67% goes to the top 5% of players.

Occupy MLB!

Its tough to make any changes though when 47% of the players are just looking for free stuff.

Interesting, I suppose, but painfully trying to compare apples and oranges in an attempt to declare the peach downtrodden.

Minor league players (OK, 98% of minor-league players) are not, at any given moment, qualified to be MLB players. Suggesting that they should be part of MLB's compensation system is roughly akin to suggesting that hospitals should pay pre-med students. Will some of them be doctors some day? Sure, but they provide zero current value.

Suggesting that they should be part of MLB's compensation system is roughly akin to suggesting that hospitals should pay pre-med students. Will some of them be doctors some day? Sure, but they provide zero current value.

Do you mean medical students? If so, then they might provide some current value, accompanying the doctors on rounds, looking pretty and so forth.

Do you mean medical students? If so, then they might provide some current value, accompanying the doctors on rounds, looking pretty and so forth.

In my particular B.S. construction, actual medical students would be akin to the guys in AA or AAA who are "on call" in case of injury, maybe the equivalent of 40-man roster guys.

In the larger point, I'd include pre-med undergrads. Most guys in rookie or short season never sniff high-A, much less provide major-league value. Those 40-man guys have survived many weedings to get to that level - they are a tiny and extremely successful minority.

From TFA:

The majority of professional baseball players, youngsters toiling in the minor leagues, “are effectively still bound by the old reserve clause”

Imagine if it were not so. You would have to go back to old-school independent leagues, which seem economically non-viable in today's entertainment landscape.

That's actually not an accurate summary of my original article. I was saying that pre-arb MLB players are still bound by the old reserve clause, not minor leaguers.

Boehner Refuses to Allow Debt Ceiling Vote Without Tax Cut for Alex Rodriguez

Imagine if it were not so. You would have to go back to old-school independent leagues, which seem economically non-viable in today's entertainment landscape.

Why would it affect anything in the minor leagues? If youngsters were free to sign with whoever they wanted the Yankees would still need somewhere to stash all their 400 prospects.

No doubt several of Rosenheck’s points are imminently debatable,

Yes, any minute now they will be debatable, perhaps even eminently debatable.

Although, for all intensive purposes, they have already been debated.

Ah, Srul, I think you're post honed in on there true problem.

If youngsters were free to sign with whoever they wanted the Yankees would still need somewhere to stash all their 400 prospects.

Guys were free to sign with whoever they wanted, with the reserve clause, prior to the draft being created. If, after signing or being drafted, the player wasn't both tied to the team and cheap, there would be no reason for the teams to pay to train a worker who is no more likely to ever work for them than their competitor.

That's actually not an accurate summary of my original article. I was saying that pre-arb MLB players are still bound by the old reserve clause, not minor leaguers.

Then I guess I'm quibbling with the intermediate author's distortion.

Sports is a business and is also entertainment. I'm shocked to find out that in a part of the entertainment world, stars with drawing power make significantly more than bit players.

If minor leagues were unionized, a number of teams would be dropped by MLB organizations due to higher costs. They would then attempt to use college programs in the same way that the NBA and NHL do as a free source of talent.

If minor leagues were unionized, we'd also see very hastily organized summer leagues in Mexico, the Dominican, etc.

When disparities such as these exist, it is not evidence of a rational or efficient marketplace

When quotes like these exist, it is evidence of the author's simple mindedness and pre-existing bias.

If minor leagues were unionized, we'd also see very hastily organized summer leagues in Mexico, the Dominican, etc.

... where labor rights make 1950s MLB seem like the height of worker empowerment.

... where labor rights make 1950s MLB seem like the height of worker empowerment.

Yup. Globalization, baby! Those Little Leaguers from Taipei will be next.

When quotes like these exist, it is evidence of the author's simple mindedness and pre-existing bias.

