Kristen, I don't think there's any question Stanton has more value. The trade the Braves made for Teixeira was for one year and two months (or about $15.5M of the $21.5M Teixeira made in 2007-'08).

Stanton, on the other hand, still has one pre-arb season plus his three arbitration years remaining. The problem from the Marlins perspective is the Mark Teixeira trade effectively altered the landscape of baseball in that GM's are much more stringent on holding onto their prospects. Add in the new collective bargaining agreement, and teams are ever more unwilling to part ways with their cheap, controllable talent.

So even though that's the type of package they desire, I'm not sure they'd be able to get it even if they wanted it.

One thing I find reasonably interesting is that the Marlins are shopping the shortstop they just acquired from Toronto, Yunel Escobar. Shall we put on our conspiracy theory caps for a second? Who in the Major Leagues has more depth at shortstop than the Rangers? Could the Marlins potentially move Escobar to a team in need of a SS and eventually make a move at acquiring one of the two Rangers shortstops?

Probably not, but it's worth considering, especially this time of the year. The Marlins really only have one commodity that the Rangers would be interested in, and we know Jon Daniels is going to do something this offseason. Probably something big.

There's no way to create a real-life formula for "fair trade" because there is no such thing as "objective trade value" for a player. It's all subjective to the whims of his front office's point of view, and always shifting as well any time things happen regarding his team..

So the same player's Trade Value would be different on different teams, even though he's the same player.

With Miami, while on your scale Stanton might be "worth" 150 or 175 or whatever, in real life the cost to acquire him might be 1000. In other words, to them he's valued at some ridiculously high level, like "Babe Ruth in his prime," even though he is far from that value on the field.

It would be interesting to see a reverse engineering of the trade that already happened, the mega-swap between Miami and Toronto, where you weigh the values placed on both sides that made them decide that was a value proposition. The finances in these swaps have way more weight in Trade Value than fans realize.

Here's my crazy 3 team ridiculous trade:

Marlins get:
Profar
Perez
Grimm
Sardinas
Alfaro
L. Jackson

Arizona gets:
Escobar
Cruz
Olt
Buckel
$5 mil

Rangers:
Stanton
Upton

Then we sign Greinke, sign LaRoche, extend Elvis, extend Harry, and go win a World Series.

Gentry/Martin
Elvis
Stanton
Beltre
LaRoche
Upton
Murphy (DH)
Kinsler
Soto/Vet

Darvish
Greinke
Harrison
Ogando
Holland

I could live with that roster for 3-5 years while we restock the farm...

Now everyone tell me how crazy I am.

Not here to tell you how crazy you are. This type of thing is fun to think about, so how about we just run through the whole scenario?

Marlins get:
Profar
Perez
Grimm
Sardinas
Alfaro
L. Jackson

This is worth consideration. Getting Profar and Perez off the top is a nice combination. Grimm and Alfaro are each tier-2 players, well, Alfaro at least potentially is. Sardinas would make a nice 2nd base prospect. Luke Jackson will probably end up in the bullpen, but he's not a bad prospect himself. Interesting package overall.

Arizona gets:
Escobar
Cruz
Olt
Buckel

Escobar probably isn't the franchise SS Arizona is looking for, but he's cheap for what he provides. The inclusion of Olt makes this more feasible. Buckel could crack the Arizona rotation in a couple years and projects as a MORP. Cruz is a stopgap outfielder before hitting free agency, but could give you a lot of bang for your buck in a contract year.

Then we sign Greinke, sign LaRoche, extend Elvis, extend Harry, and go win a World Series.

This may be where we hit a couple land mines. As far as the 2013 payroll is affected, we'd be getting Cruz's $10.5M off the books and adding on Justin Upton's $9.75M, so it's basically a wash.

Adding Greinke would be another $20M-$23M AAV, so that pushes the loose $130M limit to its threshold. You figure after a career year Adam LaRoche is going to want something in the 3-year range between $10M-$12M AAV, putting the figure up in the $140M range. That's where the roadblock is for me.

With Profar on the Marlins, extending Elvis then become a necessity (perhaps with the 2015 TV contract money) and Harrison a little less of a priority, at least right away. For what Mitch Moreland is -- a slightly above replacement-level 1st baseman -- I'd rather pay him the league minimum next year and get into arbitration after 2013 than to pay LaRoche.

