The 20-21 Club
Here’s how one small group of hitters progressed through their 20s:
Those lofty stats were logged by Ty Cobb, Rogers Hornsby, Jimmie Foxx, Mel Ott, Ted Williams and Mickey Mantle. The modern era’s first six decades produced, at regular intervals, these six players with a qualified 140 OPS+ at age 20 and 21. But for almost 60 years after The Mick, it was a closed society; even dropping the threshold all the way to 15 games per season, no one else notched a 140 OPS+ at 20 and 21.
Last year, Mike Trout crashed that club with a 179 OPS+ at 21, after 168 at 20. For 20-21 combined, his OPS+ and Offensive WAR topped everyone, even those legends. His oWAR is 30% more than #2 (Alex Rodriguez), and he’s #1 in two of oWAR’s main components, WAR Runs from Batting (Rbat) and from Baserunning.
Enough has been said of Trout’s two great years. But a stats-history buff can’t help wondering: Will he get even better, as a hitter? What pattern might his growth take?
The stock approach is to track players who had comparable starts. But comps are scarce for one so good, so young. As impressive as Trout’s two-year totals is the consistency of those seasons: Out of all seasons through age 21, he owns nos. 1 and 6 in OPS+, nos. 1 and 9 in Rbat, #1-2 in oWAR and #1 and 4 in WAR. No one else owns two top-12 marks in even two of those categories. And his bat’s trending upward.
Despite the small pool, I looked at those who broke in with a bang at 20-21, and how they fared in the rest of their 20s. Among modern players who have completed their 20s, here are the top 25 in OPS+ for age 20-21 combined (900+ PAs), followed by their stats for age 22-29. Each had at least one year with a 130 OPS+ and/or 4.5 oWAR at 20 or 21:
Outlining what I’ll detail below:
- Twenty of the 25 raised their OPS+ in 22-29, fifteen by 10 points or more.
- They peaked a good two years younger than normal.
- They were extremely durable through their 20s.
I’ll take them in two separate groups, starting with the Big Six who were most like Trout at 20-21, and then their composite. (If you’re uneasy with these groupings, you might want to jump to the last section, “Alternate Groups.”)
Group 1: 140 OPS+ at 20 and 21 (Cobb, Foxx, Williams, Hornsby, Ott, Mantle)
These greats built castles atop their rock-solid foundations, hitting much better through the rest of their 20s: Group OPS+ rose from 160 to 183, Rbat from 44.0 to 61.9. Each one’s OPS+ was better from 22-29 than 20-21, four of them by 25 points or more. And the growth came quickly: By age 25, all had logged at least two seasons topping their 20-21 best in OPS+.
Here are their combined averages, for age 20-21 and for various spans in the rest of their 20s. Games are pegged to a 162 schedule, just to show their durability; other stats are not adjusted:
- Age 20-21 … OPS+ 160 … Rbat 44.0 … oWAR 6.3 … WAR 6.6 … Games 150
- Age 22-23 … OPS+ 176 … Rbat 52.6 … oWAR 7.4 … WAR 7.9 … Games 154
- Age 24-26 … OPS+ 190 … Rbat 68.9 … oWAR 8.9 … WAR 9.2 … Games 156
- Age 27-29 … OPS+ 181 … Rbat 61.0 … oWAR 8.0 … WAR 8.1 … Games 151
- Age 22-29 … OPS+ 183 … Rbat 61.9 … oWAR 8.2 … WAR 8.5 … Games 153
As seen here and the opening graph, this group peaked at 24-26, about two years younger than normal. And that’s with Williams in the service all three years; he set age-group OPS+ records on both ends of that gap.
They were remarkably healthy through their 20s, missing about 9 games per year from age 22-29 (not counting Ted’s service years). Just three of those 45 seasons had more than 17 missed games, none more than Cobb’s 56 in 1914, when he broke a thumb in a butcher-shop brawl.
