According to my research, the first major league trade involving only players and not cash was no Frisch-for-Hornsby affair but an Elmer Cleveland for Arthur Whitney swap in June, 1888. Neither of these players would have been worth much if merely sold. If Cleveland had his 15 minutes, it was in the comfort of the 1884 Union Association, the league that Fred Dunlap had pulverized to the tune of .412. Whitney, for his part, was holding out at the time of the trade, and probably was not greatly missed. But, in reviewing these players, I noticed a fun statistic in Whitney’s record: in that season of 1888, he would double only one time in 328 at-bats.
I’m not at the point that I can mentally calibrate for 1888, but putting this into Stathead, I found that it made for true trivia: in all the years of major league baseball, there have been only five players to hit exactly one double in a season with at least 300 at-bats. Whitney had the 3rd-most at-bats. Here is the story on each of them, if my flat, statistical stories can be deemed such.
Most At-Bats in a Season with Just 1 2B
Bill Bergen, 346 AB, 1909
Well, this is the guy, the catcher, renowned for being the very worst-hitting position player in all of major league baseball history. 1909 was his worst year: he hit .139. He did keep his OPS+ in positive territory: it was 1 (that’s 001 to you). He had a career OPS+ of 21. So right away, seeing Bergen at the top made me go “huh.” I had regarded doubles as an oddity of the record, maybe the oddity, and didn’t think hitting them really meant that you were good, or that not knowing them personally made you bad. But when Bill Bergen leads off a list, it may be a list of very bad hitters.
Curt Welch, 331 AB, 1892
While he may not have been a great player (career 20.2 WAR in 1,107 games, 108 OPS+), Welch had some special skills, principally in his play of center field. It would probably be accurate to characterize him as a star (judging by the Baseball Reference Bullpen entry). He excelled in getting hit by pitches, leading his league in that three times. Everyone stole bases, but he racked up 89 and 95 in consecutive seasons, and the 95 was good for 3rd in the American Association in 1888. But here’s the really shocking thing: Welch once led his league in doubles — with 39 in 1889!
In 1892, he was just 30, but obviously a far cry from the player he had been. He began the year with the Orioles, who had transferred over to the National League from the defunct American Association. The Orioles had a miserable season, going 46-101, but Welch was not in Baltimore by the end of it, getting released on July 13. The Reds picked him up, but then they released him, too, in September. He might have fallen prey to the reduced patience with veterans; while his batting average was just .227, the league hit only .245. Welch was still second in the league in hit by pitches, despite being more than 300 plate appearances behind the plate appearance league leader in the National League’s first 154-game season (or thereabouts). He had a .345 OP and an 87 OPS+.
He only lasted for 14 games in 1893 and died on August 29, 1896, apparently of tuberculosis. B-R Bullpen says that his play in minor league stints post-1892 was compromised by consumption.
So now things come into focus, or at least a narrative does. Did this one-time doubles leader fall off so badly in 1892 because he was sick? We are probably guilty of carrying the Lou Gehrig example too much in our minds, but one wonders.
Arthur Whitney, 328 AB, 1888
The most interesting things about Whitney appear to be, first, his advocacy of fielding gloves (see my Substack Note) and second, the business conflict that prompted his trade, but this is not the place for those. He was a third baseman who compiled 4,001 career plate appearances and was 30 years old in 1888. What I want to note about his play is that he fell off the table in 1888. He had a 59 OPS+ in 1888 and, in four subsequent seasons with at least 90 games played, did even worse in all of them. In 368 games prior to coming to the Giants (which is where he was traded in 1888), he’d had an 87 OPS+. So his struggles weren’t just confined to doubles, and they maybe underscored a larger issue with him.
I should also note that, while Whitney only hit one double in 1888, he did hit 4 triples. This points out the fundamental problem of rating and classifying with a hierarchical statistic not at the top of the pyramid, and is also the problem with “cycles.” The weakness in the skill that doubles may represent is best seen by taking doubles and triples together, if one is taking this super seriously1. We don’t want to be penalizing players because they were fast, for instance.
Jack Burdock, 325 AB, 1888
Yes, Burdock matched Whitney single for single, and out for out, in 1888. This was a year truly worthy of Bill Bergen. A .142 batting average and an 11 OPS+. Just 2 triple and a home run to go along with the one double. A WAR of -2.0 during his time in the American Association, only 0.1 ahead of Harry Lyons, who played 53 more games, or he would have been worst in the league.
What went on here? Burdock got most of his action with 2nd-place, 88-52, Brooklyn, and so I am reminded of Joe Gerhardt, 1885 (a 29 OPS+ for the 85-27, 2nd-place Giants). Both players were second basemen, too. Burdock’s SABR Bio quotes an article from this season whose writer was convinced that his poor hitting couldn’t possibly outweigh the benefits of his defense (although BR Fielding Runs doesn’t even buy the good defense, either for this season, when it was 0, or really for his career, when it was +42). One is also reminded of Bill James’s incredulous reaction to The Sporting News article claiming that “Wes Parker saved a hit a game with his glove.” James, Parker evidence2
But I did look behind door number two, and even by the standards of 19th-century baseball players, Burdock was apparently a drinker with few peers. Kathy Torres’s sharp SABR Bio of him includes the following.
“Sadly, Burdock’s considerable skill and athleticism were ultimately undermined by alcoholism.”
“[Alcoholism] ultimately destroyed his career and permanently damaged his family relationships.”
He was “hit by suspensions for his drinking.”
She also documents that Burdock’s drinking was reported by the press.
I want to give the proper impression of Burdock, though. He had real longevity, not only living to 79, but playing 18 years in the big leagues. His career batting average was .250, and his career OPS+, 86. Except for 3 more games with Brooklyn in 1891, his disastrous 1888 season was his last. He had given some warning in the doubles department the year before, managing just 6 in 237 AB, but he had hit .257, although with just a 70 OPS+.
I was struck by the contrast in his 1882 and 1883 seasons. Playing enough to be a rate qualifier in both, his batting average jumped from .238 to .330, and his OPS+, from 79 to 145. Even by the standards of career years, Burdock’s 1883 stands out. Might he have cut down on the booze then?
The SABR Bio doesn’t give this impression. It more gives the impression that his 1883 success occurred in spite of his drinking.
In any event, it is obvious what I’m driving at with Burdock. His one double in 1888 wasn’t just a statistical oddity. He hit so poorly, the 3-year-old Ring Lardner was probably composing ditties about it, and he didn’t just happen to hit poorly, he was drinking. His is another exceptional case on this list.
Tommy McMillan, 322 AB, 1910
McMillan was just 22 years old for this year, and perhaps essentially culled from the majors at a young age subsequently, only having occasion to add to his MacMillan Encyclopedia record when he had 41 games with the Highlanders in 1912 (and he hit only 2 doubles in 149 at-bats). He had his fun, though, and made a good living playing ball: He was in the minors until 1929, and racked up over 7,000 minor league plate appearances.
The impression I have is that McMillan got the bat knocked out of his hands. A shortstop, he is listed as having been 5’5”. He hit no home runs in 991 major league at-bats. At least in these extreme cases, doubles do tell you about power. He also had only 6 minor league home runs.
The real delight is that he was a teammate of Bill Bergen’s during their “historic” years. 1909 had gone marginally better for McMillan at the plate than 1910, particularly in the doubles department: he hit 15 of them (20 led the team)3. He hit .212 with a 61 OPS+.
He had a 42 OPS+ in 1910, last of 66 in the NL (Bergen’s OPS+ was 6, but his 299 PA didn’t qualify him). His .183 batting average did beat one player. McMillan was waived by Brooklyn on June 5 (picked up by the 75-79 Reds), and the man who replaced him as the Superbas shortstop, Tony Smith, was the one who did worse, finishing the year at .181. (This was really Mac Jones and Bailey Zappe; I have the sense the Superbas knew what they had in both of them, maybe delaying going to Smith the same way the Pats delayed going to Zappe.)
