Username:	Password:

Remember me

Forgot your username/password?

Print Email

Home>Articles

Soto and Acuna

By Bill James

September 1, 2018

Soto and Acuna

On who is likely to have a better career, Acuna or Soto:

I don’t remember the specific context, but years ago you wrote about a player (might have been Frank Thomas) that they came to the majors with old player skills, and thus weren’t able to improve as much as other players.

Soto has displayed ridiculous plate discipline, so he may be less likely to "develop" as a hitter. Acuna may be able to improve that skill over time a bit more since it’s not as developed, and his slugging percentage is insane for his age.

If I had to hazard a guess, I’d say Acuna. Barring major setbacks they’re both going to be outstanding. The reality is at least one of them is liable to fall off due to injuries or some other unpredictable factor, which is probably why you answered "no idea"?

I’d love to have either one on the Dodgers. I spent a bunch of time explaining that Max Muncy, being 27, isn’t likely to get much better than he currently is.

--Christopher (Hey, Bill Question)

I did a brief study related to the issue. I took all players in history who earned at least 10 Win Shares and had at least 300 Plate Appearances in a season at age 21 or younger, up to 2009; no players from the current decade. This left a group of 361 players, but with some duplicates because some players had multiple seasons qualifying.

In that group, I tried to find players who matched as nearly as possible in terms of Win Shares, Runs Created, Outs Made and Age, but who were different in walk rate. Also, I put in a time-line separator so that a player from 1885 would not tend to be matched with a player from the 1960s, and then I made a hard rule that a 19^th century player would not be compared to a post-1900 player.

To site one of the better examples, Joe Medwick in 1933 was matched with Harlond Clift in 1934. Each player was 21 years old, and each player made 431 outs. Medwick created 97 runs and was credited with 24 Win Shares; Clift created 92 runs but for some reason was credited with only 18 Win Shares. Besides being separated by only one year, both players played in the same park, Sportsman’s Park in St. Louis, Medwick in the National League and Clift in the American, but the teams shared a park. Clift, however, drew 84 walks while Medwick drew only 26, which of course means that, to be even overall, Medwick was stronger in some other areas. In this case, Medwick probably went on to a somewhat better career—not much better—but that would be consistent with your thesis.

A counter-example: Joe Morgan, 1965, was matched with Garry Templeton, 1977. They were the same age, made about the same number of outs (458-461), while Morgan had advantages in runs created and Win Shares based in part on his having drawn 97 walks to Templeton’s 15. Morgan, of course, went on to a much better career than Templeton.

Greg Gross, 1974, was matched with Rocco Baldelli, 2003. Gross, 21 years old, had 85 runs created, 16 Win Shares, made 447 outs; Baldelli, also 21 years old, had 86 runs created, 18 win shares, made 482 outs, but Gross drew 76 walks while Baldelli drew only 30. In that case, neither man went on to a highly successful career, Baldelli because of injury and Gross because the teams of his era did not recognize the value of a slow player with no power having a .375 on base percentage.

Rickey Henderson, 1980, was matched with Cesar Cedeno, 1972; same age, same value, Henderson walked a lot and Cedeno not as much. Ted Williams in 1939 was matched with A-Rod in 1996, and Ted Williams in 1940 was matched with Frank Robinson in 1957. Al Kaline in 1955 was matched with Ty Cobb in 1907. That was neat. Some of you who are older will remember that Cobb was the youngest man ever to win the batting championship, until Kaline. Cobb was born on December 18, Kaline on December 19—thus Kaline was one day younger than Cobb when he won the batting title, thus broke the record. That was a famous fact back in the day; neat coincidence that they popped up as a match in the study. Mel Ott, 1929 (113 walks) was matched with Joe DiMaggio in 1936 (24 walks), while Mel Ott, 1930 (103 walks) was matched with Sam Crawford, 1901 (37 walks). Jimmie Foxx, 1929 (103 walks) was matched with Hal Trosky, 1934 (58 walks). Gary Carter, 1975 (72 walks) is matched with Ted Simmons, 1971 (36 walks). Jason Thompson, 1976 (68 walks) is matched with Harold Baines, 1980 (19 walks).