Obviously, each rates quite highly among all players age 22-29:
- OPS+ (3,000 PAs) — #2, 3, 4, 5, 13 and 19 (Williams, Cobb, Hornsby, Mantle, Foxx, Ott)
- Rbat — #3, 4, 6, 7, 9 and 12 (Hornsby, Mantle, Cobb, Foxx, Williams, Ott)
- Offensive WAR — #1, 2, 4, 14, 16 and 20 (Cobb, Mantle, Hornsby, Foxx, Ott, Williams)
- WAR — #2, 3, 4, 14, 16, 26 (Hornsby, Mantle, Cobb, Foxx, Ott, Williams)
The Splinter’s ranks in 22-29 WAR stats are impressive for pitting five years against the others’ eight. He’s #1 in WAR & oWAR for 22-23 combined, 2nd to Ruth for 27-29.
Group 2: OPS+ 113 to 150 for 20-21 combined (900+ PAs)
These 19 hitters also grew from their 20-21 average, but less so than the Big Six; they had a few more injuries, most notably to Conigliaro. Still, 16 of these 19 players rank in the top 100 for oWAR age 22-29, with three in the top 10 and eight in the top 30.
Group 2 averages in their 20s, counting zeroes for the four years Conigliaro sat out (with games again pegged to 162):
- Age 20-21 … OPS+ 132 … Rbat 23.9 … oWAR 4.6 … WAR 4.7 … Games 148
- Age 22-23 … OPS+ 140 … Rbat 29.0 … oWAR 5.1 … WAR 5.6 … Games 149
- Age 24-26 … OPS+ 141 … Rbat 30.0 … oWAR 5.2 … WAR 5.5 … Games 150
- Age 27-29 … OPS+ 140 … Rbat 26.3 … oWAR 4.7 … WAR 5.0 … Games 135
- Age 22-29 … OPS+ 140 … Rbat 28.3 … oWAR 5.0 … WAR 5.3 … Games 144
Putting it graphically:
This group peaked around 23-26 — even sooner than Group 1, though not as dramatically. By Rbat alone, their age-21 was as good as any other year in their 20s save 26, a hair above their 23-26 average, and 2 runs above their 22-29 average.
Outside of Tony C., they were almost as durable as the Big Six. Just seven of the others’ 144 seasons from 22-29 missed more than 35 games: Griffey and Kaline each broke a bone on a great catch; Cepeda and Cedeno each wrecked a knee on the field; Vaughan missed a month from a beaning; Hoblitzell slumped and got benched while awaiting WWI Army induction; and Henderson missed two separate months with hamstring injuries. (Note that Conigliaro’s absence at 23 and 27-29 impacts the group’s counting stats, but not their OPS+.)
Their drop in Rbat at the prime age of 27 isn’t just from Conigliaro being out. The noted injuries to Cepeda, Cedeno and Kaline all came at 27 (as did those to Cobb and Hornsby in Group 1), while five more just had off years, by their standards (Magee, McInnis, Vaughan, Robinson, Henderson).
Thirteen of these 19 raised both their OPS+ and oWAR during 22-29. For the other six, a quick synopsis:
Al Kaline (150 OPS+ age 20-21; 134 for 22-29): After an epic age-20 (top-4 for that age in OPS+, oWAR and WAR), Kaline didn’t have quite the offensive career that his few comparables did, partly due to a congenital foot problem. Hit 27 HRs at 20 and 21 (the first to hit 26+ both years), but never did reach 30. But the perception that he never played so well again is inaccurate; Kaline set a career high in WAR at age 26, in OPS+ at 31, with his best 2-year oWAR at 31-32. And while his 134 OPS+ for age 22-29 was a decline from his early peak, we should all be so blessed in our declines; he still ranked in MLB’s top-10 in OPS+, oWAR and WAR during that span (1957-64). Despite nagging injuries in his 30s, his bat aged brilliantly, averaging a 145 OPS+ from 30-37, helping him to career ranks of 27th in career WAR and 31st in product of total WAR and WAR per game.
Stuffy McInnis (130, 108): The first baseman for Connie Mack’s “$100,000 infield” had a third star season at 22, helping that quartet earn Bill James’s “best-ever” nod, then slowly faded through his 20s. Stuffy still hit for average, ranking 10th in MLB BA during his 22-29 years, but his walks and long hits shriveled, perhaps affected by thin lineups around him after Mack’s big sell-off. I found no injury reports; he played full-time through age 33, but was an average player at best after age 23. Out of 22 players with two 5-WAR seasons by age 22, only McInnis never reached that level again.