Most At-Bats in a Season with 0 2B
Arthur Whitney’s 1888 stat line was my departure point, but one could argue that a list of players with the most at-bats who didn’t hit any doubles would better capture the spirit of doubles futility. The truth of that depends on just how many at-bats those players had. Without pulling the data, there was no way to know if this stricture was too demanding to capture any players except those who had merely a cup of coffee and might not be of much interest. The same criticism could of course be levied at the approach when it is predicated on 1-double players. Regardless, this is all silliness. These are valid examples of low-double players, not serious attempts at generating the most remarkable seasons in that regard. Just like a child, a man must be free to explore, and always enlisting the most comprehensive method can be stultifying.
I’ll go through the five 0-double players with the most at-bats in the same way I went through their 1-double counterparts.
Rafael Belliard, 286 AB, 1988
It was nice to see a more recent name….I was both astonished that a no-doubles season of this significance was possible so far out from the dead-ball era, and surprised that it did not make a big enough impression on me at the time that I was able to remember it, even 35 years later.
Belliard may not need an introduction. I always lodged him in my brain because my Mets had their own light-hitting Rafael who played shortstop and who came up at about the same time, Rafael Santana (Santana was actually quite a bit better hitter, and quite a worse fielder, than Belliard, quickly taking inventory). Belliard was like Tommy McMillan in that he was diminutive and devoid of home run as well as doubles power: he hit just 2 home runs in 2,301 career at-bats. His career OPS+ stacks up with some of the guys I have been bandying about; it was just 46. Another future comparison point is that he hit 13.7 doubles per 575 at-bats in his career. The most doubles he hit was 9, with the ‘91 National League champion Braves, but in his defense, he never even notched 400 plate appearances in a season.
For my money, nothing can detract from 286 at-bats without a double, but like Art Whitney did in his 1-double season, Belliard hit 4 triples in 1988. If Belliard cared one way or the other about his zero on the stat sheet, he probably wanted a double, but we can’t give him credit for selflessness the way we can Gunnar Henderson for running out a double when a single would have given him a cycle last season: Belliard’s last triple came on just July 21. Four triples and zero doubles was an atypical ratio for Belliard: for his career, his numbers were 14 and 55. However, through 1988, in 843 AB, he had 9 triples and 9 doubles; in his last 1,458 AB, his ratio was 5/46.
That at-bat breakdown points out that Belliard was somewhat unusual in that his historic season didn’t come near the end of his career. He was 26 years old in 1988, and would play to 36. He was just a guy who couldn’t hit, and he might have had a slanted season at any time. Even Bill Bergen was on about the 15th hole when he had his 1-2B season.
It might be worthwhile to take a step back and ask why a lack of doubles power and home run power go together in these extreme cases. Exit velocity must be the key. You’d think anyone could get lucky in terms of hitting balls down the line that then went for doubles, no matter how little power he had. But that is only one species of double.
Herman Pitz, 284 AB, 1890
If researchers somehow unearth 3 more double-less at-bats for Pitz from this season, I’ll be bummed, because no one will be interested if he’s the answer to the trivia question, but he actually has an interesting profile. To start with the not interesting but helpful stuff: he played for Brooklyn and Syracuse, two American Association clubs, in 1890, and that was his only major league season. The streak of small guys continues: he is listed as having been 5’6”. Pitz not only didn’t hit a double in 1890, he didn’t hit a triple or a home run, either. His batting average was .165.
However, his on-base percentage was .315. He walked 58 times. Weird. Maybe he had the Eddie Gaedel approach. Who knows?
That really was a high rate of walks. He had the 17th-most in the American Association in 346 plate appearances. Of the 16 guys with more walks, only one had fewer than 500 plate appearances. He didn’t walk as often as Jim McTamany, who had a 16-walk edge on everyone else, but he did walk more than Jack Crooks, who had a 16-walk edge on the third-place walker.
Pitz also was exceptional as a majority catcher who stole bases. His 39 steals ranked him only 28th in the AA, but it was the highest total for a majority catcher in the league.
Baseball Reference has statistics for Pitz for six other seasons, sometimes described as minor league, sometimes as “other league” (the Pennsylvania State League being an example of the latter). Unfortunately, they don’t have his walk totals, but the exceptional lack of power is not evident. He had 52 2B, 12 3B, and 4 home runs in the 961 at-bats on record. This encompassed seasons after he left the American Association, where, you remember, he had no extra-base hits.
One reason Herman Pitz would not have been a good answer (or even necessarily a right answer) to the trivia question is he might not have been Herman Pitz at all, as he is another 19th-century player who appears with a question mark after his name on Baseball Reference’s “Player Standard Batting” table.
Tom Barlow, 269 AB, 1873
The National Association was the forerunner of the National League, and this is our first NA player. Don’t let that at-bat total fool you into thinking Barlow was a part-time player; his team, the Brooklyn Atlantics, played just 54 games (yes, this is another Brooklyn player). He was a respectable hitter: .290 for his career (which lasted from 1872-1874, plus 2 games in 1875). The average National Association average from 1872-1874 was .283.
Barlow indicated that he was a mere singles hitter in 1872, only managing one double, and no triples or home runs, in 171 at-bats. This was unusual: only 4 of the 63 players in the NA with 100 or more at-bats had 0 or 1 double, and Barlow had the most at-bats of them.
The intrigue is that Barlow’s BR Bullpen page says, “Barlow is credited by some sources as the most likely creator of the bunt, which was popularized by teammate Dickey Pearce.” Maybe Barlow essentially only hit singles because he was always bunting? Since your bunts were live even if they rolled foul, the play may have had a prominence completely foreign to us. I can envision hitters who were really full-time bunters, the way that later some pitchers only threw knuckleballs.
Dickey Pearce’s statistics also bear the stamp of a guy who was mainly hitting in the infield. Behind Barlow’s 1873, do you know which 1870s player had the season with the most at-bats with 1 double or less? That would be Dickey Pearce, 1874.
It is odd to think of Barlow and Pearce as fellow innovators. Barlow was just 21 in 1873, and Pearce, at 37, perhaps old enough to be his father.4 I have only skimmed Pearce’s SABR Bio, but I gather that Pearce actually was pivotal in developing a whole style of play, while Barlow’s contributions, such as they were, were limited to the bunt. I’ll defer to the judgment expressed in B-R Bullpen, but one would have thought Barlow would have popularized a play Pearce invented, not the other way around, if for no other reason than Pearce started playing ball much before Barlow. I think what the Bullpen entry may be getting at is that Pearce had to be the one to popularize the bunt, even if Barlow was the inventor, because Pearce was the famous one.
Anyhow, returning to super-low doubles, the larger point is that once again there may be a catch here. Not only are the seasons often from likely culprits, they also often signify special circumstances that surrounded players.
Choo-Choo Coleman, 247 AB, 1963
Of course I know the name, but Choo-Choo predates me, so I’ll have to rely on others to explain the grip he had on people. Inasmuch as he achieved individual notoriety competing for and with one of the worst and (at least according to lore) least fundamentally sound teams ever, at first I considered him a highly promising member of this list, but I found I had little to say about him. He hit for good power at times, both with the inaugural ‘62 Mets and in the minor leagues. He wasn’t lilliputian, ill in 1963, or a drunk. He did only hit .178 in 1963. In any event, Uncle Marvin5, now that I’ve mentioned Choo-Choo, whose passing you brought up to me unbidden in 2016, it’s time to subscribe!
Billy Holbert, 244 AB, 1879
Holbert was 24 in his historic season, which he spent primarily with the National League’s Syracuse Stars before the team folded. A catcher, he logged between 176 and 300 plate appearances consecutively for a decade in shorter schedules than we have now.