Some of these make intuitive sense and some of them don’t. Willie Mays, 1951, is matched with Stuffy McInnis, 1911, which seems jarring and inappropriate, but Mays and McInnis were both 20 years old and created 79 runs each. Mays did this while using 353 outs; McInnis, only 338, and McInnis had more Win Shares, 21 to 19. It meets the standards of the study. It seems jarring, I think, because we remember Willie Mays, but from Mays’ rookie season backward to McInnis’ is only 40 years, while from Mays’ rookie season forward to now is 67 years, so he is actually more a part of McInnis’ time than he is of our own.

An example of a case that seems to "prove your point"—understanding that we are all sophisticated enough to know that a sample of one proves nothing—would be Ed Yost, 1948, and Roberto Clemente, 1956. Both players were 21 years old; Yost created 73 runs with 442 outs, Clemente created 71 runs with 406. Each player was credited with 14 Win Shares, both of them playing for bad teams. Yost (The Walking Man) drew 82 walks, while Clemente drew only 13, although Clemente was far better in other ways. Obviously Clemente went on to a much better career, which seems to prove your point: he had a pathway for growth open to him.

Another case of a 1940s/1950s comp with a similar message is Whitey Lockman, 1948, and Orlando Cepeda, 1959. Both men played for the Giants, Lockman in New York, Cepeda in San Francisco. Three years later (1951) Lockman would be the first baseman for a New York Giants team that went to the World Series, losing to the Yankees; three years later, Cepeda would be the first baseman for a San Francisco Giants team that went to the World Series, losing to the Yankees. Lockman had an outstanding season as a 21-year-old, scoring 117 runs. Lockman, 21 years old, created 102 runs with 427 outs and was credited with 20 Win Shares. Cepeda, also 21 years old, created 110 runs with 436 outs, and is credited with 23 Win Shares. Lockman, however, walked 68 times while Cepeda walked only 33 times. Consistent with your thesis, Cepeda went on to a great career and is now in the Hall of Fame, while Lockman faded away after a brilliant start to his career.

Donie Bush, 21 years old, was the regular shortstop on the 1909 Detroit Tigers, who won the American League pennant for the third consecutive time but lost in the World Series. Travis Jackson, also 21 years old, was the regular shortstop for the 1924 New York Giants, who won the National League pennant for the fourth consecutive time but also lost the World Series. Each player made 439 outs. Bush created 83 runs, Jackson 81 runs, although, mostly because of offensive context, Bush was credited with 28 Win Shares to Jackson’s 20. Bush, however, drew 88 walks to Jackson’s 21. Again, Travis Jackson went on to a Hall of Fame career and Bush did not, although there is not a huge discrepancy in terms of career value, but it can accurately be said that Bush, who was almost the Eddie Yost of his era, never improved as a hitter after his fine season at age 21. His career high in WAR was at age 21.

I made 120 matched sets of players. From this, I eliminated those in which the difference in walks was less than 20; for example, Jose Canseco in 1986 was matched with Eddie Murray in 1977, an outstanding match in terms of similar age, runs created, outs made and win shares, but the difference between them in walks is only 17 (65 to 48), and it doesn’t seem reasonable to include them in this study with such a small difference in the critical area. This reduced the number of matched sets in the study to 88.

Conclusion

The study argues fairly convincingly that the Frank Thomas Thesis, as applied to walks, is not true. While of course no one study of the issue is exhaustive or perfect, there appears to be little possibility, based on this study, that young players who walk a lot have less room for development than players who walk less. If anything, the opposite appears to be the case.

We will call the high-walk players the "Ted Williams type" players and the low-walk players the "Joe DiMaggio type players". In the base year of the study, the Ted Williams-type (high walk) hitters hit more home runs (983-860), drew more walks (5577-2346) and scored more runs (6830-6357), but the Joe DiMaggio-type (low-walk) hitters hit for a higher average (.298 to .279), had more doubles and triples, drove in more runs and had (cumulatively) 1,000 more total bases (18,518 to 17,526). The Joe DiMaggio-type (low-walk) players had a .334 on base percentage, .440 slugging, .774 OPS; the Ted Williams-type (high-walk) players had a .363 on base percentage, .421 slugging, .784 OPS. The two groups were identical in terms of age (average 20.7614 in both cases), and very near to one another in terms of the controlled factors of runs created, outs made and win shares. The Joe DiMaggio group was actually 1% ahead in terms of Win Shares, 1,635 to 1,614.