Cesar Cedeno (129, 132): His age-21 was among the best ever, ranking 7th or better in OPS+ and oWAR, and age 22 was almost the same. But without one clear cause, Cedeno just didn’t maintain that output. Some have blamed injuries, though he missed 31+ games just once from 22-29; others his guilt in the shooting death of his girlfriend after his second big season. (A wrist-slap sentence after pleading to involuntary manslaughter made him a target of opposing fans’ vitriol.) Teammate Bob Watson said that “he altered his swing trying to hit homers … pitchers adjusted, and he hasn’t readjusted himself.”* Cedeno still played like an All-Star from 23-26, averaging a 130 OPS+ and 92 runs in a low-run environment. Started 27 the same way, heading towards 60 steals for the second straight year, but in June he tore up a knee sliding, ending his year. He recovered enough for one last star turn at 29, batting .309 with a 147 OPS+ and 48 steals, helping the Astros to their first playoff berth. But in game 3 of that NLCS, he broke an ankle running out a DP; Philly won the last two games in extra innings to claim the pennant. Playing with chronic pain in his 30s, Cedeno was a shadow of the player once hailed as “the next Willie Mays.” Last run of glory came as a stretch-drive pickup for the ’85 Cardinals, including a 9/11 blast that Mets fans still rue.
* I checked some points that might bear on Watson’s theory: Cedeno’s contact rate was generally good, and actually improved quite a bit after age 23. On the other hand, his rate of flyouts did rise after his first four years, and his average on balls in play fell from .326 to .294.
Tony Conigliaro (127, 113): One of the saddest sagas of “what might have been.” A hometown hero, Tony C. slugged a teen-record 24 HRs in 111 rookie games, then averaged 30 bombs at 20-21. Having his best year at 22 (third to reach 100 HRs so young), when a gruesome beaning changed his life. Out more than a year, and never regained full use of his left eye, but he did make it back on guts and guile to notch two full, productive seasons, setting career highs with 36 HRs, 118 RBI and 89 runs at 25, even improving his contact rate. But the strain of reading pitches with one good eye led to headaches, which combined with other ailments (and a heartbreaking trade to the Angels) to halt his career midway through the next year. After 160 HRs by age 25, he hit just 6 more; no one else with even 100 HRs that young hit less than 28 thereafter.
Dick Hoblitzell (125, 106): Hit .308 with a 143 OPS+ at age 20, in 1909, but never above .294 or 126 in his other eight full seasons, in which he averaged a steady 108 OPS+. That first year may have been flukey, with a .333 BAbip that was 62 points over the NL average and 41 more than his next 8 years. Averaged 10 triples over his 9 full years, 18th in the majors for that span. Was batting .210 halfway through 1914 when the Reds waived him, a turning point for both parties: Cincy spiraled from 39-40 to 21-54, while Hoblitzell landed with the dynasty Red Sox, hit .319 that second half, then won back-to-back titles as their starting 1B. Went in the Army in 1918 (with McInnis taking over his job); had a near-fatal bout with the Spanish influenza, and never got back to the majors. Hit well in the minors for many more years, batting .329 at age 41 as an everyday player and manager of the Charlotte Hornets. (And now I know where the hoops team got its name.)
Vada Pinson (124, 118) — After two superstar years in the minors, Pinson at 20(?) was all over the NL leaderboards, and his 6.7 oWAR still rates 5th all-time for age 20 or younger. Held near that mark for six more years, without reaching the next level; slipped a rung at 27 and never climbed back. Not quite the hitter implied by his .303-21-88-105 average from 20-26 (124 OPS+), nor as mediocre as you’d think from his .276-13-61-67 for 27-33 (103 OPS+). Conditions helped his early stats, but his age 27-29 were the heart of the dead-ball ’60s; and after being traded off the Reds, he played mostly for poor offenses in tough parks the rest of his career. His arc teaches caution in projecting counting stats: Pinson had 2,453 hits through age 33, beating the pace of all but three 3,000-hits men, and ranked top-15 in XBH and total bases through that age to that date; and he was still productive at 33. But he slipped another rung at 34, and was done by 36. May have been two years older than his listed age, which could help explain his failure to advance from his early breakthrough, but I’ve found no published evidence.