My lineup would read (with WAR Projections):

Kinsler - 2B (4.2 fWAR)
Andrus - SS (4.4 fWAR)
Upton - LF (4.0 fWAR)
Stanton - RF (4.6 fWAR)
Beltre - 3B (3.9 fWAR)
Murphy - DH (1.6 fWAR)
Soto/Shoppach - C (average 1.1 fWAR)
Moreland - 1B (1.4 fWAR)
Martin/Gentry - CF (average 2.2 fWAR)

That's an imposing top-half of the lineup, overall certainly a top-5 lineup in baseball. The team would be particularly solid on defense. I might even be selling Upton, Stanton and Beltre short. I think Adrian and David are due for falloffs from great 2012's, and that Kinsler and Upton are going to have big rebounds.

Anyway, 98% chance that wouldn't happen, but pretty neat idea.

Not bad, eric. That would be a ridiculous lineup. But I disagree with your batting order (at least, I would do it differently to start the year and see if things regress, especially with Murphy and Kinsler), and based on last year, I'd rather have Murphy in LF more often than Upton. But I daresay, I don't think there would be a better 3-4-5 in baseball than that.

I guess part of it is that I'm a still a big believer in Kinsler, and Ron Washington already saying Kinsler will lead off kind of ends that discussion for me. But you're right, if Kinsler proves the first two months that 2012 is basically the player he is now (which I don't believe), and Murphy performs closer to 2012 than the rest of his career (which I'm also skeptical of), then maybe you look at a 1-2 of Andrus and Murphy with Kinsler in the 6-hole.

Also, for anyone observing, yes, the lineup is extremely right-handed heavy, which seems to be something of a slight disadvantage in a park like ours. I tend to think if you have too many lefties bunched up it can be an issue, but, really, how many pitchers are there that are right-handed stoppers? I can think of sidearm reliever types like Darren O'Day, but as far as starters are concerned I think it's a moot point. I feel the same way about ordering your rotation righty-lefty-righty-lefty-righty just for the hell of it.

Point is, there are a lot of righty hitters, but with the exception of the catching slot, they are all good righty hitters.

*Disclaimer: I stopped reading the posts after a while so this may have already been covered. If so, it needs to be said again.

The biggest flaw in the formula that you guys have been working on (and It's a flaw that probably can't be rectified) is that it doesn't take into account limited roster space and opportunity.

Using some of the logic here, we could use the more familiar WAR values. You wouldn't trade Stanton (say 6 WAR) for a dozen .5 WAR guys because 1) those 12 guys obviously aren't very good, and 2) you don't have the roster spots available to keep 12 guys around to play in situations that maximize their abilities.

In short, quantity =/= quality as this formula can be manipulated to suggest. Back to the drawing board.

This not as much fun as figuring how to get Stanton--I would just pay cash to Lorio directly--but it was on my mind. Sorry for the formatting.

I’ve been thinking about WAR and my first impressions were that there are different categories of WAR. First, WAR should vary depending upon the competitive nature of a player’s contract, i.e., WAR should cost more for an elite free agent sought by multiple clubs than for a salary controlled player who has not achieved arbitration or free agent status. Second, WAR would be more expensive for high salary teams, i.e., teams with more revenues can afford to pay higher prices for players through extensions and free agency offers than can lower revenue teams with more restrictive budgets. Last, the price of WAR should be higher for players at positions that require higher degrees of skill and therefore have fewer players available for these skilled positions, e.g., C, SS, 2B, and CF should be more expensive per unit of WAR than COF, 1B, and 3B, and Starting pitchers should be more valuable than closers, who should be more valuable than middle relief pitchers.

I decided to test the first two of these propositions by comparing the price of WAR by team based on team WAR and team payroll, and to test the second proposition by comparing the salaries of the 2011 free agent class by the WAR of each player. This last exercise requires a fair degree of subjectivity; should WAR be 2011 WAR, an average or weighted average of the last x number of seasons, and should it be adjusted for the age of the player? While my first inclination was to use 2011 WAR—it is the easiest option—I chose to use the average of the last three years.

