A different way to look at the problem is to establish probabilities for various potential career paths. You can then determine a level of production and value for each potential career path and come up with expected values based on the player’s probability of attaining each career path. I think it works pretty well, but we’ll see.
Anyway, it’s not too difficult to modify those approaches to try and get a sense of potential productivity and value for a prospect.
Using players from the 1987-1994 drafts I created five different potential career paths – 20, 40, 60, 80, 100+ WARP3. As a quick shorthand guide, players that exceed 20 WARP3 are “useful”, players that exceed 40 WARP3 are “good”, and players that exceed 100 WARP3 are “great” or at least very good for a very long time. As you’ll see the 60 and 80 WARP3 groups aren’t especially distinct – at least in pre-FA service time productivity – but they are generally “very good” players.
For each player I tried to determine how much of his career productivity occurred prior to his ability to become a FA. Looking at the whole career is interesting, but for draft picks and prospects we’re mostly interested in the years that they are under control by their team. Since service time information is not widely available, I’m sure that I made a decent number of mistakes, but I did look at each player using a few different references to make it as accurate as possible. Generally, players with longer, better careers are easier to get right. I’m probably correct for each of the players in the 100+ WARP3 group, but I was definitely guessing on a number of 20 WARP3 players.
I was then able to generate an average WARP3 pre-FA productivity career path for each of my five groups. Unfortunately, the value method I intended to use required WARP1 data. That makes a big difference so I generated a quick WARP3 to WARP1 converter the details of which aren’t very important. So the potential career path groups are in WARP3 and but the specific yearly totals are in WARP1. It’s a little confusing and there are obviously little systemic estimation errors all along the way, but I don’t think that’s a big deal. This isn’t designed to precisely differentiate player value so much as it’s a way to get a handle on the relative size of the differences between the groups.
The valuation method was developed by BP’s Nate Silver in his article valuing first round draft picks. You might remember that I redid Silver’s original article when he updated some of his player valuation methods. If you want to read more about the entire method you can do that here (the link to Silver’s original subscription article is in there too):
Re-valuing 1st rd draft picks
For this post I’m really only interested in the yearly WARP1 line and the final Net Present Value line. The WARP1 line obviously represents on field production and is the most important thing to look at. Net Present Value is the difference between the player’s Market Value and his cost to the team. It’s a kind of “value” measurement, but not every team can fully take advantage of that value. In order to take advantage of this value, the team has to be able to re-invest that money in the free agent market. This shouldn’t be a concern for a high revenue team like the Sox, but it is an issue for low and medium revenue teams.
One other thing to notice is that I replaced the original “Signing Bonus” line with “Acquisition Cost” so that we can see how Marte’s hefty 11M acquisition cost effects his ultimate value. There’s a reasonable argument that the money is really a sunk cost that should be at least partially associated with Renteria, but conceptually I think it’s simpler and more interesting to dump the whole weight of that 11M on Marte and see how it effects him.
Just for some perspective on what a WARP1 season of X really means here are the WARP1 totals for the Sox main hitters last year:
8.0 – Ortiz 6.6 – Ramirez 6.2 – Vartek 5.5 – Damon 5.2 – Mueller 4.3 – Nixon 3.9 – Millar
Loosely, anything 8 and over is a solid MVP candidate, something in the 6s is probably All Star level, something in the 5s is probably a good regular and below that you’re looking platoon players or mediocrities.
We’ll start at the top and take a look at an average pre-FA career path of a 100+ WARP3 player. The 100+ WARP3 players in this study are Griffey, ARod, Frank Thomas, Bagwell, Biggio, Manny, Mussina and Olerud. It’s mostly HoF caliber players who were great right away with a couple very good for a long time players.