But it's impossible to argue that MLB is a "rational" or "efficient" marketplace since it's explicitly designed to not be a "rational" or "efficient" marketplace. The draft, the minor-league FA rules, the rule 5 draft, the ML FA and arb rules, waiver rules, trades, guaranteed contracts, 10 & 5 rights and whatever I've forgotten are most definitely not "rational" or "efficient" as those terms are used within the context of markets. Nobody from Ayn Rand to Karl Marx would consider it so.

Of course they might be perfectly rational in the sense that it's a system that works pretty well and probably better than its "rational and efficient" counterpart but for me to suggest so would be evidence of my pre-existing bias.*

*Presumably I am not best-suited to judge my own simple-mindedness so that is left as an exercise for the reader.

Presumably I am not best-suited to judge my own simple-mindedness

The East German judge gives your simple-mindedness a 2, giving the gold medal to the Russians.

Do you mean medical students? If so, then they might provide some current value, accompanying the doctors on rounds, looking pretty and so forth.

A Grey's Anatomy hi-jack at BTF. Never thought I'd see the day...

In truth, the more efficient it is the more extreme the salary spread will get. Baseball talent is not distributed normally, it is at the far end of the bell curve. For every superstar there are many all-star, for every all-star there are many regulars, for every regular there are many backups, for every backup there are hundreds of minor leagueres and for every minor leaguer there are hundreds if not thousands of 'just for fun' players.

The peak level, superstar players are so rare that they are worth tens of millions to a team, while utility/backup guys are lucky to get $1-2 million. The problem with salary by WAR is that is isn't a linear relationship - a guy who provides 2 WAR of value over a season is not worth 1/2 of a guy who provides 4, but far less as you can find more guys who can provide 2 WAR than you can find who can provide 4 WAR over a season.

To smooth it out the players union could push for a higher minimum wage, earlier maxing out of retirement benefits, and higher wages for the minors. Don't see the last point happening though as it is MLB players who run the union, not minor leaguers.

It's always been my impression that screwing over the young players also provides the added benefit of mitigating the struggles of smaller market teams as they get to hold on to cheap young talent for longer.

Guys were free to sign with whoever they wanted, with the reserve clause, prior to the draft being created. If, after signing or being drafted,

Yes, but they were signing up for possible lifetime service at uncertain compensation, hence they took a large bonus upfront if they could.

the player wasn't both tied to the team and cheap, there would be no reason for the teams to pay to train a worker who is no more likely to ever work for them than their competitor.

Presumably teams would take such things into account when they bid on untested players.

And the players would be bound by the terms of the contract they sign with the team whose offer they accept. There is a difference between having a market and anarchy.

#21 is a very perceptive comment. Some of the strongest unions anymore are in entertainment industries where of course the superstars get paid more than the rank-and-file; in movie acting, the discrepancy between star salaries and scale is far greater than that in baseball, for similar reasons. But for many other good reasons analogous to baseball (including the fact that many stars come up the hard way, and you can't make a movie without character actors and minor roles), unions flourish and members are mutually supportive. The guy playing Cop #2 for $800 a day (and one week on the set) does not automatically resent the fact that his fellow union member Brad Pitt is making $20M on this picture, and it's not a problem at all for his union as a whole.

Agreed on both #21 & #24.

Efficient is not the same as optimal.

Efficient is not the same as optimal.

Optimal isn't even the same as optimal. As soon as any 2 parties have different goals or priorities, optical goes right out the window.

In truth, the more efficient it is the more extreme the salary spread will get.

Maybe, but the distribution would be changed even more dramatically with younger players raking it in and older vets hanging on. (Heck, you can do the Hollywood analogy here as roles dry up for most actors/actresses as they age.)

Some of the strongest unions anymore are in entertainment industries where of course the superstars get paid more than the rank-and-file; in movie acting, the discrepancy between star salaries and scale is far greater than that in baseball, for similar reasons.