The team WAR exercise was surprising. I assumed the Yankees would pay more per WAR than the Astros, who actually paid much more for WAR in 2012 than any other team except the Marlins, spending $12.7mm per WAR for the teams 5.9 WAR. Median 2012 WAR came to $4.8mm, while average WAR was $5.4mm.

I next compared the cost of WAR to the nine teams with aggregate salaries in excess of $100mm. These are set forth below originally formatted to display Team/WAR/Salary/$ per WAR:

NYY /27.2/ 190,467,425/ 7,002,479
PHI /15.0 / 171,501,558/ 11,433,437
LAD /12.4 / 153,324,600 / 12,364,887
LAA /37.9 139,619,500/ 3,683,892
TEX /18.6 121,388,900 / 6,526,285
DET /13.7 / 118,394,000 / 8,641,898
CHW /16.0 / 118,208,000 / 7,388,000
STL /27.7 / 111,221,000 / 4,015,199
SFG / 28.9/ 111,157,350/ 3,846,275

The average price per WAR for these 9 teams is $7.2mm and the median is $7mm.

The preceding exercise shows what teams received for what they paid. Free agency should impose a different equation based on what teams paid based upon what I assume they expected, that is the annualized price paid for a player’s average WAR based on the last three seasons. Note that I am using annualized numbers based upon the aggregate salary commitments divided by the number of committed years. This is likely very different from the team numbers used above as I am using straight line amortization whereas I suspect the team numbers are based on actual annual salary, but I was not going to parse through the player details, especially since NateAggie is going to kindly explain why this good idea is put together wrong and someone else will refer me to a Nate Silver article that should have spared me the effort.

The following are player WAR values formatted by Player/Contract Years/Gross $/Annual Dollars/3 Yr WAR Avg/ and $ per Avg WAR

Aggregate Annualized 3 Yr Avg $ per WAR
Player Years Dollars Dollars WAR
Albert Pujols/ 10/ $250,000,000/ $25,000,000 / 7.27/ $3,440,366/
Prince Fielder/ 9/ $214,000,000/ $23,777,778/ 3.87/ $6,149,425
Jose Reyes/ 6/ $106,000,000/ $17,666,667/ 2.57/ $6,883,116
C.J. Wilson/ 5/ $77,500,000/ $15,500,000/ 3.57 / $4,345,794
Mark Buehrle / 4/ $58,000,000/ $14,500,000/ 4.03/ $3,595,041
Jonathan Papelbon/ 4/ $50,000,058/ $12,500,015 / 1.63 / $7,653,070
Jimmy Rollins/ 3/ $38,000,000/ $12,666,667/ 2.00 / $6,333,333
Aramis Ramirez/ 3/ $36,000,000/ $12,000,000/ 1.03 / $11,612,903
Michael Cuddyer/ 3/ $31,500,000/ $10,500,000/ 1.33/ $7,875,000
Carlos Beltran/ 2/ $26,000,000/ $13,000,000/ 2.87 / $4,534,883

The average and median dollars per WAR are both about $6.2mm

The surprises for me are not the Fielders and Pujols, but why did Rollins, Cuddyer, and Ramirez receive so much?

@Eric
I'm just glad you like my trades. I'm more than fine with Moreland at first. Someone needs to post the point value of my packages bc that's what a realistic trade for Stanton is.

@ P1 Stephan
We would still have the Fab Five but yeah we would have a much larger transition period after this core finished but we might win 3 or 4 straight. If we are that desperate than let's do it,

Primi, it appears to me there's a huge mistake in your charts. Maybe the problem is with WAR itself, or maybe it lies elsewhere.

Isn't a "replacement player" by definition one who produces the league average at his position? Therefore, if you have a whole team of players who are "replacement level" you have an average team, right? And mathematically the average team wins 81 games.

So when you calculate the Angels as a team credited with +37 Above Replacement Wins in 2012, ummm no that didn't happen. Their players, all together, only created a +8.

In any event, simply calculate dollars per win, using total payroll above average and total wins above average, and you probably get a better analysis of what each team paid for WAR. And the Angels, with about $140M in payroll and 89 wins, paid through the nose for their extra wins, contrary to what your chart says.