The Joe DiMaggio group was slightly FASTER, on average, than the Ted Williams group, and also was slightly larger. The Ted Williams group had six more left-handed hitters.

In the rest-of-their careers performance, however, the Ted Williams-group (high-walk) hitters did slightly BETTER than the Joe DiMaggio-type (low-walk) hitters. The two groups were almost identical in terms of career at bats (558,689 to 553,954) and in terms of career RBI (79298-79279). The difference between them in walks relatively narrowed in rest-of-career performance (71864-44617); the difference between them in walks was 138% in the base year study, but only 61% in rest-of-career performance. (I should alert the readers that I am sometimes citing rest-of-career performance and sometimes final career totals, which is nearly the same thing but not EXACTLY the same.)

However, the advantages of the low-walk group narrowed as well. The rest-of-career batting average for the Joe DiMaggio group went down from .298 to .294; for the Ted Williams group it went up from .279 to .285. The Joe DiMaggio (low-walk) group had a 9% advantage in doubles, as young players; in the rest of their careers it was 5%. The advantage in triples stayed at 19%.

What did NOT narrow, however, was the home run gap. The Ted Williams group had a 15% advantage in home runs, as young players. In the rest of their careers, it was 32%. Thus, while the Joe DiMaggio group drove in more runs as very young players, the Ted Williams group drove in slightly more in the rest of their careers, winding up almost exactly even, as noted before.

In the rest of their careers, the Joe DiMaggio group had a .801 OPS and created 6.11 runs per 27 outs. The Ted Williams group had an .842 OPS and created 6.658 runs per 27 outs. The Ted Williams group had 29 Hall of Famers in the 88 players; the Joe DiMaggio group had 27 Hall of Famers. The Ted Williams group won 38 MVP Awards; the Ted Williams group won 27. The Ted Williams group played in 249 All Star games cumulatively; the Joe DiMaggio group, 149.

These numbers do not prove definitively than the Juan Soto high-walk rookie is a BETTER bet than the Ronald Acuna low-walk rookie, and they do not address the related but more general question of "old player’s skills" versus "young player’s skills." But it would be difficult to reconcile this study with the theory that the a low-walk total represents a growth advantage for a younger player.

COMMENTS (30 Comments, most recent shown first)

MarisFan61
At least somewhat negating what I just said.... (I seem to have to do a lot of that) :-)

You could say it completely negates it. Certainly it further supports Guy's notion.

I went and looked at those 10 players on Win Shares -- i.e. the 10 current players who I think meet all of the criteria I wanted to observe.

The reason I went and did it wasn't because I really thought it would show differently than how I did it before, but because I realized it wouldn't be hard, and I just enjoy doing stuff like this.
But it does show differently.

I got the Win Share data from Baseball Gauge, not because I wouldn't rather have used the data from here but because it's easier.

Here are the 10 players, together with my finding from each method -- first, in italics, the finding from before (just OPS+), then, bold,the new finding, by Win Shares.

("yes" means it fits Guy's notion; "no" means it doesn't -- all by my judgment; others might judge otherwise)

Starlin Castro yes yes
Lonnie Chisenhall no yes
Freddie Freeman no yes
Jason Heyward yes yes
Eric Hosmer either way no
Wilson Ramos no no
Anthony Rendon either way either way
Anthony Rizzo yes yes
Giancarlo Stanton yes yes
Justin Upton no yes

The results from before (by OPS+): 4 do fit, 4 don't, 2 "either way"
(a little different than I said before; dunno if I miscounted before or if I'm changing my mind on how I assess 1 of them)

The results from Win Shares: 7 do fit (!!), 2 don't, 1 "either way"

Chisenhall goes from "no" to "yes" because of much-reduced playing time since age 25.
Freeman goes from "no" to "yes" because....I'm not sure why.
Upton goes from "no" to "yes" in part because -- score one for Guy -- poorer defense (as shown by Baseball Gauge's Win Shares) from ages 27 to 29 (he shows better again on defense this year, age 30), but mainly because of some generalized difference between how he shows on OPS+ and Win Shares.

