What is the value of an MLB player in trade?

Back when trades started, it was pretty straightforward: a player’s trade value was the aggregate value of his talent over the course of his career. That was a reasonable approximation to the truth up until free agency began in the mid-1970’s. But since then, things have been different. Salaries and years remaining have become critical. The defining moment of the current era perhaps came when Manny Ramirez was placed on waivers and nobody took him—the man who was arguably the best all-around hitter in his league had negative trade value. That’s how much things have changed.

But what factors define trade value?

Here are some important things to consider:

1) Talent. A player’s current talent—offensive, defensive, and pitching—is an important aspect of his trade value.

2) Salary. A player’s current and future salaries are important to considering trades.

3) Age. To a limited extent, older players are worth less in trade than younger players, all other factors considered.

4) Injury. A player’s injury history and how it forecasts to the future affects his trade value.

5) Chaining. The talent of the next-best player available for a given role affects a player’s trade value for a given team.

6) Marginal Revenue Product (MRP). The MRP for a player will vary with a team’s competitiveness in a given year. Regardless of a player’s talent, the MRP for that player will be low if the team is not in contention for the postseason. In close pennant races, a difference of one win can have large MRP, and in postseason play, the MRP can be even greater. There are four levels of competitiveness that result in big paydays for a ball club:
a. The Pennant Race
b. The Divisional Series
c. The League Championship
d. The World Series

7) Park Factors. Some players are better suited for particular home fields because of their particular talents and skills.

8) League Factors. Some players thrive in one league and struggle in another.

9) Intangibles. A player’s personality will mesh better with certain clubhouses, managers, sportswriters and fan bases than with others.

That list may not be comprehensive, but it is certainly already very complex.

I know of no all-inclusive study that has attempted to look at past trades in an attempt to form a model considering several of these factors. My understanding is that modeling free agent market salaries is where the current research is being done, with models such as BP’s MORP doing a relatively good, but incomplete, job of modeling salary with respect to talent. I don’t know of studies matching dissimilar talent to talent, in the context of factors such as those listed above. If there’s research already done in this area, I’m eager to hear about it.

If not, I may work on it myself. We’re at the level of knowledge where good posters here are absolutely sure that trades are unreasonable because one player is worth more than another, but with some good posters on each side of the same proposal. The higher salaries get and the longer contracts get, and the more minor league prospects are involved, the more there seems to be disagreement. The community knowledge here is pretty good. If there are still disagreements, maybe there hasn’t yet been enough research done.

Great idea for a thread.

One thing that I would add is reputation. Maybe it belongs under talent (perceived talent anyway) or intangibles but Alex Gonzalez is my case in point. It seems like everyone under the sun is blown away by his defense, except for people who have crunched the numbers (what's available anyway) and are unimpressed. A disconnect between someone's actual talent and their perceived talent can have a huge impact on their value. And of course the problem with trying to measure that is that it is so subjective. Theo might be in love with a player that someone like Brian Sabean might have no use for (to pull a very random example out of my ass).

What about potential? I mean Wily Mo is a great example. His "potential" as a power hitting, middle of the lineup guy makes him more valuable than say a Youkilis, who may (or may not) put up better numbers now, but doesn't project to be the player that WMP does...i.e power is more valuable than avg, because it's more rare.

On the pitching end, the ghastly Mets trade of Kazmir for Zambrano..From an unenlightened view, that trade was a bad one because the Mets were trading an unproven, but unlimited potential guy for a 4-5 starter that they thought would put them over the top.


Also, and this may be covered in chemistry or intangibles... the Red Sox were forced to take a hit when trading for Mirabelli. His trade value to other teams was NOTHING, but to us it was High because of Wakes. Trading a future starting catcher and possibly a closer for a washed up catcher who can miraculously increase the benefits of one pitcher...The Padres knew they had us grabbing our collective ankles on that because we were desparate...and they took advantage.

One thing I would like to add... Often a trade is reviewed after the fact and judged by what the actual outcome was. That method is obviously hard to argue with. Ive always felt it was somewhat unfair. The actual results are important yes, but no one who goes all in with pocket aces is ridiculed if he has a bad beat. In poker, there is a recognition that long odds coming in didnt change the intelligence behind the initial decision.

For instance, look at the Bagwell trade. It looks terrible on a historical scale if you look at the production each team received from the players involved. What exactly were the odds that Bagwell developed into a top 10 all time first baseman though? Anderson was also made a free agent a year early in a collusion suit, but thats a minor point... The trade was always laid at the hands of poor scouting and talent evaluation, but Im not sure I buy that. Bagwell had some promise but hadnt exactly demonstrated much power in the minors.

Does the fact that Bagwell came in for Houston on extremely long odds really make this one of the worst trades of all time? IMO there have probably been 3 or 4 worse trades the last couple years. Maybe its b/c we cant see the long odds like we can in poker, but Ive always felt that this method of evaluation trades(ie, career stats, career win shares, etc...) is incomplete. Something can be a good decision at the time and turn out terrribly wrong.