He is actually the fourth catcher in a row on this list. I don’t know about Holbert, who is listed as having been 197 lbs., and who played most of his career before stolen bases were recorded, but the three previous catchers really were not archetypal, in that they were fast. (For instance, Tom Barlow was a bunter, and led the National Association in steals in 1874. And where do you think the name Choo-Choo comes from?) So, I’m betting first and foremost that catchers are here coincidentally, and second, I’m betting they are here because they are often some of the weakest hitters — not because of Ernie Lombardi syndrome.
Holbert is a fitting member because he compiled more at-bats without a home run, 2,335, than anyone in major league history6 . Belliard’s numbers are highly similar: a 46 career OPS+ to Holbert’s career 47; 2,301 AB and 2 HR to Holbert’s 2,335 and 0. Holbert’s career rate of doubles, 10.1 per 575 AB, was worse than Belliard’s. (Yes, it is absurd to be comparing an 1870s player to a 1980s one, but it’s still fun to see such similar numbers).
Longest Streaks Without a Double, in Season
My attachment to my original design notwithstanding, this is a wider way to look at doubles futility, and probably a fairer one. Here, playing time is actually a virtue. We don’t need to rue those extra at-bats, each a doubles opportunity, that might knock the player off the scoreboard entirely.
Stathead’s “streak finder” only goes back to the beginning of the American League, 1901, so we are forced to remove those 1800s players who so frequently stood out in terms of not having doubles. The tool also required a bit of a workaround, as it won’t directly search for streaks of at-bats or plate appearances (I don’t think). It automatically searches for streaks of games7. However, it provides the complete statistics for the games in the streak. By this means, and by setting the games threshold low enough, I could find all notable streaks of at-bats without a double. The quoted at-bats do ignore those that occurred in the game after the last double before the streak began, as well as those that occurred before the double in the game in which the streak ended (these never amount to much more than about a game’s worth of at-bats, anyway).
Turning to the results….As it turned out, all four of the seasons with just 0 or 1 double that occurred after 1900 also featured a historically significant streak. Bill Bergen, 1909, tops the list again (Or perhaps, I could say, Bergen attained a larger championship; one could think of him as a pennant winner, here beating his fellow pennant winner of the zero-double bracket, Rafael Belliard). Since he had no doubles in 1988, Belliard had to have a notable streak; it was just a matter of how many guys had topped him in seasons where they hit doubles. Only four had, and the year rated 5th. The same logic applies to Choo-Choo Coleman, 1963, who is 18th now that we are using streaks. Tommy McMillan’s 1910, in which he hit 1 double, rates 11th for the streak present in that season.
Oddly, counting the streaks in terms of games and not at-bats, the most notable 0- and 1- double seasons by at-bats basically just migrate over to the same spots on the streak leaderboard. You would think you only needed to consider 0- and 1- double seasons if you didn’t bother to figure the actual number of at-bats in streaks. In terms of games without a double, Bergen’s 1909 is #1, at 104 games, Belliard is #2, at 100 games8, Choo-Choo Coleman is tied for 3rd at 91 games (with a streak of Ichiro’s from the start of 2015, wonderfully), and McMillan is 5th with 89 games. That these players rated higher in their games streak than their at-bats streak tells you they had low at-bats per game, which probably mostly signifies that they were part-time players.
A critical appraisal of the streak lengths does render some of the differences trivial. In other words, the 60th-longest streak naturally gets lost in the shuffle, while the 10th-longest streak seems more notable, but the lengths of those streaks cluster. The 2023 rate was 20 at-bats for every double (actually, 19.98), so how different are streaks differing by 20 at-bats, really? This needs to be kept in mind. Four streaks really stand out by number of at-bats.
Bill Bergen, 1909, 334 AB
Pat Newnam, 1910, 322 AB
Al Bridwell, 1907, 313 AB
Jimmy Slagle, 1906, 293 AB
Belliard’s perfect season was 286 at-bats, and any season with no doubles really deserves a bonus because the true length of the streak is unknown. (It is at least the number of at-bats compiled.) This provides a justification for allowing streaks that span more than one season. (Maybe my prejudice against them is benighted?) And, to be clear, the distinction is not between zero-double and some-double seasons, but between seasons ending with the streak intact and those in which it had already been put to bed by season’s end. Jimmy Slagle’s 19069 is the only other one of those above whose streak was alive at season’s end.
To show how much more dense the territory gets, #18, Coleman, and #33, Roy Thomas (1904) range from 247 to 238 at-bats. So, the average at-bat total in that range had 1.6 players. Continuing, there were at least 23 guys with streaks of between 200 and 209 at-bats10.
Studying an expansive list, rather than just focusing on the top-five players as I did before, the phenomenon that was Belliard’s season stood out more. Of the top 10 here, eight of them comprised the 1906 to 1910 seasons. Just because 1906-1910 was the peak period, it doesn’t necessarily hold that the farther we travel from 1910, the less likely double outages were, but it’s a hypothesis that basically holds up, as we will see.
Belliard, however, was joined by another comparatively recent player in the top 10, or at least one whom many serious fans will see as more than just a name in a record book. Pete Runnels did not hit a double from April 16 through August 12 of 1963, a span totaling 275 at-bats. He was finishing up for the Houston Colt .45s in his native Texas after five straight seasons of hitting over .300 for the Red Sox. In fact, in just the season before he joined the Colt .45s, Runnels had won the batting title11. While Houston presumably expected much more of Runnels, it would be exaggerating to say that his overall decline merits case study: he was 35 years old and recorded a .628 OPS in a .669 league. But it would be fair to say that this doubles drought did signify he’d reached the end of the road. His career ended when he was released by the Colt .45s the following May 19.
He followed his August 12 double with ones in his next two games. So he went 88 games without a double, then doubled in three straight games.
Below are the longest streaks by decade.
1900s: Bill Bergen, Superbas, 1909, 334 AB
1910s: Pat Newnam, Browns, 1910, 322 AB
1920s: Whitey Witt, Yankees, 1922, 245 AB
1930s: Jo-Jo Moore, Giants, 1933, 241 AB
1940s: Eddie Mayo, Athletics, 1943, 252 AB
1950s: Andy Pafko, Braves, 1955, 232 AB
1960s: Pete Runnels, Colt .45s, 1963, 275 AB
1970s: Wayne Garrett, Mets, 1974, 267 AB
1980s: Rafael Belliard, Pirates, 1988, 286 AB
1990s: Eric Yelding, Astros, 1990, 264 AB
2000s: Edgar Renteria, Tigers, 2008, 215 AB
2010s: Jean Segura, Brewers, 2015, 239 AB
2020s: Carlos Santana, Royals, 2021, 189 AB
Rather than a mark of ignominy, not hitting a double for a long stretch here seems a badge of honor, because this is quite a good group of players, and particularly a group of players who had very long careers, on average. I don’t think this observation is important enough to bother and check when their long double-less streaks occurred — whether they were near the end of their careers, as often seems to happen. At least from the 1920s on, the trend is certainly not of a bad hitter’s having the longest double-less streak of the decade, but it is probably toward pure contact hitters showing up, reinforcing the notion that doubles mean power, at least of a certain kind.
Again, sticking with quick impressions, Witt, Moore, Pafko, Runnels, Renteria, and Segura strike me as oddly similar. There is nothing odd that a certain type of player dominates a list; that is in fact what you trust you will see, and what you hope you will see, when you generate a list. The oddity is these are not players I would have thought would have been low in doubles. Believe it or not, I don’t typically have a working idea of a player’s doubles record in my head, but the traits I do associate with the Edgar Renterias of the world don’t seem to fit “low doubles.”
Because he is a current player, I was most confident in my understanding of Jean Segura, and thus most surprised by his presence on the list. In fact, in the unlikely event a focus group on him were convened, one of my contributions might have been “gap power.” I therefore decided to review his statistics and see if I was completely off in my impression.
