Prince #2?

Huh.

What is young? 27 or under this season? And what is value? And who is eligible? Anybody likely to make the team in ST?

Boston is #4 with Pedroia, Lester, Buchholz and Bard? If eligible players include anybody who played in MLB last year, the Red Sox pick up some talent not likely to make an impact this year.

Jurrjens over Kershaw?

Sandoval #5?

Yeah, I've got some young talent...in my pants!

Really, though, Bill James seems addicted to these toys these days. This one seems kind of arbitrary but the link only goes to a blogger talking about James' list and not James' methodology.

Really, though, Bill James seems addicted to these toys these days. This one seems kind of arbitrary but the link only goes to a blogger talking about James' list and not James' methodology.

If it's the same methodology he's used in the past, it's just about entirely worthless - most notably, there's no actual position adjustment. (There is a speed adjustment, which is like a position adjustment except that it ignores catchers.)

I don't mind the list, but the order is ummm, bad. Is that seriously the order? Among other problems Joe Mauer is way too low. But hey whatever, typical spring training time waste article.

Is James still working for the Red Sox? I guess he's keeping all of his good stuff for them.

Prince #2?

This list has been a fan of Prince for a couple years now. I believe he was #1 on it last year or the year before. It seems that James is more bullish on Prince than most analysts, many of whom seem to disparage him because of his position and weight. I hope James is right about Prince's future.

I'll trade you Matt Kemp for Troy Tulowitzki, then, Rockies fans.

Tulowitzki just seems low in general, but then again maybe that means I'm overrating him.

How many MVPs do you know who can't even crack the top 10 of player value for their age group?

Here's the description of the method from the '08 Bill James Handbook:

First, I eliminated from my study all players who were 29 years old in 2007 or older, since 29-year-olds in 2007 are now 30, and 30-year-old baseball players are not young... Second, I figured runs created by each player. Third, I made a "speed adjustment," since speed correlates strongly with defensive value, and defensive value is more difficult to measure. Fourth, I divided that total by the runs scored/runs allowed per game by the player's team, thus building in context adjustments. Fifth, I multiplied that by the number of years the player had left before he was 33 years old.

That reads to me like something that would have been acceptable maybe 20 years ago. It's not obvious that there's any adjustment for outs made, so the baseline is 0, which is way too low. No position adjustment outside of the nebulous speed factor (James's method would presumably rate Lou Brock as a better prospective defender than Brooks Robinson), and a park adjustment that's distressingly rough, considering park adjustments are not very difficult to get right (James does them perfectly well in the Handbook itself, and then ignores them for this). It looks to me like James just threw this together, and nobody would take it seriously at all if it didn't have his name attached to it.

Fifth, I multiplied that by the number of years the player had left before he was 33 years old.

This is the kicker. At best, it's an "esimate" of "expected value produced between now and age 33."

Wow. No Grady Sizemore?

I must be drastically overvaluing good defensive center fielders who put up 75 XBH a year, pop 25 dingers, walk 90 times, and steal 25 bases at an 80% clip.

Adam Lind is 25? For some reason that makes me feel young...

No Franklin Gutierrez?

Wow. No Grady Sizemore?

The method is backward-looking, not forward-looking, and based only on last year's production. (Again, James goes to the trouble of doing projections, and completely ignores them for the sake of this list.)

No Franklin Gutierrez?

No real attempt to account for defensive value.

At best, it's an "esimate" of "expected value produced between now and age 33."

I'm actually kind of OK with this part... it's arbitrary, of course, but it's not nearly as bad as the context adjustment that penalizes a player whose team happens to have a great offense and bad pitching, or the "speed adjustment" that may well give, say, Bobby Abreu more expected defensive value than Yadier Molina (if Abreu was young enough to be on the list, anyway). If it was some kind of weighted average of the last 3 years' Win Shares, with allowances made for players who haven't been in the majors that long, multiplied by the number of years between now and 33, then it might produce something that puts a shortstop who's a few runs below average defensively, has a 145 OPS+ over the last three seasons, and steals 30 or more bases a year ahead of a first baseman who's the same age, is a bit more below average defensively, has a 152 OPS+ over the last three seasons, and has the running speed of a turtle fighting its way through Steve Carlton's rice vat. (Fine, so Prince has 8 career triples and 14 career steals. Hanley's speed still puts him at least even on offense alone.)

It's a toy method that's not even worthy of being compared to decent toy methods. It certainly doesn't deserve the authority of being interpreted and debated as if it were some kind of definitive list.

Bill James calls Dodgers #11 in Young Talent Inventory

Obviously...this was before the Garrett Anderson signing.

Adam Lind is 25? For some reason that makes me feel young...

He's actually currently 26, being born July 17, 1983. So he's still young, but not quite as young.

The method is backward-looking, not forward-looking, and based only on last year's production.

That's just bizarre. Basing a list of the Most Valuable Properties In Baseball solely on one year of data?

I have the utmost respect for Bill James. He's done as much to help people understand baseball as anyone has with any sport, ever. He's an original, a visionary, and a brilliant guy. But this is plain silly.

The list is also based entirely on performance in the majors, so young players who have a lot of upside but haven't yet put it together (Wieters, Price, Bruce) don't count for much. That alone makes this list worthless.