10:29 PM Sep 12th

MarisFan61
Guy: I think you should have quit when you were about even. :-)

I agree that nothing is anywhere near definitive with such small samples, and I said so (repeatedly).
But, there AREN'T any more examples! What I wound up looking at was the total field of players that I thought would meaningfully address what you said. (I assumed you meant specifically the quite recent period. If not, depending on how far back you meant it to be relevant, of course there could be more.)

I doubt that the specific things you're mentioning there would help further.
First of all, what you said about how only looking at players who are still in the majors at age 28-30 biases the results is very weak. Theoretically it could be a factor (and I said so), but I doubt it affects even a single example, because....
If you want to retain the idea of only considering players who were already above average by age 24-25 (remember, if we don't do that, there isn't a real impressive percentage who fit what you said), it's highly, highly, highly unlikely that ANY of them would no longer be in the majors by ages 28-30.
As I explained, the reason I didn't look at players who weren't already at least 28 was that there would be too much of a chance to miss subsequent improvement after age 25 (because they haven't yet had enough years past age 25). I think there's little doubt that this would be a far larger error than the above, which frankly I think is zero error.

And second, I don't agree that fielding and baserunning are any less likely to improve after age 25 than hitting. I can't deny that you might be right, but I don't think so -- and anyway, look: I was conservative in saying which players met the opposite expectation from yours; I didn't say a player did improve after age 25 unless there was a very palpable gap -- a degree of clear improvement which is very unlikely to be undone by any possible trends or non-trends in defense and baserunning. I would suggest that if anything, those additional factors would be more likely to produce more examples of improvement after age 25, not fewer.

-----------------------------

BUT, so as not to have this main thing be lost in the shuffle:

YOU MADE A GOOD POINT.
The stuff I looked at convinced me that there's something to it: More players than I imagined do essentially reach their peak by age 25, and I do think it's more common among very good or great players than among lesser ones. Even though, as we both keep saying, not a lot can be told from small samples, to me it was enough to be convinced.
I just think you're going a bit too far about the degree of incidence.
9:39 PM Sep 12th

Guy123
MF61: If you want to do this, you need to consider fielding and base running too, which are less likely to improve after 25 -- using WAR or WS is probably the easiest way to do that. More importantly, you can't look only at guys still playing at age 28-30, as that creates a huge bias toward post-age-25 improvement -- many of the players who crashed and burned after 25 won't even be playing in MLB at these ages. And most important, you need *much* larger samples.
3:51 PM Sep 12th

MarisFan61
......Looks like there are just 2 other players who meet all those criteria including the part about "at least 110 OPS+" at either age 24 or age 25 (or both):

Giancarlo Stanton 164, 159; since then, 120,169,122
Justin Upton 110, 124; since then, 133,119,106,136,116

Stanton clearly fits what Guy said; I'd say Upton doesn't although you could argue he does.

So -- Overall, IMO that makes 5 out of 10 from among these players who meet all the criteria I indicated who fit what Guy said; you could argue for as many as 7 out of 10.

Granting, again, that we can't conclude anything from such a small sample -- which, if we include all those criteria, is just 10 current players -- it's much more of a fit to what Guy said than I would have expected.
9:07 PM Sep 11th

MarisFan61
OH -- forgot to mention, I abandoned the criterion of "only take players who reached an OPS+ of 110 at age 24 or 25" because it was hard enough to find players who met the other criteria at all.

Just counting the ones from among those who did reach OPS+ of 110 at age 24 or 25, it's these 8 guys:

Starlin Castro 115, 84; hasn't reached higher than 115
Lonnie Chisenhall 86, 115: since then, 77,98,126,128
Freddie Freeman 139, 132; since then, 157,157,141
Jason Heyward 109, 117; worse since
Eric Hosmer 99, 122; since then, 102,133,98 (could argue either way)
Wilson Ramos 104, 110; since then, 91,66,121,98,136
Anthony Rendon 125, 93; since then, 108,138,124; either way, I suppose)
Anthony Rizzo 152, 146; less good since]

Among these players, Guy's thing does fit better: 4 out of 8, could say 5 of 8.

I might work a little more later to see who else I might find by going further down the alphabetical list.
5:58 PM Sep 11th

MarisFan61
".....most MLB players will not improve appreciably after age 24-25. I think that's especially true for those who already have significant MLB experience and are above-average players by that age."