Yes, but any given movie has dozens and dozens of cop #2s. If we want to push this analogy, those guys are the Dominican summer-leaguers (who aren't getting $800 per day). John C Reilly is Nick Swisher and the chief's wife on Gray's Anatomy is Miguel Cairo. I am generally stunned at how much "small-time" actors get paid for movies or as TV regulars when they seem eminently replaceable.

The peak level, superstar players are so rare that they are worth tens of millions to a team, while utility/backup guys are lucky to get $1-2 million. The problem with salary by WAR is that is isn't a linear relationship - a guy who provides 2 WAR of value over a season is not worth 1/2 of a guy who provides 4, but far less as you can find more guys who can provide 2 WAR than you can find who can provide 4 WAR over a season.

s

John is just doing a much better job than I of making the point I wanted to make.

I think the missing pieces are roster size and positional scarcity. If you could play 100 players at once, price per WAR would be far more closer to linear, teams would just accumulate as much WAR as they could at reasonable prices.

But since you can only have 8 positional players and 1 pitcher on the field at once, and only a limited number on the roster to serve as backups, you will pay a premium for players who are likely to give you lots of WAR in one position or pitching slot.

While the author and Walt are obviously right that the specific limits and restrictions in the MLB talent market affect it, those limits don't mean that the result is anything greatly different than what it would be if the market had no limits. The best players would still make many times more than the worst, the authors bias is to think somehow "wealth" would be redistributed from highest paid players to lowest paid players, and I think it's unclear whether it would true, it would just accrue to a slightly different set of highest paid players.

Optimal isn't even the same as optimal. As soon as any 2 parties have different goals or priorities, optical goes right out the window.

In an economics sense, optimal just means that you can't make one party better off without making someone else worse offer. It has no connotation of fairness.

Example:

Let's say I own two farms and generate $125,000 p.a. in income by working both (state A). You are unemployed.

If I could hire you for $50,000 p.a. to help, and generate $200,000 in income (state B), state A is not optimal. We are both better off making the switch.

I could also pay you $60,000 (or $40,000), to generate that $200,000 total income. Those are also optimal, but no more optimal than state B, b/c it doesn't grow the pie any more. How we split up the total doesn't impact optimality.

The best players would still make many times more than the worst, the authors bias is to think somehow "wealth" would be redistributed from highest paid players to lowest paid players, and I think it's unclear whether it would true, it would just accrue to a slightly different set of highest paid players.

Possibly, possibly not. I would guess not (well, ignoring minor-leaguers). The average salary is still around $2, maybe $2.5 M per year I think. The highest salary is 10 times that but the highest WAR is only 5 times average. $/WAR might be non-linear but I don't think it's that non-linear. And of course the guy with 10 WAR made the minimum salary (well, 5/6 of it really). And $/WAR is only as high as it is because of all the artificial $/WAR bargains among pre-FA players.

There are 109 position players who had at least 6 WAR from 2010-12. Somewhere between 43 and 55 of them had not reached FA until this offseason at the earliest. That frees up a lot of money to be spent on a smaller pool of players.

Really all you need to do is graph the distribution of WAR and the distribution of salary on top of each other. I'm guessing there's a bigger peak in the low salary range than there is in the low WAR range and a longer tail in salary than there is in WAR. But I could be wrong.

The thornier question is what a "natural" system would look like. A "Charlie Finley" system of everybody is an FA every year is also not "natural". That is, in the absence of a draft and the reserve clause, it still makes sense for some team to offer Bryce Harper a lot of money for a lot of years. It still makes sense for the Rays to offer Longoria long-term security for a "reasonable" salary. And, who knows, if he's the best hitter available one offseason, maybe it still makes (as much) sense to sign Soriano for 8/$136.

But I think you do have to show that a "natural" system would lead to a similarly restricted supply and movement for it to lead to a similar distribution. I would guess that's unlikely.

But I think you do have to show that a "natural" system would lead to a similarly restricted supply and movement for it to lead to a similar distribution. I would guess that's unlikely.