Note: Three active players met the 20-21 standard for this group, but are still age 26 or under in 2014:
- Giancarlo Stanton: 132 OPS+ for 20-21 … for 22-23, 144 OPS+, with 26.0 Rbat and 120 games/yr.
- Justin Upton: 120 OPS+ for 20-21 … for 22-25, 121 OPS+, with 15.5 Rbat and 148 games/yr.
- Jason Heyward: 115 OPS+ for 20-21 … for 22-23, 115 OPS+, with 11.5 Rbat and 131 games/yr.
Combined Group: 25 men with OPS+ of 113 or more for 20-21 combined (900+ PAs)
- Age 20-21 … OPS+ 138 … Rbat 28.7 … oWAR 5.0 … WAR 5.1 … Games 149
- Age 22-23 … OPS+ 149 … Rbat 34.7 … oWAR 5.7 … WAR 6.2 … Games 150
- Age 24-26 … OPS+ 151 … Rbat 38.1 … oWAR 6.0 … WAR 6.3 … Games 151
- Age 27-29 … OPS+ 151 … Rbat 34.6 … oWAR 5.5 … WAR 5.7 … Games 139
- Age 22-29 … OPS+ 151 … Rbat 35.9 … oWAR 5.7 … WAR 6.0 … Games 146
The main points noted for the separate groups still hold for the composite: Substantial growth after 20-21; early peak; highly durable.
Looking beyond their 20s to age 22 onward, most of these 25 land in the top 100 for the stats we’ve been looking at:
- OPS+ (5,000 PAs) — 8 top-20 … 12 top-50 … 17 top-100 … 20 top-150
- Rbat — 8 top-20 … 13 top-50 … 16 top-100 … 19 top-150 (Bench is #156 overall but #5 for catchers 22+)
- oWAR — 8 top-20 … 12 top-50 … 18 top-100
- WAR — 8 top-20 … 12 top-35 … 17 top-100
The seven who fall outside the top 100 in oWAR are Magee (#178), Cepeda (#215), Pinson (#219), Cedeno (#245), with McInnis, Hoblitzell and Conigliaro outside the top-400. FWIW, the Hall of Fame currently has 147 position players elected on the basis of their MLB careers.
Does regular play at 20-21 predict durability?
The question is raised by this group of 25 averaging 146 games (per 162) from 22-29. For reference, just 13 players of any age maintained such an average over the last 8 years. But with a small sample, that durability could be coincidence. A health predictor from just playing regularly at 20-21 makes sense in theory, just because such a player most likely hasn’t had a serious injury yet, while one debuting later might have done — and past injury does predict future injury. But it would be hard to pin down such a factor from playing data alone, without injury data. Talent is also crucial to playing time. A typical 20-year-old regular has more innate talent than one who breaks in at 22; even if their “health skills” are equal, the one who broke through at 20 is likely to play more in the future because of that talent gap.
There’s some small sign that age 20-21 walk rates predict batting growth within this group.
- The top 13 in 20-21 walk (at least 52 BB/650 PAs) saw their 22-29 OPS+ rise by 14.4 points; the bottom 12 (49 BB/650 or less) averaged an 8.8-point gain in 22-29 OPS+.
- Four of the five who did not advance after 20-21 (excluding Tony C.) were among the eight lowest 20-21 walk rates of these 25, and the fifth (Kaline) saw his walk rate decline in 22-29, though it remained healthy.
On the other hand … Cobb had the lowest 20-21 walk rate, about half the median, and his OPS+ grew 25 points in 22-29. Cepeda started 2nd-lowest and barely improved, but gained 17 points in OPS+. And Aaron started with the 5th-lowest walk rate, then gained 40 points in OPS+, second-biggest growth in the group.
Certainly, the walk rate after 21 is meaningful, and not just because of its direct role in OPS+. Rating the change in BB/650 PAs from 20-21 to 22-29, the worst eight averaged -5 BB and -7 in OPS+, while the other 17 averaged +22 walks (all +9 or more) and +21 OPS+ (all +3 or more). But whether walk gains are more cause or effect of improved batting in this group, I cannot say.
