This is a follow up to my last post on "how good was he?" which you can read below. He batted .363 from ages 22-29 but only .309 from ages 30-37. That is a very big dropoff. I wonder if he had nagging injuries that eventually caught up to him (he missed about 140 games from ages 25-29). There are a couple of reasons mentioned below from his SABR bio.
To see how his declined compared to others, I found every player who had 4000+ PAs from ages 22-29 (390 players) using the Lee Sinins Complete Baseball Encyclopedia and every who had at least 3230 PAs from ages 30-37 (also 390 players). Then I found all the players who were in both groups (143) and the difference between their "young" average and their "old" average and ranked them from highest to lowest. The top and bottom tens are in the table below.
As you can see, Simmons is very near the bottom. He did not have the biggest decline, but it was still very big. As for Keeler and Davis, those two guys both had their careers cover both a high average period and a low average period. When Davis was 22-29, the NL league average was .287 (Keeler's ages 22-29 are almost identical to these years). Then from 1902-08, when Davis was 31-37, the AL league average was .250 (Davis switched leagues). Keeler spent most of his 30s in the AL, too. From 1903-09, when he was 31-37, the AL league average was .246. But the AL league average when Simmons was "young" was .293 and when he was "old" it only fell to .288. So his dropoff is more surprising. I also calculated the average park factor for the parks he played in when he was "old" and it came out to about 101, meaning average. There might be some lefty/righty issues in those parks, though, that I don't know about it. But it looks like his decline cannot be explained by a changing league average or tougher parks. The A's park had a park factor just under 104 when Simmons was aged 22-29.
The simple average of the change in average from "young" to "old" was about .007 (if I only used "old" guys who had 4000+ PAs,it was .006). So if Simmons could have had an average drop off, it would mean that he would have hit aroung .356 in the second half of his career. That would put his career average close to .360, still second to Cobb (I have not quite split his career evenly in two, so I could have overestimated this by a couple of points). Anyway, Rogers Hornsby is 2nd with a .358 average. Simmons is currently 21st all-time in average for players with 5000+ PAs.
I wonder if he had made it to 3000 hits if he would be better known today (he would have made it with the higher projected average). Also, he batted .390 or better twice. He was at .401 on July 17, 1927. He played only 5 more games in July and then did not play again until Sept. 6. Maybe some injury cost him .400. In 1931, when he batted .390, he only played 27 games in the final two months of the season. Yet he batted .465 & .426, in those two months, respectively. His combined average then was .442. Another 100 or so ABs at that rate would have gotten him to .400. A .400 season might have made him more famous. Maybe playing for the Yankees would have helped (although being a right handed batter would have hurt). Data in this paragraph came from Retrosheet.
His bio by Fred Stein at the SABR Baseball Biography Project suggests a couple of reasons for his decline, but I don't think it explains enough.
The bios says "In 1926 Simmons “slipped” to .341 and 199 hits. He was hampered by injuries the following two seasons although he hit for high averages." In 1930, at one point, he had a "swelling knee." He missed 16 games that year. Also, "Years later, Simmons admitted to a writer that he had accepted the White Sox’ second-division attitude and had slacked off in his customary strenuous practice habits." (but he actually did better in OWP, discussed below, in his first two years with the White Sox, .617 & .675, than in his last year with the A's, .590).
One curious thing it reports is that Simmons said "I’ve studied movies of myself batting." I wonder how many players did that back then?
The table below shows how Simmons declined in offensive winning percentage (OWP) when he got "old." OWP is the Bill James stat that says what your team's winning percentage would be with 9 identical hitters when you give up an average number of runs. In this case, it is park adjusted. As you can see, his dropoff was great. I think Sisler had some kind of eye problem and even missed a whole season. I don't know anything about Bottomley. Maybe sports medicine was not very good back then.
Simmons finished with a career OWP of .644, which is now ranked 117th among players with 5000+ PAs. The typical dropoff when getting "old" was about .024. If Simmons had a .698 OWP when he was "old" that would give him a career OWP of about .710. That would currently rank 35th, a big improvement over 117.