In Segura’s decade-topping season, 2015, he hit 16 doubles in 560 at-bats. Outside of his 239 at-bat slump, then, he was right on that rate of 20 at-bats for each double. That was his third year as a regular with the Brewers, however, and it would have been accurate to cite him as an example of a player who didn’t hit doubles. For his Brewers career, he hit one per every 33.5 at-bats.
In 2016, he came to the Diamondbacks. Not only did he hit 20 home runs, about doubling his career rate, but he emerged as a doubles star, hitting 41. He followed that with doubles totals of 30, 29, and 37.
Over the last two years, however, Segura has hit just 14 doubles in 655 at bats. I missed that he was released last August. His very low recent doubles rate could just be another sign that he’s done, or it might not have than the usual significance, since he’s metamorphosed in the doubles categories once before. A low doubles rate from an old player seems more ominous than the same from a young player, however….
Although his name sounded familiar, one of the decades leaders I couldn’t have told you anything about was Eddie Mayo. Studies tell you that we don’t generally just concoct a sense of familiarity, but I sometimes feel I’m an exception to the rule. But I was fortified in the thought that there might be something especially interesting about Mayo when I noted that Mayo also didn’t have an extra-base hit during his streak, giving his case another layer of interest. Maybe I hadn’t heard of him because he was notoriously inept at bat? Or could he have been the guy who led the ‘68 World Series champion Tigers from the dugout? I thought this was even a better possibility when I saw that Mayo eventually went from the Athletics to the Tigers (managers tend to come from teams they played on12 ).
Sinking in to his Reference page, while he was no Hall of Famer, Mayo was in fact a respectable hitter and not someone one would expect to turn up on a list for futility. Just two years after his streak, he was not only the second baseman on the Tigers ‘45 World Series winners, he was second in league MVP voting behind Hal Newhouser (although a 112 OPS+ and 4.8 WAR presumably wouldn’t allow such a high finish today). I’m guessing this is why his name was familiar to me.
Since Mayo could hit, had bright days ahead of him, and was facing war-depleted competition, how then did he go 252 at-bats without an extra-base hit? After a slump of our own, it seems another of these cases of a player who went on a doubles drought not because he was unlucky or bad but because of special circumstances. To wit, his SABR Bio says he injured his eye in spring training of ‘43, which left him with a literal blind spot. The blind spot didn’t go away until the next season. Since Mayo not only didn’t hit any doubles in his slump, but didn’t have extra base hits, period, it would be disingenuous to say that the doubles slump was the unique key to perceiving his deeper issue. Maybe the lesson is that any kind of alarming bad performance can signal special circumstances. Our lives are dotted with asterisks….
Spike Shannon, who played every game for John McGraw’s 1907 Giants team and led the National League in runs scored that year, had no reason to hang his head about the quality of his career in general, but he certainly did not hit doubles. In fact, there is a compelling case that he was the worst doubles hitter of all time, if we lean on the streaks list. Shannon was active in just five major league seasons, but managed the 8th-longest streak of doubles futility ever, the 10th-longest, and the 12th-longest. Those streaks came in consecutive seasons: 1906, 1907, and 1908. Going by his listed height and weight (5’11”, 180 lbs.), he wasn’t small, but lack of homerun power is applicable here: Shannon hit a home run in just 3 of his 2,613 at-bats.
When I took gymnastics, if you were threatening to be able to execute a challenging move the coach had introduced you to, he would put you in your place by saying, “Once is just a coincidence, twice is a fluke, but three times is a trick. Show me again.” The streaks list seems to confirm the wisdom of this. Only a few names repeat, and of those, a few seem to dominate the list. Shannon is a case in point, and this criterion can perhaps be effectively used to find the contenders for the worst doubles hitter post 1900.
Among the 386 cases of streaks of 51 games or longer13, just five players make at least three appearances. These are Bergen, with 5; Glory of Their Times interviewee Al Bridwell, also with 5; Roy Thomas, with 4; Shannon, with 3; and a more modern player, Dal Maxvill, with 3 (I couldn’t even have told you Dal Maxvill played. I knew him as Whitey Herzog’s general manager).
Given the greater rarity of extremely-low doubles seasons in modern times, Maxvill may well belong with these others, but I am simply asking a narrow question and not adjusting for era norms. I am just concerned with double rates, and when you account for how many of Maxvill’s games without a double were ones where he didn’t start, or ones where maybe he started but hit at the bottom of the lineup14, a lot of the starch goes out of his streaks. The longest was just 184 at-bats (all of Shannon’s streaks were at least 270 at-bats).
To repeat, my ideal list would have been an inclusive one that rated players by at-bats, not games, but getting such would have been a somewhat extended process. The games list, then, contained lots of garbage in it, but I could remove it on the back end. In that vein, for this exercise, I decided to only count the streaks of players if they consisted of 175 at-bats as well as 51 games15. 175 at-bats doesn’t seem too much to ask. It doesn’t even represent 44 4-AB games. But when you put this stipulation in, Maxvill drops out, and we are left with four notables. They lose three streaks among them, with the big change involving Bridwell, who drops from 5 to 3. The tally is now.
Bill Bergen, 4 streaks; Roy Thomas, 4 streaks; Al Bridwell, 3 streaks; Spike Shannon, 3 streaks.
Thomas did Shannon one better, compiling a streak in four consecutive seasons (1902-1905). With 477-500 at-bats in each of these seasons, across them he amassed what we would think of as about a season’s worth of doubles: 26. His worst doubles season was actually 1900, the year before his first known streak. We actually don’t know what kind of doubles streak might be contained in there; it falls outside Streak Finder’s window. Since he had 4 doubles in 531 at-bats in 1900, one every 132.8 at-bats, it is very likely he had a streak of at least 175 at-bats that season.
Who was Thomas? My knowledge doesn’t go beyond the superficial, but I rate him a hell of a ballplayer. I learned of him because he was one the 25 nominees Bill James chose before picking the 10-best center fielders of all time in his 1985 Historical Abstract. James and admiring guest columnist Jim Baker talked about how remarkable Thomas was: he hit the ball so gently, yet played the game so well. He had a .413 career on-base percentage — absurd for that time.
He was 5’11”, the same height as Spike Shannon, so not a short baseball player for the time. One can’t help noticing the general similarities among Thomas, Shannon, and Jimmy Slagle, who had a 293-AB streak of his own (although Slagle was 5’7”). The common factor that is the least interesting but most important to their overall ranking is certainly just that they were contemporaries during the seasons in the study where doubles were hardest to come by. But beyond that, all basically profile as having been leadoff hitters. First, they were center fielders or left fielders, so more blessed with running speed than good arms, presumably. No power, but walks, walks, and more walks. Except for in 1905, Thomas led the NL in walks every year from 1900-1907. Slagle was three times in the top five in walks, and five times in the top ten. Shannon was third in the NL in walks in 1907. Thomas was another class from Slagle and Shannon, but my point is that despite the dearth of doubles, they were good players.
You could be a very good player without power in the dead-ball era. That game seems generally a duller one with less dimensions than its successors, but there is something compelling about a game where leadoff men, and I don’t mean your Ronald Acuna type leadoff men, were some of the better players. A game where brawn seemingly played a minor role.
A couple other things. Since we found a possible link between rampant bunting and low doubles, maybe the significance of these leadoff types dominating my lists isn’t that they were the guys who innately lacked power, but that they were the guys who bunted the most often. Just as Bobby Bragan wrote a book, You Can’t Hit the Ball With the Bat on Your Shoulder, you generally can’t hit a double without swinging.
Sacrifice bunt totals could shed light on how these players bunted, and Jimmy Slagle did once lead the National League. Spike Shannon was once 2nd. On the other hand, while I don’t quite know how sacrifice hits were counted then, their totals certainly don’t indicate they were bunting a sizeable percentage of the time. I also have a reflex to think that many bunt attempts do not go with the kinds of high-walk totals that these players put up, but maybe I’m wrong about that, as Brett Butler had 90 walks or more five times in six years.