At first glance, Cedeno looks like an outlier here, his walks growing by nearly half in 22-29 while his OPS+ remained static. But actually, his leap forward in walks came at age 21, along with his first big batting numbers. Gauging his 22-29 against just age 21, his walk rate held steady, but his OPS+ declined — which I suspect is far more common than a spike in walks with no gain in OPS+.
“Put me in, Coach!”
Opportunities for hitters age 21 or under have been historically low in the last 30+ years:
I don’t know if this is more cause or effect of the dearth of baby-faced stars. But I think it’s clear that expansion, free agency and long-term contracts have created more chances for “established” players, even when their established level isn’t good. There are 73 modern players with at least 4 years in their 30s of 300+ PAs and less than 1 WAR per 600 PAs; 45 of them did it since free agency, and 27 in the last 20 years.
It’s said the game’s been getting younger in recent years, and there has been a small drop in the average age. But even the most generous view of the trend in regulars 21 or younger shows them still far below historic norms. The last two seasons brought the heralded Trout, Bryce Harper and Manny Machado, but only two others got 200 PAs at 21 or under. With other factors arrayed against the very young — the arbitration structure, the growth of college as an MLB training ground (and their use of aluminum bats) — the status quo seems likely to hold for some time yet.
Alternate Groups based on 130 OPS+
(1) Had I drawn the “elite” threshold at 130 OPS+ (not 140) for both 20 and 21, four players would have shifted from Group 2 into Group 1: Magee, Kaline, Robinson and Griffey. That would lower the heights reached by Group 1 during 22-29, since the additions all had OPS+ from 134-154 in that period, while the Big Six ran 162-211. But it doesn’t really change the shape of the graph, or the fact that they all remained outstanding hitters throughout their 20s.
(2) Setting the threshold to 130 OPS+ for 20-21 combined lands 15 total: the Big Six, plus those noted in (1) above, plus Mathews, A-Rod, Crawford, Vaughan and McInnis. Four of those last five had 22-29 OPS+ of 144-156, with McInnis limping in at 108.
Forming any group of “Trout comps” requires compromise, but this might be the soberest approach. Critiquing my original groups, I made two cherry-picking accusations:
- Setting the “elite” threshold at 140 OPS+ to exclude Kaline — 162 at 20, 139 at 21, and 7th in 20-21 combined; and
- Including in the big group Aaron and Cabrera, who ranked 20th and 23rd in 20-21 OPS+ and both improved greatly in 22-29.
So, this in-between-sized group includes Kaline and drops Hank & Miggy. Also, cutting off at 130 includes McInnis, who declined sharply, and drops the 129’s of Cepeda and Cedeno, whose combined 22-29 OPS+ grew 7 points.
Anyway … This group peaks at age 25, but still were collectively great throughout 22-29. Here’s how many players outside the group had Rbat as good as the group average for each age; the numbers reflect the group’s early peak:
- Age 22 (40.0 Rbat) — 11 players
- Age 23 (43.6) — 20 players
- Age 24 (45.1) — 26 players
- Age 25 (45.9) — 37 players
- Age 26 (44.6) — 43 players
- Age 27 (40.6) — 62 players
- Age 28 (44.4) — 44 players
- Age 29 (43.7) — 49 players
In total, 159 players outside this group of 15 had any 22-29 season Rbat as good as the group average for that age. Nineteen in all had 5 or more such years: 8 within the group (including all Big Six), 11 outside the group (in bold):
- 8 years — Pujols
- 7 years — Cobb, Gehrig, F. Thomas
- 6 years — Hornsby, Foxx, Mantle, Collins, Speaker, Ruth, Musial, Aaron
- 5 years — Ott, Williams, Mathews, A-Rod, Greenberg, Mays, Cabrera
In conclusion, the strongest point I’d take from all of this: Even if Trout puts up bigger years from 22-25 than what he’s done so far, I would not expect further growth based on the normal aging pattern. Of the 25 players studied here, 17 logged their highest Rbat at 25 or younger, and eight had their two best Rbat marks by 25.
My thanks to you readers who stayed with this long post. I meant to get it done before Opening Day, but Michigan’s NCAA run cut into my blog time. As always, I welcome your feedback — even if you only want to say that projecting is a waste of time. For the record, I agree with you; I just can’t help myself. Happy Opening Day, everyone!
Subscribe to: RSS feed