thoughts on the Mariners, MLB draft, and more homelinksdraftabout me

Nelson Cruz and Compensatory Draft Picks

Nelson Cruz recently signed with the Orioles for 1 year, $8 million. This is a bad move for him, given that he could have accepted the Rangers' 1 year, $14 million qualifying offer in December. He also reportedly had some multi-year offers he rejected as he waited for what he thought he was worth. Cruz played the market poorly, or at least as poorly as someone who is $8 million richer can play it.

The Mariners were linked to Nelson Cruz throughout the offseason. I hated the idea of signing him the whole way - until he signed with Baltimore. $8 million seemed downright reasonable, if not a bargain. I started to wonder if the Mariners missed out on Cruz.

The main issue with Nelson Cruz from the start was the qualifying offer, which meant a team would forfeit a draft pick to sign him. Some other day we can argue the absurdity of free agents being tied to draft picks, but for now it is the system teams live in. When a player like Nelson Cruz hits the free agent market a team must weigh if the free agent is more valuable than the draft pick they would lose. Teams seem to treat their draft picks like gold these days.

So I got to thinking and researching. How valuable are these draft picks?

Draft picks, like any players, gain value primarily based on how good they are. However, MLB players are virtually guaranteed to make less money than they would on the open market for their first six seasons (before they can become free agents). They are essentially guaranteed to earn the league minimum their first three seasons, and then go through arbitration for three more years to gradually escalate their pay towards what they are expected to make in free agency. So, a draft pick's value isn't simply wrapped up in their potential talent; it is also wrapped up in how cost effective they are. The threshold seems pretty low to get a positive return on the investment.

The more I thought about draft picks, the more I thought about a stat that I am going to call "effective WAR." It works like this: There is a fair market price for WAR on the open market. In a perfectly efficient market, a team pays for exactly how much WAR costs. But teams aren't perfectly efficient. A player who is a bargain brings two valuable traits: not only are they performing better than they should (based on their cost), but that money saved can be invested in the market to obtain more talent. This is the driving force behind effective WAR.

A hypothetical will help clean up this idea. Let's say 1 WAR costs $3.5 million in today's free agent market. A player who produces 1 WAR would have an "effective WAR" of 1. There is no change - they did exactly what was expected of them, based on the team's investment. However, let's say the player is actually worth 2 WAR. That should have cost the team $7 million dollars. So, not only is the team getting 1 WAR essentially for "free," that money they didn't spend getting that WAR can now be spent to get 1 WAR of talent in the marketplace. Therefore, the 2 WAR version of a $3.5 million player is worth 3 "effective WAR." The player isn't going to generate that extra 1 WAR on their own, but their production and contract has given the team the ability to go acquire an extra WAR that they shouldn't be able to get with a perfectly efficient team. Conversely, players who are overpaid get dinged by effective WAR. Not only do they eat up payroll space they shouldn't, but the cash they eat up is prevented from being spent efficiently in the open market.

Let's take this hypothetical into real life with Robinson Cano. The Mariners gave up a draft pick to sign him, though not their first-rounder because it's not protected. It's obvious that whatever draft pick the M's gave up is extremely unlikely to be as good as Cano because Cano is a bona fide star. However, Cano also signed a massive contract. What does effective WAR have to say about Cano?

For now, I'm estimating that 1 WAR on the open market costs $3.5 million.* This can be debated, but just roll with me. I went to Cano's fangraphs page and then averaged the three projection methods they offer (Steamer, Oliver, and crowdsourcing), which gave a projected 5.2 WAR for 2014. Cano, even with a rather lofty projected WAR total, is not a cost efficient player because he cost so darn much. However, he's still valuable. His effective WAR is estimated at 1.9, which in essence means he was penalized 3.3 WAR because he was so expensive. However, he is still worth something because he is a really good player.

*In practice for this model, I'm going with whatever the average salary of an MLB player is, to the nearest hundred thousand. Last year it was $3.4 million so I've got a bit of a bump built in, though probably not high enough. By assuming that the average cost of an MLB player is worth 1 WAR I am saying the average MLB player is worth 1 WAR. An average team will win 81 games, and subtracting 25 WAR from 81 gets us to 56 wins - which is around the theoretical threshold for how a replacement level team (all 0 WAR) would do. Again, roll with this for now. This might be improved if I get serious about this stat. It's good enough for a starting point.

Now, let's talk about draft picks. Effective WAR gives us a system to figure out how good a draft pick would have to be to be better than Cano when cost is figured in. The Mariners have to give up their second round pick because of Cano. Last year, the Royals drafted in the same slot as the Mariners would have in 2014, and the MLB slot recommendation was about $1.2 million. The 2014 recommendations aren't out yet (at least on the internet), so I will project a slot of $1.3 million. This is important to consider because it is part of the cost of acquiring draft talent.

Long story short**, a draft pick with a $1.3 million bonus would have to be worth about 1.0 WAR per year their first six seasons to be worth an effective WAR of 1.9. So, if the expected value of the M's draft pick is less than 1.0 WAR, then Cano is the better resource, even at 10 years, $240 million. I haven't gone through and figured out the expected WAR of the M's pick yet, but my gut says that it will be lower than 1.0 WAR. Cano is an upgrade, though an inefficient one.

**Here's the long story. I took the signing bonus and divided it over a player's first three pro seasons, and then added $500,000 (the league minimum salary). After that, I assumed a player would earn 40% of their open market value in year 1 of arbitration, 60% in year 2, and 80% in year 3. I felt this was the best way to estimate a player's cost in all their team-controlled years before hitting the open market, and take those six years as a whole and consider them for calculating the effective WAR for a player. I also considered inflation and set it at 6% annually in baseball, based on how the average price for a player has changed over the past 25 years (6% is the median). I also assumed a player will spend three seasons in the minors before making the majors, which mattered for inflation purposes. It turns out inflation doesn't change the values much anyway.

But what about Nelson Cruz? His projections suggest he will be about a 1.7 WAR player, which at $8 million gives him an effective WAR of 0.5 (so he's still an inefficient value). The Orioles will give up their second round pick because they already signed Ubaldo Jimenez***, and I loosely project the recommended bonus for that pick will be around $1 million. That pick would need to produce 0.3 WAR per season the first six years to match Cruz's 0.5 effective WAR. That's a pretty low WAR, low enough that it's debatable whether Cruz is more valuable than the draft pick before other factors are considered - namely if WAR now, like Cruz provides, is more valuable than the potential for WAR down the road (and in Baltimore's case, I would unequivocally say that it is).

***Ubaldo's effective WAR for next season is 1.1, though will likely sink quickly as he declines and his salary increases. Still, it seems like there's a reasonable (though not a definitive) case that signing him was better than keeping the draft pick. A first-round draft pick would need to be worth 0.6 WAR to get to 1.1 effective WAR.

The money involved is really important though. Cruz's effective WAR sinks to 0.0 if he were to be paid $9 million, so it would be very hard to justify any amount beyond that. MLB teams seemed to intuitively understand this (or perhaps run similar calculations to mine) with Nelson Cruz, but Nelson Cruz did not.

Effective WAR is far from a perfect stat for several reasons. It's unreasonable to think that we can predict a player's WAR with precision, the going rate for WAR on the open market, or how inflation will change over time. The draft is also more of a crapshoot than most are willing to admit. However, effective WAR is better than nothing, and gives a starting point for discussing free agents and draft picks.