Slagle, Shannon, and Thomas were also all part of the famous NL pennant race of 1908, which I have fond memories of learning about when I read G. H.’s Fleming’s The Unforgettable Season. Both Shannon and Thomas joined the Pirates, who ultimately tied for 2nd with the Giants, one game back of the Cubs, in mid-season (and Shannon had actually been waived by the Giants). For his part, Slagle was with the pennant-winning Cubs, playing 104 games, but only good for a 71 OPS+, confirming his 1906 downturn. The microcosm of 1908 speaks to one of my broader points, that it’s ironic that these great teams had these outfielders historically lacking in power. But truth be told, Slagle, Shannon, and Thomas would be rated very differently if their whole careers looked like their 1908 statistics. They were such neat contemporaries that they were all at the end of the line then, and there was less than two years separating the three of them in age.
The Historic and Fun
I’m sure a deep dive into the streaks of other players could be fascinating, but you will excuse me if I give Rogers Hornsby and Mickey Mantle preferential treatment. Beyond their names, their streak statistics contained a big hook of a similar nature. Hornsby had a 206 AB double-less streak in 1916, his rookie season, but he hit 12 triples during that streak. In 1958, Mantle went 177 at-bats without hitting a double, but he hit 22 home runs during that streak. What more can we learn?
Rogers Hornsby
If there were a time when I would expect a player to have a high triple-to-double ratio, it would be his rookie season. That is when players are doing more than the obligatory hustling, when they are looking to taking the extra base. That is when football players are foolhardy but inspirational and refuse to go down, even when the play is a lost cause. But Hornsby’s 1916 3b-to-2b ratio, considering the era, does not quite put him in this light. It was 15-to-17. Since he was tied for 9th in the NL in hits, and he was Rogers Hornsby (his 38th ranking in career doubles doesn’t make for easy interpretation, but he led the NL in doubles four times in five seasons from 1920-1924), it’s somewhat surprising that he didn’t hit more doubles, period, as a rookie (he was tied for 30th in the league).
Anyway, his overall 3b-to-2b ratio suffered because outside of his streak, Hornsby had just 3 triples and 17 doubles. The split between his numbers during his streak and for the rest of his season is downright bizarre. With all due respect for statistical coincidence, it seems he must have been up to something, and could not naturally have had 12 triples and 0 doubles in a 56-game stretch. My image is that he took it into his head that he was not stopping at second base, come hell or high water (but, note that if he had been thrown out at third attempting a triple, he would have been granted a double, putting a bit of a crimp in this theory).
That is the knee-jerk response. It’s likely wrong, but there it is. Trying to get some context, we just heard during an Eagles-49ers game that the Eagles had scored touchdowns in 12 straight red zone possessions. “Have you heard of a streak like that?” Kevin Burkhardt asked. The streak does seem worth noting, as the Eagles score just 62% touchdowns in the red zone on the season. But Hornsby was 47% triples from doubles on the season, yet had the same streak of 12 with his triples. And, even outside the higher base rate, the football streak is much easier to come by, because 12 straight touchdowns only consists of a few games, meaning only against a few opponents. The odds are very good that it was achieved against teams the Eagles could have been expected to score a touchdown on over 62% of the time. Triples versus doubles is less opponent dependent, and generally takes as many games as hits to achieve, meaning against probably two or three times as many opponents.
I studied Hornsby’s game logs, and here’s what I found. First, his streak ran from May 14, his 25th game of the season, through July 28, his 90th. But it turns out he also didn’t get out of the blocks well. While he’d hit 2 in 18 games the previous year, it wasn’t until his 21st game of 1916 that he managed his first double. He had hit a triple in his 7th game. At the conclusion of that 21st game, Hornsby was hitting .275 with a .319 Slug. Ruining our story a bit, Hornsby then doubled in both his 23rd and 24th games. The next day, May 14, he hit his first major league home run (off Jeff Pfeffer, who was about as good as anyone in the league outside of Pete Alexander). That was the beginning of his “no doubles” streak. With the strong four games, his slugging average had already risen to .400.
So, we know Hornsby hit 12 triples over 56 games. To further illustrate the spree, I’ve compiled some nuggets.
The run included 10 in 23 games, June 20 through July 11.
He tripled twice on June 20 and twice on June 28.16
He hit triples in three consecutive games: July 8, 9, and 11.
In one of the two-triple games, he also hit two singles and a home run. So he could have converted one of those triples into a double and had the cycle. (Wonder what the situation was, when he hit the last triple?) The cycle is a minimum of 10 total bases. Hornsby’s game gave him 12. I don’t know if that’s a sensational number (it’s three home runs), but in 9-inning games where a player has hit less than 2 home runs, only Ian Kinsler, 4/15/09, has beat it.
In the game in which Hornsby finally hit a double, it was his only extra-base hit, but he was 4 for 4. He then went another 15 games without hitting a double.
During the streak as a whole, again 206 AB, Hornsby hit .320 and slugged .510. Five of his six home runs on the season came during the streak. Remember, he never reached double digits in home runs in the dead-ball era.
Mickey Mantle
In 1958, Mantle had a 177-AB streak without a double, but he offset that with 22 home runs. He also didn’t have a triple. One wonders if the lack of doubles wasn’t a sign of his greatness: that when Mantle got his pitch, he didn’t miss it, at all. Ballparks couldn’t hold him.
Somehow, in my mind, I thought the low-doubles notables would include some Giancarlo Stantons, some “three true outcome” players, but these are harder to find, just browsing through putative candidates, than I would have imagined. Of the 12 seasons of 50 or more home runs in the last 20 years, the lowest number of doubles is 24. It was singles that Barry Bonds basically stopped hitting, not doubles. His 73 home runs were supplemented by 32 doubles. So Mantle’s period seems anomalous. It doesn’t seem to track with a larger phenomenon.
I’m also guessing that “Mantle, 1958” doesn’t mean anything to you. (I wonder if it means anything to Bob Costas? Not that he’s not capable of a little bit of historical knowledge, but he was born in 1952, so maybe not.) Mantle 1956, Mantle 1957, and Mantle 1961 mean something to me, and I have a concept of him in other years, but not 1958.
This appears to be a year when he got off to a slow start, perhaps accounting for an eventual lackluster 5th in the MVP voting behind Jackie Jensen, Bob Turley, Rocky Colavito, and Bob Cerv, despite ultimately clearly posting the best numbers. Through May, in 35 games, he had just 4 home runs. He then hit 10 home runs in June, but while hitting only .245. He had a monster July: .349 with 14 home runs. The high home runs, no-doubles streak ran from June 2 through July 25. He finished the year with a .304 average, a .442 on-base average, and 42 home runs, the latter the best in the league.
After the streak, Mantle just seemed to go back to being Mantle. He hit 11 doubles and 16 home runs in 212 at-bats for the rest of the season. The next season, he hit only 31 home runs. The streak doesn’t seem to have signaled a change his game. The way I read his statistics, he unlocked his full home run talent in either 1955 or 1956. Then I don’t see another upward move, but he probably didn’t lose any ground as a home run hitter until 1965.
It may be a leap to be talking about a streak with a switch-hitter, anyway. Performance from the two sides is consistently distinctly different for switch-hitters. That Mantle was a switch-hitter does raise the possibility that he had an unusual percentage of at-bats from one side or the other during his streak, though, and that this caused the odd statistics.
I was unable to get the data on this. (No great loss; it’s an extremely weak theory, anyway.)
