Fletcher and Mickey Vernon.
2B
In the triple universe, if second base were a player, he would be lovable and uncontroversial, perhaps bordering on invisible. Not a threat to anybody, never taken seriously as the best, for all his competence. The trend for triples at second has been decidedly up, however. The three lists I led off with do tell a lot of the story. The second base share was .114 from 1924 - 1949, compared to .132 for the whole 99 years, and .145 for 2010 - 2022. The rate has not gone continuously up in the same way that first base has gone continuously down, though. A breakthrough was in the ‘90s, when second basemen hit the second-most triples, and 15.1% overall.
The only time in my sample when second basemen led baseball in triples was in 1924, the very first year of the sample. As second basemen hit only 12.1% of triples from 1924 - 1929, this is not the year I would have guessed for them. Triples were so much across the board in the ‘20s, however, that aberrations were possible.
Then, the next year, the second baseman share was down 30%, and second base ranked 7th of the 8 positions. What happened in ‘24, and what in ‘25?
There were four big hitters of triples in 1924 at the second sack: Rabbit Maranville with 20; Frankie Frisch with 15; Rogers Hornsby with 14; Andy High with 13. Hornsby escaped it, not only manning second again and hitting 10 triples but repeating a .400 average, but a plague seemed to fall on the other three players in ‘25. Incredibly, they combined for only 55 games at second! Maranville switched teams and moved completely back to short (which pleased him, I imagine). Frisch played as much short and third as second. Andy High, a strong offensive performer in ‘24 and 12th in the MVP, found himself released after just 131 plate appearances with Brooklyn. He went on to play 60 games at third base for the Braves. If these players had stayed at second, they would have contributed little to the cause, anyway. Frisch had 6 triples, Maranville just 3, and High just 2. Part of what we are witnessing is just a classic regression-to-the-mean effect.
Maranville had other good triple seasons. He hit 177 in his career. The all-time triples leader, Sam Crawford, was probably more of a power hitter than a speedster. He got triples by hittin’ ‘em where they weren’t: over their heads, but in the ballpark. That Maranville, who was all of 5’ 5”, could hit so many triples suggests that after the dead-ball era, power was a much smaller part of triples. It’s natural to see first basemen on a level with outfielders hitting triples and think it was still largely about power, but Maranville’s triples record belies this.
3B
Third base and triples are Boring with a capital B. If you have nowhere else to play, you play third base. Of the 11 periods, third base has finished 5th with its average twice, 6th eight times, and 7th once. Through 1989, every decade’s share was between .109 and .118. The decimal dipped to the .09s in the three decades subsequent. It was .140 in 2020, but that was presumably largely a sample size issue. There clearly is a slight upswing at third, though, or perhaps a move to the percentage before the ‘90s: the last seven seasons have been .099 or higher. The worst days for third are over.
The best moment was in 1964. Third basemen out-tripled everyone that year! With a 14.0% share, it is the lowest leading share for any season and thus confirmed as a fluke. However, there is one heartwarming fact that made third base a worthy winner. Rookie Dick Allen, playing 162 games at third and putting together a superlative season, was the co-leader in MLB in triples with 13. He was joined by Ron Santo. Since third base is not and never has been a haven for triples hitters, that the two triples leading lights in MLB in 1964 played third deserves to be set aside as a trivia question.
SS
Other than center field, no position better demonstrates the shift in triples toward what are typically thought of as more athletic positions than shortstop. When you look at the 1924-1949 list, where shortstops were 7th in number of triples, and then you look at the 2010-2022 chart, where they were second, you see that baseball has really changed in this respect. Note, however, that the difference in triple share for SS between 2010-2022 and 1924-1949 (.043) is still far less than the increase in triple share for CF comparing those two periods. Also note that the correlation between the triple rates for SS and 2B over the 11 periods is .89, which seems extremely high. Since we are dealing in zero-sum shares, the average between any two positions is naturally 0, with negative correlations as likely as positive correlations. So the move to more triples at second base has been nearly as decided as at shortstop, and it is satisfying that these two positions have moved together. You probably think of shortstops as more athletic than second basemen (they at least have stronger arms), but we think of the players who play them as similar types, and their triple shares have been similar as well.
There are several differences in the data for the positions, however. The first that seems to merit comment is that shortstop triples really took off in the ‘00 - ‘09 period. While beaten out by center field’s .188, shortstop’s .170 would have topped any decade before the 1980s.
