8 replies to this topic

[Todd Benzinger]

The phrase "old player skills" keeps popping up on SoSH, and the way it is being used here and around the internet is often misleading or confused. Sorry to be pedantic, but it is a useful/ intriguing concept that is being rendered meaningless by misapplication.

The concept is that YOUNG players who set out with the skillset of an (effective) old player are likely to decline SOMEWHAT faster than young players with young player skills.

The "old player skillset" is: power, walks, low average, lack of speed. It does not involve defense; as a young player, as James points out, Tom Brunansky, the poster boy for the concept, was a very good defensive player.

Hence, Trot Nixon, Kevin Youkilis and Mike Lowell don't really have "old player skills." They already hit for average, and aren't overly dependent on the HR. Perhaps Ortiz could be put in the category, although he hasn't really been a low average hitter. Adam Dunn is probably the current poster boy for the young player with the old player skill set. What James observstion tells you, though, is that you don't want to gamble on Dunn in his mid to late 30s--so even if he follows the pattern he's got 5 years or so of peak play left.

Jason Giambi did not have "old player skills" when YOUNG, he developed them as he aged. Having old player skills when you are old is not a sign of anything, it's normal.

James gives a useful brief overview of the concept in the New Historical Baseball Abstract in the entry for Tom Brunansky. The page is avail for reading on google books, although you need a google/gmail account to read it:

James' Definition (Brunansky entry)

Bruno's Stats via BB Ref

[Vermonter At Large]

I tried to ignore this topic, but while doing research this morning I happened upon and old Baseball Magazine article from the 1910's in which the writer shows a bunch of photograph's of players' hands to support the thesis that a player's hand shape is linked to their optimal position in the field. Double sigh ...

I love Bill James' approach to baseball analysis, but there are times when he overly generalizes certain theses and this is a horrible example of this, along with his "pitching style" analysis also found in the new abstract. Its nice to categorize things, but he grossly ignores complexity in his categorizations.

There are dozens of players with "old player skills" that had nice long careers, and lots of reasons why both multi and single-tool players rapidly decline in unexpected fashion. In Bruno's case, he just wasn't all that good to begin with.

I'm not saying that these theses have no merit whatsoever. In the "old player skill" thesis, you can pick out that certain skill sets are more likely to decline than others, and that single skilled players are naturally more vulnerable.

Its great to try to identify certain axioms, but the basic human tendency to try to reduce the world to general theories of everything pretty much sucks in baseball. There's just so much more to the game ... embrace the complexity.

[Worst Trade Evah]

What James offered was a note in a player comment about a general observation. He set up a simple study and offered an interesting and plausible result. It's not exhaustive nor definitive, and it wasn't intended to be. Criticizing it on those grounds seems unfair.

Even so, I don't see why he's wrong or why this comment is "horrible" analysis. There may be an underlying truth here about player development he's found -- part of the analyst's job is to interpret and understand the deep patterns that might ground complex topics. No one does that better than James -- his targeted little studies of similar players actually are generally very robust.

Dozens of young players with old player skills having very long careers? Which ones? Got a list? How does that list compare to a comparably valuable set of young players who do not have old player skills? I'm also not sure why you're introducing single- vs multi- skill set distinctions, since the "old player skills" notion reflects a suite of player performance attributes.

[Todd Benzinger]

I have to admit that James actual concept could use some help--what it doesn't need is people vaguely remembering the phrase and throwing it around like it meant something.

For instance, one of the things that spring to mind after reading James' Bruno entry is "Hmm, Dewey--so wait, did he have young player skills or old player skills?" He actually seems like a guy who approriately developed old player skills as he aged--he started showing more patience and power. But his average also went from mediocre to above average.

So 4 problems with James' quick and dirty study are 1) he seems to be using only his own top 100 guys as his sample, which eliminates people with really quick declines and/or utterly mediocre performances; 2) he himself applies his definition rather casually; 3) his comparison group is guys with "opposite" skillsets from the same group... But obviously, speedy guys with decent average,low OBP and low SLG who made it into the top 100 had to either develop or at least last a long time, so it's no surprise they were more productive as older guys, and they don't tell us anything about guys with (say) high average, SLG, and OBP as they age and 4) his definition is about young guys with these skills, but he seems to have picked the guys for his sample on their whole careers, rather than their first 5 or so years.

I would like to see a study of guys fitting a very specific profile, low ave, high OBP and SLG, VS speedy/toolsy guys/high draft pick guys, VS guys with median ave, median OBP and median SLG all in a certain window, like first five years up to age 28 or something. I think that would do more to investigate James hypothesis than his actual study.

But the flaws of the study don't make it better to worry that Trot is developing "old player skills" becuase he's singling too often, as if that were indicative of a quick decline, when that;s got nothing to do with nothin'. (Sorry Rudy--I appreciate the thread, it's just that people have been throwing the term around and I noticed your thread).

Edited by Todd Benzinger, 14 June 2006 - 01:06 PM.

[Rudy Pemberton]

No offense taken, it was a crappy title and didn't appropriately address what was going on with Trot (or what I thought was going on with Trot) at all.

[bakahump]

I always thought BJ missed the bus on this one.

OS=Old Skills
YS= Young Skills

If a player has "Old Skills" wouldnt that put him ahead of player with young?