I'm interested and happy to look into that too.
Not because I'm as confident as I was about the other thing that it'll show otherwise (although I'd guess it will), but just because I'm interested -- and I'll find it interesting and informative if it turns out to be so.

I'll use a similar approach to what I did on the other thing. I'll go down the alphabetical list of current players -- this time starting not at the top but where I left off, so as not to have any redundancy nor to get any suspicion that I'm trying to bias it -- i.e. it might look like I'm doing this thing and approaching it that way since I already know that the basic impression from that portion of the list was what I expected.

Again, I'll only take players whose ages as shown on baseball-ref.com are between 28 and 30, and from those, I'll try to select the ones who fit what you just described.
This time, to make it more standard and duplicable, I'll look only at hitting, and the way I'll draw the line of "above average" is to only take players who reached an OPS+ of 110 at age 24 or 25.
Also, to make sure they fit your criterion of "who already have significant MLB experience," I'll only take players who had at least one season equivalent prior to age 24.
(BTW, in the name of throughness and transparency, here are examples of guys I didn't include, due to that criterion: Freddy Galvis, Aaron Hicks, JD Martinez, Brad Miller; did include Anthony Rendon. I don't think those are real questionable, but I figured I'd better say.)

Again, I'll take the first 15 that appear.
I'll indicate their OPS+'s at age 24 and 25, then indicate what they have reached thereafter.

Here we go....
The first two numbers after each name are their OPS+'s at ages 24 and 25.

Starlin Castro 115, 84; hasn't reached higher than 115
Lonnie Chisenhall 86, 115: since then, 77,98,126,128
Freddie Freeman 139, 132; since then, 157,157,141
Didi Gregorius 81, 89; since then, 97,105,114
Jason Heyward 109, 117; worse since
Eric Hosmer 99, 122; since then, 102,133,98 (could argue either way)
Jose Iglesias: 101 at age 23, 99 at age 25 (not in majors at 24), worse since
Logan Morrison 89, 95; since then, 111,92,101,134,73
Mike Moustakas 91, 77; since then, 75,119,110,117,110
Salvador Perez 91, 88; since then, 91,106,91 (could argue either way, I suppose)
Wilson Ramos 104, 110; since then, 91,66,121,98,136
Anthony Rendon 125, 93; since then, 108,138,124; either way, I suppose)
Anthony Rizzo 152, 146; less good since
Jean Segura 70, 68; since then, 122,111,114
Andrelton Simmons 75, 84; since then, 91,102,108

--------------------------------

Granting, again, that a sample of 15 doesn't prove anything, these data aren't well supportive of what Guy said, even with his refined definition, although they're not far off.
It's less unsupportive of it than I would have expected. :-)

By my assessment, only 6 out of the 15 fit it, even giving some benefit of the doubt, although you could argue for perhaps one more.
5:51 PM Sep 11th

Guy123
Jeff Zimmerman has written a lot about aging curves, for example:
https://www.beyondtheboxscore.com/2011/5/31/2199146/hitter-aging-curves
https://www.fangraphs.com/blogs/hitters-no-longer-peak-only-decline/
If you google 'baseball aging curves' you will find plenty of other research.

My point is not that age 25 > age 27, but simply that most MLB players will not improve appreciably after age 24-25. I think that's especially true for those who already have significant MLB experience and are above-average players by that age. I suspect there is some performance gain from having experience at the MLB level, separate from aging per se -- so 25-yr-old rookies may see some improvement on average -- but that's just speculation on my part.
12:10 PM Sep 11th

Arrojo
Yikes typo - 29, not 39.
5:21 PM Sep 10th

Arrojo
Thanks MarisFan61. Regarding peak age, what you thought, and what I thought, are both in line with what Bill James has published, namely that the peak is 26-39. I’m not aware of any published studies that he’s done in line with what Guy123 is claiming which is that position players peek at age 24-25. He states it like it’s common knowledge. I’d like to see such studies.
5:20 PM Sep 10th

MarisFan61
.....About the age at which position players now reach their peak, not to belabor the point but....... OK, to belabor the point :-) ....actually to look at it further....

