Endangered Species: The Three-Base Hitby John Walsh
May 10, 2006
The triple is disappearing from major league baseball. Not suddenly or catastrophically, but slowly, inexorably. In 2005, the average team hit fewer than 30 triples during the course of the season, the lowest number in the history of the game. Here is a plot of the average number of triples per team over the last 100 years or so. I've scaled all seasons to 162 games to put each year on equal footing. You can see that since about 1930 triples have been declining.
So, what's happening? Well, one obvious reason for the decline in triples is the increase in home runs. About half of all triples come off fly balls, and presumably a lot more of those fly balls are clearing the wall in 2005 than in 1915. So the decrease in triples is tied to the increase in homers. Much has been written about the increase in home runs through the years, that's not a subject I'm going to tackle in a comprehensive way, but let's just examine for a moment one aspect: ballpark dimensions.
Consider Forbes Field, home of the Pittsburgh Pirates, in 1912. The distance to the fence down the left-field line was 360 feet, to center field it was 442 feet and the right-field line distance was 376. The left- and right-field gaps were both more than 400 feet from home plate. This is one huge playground, readers. Since outfielders in this era generally played shallow (by today's standards), they had a long way to run for balls that got past them. This resulted in many more triples, even for batters that were not all that fast.
Other parks had similar dimensions, although Forbes Field was one of the bigger ones. Many parks had one short fence but were compensated with large dimensions in other parts of the park. For example, Griffith Stadium in Washington had a short right-field fence (328 ft), but it was 407 ft. (!) down the line in left and 421 in center.
Of course, it isn't just smaller ballparks that have contributed to the near demise of the triple over the years. Another important factor is improved fielding. Today's outfielders are generally bigger, stronger and faster than the old timers. They run down balls faster, and they rifle them back to the infield more quickly. Another important aspect is that in today's game, when runs scored per game are historically high and there is a large emphasis on the home run, there is not that much additional value in getting to third base, compared to standing on second (especially with two outs). It makes little sense risking getting thrown out at third when there's a good chance you'll be driven in from second.
It's amusing (sort of) to compare the triples-hitting ability of modern players with players of the distant (or not so distant) past. Eddie Collins was a speedy second baseman who played about a billion games from 1906 to 1930 and amassed a total of 187 triples in his Hall of Fame career. Joe Morgan, also a Hall of Famer, was a speedy second-sacker from the modern era who played about the same number of games. His triples total, 96, was about half that of Collins' total. Jimmy Foxx was a slugging first baseman in the '20s and '30s who hit 124 triples in his career. Jeff Bagwell was also a slugging first baseman, but he was also was more of base-stealer than Foxx. His triples total: 32, about a quarter of Foxx's total. Darin Erstad is a pretty fleet center fielder who has hit 29 triples in 5,600 plate appearances. Walter Johnson was a pitcher who in half the plate appearances hit 50% more triples: 41 three-baggers for the Big Train.
Smoky Burgess was a short squat catcher who became a pinch hitter when he got too fat to catch. In 18 seasons (1949-1967) he stole only 13 bases. Still, in his career he hit more triples than the following players (despite logging fewer plate appearances): Don Mattingly, Eric Davis, Gary Sheffield, Alex Rodriguez, Chipper Jones, Andruw Jones, etc. Ok, I'll stop here before Jayson Stark writes me a nasty note for copying his shtick. But you get the picture.
Triples Hitters of Today (Such as They Are)
At the start of this season, Steve Finley was the only active player to have hit 100 triples in his career. Kenny Lofton has joined him in the 100 Club with four triples already this year. Next in line is Johnny Damon with 80 and Barry Bonds with 77, but I have a feeling Barry will not be adding much to that total before he retires.
The most prolific triples hitter of the post-war era was Roberto Clemente with 166. It's curious how the guy with the arm who cut so many runners down at third base was so adept at getting there himself. Other three-bag experts from our times: Willie Wilson (147), Lou Brock (141) and Willie Mays (140). If we consider triples per plate appearance, the top modern guys are (min 5000 PA): Lance Johnson, Wilson, Clemente, Omar Moreno, Juan Samuel and Vince Coleman. Aside from Clemente (and to a lesser degree Samuel), these guys are all about speed and not so much about power.