100+ WARP3 Player
y1 y2 y3 y4 y5 y6 Total WARP1 5.3 6.9 8.2 10.1 7.4 9.4 47.3 MarkVal 8.1 12.9 17.6 25.8 14.6 22.6 101.6 MargCost 0.5 0.75 5.5 11.3 8.9 14.5 41.4 NetVal 7.6 12.2 12.1 14.4 5.7 8.1 60.2 GrossPresVal 7.2 11.5 11.5 13.7 5.4 7.7 57.2 AcquisitionCost 11.0 NetPressVal 46.2
You can see from the yearly WARP1 totals that these players are very, very good, very quickly. They’re capable of contending for MVP awards just as they start to become arbitration eligible. These players really push their teams towards championships and as a result are tremendously valuable just for their on field production.
In addition to that, they are tremendous values because the structure of the CBA allows a team to pay one of these great players dramatically less than their free market value. In this case and even including the 11M acquisition cost the resulting difference would be about 45M.
Instead of thinking about that value as a huge lump sum, I think it makes more sense to think about it in yearly increments and how a team might be able to use it. Because you can pay a great player much less than his free market worth, you effectively have an extra 7-8M per year (assuming you can afford it) to add free market talent at other positions. If an expected win costs 2-3M, then you might be able to bid on a player worth an extra 3-4 wins. With the ever present caveat that you have to be able to afford to spend that “additional” money, that’s a pretty sizable secondary advantage on top of the great productivity that the 100+ WARP3 player provides. Great pre-FA service time players are the greatest market inefficiency in baseball.
Some of the 80+ WARP3 players in this study are: Albert Belle, Steve Finley, Ventura, Piazza, Edmonds, Lofton and Kent.
80+ WARP3 Player
y1 y2 y3 y4 y5 y6 Total WARP1 4.2 5.2 6.5 7.7 6.5 6.8 36.9 MarkVal 5.4 7.8 11.6 15.7 11.6 12.6 64.8 MargCost 0.5 0.75 3.6 6.9 7.1 8.0 26.9 NetVal 4.9 7.1 8.0 8.8 4.5 4.5 37.9 GrossPresVal 4.7 6.7 7.6 8.4 4.3 4.3 36.0 AcquisitionCost 11.0 NetPressVal 25.0
There’s a pretty substantial drop off in pre-FA service time production. The 100+ WARP3 group averaged 47 WARP1 and this group averages 37 WARP1. Looking at the individual years, we’ve basically lost the MVP candidate years, but there’s still a nice stretch of All Star caliber production.
From a financial perspective, the gap between Market Value and Marginal Cost narrows quite a bit and as a result the big Acquisition Cost starts to take a more significant bite out of the Net Present Value. Still, the team will have enough money to buy an extra couple of wins per year on the FA market. That’s a nice bonus to an All Star.
Some of the 60+ WARP3 players from the study are: Ray Lankford, Fryman, Reggie Sanders, Salmon, Knoblauch, Shawn Green, Tino Martinez, Grissom, Ray Durham and Nomar.
60+ WARP3 Player
y1 y2 y3 y4 y5 y6 Total WARP1 3.9 6.2 5.6 6.7 6.3 6.3 35.0 MarkVal 4.8 10.7 8.9 12.2 11.0 11.0 58.6 MargCost 0.5 0.75 2.8 5.4 6.7 7.0 23.1 NetVal 4.3 9.9 6.2 6.9 4.3 4.0 35.5 GrossPresVal 4.1 9.4 5.9 6.5 4.1 3.8 33.7 AcquisitionCost 11.0 NetPressVal 22.7
There’s very little difference between the 60+ and 80+ WARPs groups. I haven’t really looked too closely into it so I’m not sure why. It may simply be that post-FA durability and productivity moreso than talent is what separates these two groups. Nomar is a 60+ WARP3 player, but the way his body is breaking down he may not get to 80 despite having been fantastically productive in his pre-FA years. There are numerous 80+ WARP3 players like Steve Finley and Jeff Kent who have been able to continue to add career WARP3 deep into their 30s. A team would much rather draft a Nomar than a Finley or a Kent.