Something is wrong in the methodology here. Or WAR is not as accurate and helpful as some claim it to be.

The replacement level team DOES NOT win 81 games. It wins closer to high 50's, low 60's.

Also, replacement level is not league average. Replacement level is serviceable filler you can pluck from most AAA teams, not LEAGUE AVERAGE MLB players.

Nevertheless, whatever level it purports to have as its baseline, its results don't accurately capture the various players' impact on winning games, because the totals provided by primi show:
the 89-win Angels at +37.9 for the year,
the 88-win Cards at +27.7,
the 95-win Yankees at +27.2, and
the 93-win Rangers at +18.6.

I didn't bother to pull out the other win totals and overlay them against wins, because we already have enough to show that WAR doesn't accurately capture what it says it does. The point being, GMs are likely to see and value different things in trade than an unreliable "ability to create wins" number from one formula or another.

A replacement level team wins 50 games. 50 not 51, not 49 but 50

Actually, we are both wrong. Usually, it has been found that replacement level teams would win in the mid 40's in any given year when taking the average value of all teams minus positional WAR and pitching WAR.

David, it seems as if you dont understand the concept and application of WAR. It would be better advised if you had a more solid grasp of it before questioning a metric that isnt going anywhere.

Also, primi explicitly stated that he created a sliding value for WAR based on his stipulations. That isnt how WAR is actually valued. No offense David, but you've twisted it all up. Start from scratch and check out fangraphs or baseball reference for more information.

TxBall, clearly my point is over your head. But thanks for trying.

WAR is supposed to tell us the impact on winning that players have. But if it actually did it accurately, then the teams who won more would have more WARs and the teams who won less would have fewer.

If we see that the WAR totals don't bear out in actual wins, and that teams with far more WARs can win less, it's clear that the WAR stat fails to accurately weigh some aspects of baseball properly. Or if, as you say, primi has "adjusted" the stats, then it's his adjustments gave a very skewed result. Because he is showing that the Angels - with fewer wins than the Rangers - with over twice as many 2012 WARs as the Rangers, and by the very definition of the metric, something's very wrong in those results.

David, you simply dont get it. Why in the HELL are you using primi's adjusted WAR Valuations as the baseline?

Have you actually LOOKED at the WAR totals of each individual team? Ill give you a moment to check them and come back with an apology.

Also, judging from the content of your posts, there is nothing that you could ever say that would qualify as being over anyone's head. You have no clue how WAR is calculated but you have enough ammunition to attack it?

Tell me how that makes any sense?

WAR = Wins Above Replacement. Start from there, David. Then go look at a teams total WAR from position players + pitchers. Simple, right? Next, read primi's post again because it's clear his post went way over your head.

TxBall, my point is - and always was - about primi's WAR totals and the calculations and conclusions he was deriving from them, and not necessarily about WAR. Again, you have failed to grasp what's being said, and retreated into an attack based on what you know (or think) about WAR, without a clue that the precise definition of WAR isn't really the issue here. But thanks for trying.

^Wrong.

Clearly, you're not trying to sell this BS, are you?

These are quotes directly from you on this very thread and on this very page:

"Primi, it appears to me there's a huge mistake in your charts. Maybe the problem is with WAR itself, or maybe it lies elsewhere.

Isn't a "replacement player" by definition one who produces the league average at his position? Therefore, if you have a whole team of players who are "replacement level" you have an average team, right? And mathematically the average team wins 81 games"."

"Something is wrong in the methodology here. Or WAR is not as accurate and helpful as some claim it to be."

"Nevertheless, whatever level it purports to have as its baseline, its results don't accurately capture the various players' impact on winning games, because the totals provided by primi show:
the 89-win Angels at +37.9 for the year,
the 88-win Cards at +27.7,
the 95-win Yankees at +27.2, and
the 93-win Rangers at +18.6.

I didn't bother to pull out the other win totals and overlay them against wins, because we already have enough to show that WAR doesn't accurately capture what it says it does. The point being, GMs are likely to see and value different things in trade than an unreliable "ability to create wins" number from one formula or another."