Not to belabor the point, but if a manager has two pinch hitters available and one has all the odds, stats in his favor and he doesnt get a hit was that a bad decision? I say no. Just as if he sends the other guy on the bench who by all measures we have has far worse odds to get a hit but manages to anyway. That wasnt a good decision just b/c it paid off.

EDIT: Sorry, meant to add in that Bagwell had 6 HRs professionally at the time of the trade and we were stocked with corner IFs at both the majors and minors. Anderson also did extremely well in his time here... but anyways... you all get the point...

I have one more....marketability. While this could be filed under intangibles, it is pretty clear that The Triumvirate clearly look at this.

This was one thing that was discussed after the Nomar trade and during the A-Rod talks. Nomar was not one to do many public appearances, corporate events, etc. The thought was that A-Rod was more marketable, and better able to bring money through the doors of Fenway Park in other ways than just for games. David Ortiz would also be a prime example: he is now the "face of the franchise", makes numerous appearances, public service announcements, etc., furthering the "Red Sox Brand". Strangely enough, if Papi were not here, or was not willing to be marketed the way he is, it is possible that ownership might have gone the extra mile to keep Johnny Damon and take advantage of his marketing ability.

Of course, in some cases, it is more about ticket sales. For example, Mets' season ticket sales jumped immediately when they signed Pedro.

One other thing is "Other Factors"

If someone isn't really doing well in a city (A-Rod, BK Kim) they are more likely to be traded. I think this can affect their value slightly negatively.

Thanks for all the feedback. Here's a synopsis of what I've learned:

1) "Talent" is composed of at least three parts:

-- Statistical talent
-- Reputation for talent
-- Potential for future talent (i.e. tools)

2) "Intangibles" includes marketability of a player and established relationship problems.

I've also noticed that nobody had a link to a study in this area...thanks to everybody who read this and didn't post--the number of hits suggests that there's not a well-known study.

Alskor, the Bagwell trade may be lopsided in retrospect, but what I'm interested in doing is evaluating trades at their moment of execution--and, as you write, it looked better then than it does now, because Bagwell gained power beyond what we might've expected from his MiLB stats back in the Steroid Era , and because Anderson went free agent a year early.

***

The challenge in trying to do a study is that so many of these things are difficult to quantify...

JHB,

How would you account for relative value due to roster construction? For instance, a hypothetical team with four Roger Clemens may want to trade for a fifth, but the incoming Clemens would have a lower relative value to them than to a rival team with no resident aces.

Another wrinkle on that...can you knock trades that are legitimately "win now" trades if the player(s) traded away have a higher future value than the player(s) acquired? If a playoff appearance in Year 1 is more valuable to a franchise than potential playoff appearances in Year 2+, can the trade be assessed without taking that into consideration?

Chaining. In either case, a team acquiring a Roger Clemens displaces another pitcher from the starting rotation. The difference in talent (probably measured by WARP1) at the moment of trade--not after the trade--is what would matter.



I can't knock them--my intent would be to try to quantify them. I absolutely would expect to find a premium for deadline trades, particularly for small-market teams getting their one chance for glory, and I'd like to capture how great that premium is.

No, no, that was my point. My point was it would be interesting to have a system to evaluate trades at the time of the trade. The way we evaluate trades now(looking at the results afterwards is terribly flawed. If we could put a number on the probabilites of a player of Bagwell quality when he was traded ever becoming HOF Bagwell... well that would dramatically change the way we evaluate trades. I think we might find very interesting results... IF you can find a way to measure it and find repeatable results that predict something useful... gonna be tough. I think MORP is a great step in this direction. Its still going to be nearly impossible to balance these elements in a useful way(ie, what is the difference in value between two similar players, one who's highest PECOTA projection is far higher ceiling than the other... which positions are more valuable to each individual team at each individual time...)

I just started to rant and forget to bring the point back home.

Im not sure I have much useful to add on the subject of a way to measure this.




Im still not sure that was clear enough... I meant... how would the Bagwell trade be viewed if we could quantify the qualities of all the players at the time, their value to their respective teams at the time, their prospective near future value, the rarity of their skills and the availability of others who could perform at their level, and the odds of the players getting better at the time of the trade(PECOTA basically).

I still dont know how you would ever find a way to properly balance these things, but I think a crude MORP/PECOTA hybrid that was adjusted for year and the market would be a nice thing to have.

Trade value is in the eye of the buyer, not the seller.

For example, trade value of a solid DH in the Red Sox eyes is probably close to zero. Angels would probably put a higher value on it.

I don't think the trade value of a player can be assessed in a vacuum, and a price tag hung around his neck. The relative needs of each possible counterparty dictate what a player's trade value is in each situation, and I expect this value would vary accordingly.

Of course--that's the whole point. In the case that you suggest, chaining comes into play. Almost nobody (if anybody) in MLB trumps Papi as a DH, so chaining would capture that.

And chaining is part of the proposed model, and it's a part that can be quantified.