Pop Lloyd played in the Negro Leagues until his 48th year. Oscar Charleston played until age 44 or longer depending on your source. Buck Leonard stuck around pro ball through age 48. Cool Papa Bell played into his mid 40s. Smokey Joe Williams? Pitched til he was 47. Satchel Paige, of course, played in the majors in his 50s.
It’s not just that these guys played into their forties. It’s how late into their forties many remained regulars and productive ones at that. As we described in our article on early-debut players, the propensity for very young and old players to hold down regular jobs and to excel at the highest level available to them result from a combination of the lack of a formal farm system and talent-acquisition process and, generally, from having a smaller overall pool of talent to work with.
Miller and I are still debating whether and how to elect Negro League players, but one thing we know: Matters such as these make the interpretation of Negro Leagues careers very complicated. Even if a Negro Leaguer was highly productive in his mid-40s, could we create a credible translation that credits him for those very late-career seasons? For the seven thousand, three hundred, and thirty-second time in the short life of our humble website, we turned to BBREF’s Play Index to help us get something truthy to report. By the way, please subscribe to the Play Index (I make no money on this, I just think it is indescribably wonderful if you’re into geekery).
Mostly the good die old
We queried the PI against the following characteristics:
- the final seasons
- of anyone 38 or older
- from 1893–1965
- who was a non-pitcher
- whose last season wasn’t 1944 or 1945.
This yielded a total of 195 players. When we looked at the debut years of very young players with similar conditions in that previous article, we pulled in nearly 600 names. We also weeded out 13 players whose final appearances could only be described as token appearances. These were all fellows whose last games occurred at least three years after their previous appearance and included fewer than a handful of contests. Hughie Jennings, guys like that. Six others had very late token appearances of this sort, but their final legit seasons were 38 or later. I’ve classified them by their legit final season ages (Dan Brouthers, Sam Thompson, Jim O’Rourke, Kid Gleason, Jimmy Austin, and Jack O’Connor). This leaves us with a sample of 182. We recorded several pieces of information about our codgerly MLBers:
- position they had the most career games at
- year of final season
- age in final season
- career PA and rBat (batting runs above average)
- PA and rBat for all seasons from age 38 onward.
Some census information, if you will, about these folks. Here’s the breakdown of late retirees by age in our 73-year sample:
- 72 38-year-olds—40% of sample, about one per year
- 40 39-year-olds—22% of sample, about 0.5 per year
- 22 40-year-olds—12% of sample, about 0.3 per year
- 19 41-year-olds—10% of sample, about 0.3 per year
- 18 42-year-olds—10% of sample, about 0.3 per year
- 6 43-year-olds—3% of sample, about 0.08 per year
- 2 44-year-olds—1% of sample, about 0.01 per year
- 2 45-year-olds—1% of sample, about 0.01 per year
- 1 48-year-olds—0.5% of sample, about 0.1 per year
From 1893–1965, the big leagues had, on average, about 1.5 players aged 38 or 39 retire each year. Another 0.8 from 40–42 also retired each year. Which leaves the rare 0.2 forty-three or older retiree per annum. Still another lens for this is to say that from 1893–1965, on average each season, 2.5 players in the league were 38 or older, and that about 1 player a year was a fortysomething. In other words, it’s a young man’s game because this all equates to under 1% of the major league population (who meet our set of conditions) being active in a given year.
A positioning statement on the baseball elderly
That’s pretty interesting stuff. Now check out this demography by position:
- C: 52—29%
- 1B: 23—13%
- 2B: 15—8%
- 3B: 12—6%
- SS: 25—14%
- LF: 16—9%
- CF: 17—9%
- RF: 22—12%
David Ross eat your heart out! There’s never enough pitching catching. Twenty-two of those backstops were 38-year-olds. Nine were 39-year-olds. That’s about the same percentage as the total player population of 38 and 39-year-olds in our sample. At age 40 there are six catchers, at age 41 there are nine. At 42 there are four, and then none until Deacon McGuire checks in at 48 years old. That’s all good news for the likes of Biz Mackey, for sure. The news is pretty good for Pop Lloyd, too, because shortstops were well represented in this pool. Not as good for the other throwing infielders, though…. Still, these aren’t the most surprising findings. The most difficult two positions on the diamond last the longest because either, they are catchers, and catchers defy all kinds of rules, or they are shortstops. Now shortstops start with the most athleticism among righty throwers, and because they have many other places they can play as they age, they can stick around, moving rightward on the defensive spectrum. Second and third basemen have a very specific skill set, and those positions were in the process of swapping places on the defensive spectrum for much of the period we’re looking at (second base growing in difficulty as its primary skill became turning the twin killing, while third base became less difficult as bunting declined in popularity). I suppose that centerfielders are the most surprising result here. With so many athletic lefty throwers, you’d think they’d age into left fielders or first basemen, but perhaps not so much.
