Researching a piece I will eventually write about trades, I came across the sale of Fred Dunlap from St. Louis to Detroit in 1886 (both teams were in the National League). Dunlap was a good player. In the one year of the Union Association, St. Louis went 94-19, and Dunlap led the league in every slash category, as well as scoring 30 more runs than anyone else in the league. However, defense was apparently the highest-rated part of his game, and his Rfield was positive in each of his first 10 seasons. When St. Louis eventually sold him, his manager justified it by saying Detroit had never paid so much for a player.
There certainly are some other interesting tidbits around that sale, but what caught my eye was that the year after Dunlap dominated, St. Louis moved to the National League and had two player managers: Dunlap, who went 30-40, and Alex McKinnon, who went 6-32. That’s quite an argument for managers making a difference, isn’t it?
Making things more interesting, I read that Dunlap compiled that record in separate stints. This sounded like Billy Martin with the Yankees, but in one season1. One tantrum does not a Billy Martin make, but it’s also evident that Dunlap could make trouble, whether bad or of the John Lewis kind, I don’t know enough to say.
I suppose I had two questions. My first was whether there was some catch to the difference in records. In that era, teams essentially tanked so often in the chaos of trying to keep afloat that I almost expected this to be the explanation. But the Maroons are generally absent from the transaction wire of 1885. Indeed, their owner Henry Lucas was particularly recognized for his desire to win, with baseball historian Joan M. Thomas noting that he “lost a large share of his family fortune” in pursuit of the goal. There may be still be some obvious reason they did so much better under Dunlap other than the same players just doing better, but I haven’t run into it. Another piece of information that makes me thing the difference in performance might reflect natural causes is that both Dunlap and McKinnon played 100+ games in a 111-game schedule. McKinnon was a capable guy, too, hitting .294, best on the team. Then shortstop Jack Glasscock, maybe the best player on that team both in 1885 and over his career, played every game.
The second riddle was trying to uncover the story behind Dunlap’s two stints. First, I just wanted to know when those had occurred. I was incredulous that the manager data were even accurate. It was hard to see how McKinnon could have gone 6-32, looking at the Maroons’ monthly records. Their worst were 3-16 in August, and 6-13 in May. You put those together and you get 6-29, and what of June and July? I wondered if maybe the two managers had alternated stints at the helm, or had managed different days of the week, sort of like the Cubs operated in the early ‘60s.
But eventually I got a break in the case. A book that usually just sits on my bookshelf, collecting dust, is the 2007 ESPN Baseball Encyclopedia. I went to the Manager Register, and not only did Dunlap’s entry confirm the 30-40 Baseball Reference has for him, but the encyclopedia breaks it down into segments: 21-29, and then 9-11. Going back to Baseball Reference’s “Schedule and Results” section for the Maroons, they indeed started 21-29, and finished 9-11. So I knew exactly when Dunlap was the manager. There was no rotation of managers. This was more or less coherent to a modern sports fan, even if the door is not now generally open for quick managerial returns the way it often is for jettisoned romantic partners. But could I find out anything about the circumstances of the changes?
I have access to New York Times archives (TimesMachine). As you can imagine, the New York teams are disproportionately covered in the baseball reports of this era. As luck would have it, the Maroons’ 50th and 51st games, presumably bridging when Dunlap went from manager to mere player, were both against the Giants, and there are newspaper accounts of both.
On the face of it, it is very odd that Dunlap would have abdicated his managerial duties at this time. The 1885 Maroons began the season 5-3 before falling to 9-19 and then 13-26, a record that exactly matched their final winning percentage. But not only did the Maroons go 8-3 over Dunlap’s final 11 games of his first stint, but they had won five in a row through July 9. But the Giants2 beat them 3-2 on July 10, and if we hue to Dunlap’s record, bam, then he was no longer manager. What happened?
Here is the lede to the story filed after that game.
“In the absence of the regular umpire Welch, of the New-Yorks, umpired today’s baseball game between the St. Louis Club and the New-Yorks. In the ninth inning, with two men on bases, Welch deliberately called [Charlie] Sweeney out on strikes, and this gave the New-Yorks the game. The contest was the prettiest played this year on the home grounds3….”