Assuming that both will improve (which I think is fair when considering young players), wouldnt it stand to reason that a player who starts out better in some catagories (Like BBs and or HR) will with yet more experience improve? IOW an "OS" player who comes into the league drawing 70BBs a years may with a few more years experience become a player who draws a 100. While a "Young Skills" player who draws 15 BBs his rookie year may someday develop into a 50BB a year guy.

Same with HR. Just because a player has power at 22 doesnt mean they cant or wont develop more as they reach 30 or 35.

Rather the say that a player with "OS" declines quicker I would counter that they Peak sooner (which BJ and I are in agreement on) and that they improve Less as they age.

They improve less not because "OS" players are incapable or bad at improving but rather that they already have some of the skills normally gained via experience. "YS" players on the other hand have more to learn so they learn more. An "OS" players learning curve is shallower while a "YS" players is steeper if you will.

All things being equal I would think that a Player who hits 30 HR and draws 100 walks(the most common things considered "OS") as a 25 yo would continue to do those things well into thier late 30s and possible 40's. While a player with "YS" may not begin to draw those numbers in those catagories until there early 30s.

BJs flaw in his "Brunansky arguement" was using "toolsy" players vs his "OS" players. Toolsy players who learn Patience (in the form of BB) and gain power (in the form of HR) are quite likely to improve more significantly then a "non toolsy" player who came up with "OS"s

To egregriously over simplify If a player comes into the league as a 5 (due to possessing "OS") on a scale of 1-10 they can certianly become an 8,9 or 10. If a player comes into the league as a 3, they too can become an 8,9 or 10. If at 33 Both are 8s who has improved more? (YS) Who is more valuable? (Probably a "toolsy YS player who has now developed what he lacked")

In closing BJ has not mentioned any reason as to why a player who can cook or paint or compose music (all skills...and in total thats all where talking about here) is going to "age faster" and have there performance suffer for it. Its not like a "OS" player is destined to suffer from arthritis or bad eyesight while a "YS" player is immune to such infirmities.

There is such a thing as an "OS" player. And they reach thier peak sooner. There is no reason they fall off the cliff faster.

[Todd Benzinger]

While I agree with the basic sentiment, I think part of what is going on is that low BA is a very important part of the formula. If you are a low BA guy, you are already depending on walks and power to give you value (think good Bellhorn) and if you slip further in BA, or in either power or patience, you are toast. In other words, the old player skillset gives you less margin for error.

[Rudy Pemberton]

Quote

While I agree with the basic sentiment, I think part of what is going on is that low BA is a very important part of the formula. If you are a low BA guy, you are already depending on walks and power to give you value (think good Bellhorn) and if you slip further in BA, or in either power or patience, you are toast. In other words, the old player skillset gives you less margin for error.

Yes...I totally agree. I think we saw this with Bellhorn- his game really depending on walks and a high average on balls in play, and a bit of power. Because of his high K rate, he was really teetering on the edge. What we then saw was a +15% spike in his K rate, which made for a miserable season, esp. since his avg. on balls in play dipped a bit too.

Jeremy Giambi is another good example of the old player skills, isn't he?

Brad Wilkerson could be another example, I think.

[Vermonter At Large]

Worst Trade Evah, on Jun 14 2006, 12:36 PM, said:

What James offered was a note in a player comment about a general observation. He set up a simple study and offered an interesting and plausible result. It's not exhaustive nor definitive, and it wasn't intended to be. Criticizing it on those grounds seems unfair.

Even so, I don't see why he's wrong or why this comment is "horrible" analysis. There may be an underlying truth here about player development he's found -- part of the analyst's job is to interpret and understand the deep patterns that might ground complex topics. No one does that better than James -- his targeted little studies of similar players actually are generally very robust.

Dozens of young players with old player skills having very long careers? Which ones? Got a list? How does that list compare to a comparably valuable set of young players who do not have old player skills? I'm also not sure why you're introducing single- vs multi- skill set distinctions, since the "old player skills" notion reflects a suite of player performance attributes.

<{POST_SNAPBACK}>

By single skill, I refer mainly to bat speed. Lots of hitters lose bat speed around the age of 30, and unless they have something else in their skill sets to counter for it, they drop off the table. I think this applies more to hitters who use their legs rather than their wrists to generate bat speed. Wrist hitters, such as Williams, Aaron, Mays, Sheffield, Manny, Giambi, Winfield, Reggie, etc seem to have longer careers, but thats a very general assertion too.

A big chunk of the guys that qualify as "old players" whose careers were shortened had them shortened by injury, particularly knee injuries. You'd have to throw out all of those guys to make your study valid.

Its interesting that of the eight players listed in the book as having "opposite skill sets," five (Pinson, L.Smith, Beaumont,LeFlore and Rivers) were essentially done by age 31, and only three (Brock, Roush and McGee) had long sustained careers. This is because these players were also basically single skill players (leg speed) which is also a fleeting thing.

As for the general quality of the study, James himself summarizes it thus, "Although the extent to which this is true is not as large as I might have expected, there is no doubt that it is true."

As you say, I don't think that even James ever intended this to be anything more than an interesting aside, certainly not a serious analysis. Its fun and points to some deeper issues, but in and of itself it doesn't hold water. I'm not criticizing the study itself, but I think maybe too many people take it too seriously.