For his career, Mantle did have quite different HR/2B ratios hitting left-handed (1.80-1) and right-handed (1.20), which you can probably trace at least partly to having Yankee Stadium as a home ballpark. In 1958, Mantle did bat more often left-handed than usual (76.6% of his plate appearances were from the left side, versus 67.0% for his career). The plate appearance ratio is in the right direction for a higher-than-average home run to double ratio (although, on the season, his ratio was just 2-1, compared to 3.38-1 for 1961, and 2.36-1 for 1956). Batting left-handed on the season, Mantle hit 34 home runs and 15 doubles; batting right-handed, he hit 8 home runs and 6 doubles….
The best slugging averages during a no-doubles streak of at least 175 at-bats (there have been 173 such streaks)….It’s such a fun list, I will take the liberty of going to 10th place.
Mickey Mantle 1958 .655
Rogers Hornsby 1916 .510
Dick Stuart 1959 .495
Tony Conigliaro 1970 .484
Deion Sanders 1992 .470
Norm Cash 1971 .460
Roger Maris 1960 .440
Ernie Lombardi 1946 .431
Lee Thomas 1961 .424
Graig Nettles 1986 .415
So I focused on the right two guys in Mantle and Hornsby. Not only were they the two best hitters in the group on an average day in an average year, but they played the best during their streaks.
Sanders hit 7 triples during his streak. For 1992, that was probably at least as impressive a triples performance, and as unlikely a triples/doubles difference, as Hornsby’s 12/0 in 1916.
The five best batting averages during a no-doubles streak of at least 175 at-bats
Roy Thomas 1901 .322
Rogers Hornsby 1916 .320
Richie Ashburn 1962 .314
Larry McLean 1907 .296
Clyde Engle 1911 .293
The Smalleys
A final bit of fun. Roy Smalley, Sr. went 58 games and 167 at-bats in 1948 without a double, and Roy Smalley, Jr. went 61 games and 198 at-bats in 1982 without a double.
Most At-Bats in a Season with 0 or 1 Double, by Decade
I’m not sure what my justification was for pursuing this line of analysis, as I am now basically conceding that the streak approach indicts with more justice. I suppose I thought that first, some of these might have the obscure and unique character of a Belliard or a Choo-Choo; second, that there might be something to be said about opening up the lists to players who were not pure and had hit a single double; third, that I could bring the project around full circle, and I needed an analysis by decade for the sake of completeness. Regardless, observations and connections are ample, whatever one’s approach.
We can skip ahead and begin with the 1920s, as the “1 2B” overall list makes explicit the winners from the 1880s (Arthur Whitney), 1890s (Curt Welch), 1900s (Bill Bergen), and 1910s (Tommy McMillan). And I let the cat out of the bag with the 1870s, already letting you know that Tom Barlow had the qualifying season with the most at-bats in the decade, even though he in fact didn’t hit a double at all.
1920s: Denny Williams: 1 2B in 218 AB, 1925 Red Sox
1930s: Hughie Critz: 0 2B in 219 AB, 1934 Giants
Critz’s OPS+ in 1934 was 18. He hit .187 with just 3 walks. I derived some satisfaction from H. Pitz and H. Critz both deserving comment in the chapter.
1940s: Tommy Butts: 0 2B in 196 AB, 1942 Baltimore Elite Giants
1950s: Ozzie Virgil, Sr.: 0 2B in 226 AB, 1957 Giants
1960s: Jim Davenport: 1 2B in 272 AB, 1968 Giants
Davenport, quite a good player who amassed just shy of 5,000 career plate appearances, and who was in the game in various capacities for many years after his playing career concluded, managed in just one season, 1985, not making it through. Roger Craig took over the Giants from him, and they started a swift turnaround the next year. 1985 was the year I became a baseball fan, so I knew that, and only that, about Davenport, before running into him here.
Davenport’s doubles hell continues the trend of players who never rebounded after their struggle. It was 1968, but remember, OPS+ adjusts for the league, and he weighed in at 63. The Giants continued on with him in ‘69. He didn’t do much better than he had in ‘68 (71 OPS+), then played just 22 games in ‘70 in his final season.
You might note the consecutive Giants, Virgil, Sr. and Davenport. Both were also third basemen, and while it’s not quite true that Davenport took over third from Virgil, Sr., it’s almost true. As you can tell from his ‘57 at-bat total, Virgil, Sr. wasn’t a full-time player, but when he did play, it was at third more than anywhere. Ray Jablonski was properly the team’s starting third baseman, though. Although he had been just 25, after his struggle in ‘57, Virgil, Sr. didn’t follow the team to San Francisco. Davenport was a Giants rookie in ‘58, starting 103 games at third, while Ray Jablonski started 50 games.
1970s: Jim Hickman: 1 2B in 201 AB, 1973 Cubs
These numbers also came near the end for Hickman, in the second-to-last year of his 13-year major league career.
A teammate on those 1973 Cubs was Glenn Beckert. The next year, Beckert would move to the Padres, get 172 at-bats, and hit just 1 double himself. If it weren’t for Hickman, I’d be writing about Beckert here (along with Elliott Maddox, who also had a 172-AB, 1 2B season, in the 1970s).
Looking at those ‘73 Cubs, I thought, man, what an old team. You go catcher through LF, everyone (Hundley, Hickman, Beckert, Kessinger, Santo, B. Williams) was in his 30s. Those were the same six starters the Cubs had at those positions in 1970. Their 77-84 record in ‘73 seems pretty good, considering, although it was the first losing season the Cubs had had since 1966 (which isn’t to say they were really good when they were posting their winning seasons from ‘67 to ‘72). The next year, the bottom fell out, as they lost 96 games.
Aside from the fact that he was at the end, I’m inclined to regard Hickman’s presence here as mostly random, but he did also have a streak of 179 at-bats without a double when he was with the ‘63 Mets. In that season, he actually hit 21 doubles, the second most he ever hit besides the 33 he hit in his all-around crazy career year of 1970 (.315, 115 RBI, 1.001 OPS). For his career, he had 24.4 at-bats per double.
1980s: Rafael Belliard: 0 2B in 286 AB, 1988 Pirates
1990s: Matt Merullo: 1 2B in 140 AB, 1991 White Sox
2000s: Willie Bloomquist: 1 2B in 165 AB, 2008 Mariners
Bloomquist also didn’t have any extra-base hits in this 2008 season, but Kansas City’s employment of him for 468 plate appearances the next year, even though he was a journeyman who had little possibility of improving, wasn’t out-and-out crazy: Bloomquist not only hit .279 in 2008, he had a .377 on-base percentage.17 Alas, I guess he caught Royals’ disease in 2009, as his OBP dropped to .308. He actually wasn’t by and large a walker (5.8% of PA for his career), and of his 11 other seasons with at least 100 plate appearances, his best OBP marks were .360 and .325.
A person I have become fond of who figures here is Doug Glanville. In his final season, 2004, he had just 1 2B in 162 AB, obviously just missing the top spot, although Robb Quinlan had an even more agonizing miss, falling 1 at-bat short of Bloomquist, also in 2008.
2010s: Sam Fuld: 0 2B in 176 AB, 2013 Rays
2020s: Luis Rengifo, 1 2B in 174 AB, 2021 Angels
This seems random because Rengifo has emerged as a pretty good player. He has 18.0 home runs per 575 AB for his career, an 89 OPS+, and even has hit some doubles (21.8 per 575 AB). He held his own as a 22-year-old rookie in 2019, posting an 82 OPS+ and hitting 18 doubles in 357 at-bats. But whether it was related to the pandemic or not, his 2020 and 2021 major league performances were miserable: .156 in 2020, and then .201 in 2021. He only hit 1 double in 90 at-bats in 2020. That’s actually the 9th-largest AB total for a 0- or 1-double hitter of the 2020-2023 period, and it was just his warm-up act.