Included in this fine decade for shortstop triples were three leading years. In addition, shortstops won titles in 1965, 1966, and 1979. Second base, if you remember, is something of the Boston Red Sox of the project, not having had the most triples since 1924. Shortstop’s winning three titles in the ‘60s and ‘70s doesn’t seem to make sense on its face. Total triples for shortstops trailed all outfield positions by a good margin in the ‘60s; shortstops, in fact, hit only three more triples than second basemen. In the ‘70s, the triple share for second basemen (.133) was similar to the triple share for shortstops (.137).
When I think of 1965 shortstops, my mind immediately goes to Zoilo Versalles. Versalles, in fact, led the American League in triples in ‘63, ‘64, and ‘65. In ‘65, he shared the league lead with another most-of-the-time shortstop, Bert Campaneris. Other shortstops with double-digit triples that year were Leo Cardenas and Hall of Famer Luis Aparicio.
The stars behind the shortstop triumph in ‘66 were somewhat different, and perhaps it could be characterized as a team effort. Bert Campaneris, though, hit 10 triples, showing up again, and Aparicio added 8.
If you told me shortstops led in triples in ‘79, I would have guessed Garry Templeton was pivotal, and indeed he was. He hit 19 triples. (Would you believe, George Brett hit 20 and bested him for the major-league lead?) Templeton’s early triples totals were downright gaudy, certainly by today’s standards. He was 25 years old when he left the Cardinals and had already hit 69. The huge ballpark and artificial turf probably facilitated triples, but nine of Templeton’s 19 in ‘79 were on the road. Other double-digit triplers who played short in ‘79 were Larry Bowa, Ivan de Jesus, and Alfredo Griffin. Sixteen players in all were in double digits in ‘79. Last year, we did have a shortstop lead the major leagues in triples, Amed Rosario, but he hit only nine.
I suggested before that 2000 - 2009 was the golden age for shortstop triples. I saw that data and immediately and uncritically thought “Jeter, ARod, and Nomar.” But Nomar peaked earlier, and ARod was off to third base by 2004. Focusing in on the 2001, 2004, and 2005 shortstop wins, Jimmy Rollins was in double-digits in triples each year. In another year, his MVP 2007, Rollins hit 20 triples and didn’t even lead major-league baseball. Tidily enough, a center fielder, Curtis Granderson, did; he hit 23 (amazing how fast the game changes!) Jose Reyes doubled with Rollins as the shortstop triples star of the decade. He was just getting going in 2005 and led the majors with 17. By the time he was 29 years old and had left the Mets, he had 113 triples (take that, Garry Templeton!)
Another shortstop who was major-league leader in a year that shortstops led all positions was Cristian Guzman, who hit 14 triples in 2001. Juan Uribe was a rookie shortstop for the Rockies then, though, and one would be excused for being more impressed by his 11 triples in 72 games.
ARod’s one year at short when shortstops were the top triple hitters says “plodding slugger” more than anything else: 52 home runs, but 1 triple. I’m sure the Rangers lived with the lack of triples.
LF
Triple percentages of left fielders have been Steady Eddie over the years. They have been between .135 and .157 in every period. In the ‘30s and ‘40s, outfield position did not predict triple rate at all; it was essentially the same for all three positions. In the ‘50s and ‘60s, left field did fall behind right field, and of course, center field, too. Left field has beaten right field in every subsequent period, but by moderate and sometimes even narrow margins. Meanwhile, we have had the ascendance in triple percentage from shortstops and second basemen, and the upshot is that left field has been just 4th, 4th, 3rd, and 4th in the four periods since the ‘80s (counting 2020 - 2022 as a period). That’s dangerously close to mediocrity, although on the other hand, an average share of triples would be .125, and left field is always over that.
Left field has been tops in triples 10 times, including in that ties with right field in 1925 and 1951. Victorious years were 1925, 1930, 1935, 1942, 1943, 1949, 1951, 1981, 1983, and 1989.
I believe commitments to the military spiked up in 1943, so it is hard to see too much commonality between ‘42 and ‘43. Stan Musial did his thing both years, hitting 10 and then 20, but the 20 came when he played over four times as many games in right asin left. Over his career, to my surprise, Musial would have had about as many games in right as left if he had had one more full season in right.
Left field was a deserving winner in ‘49, as Dale Mitchell rang up an advantage over his competitors that makes me think of Tiger Woods. No one else in the league hit more than 13, but he rolled out that Curtis Granderson total of 23.