The same forces are at work. Assume Finley's system, the Yankees would still value wins at a much higher value than other teams, and would still only have 8 starting position player spots and 5 starting pitcher spots. To give themselves a high confidence of having a winning team, they'd want to buy much more than the average amount of WAR, and that requires securing the services of more than a few well above average players. Given a $180M budget, it seems very likely they'd overbid for the high WAR players and pitchers, since they'd still have a great deal left over for bench and relievers.

For example, let's assume Cashman is slow on the trigger during an offseason in this "Finley rules" free agency system. His goal is to get at least 12 WAR above average. That requires 12 3 or better WAR starters, but if all of the 3+ WAR 3B, SS, and 2B are snapped up at market value, he now needs to compensate by finding better than 3 WAR pitchers/starters. There is only so many of those and to ensure he gets the ones he needs, he's going to be forced to "overpay".

In reality, the baseball market works exactly like this. There are very few players available who can fit your team needs, if you don't get the right ones there is no recovery, you can't make up the loss in WAR easily in other positions. Look at the 3B market this year, if your team needs a 3B, the only option is to massively overpay. Isn't it pretty clear why offering Youklis $12M for next year makes sense to the Yankees, even if Youklis might be worth less?

The average salary is still around $2, maybe $2.5 M per year I think

It was in excess of $3M.

http://www.cbssports.com/mlb/salaries/avgsalaries

The average baseball salary on opening day, based on salary studies by the the Associated Press, and the percentage increase or decrease. Figures were obtained by the AP from management and player sources and include salaries and pro-rated shares of signing bonuses. In some cases, parts of salaries that are deferred are discounted to reflect present-day values (x-adjusted for 25-man rosters from $1,073,579 figure for 28-man post-strike rosters).

  Year    Average Pct. change
  2012 $3,440,000 4.1
  2011 $3,305,393 0.2
  2010 $3,297,828 1.8
  2009 $3,240,206 2.7
  2008 $3,154,845 7.1
  2007 $2,944,556 2.7
  2006 $2,866,544 8.9
  2005 $2,632,655 5.9
  2004 $2,486,609 (-2.7)
  2003 $2,555,476 7.2
  2002 $2,383,235 5.2
  2001 $2,264,403 13.9
  2000 $1,998,034 15.6
  1999 $1,720,050 19.3
  1998 $1,441,406 4.2
  1997 $1,383,578 17.6
  1996 $1,176,967 9.9
1995-x $1,071,029 (-9.9)
  1994 $1,188,679 6.1
  1993 $1,120,254 3.3
  1992 $1,084,408 21.7
  1991 $  891,188 53.9
  1990 $  578,930 12.9
  1989 $  512,804 N/A

The highest salary is 10 times that but the highest WAR is only 5 times average. $/WAR might be non-linear but I don't think it's that non-linear. And of course the guy with 10 WAR made the minimum salary (well, 5/6 of it really). And $/WAR is only as high as it is because of all the artificial $/WAR bargains among pre-FA players. ... Really all you need to do is graph the distribution of WAR and the distribution of salary on top of each other. I'm guessing there's a bigger peak in the low salary range than there is in the low WAR range and a longer tail in salary than there is in WAR. But I could be wrong.

When a team signs a higher salary player, isn't part of the team's thinking that not only do they get higher WAR from that player than cheaper alternatives, but also that there is less variance in the WAR that player will produce each year? If teams do get less risk in performance at higher salaries, then the observed spread in $/WAR should be less at higher $.

In an economics sense, optimal just means that you can't make one party better off without making someone else worse offer. It has no connotation of fairness.

I'll buy that. I'm an engineer, and in our use, it is always "Optimal with respect to..."

I'll buy that. I'm an engineer, and in our use, it is always "Optimal with respect to..."

Do you also use optimal learning techniques with regard to data collection and actions?