Juan Cruz, Jason Varitek, and Orlando Cabrera all have something in common - they're free agents who aren't actually "free."  All three players have seen their market value depressed by the Type A compensation status attached to their names, as other teams remain hesistant to give up a top draft pick to sign them.  In essence, their prior employers are still enacting some control over them even in free agency. 

The above example is one of many problems created by the current free agent compensation system:

1. An archaic calculation method led by STATS (edit) Elias which uses less-than-ideal data points to conclude that 40-year-old Russ Springer (Type A) is in the elite tier over Milton Bradley (B), while Juan Uribe (B) is in a class above Jason Giambi.
2. The awkward "will he or won't he?" arbitration chess match that teams play with their own departing free agents. Teams are forced to either offer arbitration to players they have no intention of signing, or receive no compensation for losing a good player. 
3. Penalizing teams for spending money to improve their rosters, by forcing them to give up a top draft pick to sign an elite free agent (Type A). 
4. The inconsistent compensation that teams receive for losing top tier free agents.  In exchange for losing A.J. Burnett, the Blue Jays could've received the 17th overall pick in next year's draft (to go along with a sandwich pick at around #34 overall).  Instead, they'll receive the sandwich and the Yankees third-rounder - right around pick #100 - all because of the team that signed him and the free agent machinations around him. 

How to fix this? 

You could pick better different stats to evaluate and rank the free agents each year.  But no criteria will be perfect, and it would be difficult to find a set that pleased everyone in the game. 

Another issue is that using stats from previous years to rank players serves to compensate teams for past production. Shouldn't teams instead be compensated for the production they are projected to lose in the coming years, by virtue of not having that player's services anymore, rather than what they've already done? 

Here's my quick-and-dirty solution to the above problems, using figures that everyone can under$tand:

Let the free market dictate free agent compensation. 

The larger the contract that a player signs, the better the compensation granted to his former team. 

I'd propose creating a sandwich round that is an actual, true round - 30 picks, #31-60 - between the first and second rounds. 

The total sum of free agent compensation - gaining comp picks in return for losing free agents - is comprised within those 30 picks. 

The most valuable free agents - ones that sign nine-figure deals - are now worth 3 draft picks to their previous team, rather than two.  But no team can lose its first round-pick in this scenario, so the best possible compensatory pick is #31 overall.

From picks #31-60, teams are compensated in order of the size of the contract that their free agent eventually signed. 

Ignore for a moment that some players re-upped with their own teams, and lets look at how this hypothetical would roughly play out if it were in place for the current FA class:

• Teixeira - worth three picks to the Angels - #31 overall, 32, 33
• Sabathia - #34, 35, 36 to the Brewers
• Burnett - worth two picks #37, 38 to the Blue Jays
• Lowe - #39, 40 to the Dodgers
• Ramirez - #41, 42
• Dempster - #43, 44
• K-Rod - worth one pick (we've now dropped to deals worth $40MM or less) - #45
• Rafael Furcal #46
• Milton Bradley #47
• Raul Ibanez #48
• Adam Dunn #49
• Oliver Perez #50
• Picks #51-60: one pick apiece in compensation for free agents who earned contracts of $10-20M in total guaranteed value:  Kerry Wood, Edgar Renteria, Jamie Moyer, Casey Blake, Brian Fuentes, Pat Burrell, Orlando Hudson, Ben Sheets, Bobby Abreu, Andy Pettitte, Orlando Cabrera. 

There is a huge mass of free agents - everything from Casey Fossum to Jason Giambi - who do not garner compensatory picks for their team under this scenario.  I think that's perfectly ok - if a player wasn't worth at least $10M on the open market, his team doesn't deserve any special compensation for losing him.  The most well-run teams will be able to replace players like those internally or via trade, anyway. 

I've made the two $100M free agents of this offseason worth 3 compensatory draft picks, and free agents who sign deals of $40-100M worth two.  There's nothing scientific about how I came up with that, and it's certainly open to a more sabermetric critique - there's a better way to create a cutoff to distinguish between "three, two, and one draft pick players,"  but I'm proud of the end result: 

• teams that lose elite FAs are guaranteed a consistent, more predictable compensation system that's dictated by the free market rather than bad data,
• middling free agents like Varitek and Cabrera aren't punished by the draft pick glass ceiling, and it
• avoids penalizing teams for investing in their product on the field by eliminating the cost of a first-rounder for signing elite talent.