Changes in the Numbers Over Time
I’ve given the names. but I also want to talk about how the totals themselves have changed over time. This helps us understand how notable particular individual at-bat totals are. It is also interesting to compare the changes versus league double data: If we see fewer terrible ratios of 2B/AB now, does this change just reflect the MLB average, or does it show that there are fewer stragglers now, and major league teams have less tolerance for the kind of player who compiles this sort of data? I will analyze the trend both in terms of the highest at-bat totals for 0- or 1-double players, and in terms of the longest at-bat streaks without a double. The two measurements yield different conclusions.
To deal with the data of 0- and 1-2B players first, Tommy McMillan in 1910, with 322 AB, is far out ahead of anyone else after him, except Belliard.
Denny Williams’ 1925 total of 218 at-bats is impressive relative to what was a high-doubles decade.
Hughie Critz’s 219 AB in ‘35 does not give a realistic idea of the outer bounds for the era. He stood quite apart, and the season is an outlier. In the ‘30s, setting him aside, Wilson Redus18 of the Chicago American Giants, also in ‘35, had a 1-double season of 171 at-bats (I don’t know whether his league was a high-doubles league, though, like the 1935 National League was). But the most at-bats for a MLB player having no more than 1 double in the 1930s was just 148 (Clyde Manion, 1930; Homer Reed, 1933).
The AB totals of the ‘40s (196), ‘50s (226), ‘60s (272), and ‘70s (201) leaders make good (if not perfect) sense, given era norms for doubles.
Gotta change paragraphs when I get to my man Belliard again….The second highest AB total of a qualifying player in the ‘80s was 157, 55% of Belliard’s total (and Lenn Sakata in 1984, he of the glasses, hit a double, while Belliard in ‘88 didn’t).
Matt Merullo had the highest AB total of the ‘90s, 140. That’s the lowest for any decade, but not really much different than the decades leaders since. Sam Fuld’s double-less 2013, in 176 at-bats, is the highest AB total for a 0- or 1-double player since Belliard. I don’t know if he’s 31 lengths behind Belliard, like Twice a Prince19 was against Secretariat, but he’s at least 15 lengths behind. The ‘90s weren’t the lowest doubles decade (2000-2009 was), but the change in doubles rate has been minimal enough that it’s hardly a mortal sin to take the numbers at face value and forget about era from at least ‘94 on out, maybe from ‘90 on out. (Merullo’s weird season was in ‘91, so he didn’t really have a no-double season in the context of current norms). Removing Belliard from the equation, the last season of 200+ AB with 0 or 1 doubles was Jim Hickman’s 1973, and he squeaked over at 201.
I also had Stathead give me the number of players in each time period who qualified on the 0- or 1-double front and had at least 150 at-bats. Then, I divided this by the number of teams in the time period, and multiplied by 100 (to make it per 100 teams). I excluded any pitchers who might have made the list. I also excluded Negro Leaguers, for two reasons: 1) At the time, I wasn’t sure I could quickly get the number of Negro League teams (it turned out I could). 2) I had no idea how frequent 150+ AB seasons were in those leagues, given the limited availability of data, and some of the short schedules. Including Negro League data to gauge this trend would have been adding a disparate element, particularly problematic since the Negro Leagues operated in some eras but not in others.
The reason I added this count was that I thought it would summarize what I had merely scanned more effectively. And obviously, it makes sense to gauge a decade not just by one player, not just by its leader.
It takes some time to comb through and compare the rate of 150+ AB, low-double players in each decade to the previously cited AB totals of the decade leaders and runners-ups, but the two approaches largely coalesce. The 1901-1909 periods and 1910-1919 periods are in a class by themselves, with the first having one 0- or 1-double player for every 8.5 teams, and the second one per every 13.5 teams. Cases of 150+ AB and no more than 1 double also occurred in earnest in the ‘50s, ‘60s, and 70s, peaking in the ‘60s (the rate was actually 90% in the ‘60s what it had been in the ‘10s). Outside of these decades, however, the truth is that 0- or 1- double seasons with 150 at-bats are freakish occurrences. The total number of teams that have traversed MLB in the ‘20s, ‘30s, ‘40s, ‘80s, ‘90s, 2000s, 2010s, and 2020s is 1.9 times as great as the total number of MLB teams that have comprised the 1900s, 1910s, 1950s, 1960s, and 1970s, but the number of qualifying seasons is 27% as high in the first sample. In the low-doubles period, the rate of qualifiers is 6.5% of teams, in the high-doubles period, it is 1.0% of teams.
In the second column of data in the table below, I listed 2B-per-575-AB. The five lowest rates, in ascending order, were the 1900s, the 1910s, the 1960s, the 1970s, and the 1950s. These were also the five seasons with the highest rates of 150+ AB, 0- or 1- double men, in that order.
If you had to cite any discrepancies between number of qualifiers and overall double rate, you would start in a couple of places. Since the double rate in the ‘40s was only 24.8 per 575 AB, it’s surprising there weren’t more low-double seasons. What makes it perhaps even more surprising is that I would have thought some of the Pete Grays20 of the war years, and the pitching-oriented conditions then, might have made for a few fertile years. However, one could accurately point out that that 24.8 is also held down by the World War II seasons. The two low-double seasons that occurred were actually after the war (Skeeter Webb in 1946, Benny Zientara in 1948), however. Maybe it’s a coincidence, but from the data, you might say that the double rate in the ‘40s represents a sort of threshold, above which it is unlikely that there will be low-double seasons.
Then, that the percentage of qualifiers was as high as 1.7 in the 1920s with a doubles rate of 27.7 is a little bit out of line. One wouldn’t have thought more low-doubles seasons in the ‘20s than the ‘80s, particularly when the 25.6 rate of the ‘80s benefited from the AL DH21. It makes sense that there was less competitive balance in the ‘20s, and that this is what the rate of qualifiers in the decade, relative to the league average, is showing. But any incongruence of this kind is gone by the ‘30s, and the 1.3 rate of qualifiers from 2000-2009 seems equally elevated considering the league doubles rate.
A reality check is that the total number of qualifiers going into any of the decade percentages when low-doubles seasons were very rare is nothing to speak of (for 1920-1929, there were 3 qualifiers from 176 teams, for instance), so it seems silly to be making anything of the fluctuation within this group.
So, the two conclusions are that, first, the phenomenon of this combination of 1 double in 150+ AB is, for practical purposes, relegated to only parts of baseball history, and second, that these are the times when there have been the lowest number of doubles league wide. The data generally reject the idea that, even though we are dealing with outlier bad seasons, the trend in frequency has followed any pattern other than league double rate….
It’s nice to be able to make sense of data, but it’s also nice not to be able to, nice to confront a mystery. And for some reason, when doubles scarcity is defined by consecutive at-bats without one, league double rates do not seem informative. The list of the longest “no doubles” streaks by decade, from Bill Bergen to Carlos Santana, is much earlier in the piece, but I’ve reprinted the numbers in the third data column below. Aside from the longer streaks of the pre-1920s players, the streaks over time look flat. Leaving Santana out of it, since the 2020s is now just a partial decade, the average of the 1920s+ leaders is 252 AB with a standard deviation of 22. The ‘60s and ‘70s leaders being high, with 275 and 267 at-bats, makes some sense. But how was the longest streak in the 1950s the shortest of any period except the 2000s? The double rate wasn’t high then. How were there longer streaks from players in the ‘20s, ‘30s and ‘40s than from a player in the ‘50s? You might say with some justification that Eric Yelding’s 264-AB streak in the ‘90s shouldn’t be used to analyze the ‘90s, since baseball in 1999 was very different than baseball in 1990, but a 25.9 doubles rate in a non-DH league, which is what it was in the NL in 1990, is actually pretty high.
If the lack of correlation merely reflects that streaks are anomalous or that my effective sample size is too small (ten decade leaders), then I am overreacting, but it’s such a departure from the analysis of the low-doubles seasons, which moved in lockstep with the league double rate. Is it possible that streaks are more flukish? That my original conception of identifying players based on their season’s doubles total, even though it skewed toward part-time players, was somehow more valid? I don’t think so. I just can’t think of why that would be. The two approaches are, at bottom, fundamentally the same, and they should show the same thing.