Look, I don't need you to admit you have no idea what you're talking about because it's shown plain as day. All I'm asking is that you actually learn the subject matter, which you're attempting to pass as intelligible nuggets of conversation but is shaped like chocolate colored blob.

Oh, and "thanks for trying" but try again. Your writing style is also eerily similar to another poster, whose style usually devolves into sophomoric garbage after a few volleys.

"TxBall, clearly my point is over your head. But thanks for trying.

WAR is supposed to tell us the impact on winning that players have. But if it actually did it accurately, then the teams who won more would have more WARs and the teams who won less would have fewer.

If we see that the WAR totals don't bear out in actual wins, and that teams with far more WARs can win less, it's clear that the WAR stat fails to accurately weigh some aspects of baseball properly. Or if, as you say, primi has "adjusted" the stats, then it's his adjustments gave a very skewed result. Because he is showing that the Angels - with fewer wins than the Rangers - with over twice as many 2012 WARs as the Rangers, and by the very definition of the metric, something's very wrong in those results."

November 24, 2012 at 6:47 PM | David

The bolded blurb has nothing to do with primi's post. It has everything to do with your obvious misunderstanding of the concept of WAR. Not only that, but you pluralized WAR. Normally, I wouldn't point that out because it's an honest mistake if someone doesn't understand the concept all too lucidly, but you're a special case.

Wins Above Replacement(s)? No.

Anyways, primi's post is actually pretty clear to those who are well versed on the subject matter. It's evident he is trying to valuate the $ assigned to WAR using his own set of stipulations. Generally, WAR is valued at $4.5 million to $5 million per. Some have suggested that the monetary value of WAR has inflated to $6 million per. That's not enough for primi so he's come up with his own methods of WAR valuation in relation to a team's production based on payroll. By no means is he saying that it's a definitive measure of how a team has produced, but how wisely they've spent their money based on his adjusted measure of $/WAR.

Get it?

@David. Sorry for the late reply but I have been driving most of the last day. As far as the WAR calculations, I used the values presented in Baseball Reference. I was also surprised with the Angels' WAR values but I took them at face value. As far as my points go, the actual calculation of WAR by Baseball Reference would only be a problem if they were inconsistently applied among teams.

I used WAR as the metric because I was trying to identify possible player valuation disparities that may arise in the cases of teams willing and able to spend more on salaries and differences arising from free agent salaries versus other salaries, e.g., pre-arb salaries and contract extensions. The first example, simply payroll divided by team WAR, measures what the team received for what it paid. The second example, comparing free agent salaries with the players' previous WAR values (in my presentation the average of 2009 through 2011 WAR for the selected members of the 2012 free agent class), attempted to examine the price of WAR based upon the reasonable expectations of the buyer. This second presentation may be plagued by the use of a small sample set, being the top annualized salaries conferred to ten of the top free agents in the 2012 class.

Intuitively, at least for me, there may be a significant difference between the price of WAR arising from an auction versus the price arising from the price constraints of pre-arb, arb, and extension pricing. Of these three situations, I find arb pricing the most interesting because (1) there should be a relatively large set of arb examples, (2) the arb process in and of itself is an attempt to measure player value, and (3) there is anecdotal evidence that arb mediators in the prior incarnations of arb negotiators did not use WAR or other modern saber metrics in setting salaries.

The impetus of my original post was the very common use of a single WAR value by commentators-not just here at BBTiA but by sportswriters, other board commenters, etc.-- in evaluating player contracts as a function of player value. Greinke, for instance, may receive a contract well in excess of $4.5mm per WAR, because his contract will be negotiated in an auction environment without the structural constraints imposed by non-FA pricing mechanisms. As noted above, my intuition is that FA WAR pricing may also be influenced by scarcity of positions, participants with high revenues, and perhaps the perceived predictability of WAR results compared with expectations.

As far as my exercise went, even though I used only 2012 data to calculate differences in team WAR production versus team payroll, there should be enough data to draw conclusions that payroll may likely affect WAR production. Variations in pricing WAR expectations in a FA environment is much less conclusive, being that I looked only at about 10 free agents in one year. But I could only escape the Thanksgiving chaos and cacophony for so long without incurring the wrath of family members and guests, none of whom at my table had any interest on baseball, much less geeky sabre discussions.