The Good Old Days
You probably know that the all-time career leader in triples is Sam Crawford, whose Hall of Fame career, from 1899 to 1917, pretty much spanned the Dead Ball Era in its entirety. Although Crawford was fast (366 career stolen bases), he was known primarily as a slugger and indeed was perennially among the league leaders in slugging percentage, home runs and RBIs, as well as, naturally, triples. Crawford, a strong, left-handed batter, banged out 309 triples, edging out his teammate and archrival, Ty Cobb, by just 14.
I think "Wahoo" Sam Crawford is my favorite old-time ball player. My admiration for Crawford stems from my reading of the fabulous book The Glory of Their Times by Lawrence Ritter. Back in the '60s Ritter tracked down a bunch of old ballplayers and interviewed them for his book. Crawford, who was in his 80s at the time he talked to Ritter, comes across as a thoughtful, articulate and intelligent man. I'd recommend The Glory of Their Times to anybody who has even the slightest interest in baseball history. Actually, even if you don't care a whit about baseball history, you probably will after reading the book. Great stuff and Crawford's chapter is the best of the lot.
The single-season leader in triples is not quite so famous as Crawford: his name is Owen "Chief" Wilson, who legged out 36 triples for the Pittsburgh Pirates in 1912. Not only is that a major league record, it's a record for all professional baseball—no minor leaguer has ever equaled it.
I always thought Chief Wilson was so named because of a Native American heritage; however, it turns out that Wilson was a Texan with no Native American blood and was given the moniker "Chief" by Pittsburgh manager Fred Clarke, who said Wilson looked like a "Chief of the Texas Rangers."
Like Crawford, Wilson was a left-handed power hitter with decent speed, but he wasn't a base-stealing specialist. He finished among the league leaders in home runs, total bases and RBIs a few times, but never in stolen bases. Over his career he was a pretty good triples hitter, but his 1912 season does look a little flukey. Here are his triple totals during his peak years: 12, 13, 13, 36, 14, 12, 6. Hmm, steroids? The pattern in those numbers resembles this sequence: 13, 12, 16, 50, 18, 18, 24. (Trivia question: what are these numbers? E-mail me with the answer. Winners will receive a free complimentary subscription to http://www.hardballtimes.com.)
Wilson was undoubtedly helped by his home park, Forbes Field. We have already seen that Forbes Field had very large dimensions and in fact, Wilson hit 24 of his 36 triples at home in his record year. For his accomplishment, Wilson was elected to The Spalding Base Ball Guide's "Hall of Fame", which was a yearly distinction given out to the best fielders at each position, the batting champion, etc. It's curious that the Spalding Guide made no mention of Wilson's triples total being a major league record. One possible reason for this is that there was some confusion about the existing record holder. Some sources credited Nap Lajoie with 43 triples in 1903, although the Big Frenchman actually hit only 11 triples that year.
Anatomy of the Triple
Who hits triples and where do they go? You might be tempted to answer "left-handed batters and right-field," but things aren't quite so simple. Lefty swingers do have an advantage in hitting triples. This makes sense since a hard hit ball by a lefty has a greater chance to go to deep right field. The very long throw to third base results in more triples. Here's the breakdown for triples from left- and right-handed batters. (The data are for the seasons 2003-2005, courtesy of Retrosheet, of course.)
+-------------+--------+---------+--------+ | batter_hand | PAs | Triples | 3B/PA | +-------------+--------+---------+--------+ | Left | 239752 | 1393 | 0.0058 | | Right | 322528 | 1327 | 0.0041 | +-------------+--------+---------+--------+So, per plate appearance, lefties hit about 40% more triples than righties.