In general, it was apparently good for one’s longevity to be a shortstop, a slugging right fielder, or a slugging first baseman. Or, of course, a catcher.
As with our prior discussion on younger players, we really need to look at the better players in our sample if we hope to draw any useful conclusions about great Negro Leagues players. I once again filtered out any batters with fewer than 200 rBat so we can look at the best of the best to start with. We’ll return to good defenders in a bit. Our list now includes 38 excellent hitters. And, yeah, it’s a strong bunch. Here they are by the age of their last seasons:
- Age 38 (8): Dan Brouthers, Edd Roush, Mel Ott, Dixie Walker, Bill Nicholson, Wee Willie Keeler, Sam Thompson, George Davis
- Age 39 (11): Roger Connor, Jake Beckley, Gil Hodges, Charlie Gehringer, Ernie Lombardi, Earl Averill, Zack Wheat, Ken Williams, Gene Woodling, Gavvy Cravath, Kiki Cuyler
- Age 40 (6): Johnny Mize, Gabby Hartnett, Yogi Berra, Jimmy Ryan, Tris Speaker, Babe Ruth
- Age 41 (4): Nap Lajoie, Rogers Hornsby, Ty Cobb, Ted Williams
- Age 42 (4): Stan Musial, Cy Williams, Fred Clarke, Jim O’Rourke
- Age 43 (4): Eddie Collins, Enos Slaughter, Honus Wagner, Luke Appling
- Age 45 (1): Cap Anson
The big takeaway? No one plays past 43, and those that do are the exception that proves the rule. Heck, sixty-six percent of these guys packed it in by age 40. But what kind of playing time and production do older players generate? Let’s start by answering that question only for the year of their retirement:
- fPA = plate appearances in final season
- frBAT = offensive runs above average (rBat) in final season
- qual = qualified for batting title
Median | average | fPA frBAT | fPA frBAT | qual ==============|===========|====== 38s 79 -0.5 | 140 1.1 | 1 39s 117 0.0 | 142 1.0 | 0 40s 142 -0.5 | 178 0.3 | 1 41s 392 4.5 | 325 9.3 | 0 42s 200 -2.0 | 252 -1.5 | 1 43s 140 -3.5 | 136 -3.8 | 0 45s 497 -2.0 | 497 -2.0 | 1
Now, let’s have a look at the progression across ages for players who retired at these various ages:
- Qualifying seasons = years in which the player qualified for the batting title in his league (3.1 PA/scheduled game)
- Average seasons = Roughly speaking either 350 PAs with at least 0 rBat or a season that would obviously be thus if it were longer. Yeah, kind of a judgment call, but it’s not unobvious when you look at it.
QUALIFYING SEASONS | AVERAGE SEASONS # 38 39 40 41 42 43 44 45 | 38 39 40 41 42 43 44 45 ================================|========================= 38s 8 0 X X X X X X X | 1 X X X X X X X 39s 11 3 0 X X X X X X | 4 1 X X X X X X 40s 6 2 2 1 X X X X X | 4 3 1 X X X X X 41s 4 3 1 2 0 X X X X | 3 2 3 2 X X X X 42s 4 1 2 1 2 1 X X X | 4 3 3 2 0 X X X 43s 4 2 1 2 1 2 0 X X | 2 4 3 2 1 0 X X 45s 1 1 1 1 1 0 1 1 1 | 1 1 1 1 1 1 1 0 ---------------------------------------------------------- TOT 38 12 7 7 4 3 1 1 1 | 19 14 11 7 2 1 1 0
These amazing hitters aged amazingly well on the whole. Forty-four seasons of above-average batting by age 40 among our 38 contestants. And still another 10 thereafter. These guys were great for a reason. In general, we could summarize this table this way: The longer you hit well, the longer you stick around. Stunning, huh? That’s the kind of penetrating insight you don’t pay to get around here. But for our Negro League friends, this has pertinence when combined with the information about the number of players with late retirements.
- Most hitters are done by age 38, which includes many excellent hitters.
- Those who stick around past 40 are incredibly rare.
- Essentially no one lasts beyond age 43.
- Playing time fades but can go dark very, very quickly.