I'll be systematic this time.
I'm going to go down the alphabetical list of major league players this year -- not all the way down, need I say :-) I don't have that kind of devotion to such things, but I'll get a fair little sample. I'll look for players who are currently age 28 to 30, and take the first 15 of those (per the age shown on baseball-ref.com), and see who did or didn't reach their peak by about age 24-25.
I'm following Guy's exact wording -- "reach their peak," as opposed to looking for some center of their peak or for some tippy-top 'peaky peak," and where it's debatable, I'll try to give the benefit of the doubt to "yes" on what Guy said, i.e. I'll give what seems to be the lowest reasonable age.

As to why I'm using 28 and 30 as the boundaries for current age:
-- I figure that starting at any lower of a bottom current age than 28 has too much chance to miss a further rise after age 24-25.
-- If I use players who are older than 30, I might not be meeting what Guy meant by "now." And anyway I figure that's enough of a range to give me plenty of players pretty quick.

BTW this kind of look inevitably misses players who reached their peak early and then haven't lasted in the majors till age 28. To that extent, it probably slightly cheats Guy's assertion, but I'd guess not much, and that what we see will be indicative, to the extent that a sample of 15 can be.

I'll only look at players who have at least 100 PA's this year. Why?? I don't know :-) ....it just feels like the ones with fewer are likely to be mostly players we're not that interested in. I realize it's possible that this further cheats Guy's assertion, but I don't mean to.

I have no idea who the players will be or how it'll come out, except to the extent that I'd guess it'll show the usual peak to be closer to where we've long thought it is, than to 24/25.

So.....here they are, with the baseball age (as shown on baseball-ref) of the year in which each player "reached his peak," as best I can judge, trying hard to be fair and also taking some account of fielding and baserunning (by which I mean base stealing, for this) but going mostly on hitting.
I'll take playing time into account. Like, if a guy has similar averages or OPS+ at ages 25 and 26 but had much less playing time at the earlier age, I take the later age.
Example: That's why I say 26 rather than 25 for Jackie Bradley Jr.

Of course anybody feel free to say if you think I'm being inaccurate or unfair on any of these.

Matt Adams -- a toughie right off the bat; I'll say 25
Ehire Adrianza -- 28 (so far, and at least)
Jesús Aguilar -- 28 (ditto)
Nick Ahmed -- 28 (ditto)
Abraham Almonte - 26
José Altuve -- I would say 26, but you could argue for 24
Elvis Andrus -- 27
Austin Barnes -- 27
Tim Beckham -- 27
Brandon Belt -- 25
Justin Bour -- 27
Jackie Bradley Jr. -- 26
Kole Calhoun -- 26
Mark Canha -- 26
Curt Casali -- none really; if have to say something, 26 or 29

This fits much more with what we've long thought, than with 24-25.
What I'd thought: "Reaching" the peak at about 26, and having the peak period extend from there to about 29.
11:38 PM Sep 6th

LesLein
In the New Historical Baseball Abstract you addressed this in comments about Tom Brunansky. You compared 25 sets of outfielders based on batting average, on base percentage, slugging percentage, and stolen bases. The outfielders were picked from the top 100 outfielders.

Your conclusion was that outfielders with "old player skills" peaked earlier and declined faster than those with "young player skills."

8:20 PM Sep 6th

MarisFan61
......Granting that this isn't going to be any meaningful-sized sample, I'm going to look at some random players.
Well maybe not really random because unavoidably the ones who come to mind in this context probably will predominantly be players that my mind suspects DON'T follow the pattern you said, but it's not like I'm trying to pick guys who don't fit it.

Didi Gregorius: no; still apparently rising, in his age 28 season.
Aaron Judge: tough to categorize and too early to tell; might fit what you said
Giancarlo Stanton: no
Brett Gardner: maybe he's too old to consider for what you're saying, since you said "now" and I don't know if an older player fits "now," but, no, he doesn't fit
Aaron Hicks: no
Votto: perhaps same issue as Gardner, but, absolutely not
Goldschmidt: not really; we'd have to stretch the age range that you said (he was closer to 26)
Rizzo: YES, apparently
Lindor: don't know yet
Arenado: not far from what you said, but not really within it

WHERE are you coming from?????

6:31 PM Sep 6th

MarisFan61
Guy: That would be news to me (i.e. position players tending to reach their peak at an EARLIER age than before) and I think to most.

