A very fine Fangraphs article here that argues, strictly in terms of ROI dollars gained and lost, that the Mariners would be a little better off to offer Beltre arbitration than not.

  === Exec Sum ===

The argument reads:

1.  If Beltre accepts arbitration, he'd get $10-13m, but might be worth only $8-12m on the field.  That's a net profit value loss of "something like $4 million."   In other words, you figure on overpaying him by about $4m if he accepts arbitration.

2.  But, if he declined it, you'd get a supplemental pick -- worth a little more than $2m, per the argument.

3.  So you can view an arb offer as a wager -- $4m of overpayment if Beltre takes it, or $2m of talent gained if he doesn't.

4.  Beltre has maybe a 50% chance of doing either, so it pencils out in the M's' favor.

................

Conclusion:  most MLB teams make an incorrect value decision, usually, when they don't offer arbitration.

.

=== And Exec-Worthy It Is ===

This is the type of ROI paradigm that a GM should always have on his desk for a decision like Beltre's arbitration.  He should have 5 or 6 single-page recommendations, of different paradigms, of which this is one.

It's a great place from which to start.

.

=== Caveats ===

Al Dimond, in the comments, points out a little, easy-to-make math error that loaded the structure towards "OFFER ARB."  

Betting -4 against +2 in a 50-50 coin flip isn't an even bet.   The article remembered to multiply -4 by the 50% coin flip, but forgot to do it on the +2 side.

The kind of mistake I make all the time, especially when trying to get X number of articles done in a week.   Although usually when I do it, you guys don't gush with praise in the comments.  ;- )

..............

On paper, we might assume a sign-and-trade -- that if Beltre accepts arb, you send him and $4m to the Red Sox for a Grade C prospect. 

But in practical terms, this is much easier said than done.  The dynamics of GM relationships interfere on this one.

A real GM offers arb, he's got to at least allow for the possibility of keeping the player all year.

.................

The paradigm assumes that if Beltre's Fangraphs performance is $8-10m next year, that this "Adrian vs RLP" value is the only thing that the Mariners lose.

But remember that $4.5 to $5.0m is what you pay per win on the free agent market. 

If the Mariners have young non-arb, or arb, players available, those players will give much more than 1 win per $4.5 or $5.0m.   If Matt Tuiasosopo earns a single WAR, he'll give 10x that performance ratio.

Suppose, for argument's sake, that Tui gets 2.0 WAR for $0.4m and that Beltre gets 2.0 WAR for $13.0m.  What have the Mariners lost here?

Once again, RLP is a good way to compare free agents to each other on the same scale.  It's not a good way to compare a given club's options if those options include young players.

.........

Aside from the fact that a young player might offer more than $4.5m per 1 win ...  this free agent market may offer values better than 2 WAR per $10m. 

Marco Scutaro 3b, for example, might offer better value than 2 WAR for $13m in one season.  For $13m, two seasons, Scutaro might give you 5 WAR.

.....

And it's not just Scutaro.  $13m might buy any number of things in this market.  It might buy you 90% of John Lackey's 2010, meaning 4-5 WAR.  That $13m comes out of the entire offseason budget.

Therefore, the potential loss is much more than $4m for one season.  $4m would be more like the minimum loss.

............

Also, a #40 draft pick might well be worth $2m or $3m in a vacuum -- but take away that #40 pick, and the M's might spread that $2 or $3m across their top 10 picks and get better players up-and-down the line.

Last year, for example, the Mariners "reached" for 2nd- and 3rd- and etc-round picks higher than those players should have gone, so that they could pay those players less and pay Dustin Ackley more.

The principle works in reverse if you're not spending money on a comp pick.  You're the team grabbing the players who are sliding, and giving those players extra dollars to sign.

So, whether or not a #40 draft pick averages out to be worth $2m of performance in the bigs, you've got to consider the dominoes that fall over also.

...............

Finally, the 5-year roster plan has many variables not "captured" by the occurrence of another year of Adrian Beltre.   It might be important to accelerate Matt Tuiasosopo's development.   It might be important to keep the infield options open, to allow for trade talks -- which for the Mariners are a very real concern. 

Ossifying 3b with a player you don't want, during a transition, might have many ramifications beyond the ROI loss.

.

Part II