- Batting skill seems to linger, though diminished, until it doesn’
Two last things. I haven’t shown you these above. A rough estimate of average PA attrition for these elite hitters is about 15% a year beginning with the ratio of PAs in age 39 vs age 38. Even more crucially, the typical hitter gets about one or two seasons of negative rBat in anything more than 100 PA before he’s a goner. The exceptions to this rule: Nap Lajoie. Just Nap.
A defense against premature aging?
Now, wait just a second, you say. What about players who aren’t elite hitters but still maintain good value through defense? How do their careers progress? I’m so glad that I hypothetically asked myself that question in your voice. Among the remaining 144 players in our larger sample, I gathered up the 45 with at least 20 career rField (aka: fielding runs above average) to see what we might glean from them.
Median | average | fPA frBat | fPA frBat | qual ==============|===========|====== 38s 102 -3.0 | 152 -4.8 | 0 39s 49 -1.5 | 142 -2.9 | 0 40s 89 -4.0 | 178 -3.5 | 0 41s 50 -0.5 | 325 -1.5 | 0 42s 133 -4.0 | 133 -4.0 | 0 43s 71 -8.0 | 71 -8.0 | 0 44s 237 -7.5 | 237 - 7.5 | 0
Pretty obvious that these guys can’t hit so good. Fourteen of them exceeded 100 rBat in their careers (which is good, not bad, of course), and eight of those fourteen were fork-tender before age 40.
QUALIFYING SEASONS | AVERAGE SEASONS # 38 39 40 41 42 43 44 | 38 39 40 41 42 43 44 =============================|======================= 38s 21 0 X X X X X X | 1 X X X X X X 39s 10 1 0 X X X X X | 0 0 X X X X X 40s 6 1 1 0 X X X X | 2 1 0 X X X X 41s 4 3 1 1 0 X X X | 1 1 1 0 X X X 42s 1 0 0 0 0 0 X X | 1 1 0 0 0 X X 43s 1 1 1 1 1 0 0 X | 0 0 0 0 0 0 X 44s 2 1 1 1 1 0 1 0 | 0 0 0 0 0 0 0 ----------------------------------------------------- TOT 45 7 4 3 2 0 1 0 | 5 3 1 0 0 0 0
These gold glovers die off much faster than their big-bat relatives. They get a lot fewer plate appearances, and, of course, they don’t have nearly the level of production in their baseball dotage. The information in this table suggests these things about the time period we are looking at:
- A good glove wouldn’t keep you in the league as long as a good bat
- It was, specifically, very rare to stay in the league past 39 years old as a glove-oriented player
- Playing time diminished considerably faster for glove men than for strong hitters.
The PA attrition rate for these guys is about 40% a year. The leash on these guys is just a little looser than for the great hitters. These fielding whizzes might get three bad offensive years before they’re cut loose. But then again some of them were always bad hitters (Mr. Maranville to the white courtesy phone!). So what’s the diff?
Growing old together
So if we are coming up with a way to allocate playing time at the end of a good or great player’s career, how do we use these findings? Well, every player is very different, so we need to use sound judgment and compare a Negro Leaguer to the most similar players we can find. We can use those MLB comps as a means to chart a career path for an older player. We have to make a certain decision, though. As we translate a player’s performance and build a playing-time model do we treat the player as one single individual added to the MLB universe? Or do we assume that all Negro Leaguers are added to it? The reason this matters: If we assume the former, we have more leeway to model a unique or nearly unique career path. If we assume the latter, then not too many Negro Leaguers can be modeled that way since so few individuals in MLB end up with unusual career trajectories. There are much bigger concerns regarding that second assumption that aren’t worth diverting into now, so let’s just say that I choose the former.
Then, we have a little freedom, but we don’t want to create unreasonable translations. So we need to turn the eight bulleted statements above into actionable guidelines:
- Hitting trumps fielding when determining career length.
- To keep it realistic, don’t translate anyone past age 43 without a really good reason.
- Translate only the very best past age 40.
- In fact, most players shouldn’t be translated past age 38 or 39.
- Even if a player is very durable during the meat of his career, it’s probably wise to diminish his playing time by 15% per annum for great hitters or 40% per annum for fielding geniuses from age 39 onward…if they make it to age 39.
- And don’t keep anyone around for longer than three seasons of negative rBAT after age 38, and no great hitters past two unless they are also (still) great fielders.
That’s a pretty decent set of guidelines. It’s part of a developing toolbox for Negro Leagues translations that are starting to gel for us here. Or maybe congeal if you find the assumptions and results too far away from your own thoughts. Regardless, we’re working on this stuff pretty diligently with the hope of understanding the path ahead, should we choose to tread it.