Next, let's look at where the triples go. I don't have detailed hit location information, but I know who fielded each triple. Here's the list:
+------------+----------+ | Fielder | Triples | +------------+----------+ | 8 | 1239 | | 9 | 1166 | | 7 | 303 | | 4 | 7 | | 3 | 1 | | 1 | 1 | +------------+----------+Looking at the outfield positions, we see that many more triples are hit to right field than are hit to left field, just as we expected. I didn't expect to see center field as the most likely place for a triple to be hit though. Actually, some fraction of those "center field" triples were probably fielded in right-center or even in right field (if the right fielder fell down on the play, for example). It's also interesting to see that seven triples were fielded by the second baseman and one by the first baseman, I suppose those were pop flies in short right field. But, what about the triple fielded by the pitcher; what the heck was that? That had me stumped for a while until I discovered that it came from an inteference play: Luis Terrero flared one over the mound and Duaner Sanchez, not too intelligently, threw his glove at it and Terrero was awarded a triple. That's a play that happened all the time when I was 10 years old playing in pick-up games, but I never heard of it happening in professional ball.
A few words about ballparks. Ballpark effects for triples tend to be large. Historically, many parks were quite asymmetric and those with large pastures in right and right-center field usually yielded lots of triples. I've already mentioned the large dimensions of Forbes Field. In fact, if you look at the all-time leaders in triples, it almost reads like a Who's Who of Pirate hitters: Honus Wagner, Paul Waner, Fred Clarke, Tommy Leach, Pie Traynor and Clemente are all among the top 30 triples hitters of all-time. Actually, the majority of the careers of Clarke and Leach took place before Forbes Field opened in 1909, but the Pirates' previous home, Exposition Park, also had very large outfield dimensions. Getting back to Forbes Field, the aforementioned Chief Wilson was with the Pirates in 1912 when he set the single-season record.
In more modern times, the biggest aid to hitting triples is artificial turf. As everybody knows, articifial turf plays much faster than natural grass and balls skipping between the outfielders or, indeed, bouncing over their heads, often result in triples. And teams that play on turf tend to employ speedy outfielders who, when it's their turn to bat, usually get their share of triples.
This turf effect is clearly seen in the graph above, where there is a bit of a resurgence of the triple in the 1970s. The increase in three-base hits in this period corresponds to the building of several new stadiums that had artificial turf: Three Rivers, Riverfront, Kauffman, Veterans, Busch and the Kingdome all opened for business in the '70s. Additionally, artificial turf was installed in Candlestick Park in 1970 (and removed in 1978).
Several of the best triples hitters of modern times played their home games on turf, at least for a significant portion of their career: Willie Wilson, George Brett, Lou Brock, Juan Samuel and Vince Coleman, among others.
The Curious Case of Mr. Henderson
If you think about who should be among the top triples hitters of our time, you will likely come up with the name of Rickey Henderson at some point. The greatest base stealer of all-time and a guy with decent power, plus a 22-year career, most of it batting leadoff, and you figure you'd have one of the top triples hitters of the era.
Well, you'd be wrong. Henderson hit a mere 66 triples in his more than 13,000 plate appearances. He batted righty so that will take away some triples, but that doesn't fully explain his low total. Just how low is that 66 triples in 13,000 PAs? Looking at other right-handed batters with at least 1,000 plate appearances since 1970, Rickey is more or less average in terms of triples per plate appearance. Some other players who hit more triples per plate appearance than the greatest base stealer in history: Thurmon Munson, Danny Cater, Carney Lansford, Ray Knight, Jeff Kent, Jerry Grote, Tim Hulett, Lou Merloni, etc. You get the picture—Henderson was not a good triples hitter and considering his natural speed, he has to be considered a terrible one.
Just for fun, I tried to create a formula for predicting the number of triples a player should hit. I did a (laughingly) simple linear regression, using only at-bats and stolen bases as the input variables, on the data sample mentioned above: all batters with at least 1,000 plate appearances since 1970. I did the regression separately for lefty and righty batters (I left switch-hitters out). The resulting r-square value was around 0.86, not too bad. Note, this is an overly simple model. I didn't take into account park factors, especially playing surface, and I probably left out other variables that likely have a big effect on triples rate. This is just a quick-and-dirty exercise to see if Rickey's triples total really is a lot smaller than expected. The results are shown in the graphic on the right.