So, again, the first number in the table below after the decade is the # of low-double seasons per 100 teams, the second is MLB doubles rate per 575 AB, and the third the longest streak in AB without a double achieved by any player. I will attach an Excel sheet with the same information. It will have the columns labeled and aligned, and so be easier to read. Let me know if you can’t access it, or even if you can.
1901-1909 11.8 20.0 334
1910-1919 7.4 21.5 322
1920-1929 1.7 27.7 245
1930-1939 0.6 28.9 241
1940-1949 1.1 24.8 252
1950-1959 3.4 23.6 232
1960-1969 6.6 21.9 275
1970-1979 4.9 23.5 267
1980-1989 1.5 25.6 286
1990-1999 0 28.5 264
2000-2009 1.3 30.7 215
2010-2019 0.7 28.9 239
2020-2023 0.8 28.1 189
Conclusion
The journey here began with the fantastic and the fun. Going without doubles seems almost like going without conversation. When I see the celebration set off when a player gets his first major league hit, I sometimes think, “Oh, come on. You were going to get a hit.” Doubles is just a rung up from hits. They seem inevitable. Achievement in baseball might not be inevitable, but data accumulation is, or should be.
Perhaps the spirit of the fantastic should not be ruined by dry learnings. There is a danger here of imposing a narrative: the bunter, the drinker, the injured player, the perhaps sick one. I found those right off among low-doubles players, but I would not say they are typical of them. Certainly not in the 21st century, or even the 20th century.
And sometimes narratives can be hollow, even if they’re legitimate. But if you’ll excuse me, I think what I learned more than anything else in doing this research is that doubles are not anomalous. Granted, I was working with low outliers, but I now believe doubles are an interesting amalgam of average and power, with power probably leading the way. They are not a mere curiosity, and if they aren’t occurring, something is probably quite wrong. So continue your search for a stupid statistic elsewhere.
Which I am not. I’m being a generalist, having fun, and pursuing trivia.
I originally heard about this in the podcast interview Michael Lewis did with James, apparently in 2022 James, Lewis Podcast
I can’t help sharing something, which is that I just noticed he scored only 18 runs in 402 plate appearances! Adding up hits, walks, and hit-by-pitches, he was on base 100 times, even. That seems a very low percentage of runs scored per time on base.
The 55-98 Superbas scored only 442 runs that year (2.85 a game). The team had 5,604 plate appearances. If the whole team had had McMillan’s run scoring rate, it would have come to only 251 runs. I know that you can’t “do” that, can’t frame runs as individual efforts. I’m just making some attempt to put McMillan’s run-scoring total in perspective, and put it in context for a low-scoring team.
The dynamic on that (17-37) team must have been odd in general. Of the 10 players for whom Baseball Reference lists an age, the second oldest was 28. None were taller than 5’10”, although Bill Boyd, of height unknown, is listed as 250 lbs.! Pearce was the peanut, at 5’3”.
My uncle is not Marv Throneberry, by the way. If he had been, this would have been a Yogi Berra-saying come true, as Throneberry preceded Coleman by 22 years in death.
Source is B-R Bullpen.
The one step you can take to make the games more aligned with at-bats is to only include games where there was an at-bat. By doing this, you eliminate the very-late-in-game defensive replacements. I took the step, of course.
I should also say that one reason, and maybe the reason, why the streak total is built around games is there are a few years at the beginning of the search period where play-by-play data wasn’t available, but box scores were. With just box scores, you can’t analyze at-bat streaks.
The same addendum as above applies here. Belliard played in 122 games in 1988, but in only 100 of them did he register an at-bat.
Slagle was the starting center fielder for the 1906 Cubs, the greatest regular season National League team of all time, but he not only owned the double-less streak, he sported only an 84 OPS+. As he had 9 fielding runs in 1906 and had an average OPS+ of 114 in his four previous reasons as a regular, I wouldn’t want to contend that he didn’t live up to the normal standard of a starting player for a championship team. However, here is another case I have stumbled onto of a surprisingly mediocre player (at least where 1906 was concerned) playing a significant role on a great team. If Slagle’s double-less streak had ended in the Cubs’ World Series loss, that would have been richly ironic, but according to his SABR Bio (the solo SABR Bio contribution of Larry Slagle! Who demurely does not note his relation), he missed the Series with an injury.
This far down on the list, I will not vouch that it is complete because the divergence between games streaks and at-bats streaks becomes significant. I limited myself to the 400-longest games streaks. 201 at-bats rated 91st among those, but I have good reason to think this is on the border of where some at-bats streaks outside of the top 400 in games might start to enter in.
If you have always assumed that Runnels seemed a different and better hitter with the Red Sox than with the Senators because he enjoyed Fenway Park, you have a point: he hit .340 in home games, .299 in road games, while with the Red Sox.
The Tigers manager in 1968, as many of you know, wasn’t Eddie Mayo, but Mayo Smith. I can’t use this as an excuse for my ignorance, I certainly had never noticed it, but Mayo Smith was actually Edward Mayo Smith. He wasn’t named after Eddie Mayo, though, in case you’re wondering that. Not only was Mayo Smith not from Detroit, he was less than 5 years younger than Eddie Mayo. His SABR biography says his name is instead a nod toward the Mayo Clinic.
Ideally, I would have considered streaks of 50 games or longer, but I downloaded the 400-longest streaks, which arbitrary included some 50-game streaks but not others.
He hit 8th or 9th in 1,069 of his 1,094 starts. Since 1,050 of those starts were in the 8th position, and he made the vast majority of his starts in non-DH leagues, I think his last-in-the (non-pitcher)-order rate might be the same.
Again, it’s possible I am undercounting the number of streaks of players because streaks of 175 at-bats but less than 51 games are not included. It is also possible that other players would then have shown three 175+ at-bat streaks.
This isn’t necessarily crazy. Could a player with just 17 home runs playing a full season have, within his data, two multi-homer games in 10 games? I don’t know. At first I thought so, but the bottom line is, a guy hitting just 17 home runs on a season probably wouldn’t hit 10 in less than a month, and Hornsby hit 10 triples in 23 games. That’s baseball, Suzyn.
Bloomquist also stole 14 bases, so he had a Bill James run element ratio of 39! The run element ratio reflects that, if James’ formula connotes what it’s supposed to, you’d much, much rather have him up early in the inning than late in the inning. He walked and stole bases, but didn’t have any power, hence the score. 1.00 is meant to be an average run element ratio. Info on Run Element Ratio
He appears an interesting player. Just 5’5”, so right away, you think you understand the 1-double partial season. But then you see he hit 22 home runs in 315 at-bats in 1928 and led the National Negro League in RBI that year. He played 16 years and is all over leaderboards from 1925-1928.
This was the horse 2nd to Secretariat in the Belmont. Sham, the talented runner-up in the Derby and Preakness, ran in the Belmont, but did not finish 2nd. That he didn’t run his race has something to do with how Secretariat was able to win by 31 lengths.
I just looked it up. Gray didn’t hit a major league home run, but he did hit 6 doubles in 234 AB. He had the same OPS+ as Billy Holbert, 47. He did hit 5 home runs in the Southern Association the year before he got to the majors, and stole 68 bases that year, too.
I had to use Stathead’s “Split” function to get these data, and split does not let you filter by position, or I would have excluded pitchers so I could have apples-to-apples data. If I had had enough time to fool around with the FanGraphs searches, I think I could have gotten the uniform, non-pitcher numbers.
This is a remarkable dive into Two-Bagger Land. I won't look at doubles the same way ever again. The Deion Sanders/ Rogers Hornsby/Rafael Belliard triples-to-doubles troika is sure to haunt my dreams tonight.