Here's the description of the method from the '08 Bill James Handbook ...

The list is also based entirely on performance in the majors, so young players who have a lot of upside but haven't yet put it together (Wieters, Price, Bruce) don't count for much. That alone makes this list worthless.

Not anymore. He has changed the metric. Subscribers to BJOL can read it or you can buy the book. There are plenty of legitimate criticisms of the process, but one should at least know what they are critiquing before doing so.

The list is also based entirely on performance in the majors, so young players who have a lot of upside but haven't yet put it together (Wieters, Price, Bruce) don't count for much. That alone makes this list worthless.

Why? Doesn't actual major league performance mean a lot more than projected performance based on minor league numbers? Isn't there also a greater degree of certainty when projections are made using MLB numbers?

Which isn't to say that the list isn't flawed.

Not anymore. He has changed the metric. Subscribers to BJOL can read it or you can buy the book. There are plenty of legitimate criticisms of the process, but one should at least know what they are critiquing before doing so.

How does he account for them?

Not anymore. He has changed the metric. Subscribers to BJOL can read it or you can buy the book. There are plenty of legitimate criticisms of the process, but one should at least know what they are critiquing before doing so.

My bad. Anyone care to give an indication of what those changes are?

On the other hand: Prince Fielder is still ranked ahead of Hanley Ramirez, despite having been a worse player in each of their four full major league seasons and having the body type of an oil tanker.

Without giving away the milk for free, I'll say that one of the changes James uses is to take the ranking of the player on the MLB "Top 50 Prospects" list and work it into a supplementary score calculation. If that score is higher than the one obtained from his major league production, he uses that instead.

Why? Doesn't actual major league performance mean a lot more than projected performance based on minor league numbers? Isn't there also a greater degree of certainty when projections are made using MLB numbers?

Those are fair points, but if the goal is to figure out what kind of performance a team is going to get from its young core over the next few years, the analysis has to take into account guys who just broke in but might not have established themselves. I'm just not sure there's any way to quantify this type of analysis because there are always going to be a lot of fuzzy variables.

My bad. Anyone care to give an indication of what those changes are?

Three biggest ones are (1) inclusion of a prospect rankings weight for players who don't have significant MLB time (BJ says in the article that he needs to increase that weight; I agree) (2) use of multi-year data rather than single year data, and (3) changing the age to 35 rather than 33.

My biggest problems with it is that he has abandoned Runs Created for the even more blunt Season Score. It has no position adjustment or ballpark adjustment, a very rough tool that is designed for much broader questions. He uses this to determine which teams have the most talent at the MLB level and lists the individuals as a throw away line. At the team level some, but not all of my concerns go away.

Also, that ordinal ranking at the top of the article does not include the numeric result. In the book King and Prince are tied at #1 and there is no significant difference between their number (267) and Ramirez's (265).

(edited for clarity)

Isn't there also a greater degree of certainty when projections are made using MLB numbers?

Not that I can tell. Dan might be able to comment further.

I can say this with confidence. Gary Huckabay's Vlad was the first projection system to incorporate minor league numbers. When I looked at Vlad's projections I didn't find them any less accurate when based solely on minor league numbers. Now Vlad wasn't a particularly accurate system, so that could have hidden any problems using the minor league numbers as inputs.

James used to produce a similar type of list using AV and an age function (plus a penalty for being a pitcher). I think the old way produced a generally better list.

Also, that ordinal ranking at the top of the article does not include the numeric result. In the book King and Prince are tied at #1 and there is no significant difference between their number (267) and Ramirez's (265).

I could write a lengthy treatise on how there should be a significant difference, and it should be in Hanley's favor, because he's a young shortstop of legitimately historic quality whereas Prince is "merely" a very fine player... or I could do this:

He uses this to determine which teams have the most talent at the MLB level and lists the individuals as a throw away line.

In other words, we're spending way too much effort on this discussion. Which I guess is nothing new around here.

I can say this with confidence. Gary Huckabay's Vlad was the first projection system to incorporate minor league numbers. When I looked at Vlad's projections I didn't find them any less accurate when based solely on minor league numbers. Now Vlad wasn't a particularly accurate system, so that could have hidden any problems using the minor league numbers as inputs.

I'm not familiar with the Vlad projection system, but it would greatly surprise me if projections based partially or fully on minor league numbers weren't less accurate than those based solely on major league numbers, if only because of the additional error or uncertainty (choose whichever term you prefer) introduced by the translations needed to convert MiLB numbers into MLB equivalents - especially when the projections involve MiLB performance from multiple minor leagues, resulting in a chaining of the translations and the associated uncertainties.

I think it was one of James early tenant's that minor league numbers (when properly translated) are just as predictive as major league numbers. On the other hand, translations aren't perfect, and one might even translate them wrong (BP's Wieters projection last year), so they aren't as useful.

He uses this to determine which teams have the most talent at the MLB level and lists the individuals as a throw away line.

In other words, we're spending way too much effort on this discussion.

No, this is not a description of the list above, because it has two Giants listed, and two Brewers, and contradicts the "rankings of the Young Talent Inventory of all 30 teams" which is also in TFA.