How much, roughly speaking, are draft compensation picks worth? Back in 2005, Rany Jazayerzli did a gazillion-part study on value in the draft. He largely focused on the college vs. high school angle, which I don't especially care about here, but there were a lot of numbers there to give us ballpark estimates on draft value.

In many cases, a Type A free agent (such as Torii Hunter this year, or Santana/Nathan in 2009) will net a team one pick in the 16-30 range and one pick in the 31-45 range. In the first part of Rany's series, he tallied up how often picks reached the majors, and their career value. In the 16-30 range, you're looking at about a 60-70% chance of the pick reaching the majors--even if it is just for a cup of coffee or "Moonlight Graham" appearance, as Rany puts it--and in the 31-45 range, the rate of reaching the majors is about 40-60%. So between your two compensation picks, it's likely that one of them will reach the majors, but not terribly likely that both of them will reach the majors.

Then, he looked at draft picks' career value by measuring each player's WARP over the first fifteen years after he was drafted to determine a career worth. In the 16-30 range, you're looking at about 10-15 career wins above replacement, and in the 31-45 range, you're looking at about 5-10 career wins above replacement.

Overall, speaking very roughly, out of those two draft picks, you're going to get about 20 wins above replacement. That might not sound like a lot, but it's pretty decent for paying essentially the league minimum to get them. You're certainly most likely to get a superstar in the top 15 picks, but there are plenty of stars around baseball from the 16-45 range. Randy Bush, though blessed with a Hall of Fame name, didn't have Hall of Fame talent, but still made Aaron Gleeman's Top 40 Twins list with just 19 career WARP.

Now, let's consider some of the pieces that are in the Santana discussion. First I'd like to talk about Melky Cabrera. I've heard people describe Cabrera as a 4th outfielder with a terrible track record. I would like to know which league these people are following, because it's a lot better than the American League I know. Melky had an 89 OPS+ last year at the age of 22, and fielded his position in line with league average based on all the defensive statistics I could find. That made him essentially a league average centerfielder, and that's with absolutely zero improvement upon what he's already done. ZiPS has the following forecast for 2008:

.286/.352/.419 -- Melky Cabrera
.272/.336/.425 -- Average CF

That's not much above average, but it's above average. Being a league average CF playing regularly is worth conservatively 5-6 WARP per year.

Now for Philip Hughes. Hughes is a pitcher and Santana is a pitcher, so let's compare their ZiPS projections:

3.70 ERA, 6.8 K/9, 2.6 BB/9, 0.83 HR/9, 141 IP -- Philip Hughes
3.21 ERA, 9.2 K/9, 1.8 BB/9, 1.13 HR/9, 230 IP -- Johan Santana

Is Johan Santana better than Philip Hughes? Absolutely. Is it a huge difference? Figure that you'll have replacement level pitchers throw the 89 IP difference between Santana and Hughes, and give those replacement level pitchers about a 5.50 ERA. Over the total 230 IP, you're looking at Hughes/Scrub giving up 112 runs and Santana giving up 82 runs. That's a difference of 30 runs. That difference gets smaller if Santana pitches less than 30 innings, you can get better relief/spot starter ERA than 5.50, and/or if Hughes can pitch more than 141 innings. So it's a 3-win difference between Hughes and Santana, and roughly a 3-win difference between Hughes and replacement level (if you set replacement level around a 5.50 ERA.) Those both seem a bit low to me (the 10 runs/win scale is pretty rough), so I'd maybe set it as a 4-win difference between Santana/Hughes and a 4-win difference between Hughes/replacement. Either way, that's basically the ballpark you're talking about.

Considering that right now, the Twins are staring at replacement level in CF unless they want to pony up some money for Cameron, Rowand, or Andruw Jones, Santana for Hughes/Cabrera is arguably a wash in terms of wins for this year, and maybe even an improvement of a win or two. After that, it seems quite likely to me that Hughes and Cabrera, even without improving much, could beat the average of 20 wins above replacement that you're likely to get out of the two compensation picks. It also has the advantage of not making us wait 5-6 years for the compensation picks to mature and reach the majors.

I'd like to do the same analysis with the Red Sox' Ellsbury/Lowrie/other offer, but it is late and there are no 2008 ZiPS projections for the Red Sox yet. Suffice it to say that while I think Ellsbury is a little better than Cabrera, I don't think the difference is all that big, and Lowrie looks like a pretty good piece to me, too. I'd probably value Hughes/Cabrera a little more highly than Ellsbury/Lowrie, but they're both probably in the same ballpark.

The Hughes/Cabrera offer isn't so great that the Twins are obviously making a mistake if they walk away from it. It's also not so great that they're obviously getting a huge steal if they make the deal. At this point, I would be tempted to take the deal, even if the third player isn't all that great, just because the Hughes/Cabrera package on its own seems to be about as valuable as Santana in 2008, and will give the Twins a better chance to win in 2009 and 2010 than if they go the Santana->free agency route. And since Ellsbury/Lowrie seems pretty comparable to me, I guess I'd say the same thing about that. Both are pretty reasonable offers, but not something so compelling that I feel they need to do it.

If you made me choose one way or the other, I'd choose for them to make a deal. The timing of adding a couple of early 20s players to the team as opposed to a couple of draft picks just seems to fit better with this roster, and the roster also seems to have enough holes in it that it makes sense to address needs at multiple positions right now.