The peak for left field was certainly in the ‘30s and ‘40s, before the triple hitters ended up disproportionately going to center, but the ‘80s were also a very good time. I mentioned those three yearly titles. When we think of the speed that was patrolling left in that era, it’s no wonder. In 1981, a strike season, you had Henderson, Raines, and Willie Wilson all playing left, and all hitting 7 triples. They were no match for Gene Richards, though, who played left and had 12 triples that year.
Brett Butler made the most of his short time in left to hit 13 triples in 1983 and lead MLB. He was deferring to Dale Murphy, who played center. The next year, he escaped to Cleveland (if one could escape to Cleveland baseball in that era), and was in center for a good 12 years.
Vince Coleman was another speedy left fielder no stranger to triples. He was among the NL leaders with 9 in ‘89, and he was to triples in that era what Keith Hernandez was to Triple Crown stats, hitting 10, 8, 10, 10, and 9 in his other years with the Cardinals. No doubt power and particularly gap power are advantageous for triples, but if you are fast enough, and you have the turf, the triples will still follow. Coleman was an explosive runner but a slap hitter, with 28 career home runs in over 5,000 at-bats.
Not to be forgotten is Larry Herndon. He played left field and shows up on the triple leaderboards of ‘81 and ‘83. His first years, admittedly, were in center with the Giants, but he averaged over 9 triples a year from ‘78 to ‘83.
RF
I’m going to skip to right field, as it will be easiest to leave center field for last. When characters all know each other, so to speak, you end up talking about them without actually talking about them when you do enough talking. It’s like when I read biographies of the great race horses, if I’ve read about Alydar, do I really need to read about Affirmed? So many times, they ran in the same races. By going through the year’s leaders, I can also conclude by saying, “If you didn’t hear about a year, you can safely assume that center field led it.” So center field works best presented last.
I would just underscore that right fielders used to hit as many triples as anyone, referring you to the 1924 - 1949 list. They have actually hit the most triples in seven years since 1950 as well, although never since 1978. Leading years were 1951 (tied), 1956, 1958, 1963, 1968 (tied with CF), 1969, and 1978.
Right fielders led every year from 1937 - 1941, sometimes by healthy margins. It’s a hodgepodge of names in there, and triples in general weren’t at the level I thought. 1940 had 10 major leaguers with 13+ triples, but 1941 only had five. I continued to see evidence of players, particularly outfielders, switching positions routinely. Enos Slaughter was often part of the same outfield as Musial, and led the NL in triples in both left and right. Charlie Keller was among the league leaders at both of those positions. Jeff Heath led the AL in triples in 1938 as a left fielder and hit 20 in 1941 as a right fielder. If the selection of where outfielders played wasn’t as thought out, no wonder we’re not seeing differences in the triple statistics.
I looked at those 1968 and 1969 seasons to see who was behind right fielders’ winning them. Roberto Clemente was the right fielder with the most triples and was in double-digits both years. He only hit 12 in ‘69, but that led MLB. Second was Pete Rose with 11, and Rose played 101 games in right that year (and 56 games in center! Yikes! Well, maybe he had the speed, if not the arm. Michael Sokolove’s book will make you think he did. Baseball Reference has his “fielding runs above average” in center in ‘69 as -5).
CF
Center fielders have hit the most triples in 64 of the last 99 seasons (including shared titles). They were first eight times through 1951 and had their first three-year streak starting in 1952, which actually lasted five years. (You will notice that they had the most triples in 1954, despite no help from the White Sox and Orioles). Their only three-year-plus losing streak since a drought from 1935 - 1943 was that weird period from ‘63 - ‘66 when third base and shortstop ended up on top three times. After the three titles in five years for shortstop, center fielders have won 16 straight titles, which takes us to the present time.
So my research has really ended up confirming conventional wisdom, although the preeminence of center fielder triples was not always so. The lingering question is why they hit the most triples, or at least, why exactly? How much is speed, and how much power?
The one thing we know now is that a center fielder is never but never slow. Some speed is a prerequisite. I don’t know if that is true for another position. Power may also play a role; I remember going down second basemen in the ‘80s and ‘90s, and it seemed like you never saw a guy with less than 30 steals (my handy positional data can now be used to check all of this). But second basemen did not hit triples the way that center fielders did. If the claim that the triple is the most exciting play in baseball does not rest on the prowess for hitting them being multidimensional, it is at least buttressed by it. We enjoy not only what excites us, but what we think at bottom is impressive, and the more multidimensional something is, the more impressive we think it.