Not saying it's wrong, just that it would be news.

(Not just that it is so, but that it's CLEAR!)
4:20 PM Sep 6th

Guy123
but we know (thank you Bill) that players improve until age 27, so Soto has one more year to improve than Acuna, all else being equal.

While calculating aging curves is very challenging, it seems pretty clear now that position players reach their peak by about age 24-25. For players like Soto and Acuna who reach the majors at a young age, it may be even earlier. Mike Trout was as good at age 20 as he will ever be, and Harper did the same at age 22. Mel Ott also reached peak performance level by 20, Williams and Musial by 22, Mays at 23, Mantle 24.

Being a year younger does give Soto an edge in terms of likely future performance. But we shouldn't expect 7-8 years of improvement from these two -- 2 or 3 is much more likely (and zero is possible).
3:00 PM Sep 6th

MarisFan61
.....I did come up with at least one way to see it on which their careers weren't far apart:

All-time rankings at their positions

I don't mean that I'm asserting that they're close on that, just that I think it's fair to say that at least arguably they're close.

Like, giving it a quick/dirty look, baseball-ref.com has Clift as #37 at third base, which, while I never feel that I know enough to try to be specific about a ranking of such a player (even looking at everything available), seems low to me.
Looking at the players on there, my best guess is that he'd be better placed 5 to 10 spots higher.

They have Medwick as #16 for left field. Looking at that list -- oh I dunno, what the hell am I even trying to do here.... :-) ....we could easily see him 5 to 10 spots lower, which would still have him ahead of Clift but awfully close.
12:17 AM Sep 6th

baudib1
I wouldn't make much of the Medwick comment. I think in the context of a study like this, it's not a big deal to say Medwick wasn't THAT much better. We're not trying to project the career totals of Soto/Acuna here. Medwick is an A- star and Clift is like a B+.

10:56 PM Sep 5th

MarisFan61
I also was surprised at the Medwick-Clift thing, and, even though I've sort of loved Clift ever since he started being written about as a negleted star, it seems a real stretch to say that Medwick's career was only "somewhat better" and "not that much better." All I know is that no matter how hard I try -- and I tried pretty hard :-) -- I can't find a way to make it seem that way.
Career Win Shares -- Medwick way ahead.
Top Win Share seasons -- Medwick very far ahead.
Career "WAR" -- Medwick far ahead.
Top "WAR" seasons -- Medwick distinctly ahead.
Naked career totals -- of course, but that's not what we really want to consider for this kind of thing.
MVP voting performance, all star appearances -- of course, although ditto.
Anything else? I can't think of anything else that's real applicable.
3:52 PM Sep 5th

Arrojo
Soto is one year younger, born December 1998 vs Acuna December 1997.

Using OTS - Acuna's OTS is .20 and Soto's is .21

Seems to me that one year younger and a higher OTS means Soto is the better prospect. Yes, that's a very simplified approach, but we know (thank you Bill) that players improve until age 27, so Soto has one more year to improve than Acuna, all else being equal. And his 2018 is better than Acuna's

Barring a September collapse, I'd be drafting Soto before Acuna. :)
3:43 PM Sep 5th

hotstatrat
photon01 - I know you are not asking me, but how much is "not much"? To use Hall of Fame Monitor or HoF Ink is not the usual way people determine how much better someone is than someone else. It is a way to measure their chances of getting into the Hall of Fame. Different things. The best tools we have for a rough measure of someone's career worth are WAR and Win Shares. Both FanGraphs and B-Ref. agree that Medwick has the edge over Clift about 55 to 39. You could say that is "not much". The difference in Win Shares is larger (309 to 214) if I understand their use properly which I don't claim to, but even at that difference, I suggest it is crosses the border into nitpicking to ask Bill to defend his use of "not much better".
2:35 PM Sep 4th

phorton01
Sorry - duh - it wasn't the first paragraph, it was the third.
6:36 AM Sep 4th

phorton01
I know this is weird, but what jarred me the most about this article was your comment in the very first paragraph that "Medwick probably went on to a somewhat better career—not much better—[career]".

I'm interested in what led you to add that additional qualifier. From everything I can see, Medwick did in fact have a much better career than Clift. Certainly in terms of accumulated value, he far exceeded Clift (HOF Monitor 200 vs. 40, Gray ink 226 vs. 70, etc.). Even in peak value, it seems like Medwick's five year peak was far superior to Clift's.