The horizontal axis shows the number of triples predicted for the batter, while the vertical axis shows the actual number of triples hit. Points above the diagonal line show players who exceeded their prediction, while anything below the line represents a player who hit fewer triples than predicted by this simple model. The player who most undershot his prediction was Rickey Henderson, who you can see way over on the right side of the plot. The model predicted 130 triples for Rickey, or about twice the number he actually had.
Why? I don't know. Power was not even included in the regression and given Henderson's above-average power for a speedster, the model might actually be under-predicting Henderson's expected triples rate. The regression was done separately for left- and right-handers, so Rickey's right-handedness does not explain the shortfall.
What about park effects? Ah, now we're getting somewhere. I estimated a triples park factor for Henderson, based on the first 15 years of his career, which were spent in Oakland and New York. Those parks were not good for hitting triples. For example, in the period 1979-1984, corresponding to Henderson's first stint with Oakland, the Athletics hit 241 triples on the road and 164 at home. From, 1985-1990 the Yankees hit 151 triples on the road, compared to 119 at home. And, back in Oakland from 1990-1995, the A's hit 177 triples away and 128 at home. A rough estimate from these numbers puts Henderson's triples factor at about 0.8. In other words, he'd have hit 20% more triples had he played in neutral (triples) parks. That would give him about 79 career triples, still far short of the 130 predicted.There are additional reasons for Rickey's low triples total. He was a strict right-handed pull hitter, and we've already seen that very few triples are hit to left field, regardless of the speed of the batter. Rickey also was not particularly quick getting out of the batter's box; he hit from that low crouch and took a healthy cut and that cost him him a little in getting down the line to first. This is also reflected in his career doubles mark, which is lower than you'd expect from a speedster with gap power who played for so many years. (He's 37th on the all-time list.)
Whether these effects, taken together, fully explain Henderson's low triples total, I really do not know. However, it's a curious fact that Henderson's teammate in Oakland, Jose Canseco, also hit very few triples: 14 in his career against the prediction of 44. Maybe there was something in the water there in Oakland.
Save the Triple
The three-base hit is one of the most exciting plays to happen on baseball diamond. The legendary Luke Gofannon, the greatest hitter in the history of the game, said it best. Luke, asked by his lover Angela Trust, what it was he loved most in the world, answered simply, "Triples." Angela, crushed at not hearing her own name, managed to hide her disappointment. She said to Luke, "I don't understand. What about home runs?" Then Luke explained what he loved about triples:
"Well," he said in his slow way, "smackin' it, first off. Off the wall, up the alley, down the line, however it goes, it goes with that there crack. Then runnin' like blazes. 'Round first and into second, and the coach down there cryin' out to ya', `Keep comin'.' So ya' make the turn at second, and ya' head for third—and now ya' know that the throw is comin', ya' know it is right on your tail. So ya' slide. Two hundred and seventy feet of runnin' behind ya', and with all that there momentum, ya' hit it — whack, into the bag. Over he goes. Legs. Arms. Dust. Hell, ya' might be in a tornado, Angela. Then ya' hear the ump—'Safe!' And y're in there."So, let's move those outfield fences back. Let's deaden the ball and bring back the thick-handled bat. Let's get those outfielders playing shallow again. Friends, let's save the triple!
— Philip Roth, The Great American Novel
References and Resources
- Retrosheet, as always, is the source of play-by-play data. I also found the data used to estimate Rickey Henderson's triples park factor at the Retrosheet site.
- The Glory of Their Times by Lawrence Ritter is not only one of my favorite baseball books of all time, it's one of my favorite books period. And it cost less than ten bucks.
- Excellent articles on Chief Wilson can be found here and here.
- The Great American Novel by Philip Roth. Maybe it's not the great American novel, but it's a great read nevertheless. Many references to real events in baseball (and American) history. Good stuff.
- A selection of Spalding's Baseball Guides from the period 1889-1939 is available online. An invaluable tool for baseball research.
- Many thanks to Steve Treder, who provided valuable information on this subject.
John Walsh dabbles in baseball analysis in his spare time. He welcomes questions and comments via e-mail.