Welch was almost certainly pitcher and eventual Hall of Famer, Mickey Welch. Tim Keefe, three years older and eventually elected to the Hall of Fame nine years before Welch, might have been a more logical candidate to ump, but he was otherwise occupied, as he was pitching for the Giants that day. In any event, one can how see this arrangement would make an opposing manager want to wash his hands of things, although I think you would be just as disillusioned as a player, which Dunlap continued to be. Maybe if the managing position didn’t come with any more money, Dunlap decided they could take the job and shove it. But a resignation out of protest would make more sense in the context of grievance with one’s own team.
Another possibility is that Dunlap went ballistic and the league forced him out as a consequence. If he was thrown out of the game, we can’t know this from Baseball Reference’s database of managerial ejections, because it doesn’t go back far enough. The ejection component might not have been of any use in deciphering the situation, anyway, since the contested strikeout made for the final out of the gate (Dunlap was on base at the time, by the way).
The next day’s Times story still dwells on the absence of an independent umpire in the July 10 game, while saying nothing about St. Louis changing managers. One can infer from it, however, that Henry Lucas had complained to National League president Nicholas Young4, and Young had responded by finding a suitable and independent umpire named McCaffery5 to work the July 11 game. Although the Maroons lost 8-2, Dunlap and McKinnon both played and got hits, also occupying the same spots in the batting order, second and third, that they had the day before. So, if Dunlap did fly off the handle after Sweeney’s strikeout the day before, whatever he did wasn’t so unseemly that he got barred from the ballpark (the suspension could have been on appeal, but the notion of an appeal process in 1885 strikes me as unlikely, at least without recourse to the courts, proper). So, in conclusion, while I obtained interesting background about the Maroons’ 50th game of the season, the articles give no indication of a managerial change. I’m not sure this is peculiar; the reports seem spotty, and maybe the managerial job in general didn’t receive a lot of attention.
Surprisingly, even though a Giants game did not precede Dunlap’s reinstatement, the Times was actually more helpful filling in the second half of the story. I doubt the managerial change would have been news in itself, but it was part of a general shake-up centering around Lucas’s bowing out as president in favor of Ben Fine, whom the SABR Bio describes as “his close friend.” The Times says that Lucas was “known in baseball circles the country over,” giving some insight into why the change was news6. He and his family had enormous reach in St. Louis, and Lucas had been the president of the Union Association.7
As for our hero, the story relates that “Mr. Fine understands as well as anybody that the team has not been well coached, and at the meeting of the club, to be held tomorrow, Dunlap will be reappointed Captain. He will not only be given charge of the men on the field, but they will report to him every morning.”
If you read the whole story, there is room for confusion, because, not only is Dunlap referred to as just “Captain” (although with a capital C!), but Fine is referred to as a “manager.” I know little about 19th-century baseball, but even into the 20th century, I have noted that “captain” was not just the ceremonial position it could be said to be today. The captain and manager seemingly divided the important responsibilities. That being said, the captain was called captain for a reason, and he wasn’t manager. The manager had ultimate authority for on-field decisions. So if Dunlap had really just been captain, that would explain why his switching with McKinnon the first time went under the radar. But the detail that Dunlap is slated to supervise the team before games and on the field perhaps belies this reading. That seems to fit our current conception of manager. For whatever reason, in every source I’ve used, it is Dunlap and McKinnon who are recognized as managers, not Lucas and Fine,8 and I think that is most decisive with me.
Some of what Fine opines on in the article also seems to comport with what a manager would do. For instance, the article says, “It is [Fine’s] belief that he can reconcile certain elements in the team and get the nine to pull together and to play ball as the individual members are capable of playing.” It is clear that he at least flattered himself that he knew baseball and wanted to be what we now call general manager, but the article also reports his arranging games with the “Brown Stockings” (almost surely the American Association’s St. Louis Browns) after the season. Being manager wouldn’t preclude him from doing most everything; I think everybody had to be hands on, whether it meant promoting the team and selling tickets, or what have you, just as we might see with independent ball today. But contracting a future series sounds distinctly like the work of a team president.
It is ironic that, when Dunlap stepped down, the Maroons had been hot. Without having the full story of what happened, we still have to think there was a triggering incident, maybe related to the umpiring fiasco. Looking at the results of the preceding games ended up not being productive. But this time, the change fit the standard pattern. It would seem, and not just from the tone of the Times article, that things had boiled over. For the Maroons had compiled an eight-game losing streak, and lost 16 of 17, before McKinnon’s ouster.
The article makes it clear that Fine and Dunlap intended to raise expectations of the players and work them harder. That Dunlap was tapped to do this is consistent with my understanding of him as a fiery leader, if liable to rub people the wrong way. The suggestion is that Lucas didn’t have the stomach to do what needed to be done, and would be stepping aside for the good of the team. “The close friends of Mr. Lucas believed the club would thrive better under a sterner hand….” Fine, his friend, and Dunlap, his star player, would do the dirty work he could not bring himself to do.
This story also sheds light on the previous matter-of-fact statement that player-umpire Welch knowingly called a bogus strike on the Maroons to end the game. On the one hand, nothing can explain this, because that account still cannot be reconciled with any current journalistic standards. (How would the reporter know whether Welch thought the pitch was a strike or not? See Matt? I’m learning). But I now believe that, even though this is The New York Times, the story was written by a St. Louis reporter. If there were to be bias, it makes sense it tended in that direction. The reason I say that is that this September 7 story also has a St. Louis byline, and includes some of the same color. At the end, it is asserted that “long before [the time of the Maroons’ planned post-season exhibitions] the Maroons will be playing in their old form.” This is a voice of pure homer-ism, and we are unaccustomed to it in a newspaper story9, and not coming from a local broadcast booth.
Whether his prophesy was smart or he was just seeing things with rose-colored glasses, our anonymous reporter was quite right. The Maroons tied in Dunlap’s first game of his second stint, then won 9 of 20 to close the season.
The “tale of two seasons” (which, o.k., is more like Billy Martin’s getting 83 wins out of the 1980 Athletics than the “best” of anything) dimension to the records with Dunlap and McKinnon at the helm is of interest given the recent discussions about how much difference managers make, if any. Analyzing it also puts us in the familiar position of looking at two splits and trying to figure out if the difference is meaningful. This is seemingly something we do all the time, with no expertise other than experience.
You know how I came out on this one. If this disparity hadn’t gotten my attention, I probably wouldn’t have invested so much energy into investigating it.
One way to come at the question is to recall a post that Joe Posnanski made, “What a Month!”Posnanski post. The Braves had just gone 21-4, and he asked if that was the best month for a modern team ever. Note that the Braves were a very good team, a .642 team, so almost as good as the .333 Maroons were bad. They had a lot of good months. Posnanski says this wasn’t the best month ever, but it sounds like it was damn close: his first candidate is the 25-4 July 1941 Yankees. The Maroons were 6-32, playing .842 losing baseball, under McKinnon. That’s for a longer period than a month. We can reverse 6-32 and compare to 21-4, 25-4, etc., and pronounce “extraordinary,” indeed, it would seem.
My eye test at least says that if something isn’t going on here, ninety percent of the other trends smart people credit must be imagined, too. But I was surprised when I used the “Binom.Dist.Range” function in Excel, and asked how often a .667 team could be expected to have its more common result 32 or more times in 38 games. The answer was 1.3%. That might seem low to you, and to confirm the notion that this was a significant difference, but my interpretation of it is opposite. An old professor impressed on me that such statistical sanctioning of trends is extremely biased because we are testing a single, likely random difference that attracts our attention, and overlooking all the instances of other splits when there is obviously no difference. Implicit in my professor’s thinking is that there is really no good reason why Fred Dunlap should have outperformed Alex McKinnon. So given his tutelage and my observation through another two or three lifetimes, I would have been circumspect about this difference even if there had been a couple of more zeros in front of the ‘1’ for that probability. Note also that even the inferred 98.7% doesn’t suggest that McKinnon’s drag on the team was the full .175 in winning percentage; it suggests just that some of that might have been real.
As an aside, if you play with the parameters, and ask the percentage chance for a .667 team to lose 34 of 38, it drops dramatically, to 0.1%.
I think we also are also more swayed by the statistical data invoking the managerial factor explains the Maroons change in fortune not once, but twice. If Dunlap had managed the same 70 games, but taken over after the team had started 6-32, some people wouldn’t be quite as convinced that he was a central reason for the turnaround Regardless of the sport, I note the same extra respect accorded for coaches when there are multiple examples of their having good Plus/Minus data (to use the basketball concept). Football writer Mike Sando says this is largely how he evaluates whether a coach should be in the Hall of Fame. I’m sure he thinks Vince Lombardi and Bill Belichick should still be enshrined, but he’s skeptical of Tom Flores, who won with the Raiders but not the Seahawks.
For the most part, I think Sando’s belief that applying this method yields sufficient evidence of coaching impact is wishful thinking. If the coaching win-loss data he is parsing in his head were subjected to fair statistical tests, my hunch is they would come up woefully short of suggesting what he thinks they do. My little Excel experiment with the 1885 Maroons’ data only hardens my skepticism. A pattern with two, three, four occurrences still remains perhaps coincidental. But, all other things being equal, I think Sando’s point is correct. There is something to be said for a coach whose success is not consecutive (and maybe if it comes with different teams as well). When the stints are intermittent, they can’t coincide completely with other events, such as player injuries. The manager is the only constant.
Come to think of it, Steinbrenner and Martin did reconcile within days of Martin’s departure in 1978. But I don’t think a return in season was ever considered, with the two settling on a 1980 return (revised when the Yankees ‘78 World Championship was followed by a 34-31 start in ‘79, and thus an early Billy II).
If you’ve read my Substack note, you know how good the Giants were in 1885 (a 75.9% WPCT for the season). One wouldn’t have expected the Maroons to have beaten them, which is another odd facet of a loss to them preceding a managerial change.
I include the final sentence simply because I am amused by how quickly the writer shifts gears from his shocking opening to talking about a pretty “contest.” The tone is that this was just another game, and that the umpire circumstance wasn’t extraordinary. Also, let me acknowledge that because of where the comma is placed, the first sentence first reads as if umpire Welch was absent. But then it becomes clear that the only way to make sense of the paragraph is that “Welch of the New-Yorks” umpired.
Young himself had umpired in the National Association, which predated the National League.
Going by Baseball Reference’s index of umpires and “Bullpen” entry, this seems to correspond to Harry McCaffery, who would have been 26 years old. McCaffery played 44 games in the American Association in 1882 and 1883. He appears to have been very much a fill-in that day and not a regular umpire. But I guess I am making too much of the age of these umpires; Henry Lucas himself was just 27. Who the original umpire was, and why he couldn’t umpire the games, was not addressed in the articles.
I also can’t help him noticing that the Times seemed to care more about owners than managers, one suspects out of proportion to their actual importance to winning games, or at least their baseball knowledge. I have engaged in a very serious study of horse racing in the 1930s, reading many news accounts, and see something similar. The focus is the horse owners, who did little of the actual work with the horses but had wealth and social connections, while the men (and occasionally women) in charge, the horse trainers, were lucky to get a mention. The coverage is opposite today.
Joan M. Thomas’s SABR Bio.
An argument for taking the “Captain” designation for Dunlap to heart is that the July 11 story refers to the Giants’ manager of record as exactly that: “Manager Mutrie.” My guess is that these reporters just weren’t as disciplined in distinguishing between a captain and a manager as we are today, for whatever reason.
I read the story as news, not a column, but I suppose one could argue it was a hybrid that we don’t see today. Changes in the team’s president and manager are news, though, so subsequent opinions in the same piece come across as off-key to the modern ear.