Just wondering what led you to make that point?
6:35 AM Sep 4th

tomindc2334
Bill, my recollection is that your first reference to a young player having an old player's skills was to Tom Brunansky, who was rookie in 1982. You might have said the same thing about Alvin Davis, who came up two years later.
10:48 PM Sep 2nd

Guy123
Very nice study. It's interesting that the HR gap between the two groups grows, while all other gaps narrow. I suspect this means a high walk rate is an additional indicator of a young hitter's power, beyond his HR total in any given season. While we tend to talk of walks as a discrete skill of the hitter ("plate discipline"), to a significant degree it is the opposing pitchers who decide how often a hitter will walk. When they walk a young hitter frequently, they are telling us how dangerous a hitter they believe he is. So in a sense we can see the walk rate as a scouting report -- a forecast of a hitter's SLG -- provided by some very knowledgeable scouts on the mound.

7:20 AM Sep 2nd

baudib1
I believe Alvin Davis, not Frank Thomas, was the player Bill referred to when he first used the term "old player's skills."

The comment I remember Bill making about a young Frank Thomas was that he correctly described him as one of the best hitters in baseball, before he got to the majors. Thomas actually did improve (kinda), which is a tall order considering he was the best hitter in the AL the moment he got there. He is one of several great players who was on pace for a career year in 1994 (.353/.486/.729, all career highs) and was on pace for a 50+ homer season.
7:09 AM Sep 2nd

MarisFan61
Bill: I could well imagine (wrongly, perhaps) :-) that on something like this, in theory there could have been some factors that tended systematically to vary between lower-walks and higher-walks guys who nevertheless had similar overall production, which would make there be a difference in which group had better peaks (on the average) and which had better careers.

The main specific possible such factor that I thought of was speed. You note that in the groups you looked at, one group was a little faster. I thought it was conceivable that in general there might be a larger such difference, and that if this was so (I hope this isn't too convoluted; I know that you don't like Rube Goldberg type stuff, especially if it's wrong) :-) ......and that if this was so, the faster group would have tended to be less good on things not involving speed, like power; and that those other things, especially power, would develop further over time than would speed-related stuff, and that therefore the slower group would reach a higher peak -- but that the faster group would have longer careers and higher career value.

Anyway, I see that in these actual pairings that you came up with, the same player is higher on both peak and career, in every case where there's a clear difference on either one.
12:17 AM Sep 2nd

3for3
Got my $3 worth on day 1 of the month...the rest is butter (why should the rest be gravy, when it can be butter?)
7:44 PM Sep 1st

Steven Goldleaf
The kind of article that will keep BJOL going as long as Bill wants it to.
5:59 PM Sep 1st

bjames
Previous comment:

BTW, to me "who becomes better," which to me means mainly PEAK, would be an even more interesting bottom line than who has a better "career," but then again I'm more into peak stuff on almost any question.

These would always be the same. One GROUP of players could not be better in peak value while the other group is better in career value, unless you selected the group specifically for that trait. Otherwise, peak value and career value are so closely tied together that the group that was stronger in one would always be stronger in the other.
5:15 PM Sep 1st

trn6229
Nice article. I recall when Alvin Davis came up with the Mariners in 1984, he played firstbase and hit 27 homers and walked 97 times. He played only nine years in the majors. He was 23 years old in 1984. Wade Boggs walked a lot and he had a long career. It will be interesting to follow the career arcs of Soto and Acuna. I wish them good health and success.
3:06 PM Sep 1st

MarisFan61
Maybe those matched sets give us fodder to look for what factors, if any, do correlate strongly with who does become better.

BTW, to me "who becomes better," which to me means mainly PEAK, would be an even more interesting bottom line than who has a better "career," but then again I'm more into peak stuff on almost any question.

Probably it comes from how it was when we were kids, where the main thing we noticed was who was the best player on the field and how good they were at everything -- at that moment, and maybe for next week or next month. We weren't much taken by "career." :-)
2:13 PM Sep 1st

Soto and Acuna

COMMENTS (30 Comments, most recent shown first)

Leave a comment

Report inappropriate comment


Type of Abuse:
Comments: