The trouble with the defensive record is that it doesn’t directly answer the question of who the best fielders are. The trouble with advanced defensive statistics is that even the best ones bounce around a lot each year.
So when I see a player like Jimmy Sheckard, Bobby Veach, Sam Rice, or Harry Hooper jump from under the line to over it in my personal rankings due to high-end defensive performance, I have to ask some questions. Usually, big-time defensive seasons at the throwing positions in the infield get all the press. Sometimes catchers get pixels for gunning down a lot of runners. Centerfielders get plenty of webgems. But only rarely do corner outfielders get much love. So when they get a lot of love, I’m curious.
Life’s full of uncertainty. I’ll never have perfect knowledge of anything, so I want just enough certainty that when the time comes, I can defend a vote for Sheckard and company. Or neg them. It’s easy to defend a non-vote on anyone outside, say, the top-ten at any position; it’s feeling good about the case for a guy that’s tough.
THE STATE OF (BALLS IN) PLAY
So let me step back and ask a big question. Is there reason to believe that olde tyme outfielders could have racked up more defensive value than modern outfielders? Why yes, yes there is.
One of the massive, quiet sea changes across baseball history is the relentless trend toward fewer balls being hit into the field of play. Understanding the scope of that difference makes it clear that our modern view of an outfielder’s value may not hold throughout history.
Here’s the Balls in Play* (BIP) per 9 innings every tenth year from 1873 to this year:
*BIP means that a fieldeable ball was hit, so a quick-‘n-dirty estimate is:
(outs – strikeouts) + ( hits – home runs) + errors
minus (caught stealing + pickoffs + double plays)
YEAR NL/NA AL/AA ------------------- 1873 22.6 -- 1883 18.6 18.8 1893 17.9 -- 1903 16.6 16.1 1913 15.8 15.6 1923 16.8 16.6 1933 16.4 16.4 1943 16.0 15.6 1953 15.3 15.4 1963 14.2 14.2 1973 14.5 14.7 1983 14.4 14.9 1993 14.4 14.4 2003 13.9 14.2 2013 13.3 13.3
While the difference between 13.3 BIP (2013 NL) and 16.6 BIP (1903 NL) might seem insignificant (what’s 3.3 BIPs between pals?) it adds up to 500+ additional balls in play per team per year. Over the course of a career, Jimmy Sheckard’s teams contended with something like 8,000 more BIP than today’s left fielders do.
Kinda hard to get your head around. None of us writing or reading or ignoring this article has any frame of reference for the vast differences in the game. Here’s a few reasons why things were so different for all fielders back then, and especially outfielders.
1.) Fewer strikeouts. If they don’t K, it’s got to be in play. In Sheckard’s day, the NL struck out in ten percent of its at-bats; today they strikeout twenty-two percent of the time. This is absolutely huge.
2.) Slower fastballs. These first two points are clearly related, though the second speaks primarily to leftfielders. Today, the distribution of putouts in the outfield hovers around 30:40:30 from left field to right field. Research by Michael Humphreys (creator of DRA) indicates that until somewhere in the 1940s or 1950s, the distribution went more like 35:40:25. Once in a while leftfielders would even get as many putouts as centerfielders. In other words, left fielders were like a second center fielder. But as strikeout rates increased so did the demand for strikeout pitchers. Who gets the most strikeouts? Flame throwers. Today, on average, pitchers throw a lot harder than they did thirty years ago when a mere ninety-MPH fastball made scouts drool. What happens when batters are late on a fastball? They hit behind it. According to Fangraphs, the most playable fly balls today are hit to the opposite field (hard-hit balls tend to be pulled). Sixty-five to seventy percent of batters are right handed, so putouts have trickled away from leftfielders.
3.) More errors. For any given left fielder, errors by his teammates create additional opportunities for the opposition to put the ball in play to him. Errors have decreased dramatically over time. For perspective, in 2013, the leaguewide fielding percentage is .985. In 1903 it was .946. For Sheckard’s 1903 Brooklyn team, which fielded 5745 balls, that difference in fielding percentage represents 224 more errors than a 2013 team would make.
4.) Fewer home runs. Deadball chicks didn’t dig the long ball. At all. Until they did.
5.) Fewer double plays. Today, NL teams turn a double play in about 11% of all opportunities. In 1903, they turned two only about 7% of the time. Why have double plays increased? Because as the lively ball of 1920s took hold, teams no longer needed to play for one run. Bye-bye bunt. See ya steals. Hello, DP.
6.) More aggressive base running. For leftfielders with good arms, like Sheckard, this meant more chances to gun down the over-aggressive and save runs. Today, runners rarely advance extra bases on a ball hit in front of them, but that might not have been the case in the teeth of the deadball era, when one run meant much more. We don’t yet have play-by-play accounts to give us specific information, but the number of chances runners took was probably outrageous to our jaded modern eyes, though smallish compared to the changes in strikeouts and errors.
7.) More one-run tactics. Related to all of this, of course, is the general prevalence of one-run tactics. Between all the sac bunts and caught stealings, a 1903 outfielder may well have given back to his 2013 counterpart roughly the BIP he took from them in home runs.
As we noted above, this works out to a heckuva lot of extra balls in play per deadball team. Not all would go to the outfield, of course, and not all those balls in the outfield went to the corners, nor were those in the corners always reachable for a putout for any defender. Still, with such a massive, yet barely perceptible, change I find it difficult to imagine how much more important defensive skills were back then in building a winning ball club.
SLOW BREAK DEFENSE
An analogy from basketball helps explain the impact of all these extra BIP on individual players and teams. Even in pro hoops, where the range of talent is highest and narrowest, dominant teams regularly win 80 percent of their games—a sixty-six win season. The heat did it last year. In 2008–2009, the Cavs did it. The year before that, the Celts. The year before that, the Mavs. In 1999–2000, it was the Lakers. In baseball, only fourteen MLB teams have racked up a winning percentage of .700 or higher in more than 230 seasons of 120 or more games. Three NBA teams did it last year. As baseball analyst Tom Tango has explained (warning: math) with so many possessions (trials) in a basketball game, even relatively small differences in talent add up quickly, and the most talented teams rarely lose to the least. In the NBA, your best player might get twenty to thirty looks at the basket; in baseball he might get five at-bats or eight fielding plays. With 250 possessions in a game, basketball is much more extreme. But when more plays are available to the best fielders, they have more chances to separate from their less talented peers. Sheckard’s 1906 Cubs are one of the best defensive teams of all time. Every extra BIP was another opportunity for their Gold Glove-level talents—Tinker, Evers, Chance, and Sheckard among others—to record an out that their less talented opponents couldn’t.
So to circle back to our original question (finally!), can we believe that olde tyme outfielders racked up more value than their modern peers? Yes. In their own era, they no doubt had a much bigger influence as defenders than a corner outfielder does today. The higher number of BIP is convincing.
And can we believe a stat like DRA when it suggests that Sam Rice was worth 150 runs on defense (about the same as Ozzie Smith)? Or that Jimmy Sheckard was worth more than 200+ (more than Brooks Robinson)? Yes, we can. With more opportunities come more chances to accrue value.
But we also have to note that the range of talent was wider than in today’s game, so it was also easier to stand out. The combination of being easier to dominate and having more opportunity to do so is both earned and contextual. All of it has to be considered.
I have half the certainty I’m looking for. But the matter of the range of talent has to be addressed, and it will wait until another day.