Some of the 40+ WARP3 players from the study are: Conine, DeShields, Brosius, Clayton, John Valentin, Karros, Brian Jordan, Garrett Anderson and Rondell White.
40+ WARP3 Player
y1 y2 y3 y4 y5 y6 Total WARP1 2.7 3.3 4.1 4.7 4.5 4.6 23.9 MarkVal 2.6 3.6 5.2 6.6 6.1 6.4 30.6 MargCost 0.5 0.75 1.6 2.9 3.7 4.1 13.6 NetVal 2.1 2.9 3.6 3.7 2.4 2.3 17.0 GrossPresVal 2.0 2.8 3.4 3.5 2.3 2.2 16.2 AcquisitionCost 11.0 NetPressVal 5.2
At this level – on average – the players mostly have years befitting a solid regular player. There’s also more of a marginally productive transition to the majors. Obviously many players in this group out-produced these averages and were more than just solid regulars.
At this point the very large acquisition cost eats up most of the difference between Market Value and Marginal Cost. The team is still in the positive, but 5M over 6 years doesn’t buy much on the FA market. And arguably if an expected playoff team like the Sox missed the playoffs one year because the player had a sluggish transition, then that cost could easily outweigh the positive that is shown here.
That may be a little surprising, but because his acquisition cost is so high Marte can become a pretty solid player – a definite “success” – yet the Sox will only more or less break even.
After the trade Lajoie was quoted in the papers saying that Marte would have been in the first pick in the draft. Putting aside whether or not that’s true, it doesn’t cost 11M to sign the first pick in the draft. Last year’s #1, Justin Upton, has not signed, but I believe the negotiations broke down somewhere in the 5-7M range. Coincidentally, the second pick, Alex Gordon, is a 21 yr old power hitting 3B like Marte. He signed for 4M. At 11M the Sox are paying a significant premium for Marte (arguably the premium is worth it because Marte has proven himself at a much higher level than Upton and Gordon, but that’s not really what Lajoie is saying).
I’m certainly not in any way implying that the Sox were wrong to trade Renteria and subsidize his contract for 11M. I think huge revenue teams like the Sox should be more aggressive about buying young players when the opportunity arises. However, from a strict numbers point of view a sizable acquisition cost will eat into the overall value of the deal pretty quickly.
There are a ton of 20+ WARP3 players in the study including this unforgettable run of former Sox - Troy O’Leary, Dave Hollins, Darren Lewis, Ed Sprague, Rico Brogna, Jeff Frye, Tim Naehring, Marty Cordova, Tony Clark, Tony Graffinino, Pokey Reese, Scott Hatteberg, Trot Nixon, Todd Walker, and Jay Payton.
20+ WARP3 Player
y1 y2 y3 y4 y5 y6 Total WARP1 2.7 2.8 2.7 3.0 3.0 3.0 17.2 MarkVal 2.6 2.8 2.6 3.1 3.1 3.1 17.4 MargCost 0.5 0.75 0.8 1.4 1.9 2.0 7.3 NetVal 2.1 2.0 1.8 1.7 1.2 1.1 10.1 GrossPresVal 2.0 1.9 1.7 1.7 1.2 1.1 9.6 AcquisitionCost 11.0 NetPressVal -1.4
The team doesn’t do very well if the prospect follows this path, but we’re already seeing the “average production” approach breakdown a bit. If Marte settles in as a 3 WARP1 player he won’t continue to get tendered a contract.
Those are five different career paths that a prospect like Marte might follow. The next step in this process would be to determine the probabilities for each path. Is it – 50%, 25%, 10%, 5%, 1% - or is it – 35%, 15%, 3%, 1%, 1%? It makes a big difference. It’s been mentioned in at least one of the Marte threads that in order to figure something like that out, you would need to find a large comparison group of Marte-like prospects. What career paths have other 21 yr old corner players with AAA success and good tools scouting reports followed? Once you narrow things down like that, I think it becomes very difficult to find a large enough sample to generate any useful probabilities. It’s certainly beyond what I have access to.