MLB Farm Systems Ranked by Surplus WAR
You smell that? It’s baseball’s prospect-list season. The fresh top-100 lists — populated by new names as well as old ones — seem to be popping up each day. With the individual rankings coming out, some organization rankings are becoming available, as well. I have always regarded the organizational rankings as subjective — and, as a result, not 100% useful. Utilizing the methodology I introduced in my article on prospect evaluation from this year’s Hardball Times Annual, however, it’s possible to calculate a total value for every team’s farm system and remove the biases of subjectivity. In what follows, I’ve used that same process to rank all 30 of baseball’s farm systems by the surplus WAR they should generate.
I provide a detailed explanation of my methodology in the Annual article. To summarize it briefly, however, what I’ve done is to identify WAR equivalencies for the scouting grades produced by Baseball America in their annual Prospect Handbook. The grade-to-WAR conversion appears as follows.
| Prospect Grade | Total WAR | Surplus WAR |
| 80 | 25.0 | 18.5 |
| 75 | 18.0 | 13.0 |
| 70 | 11.0 | 9.0 |
| 65 | 8.5 | 6.0 |
| 60 | 4.7 | 3.0 |
| 55 | 2.5 | 1.5 |
| 50 | 1.1 | 0.5 |
| 45 | 0.4 | 0.0 |
To create the overall totals for this post, I used each team’s top-30 rankings per the most recent edition of Baseball America’ Prospect Handbook. Also accounting for those trades which have occurred since the BA rankings were locked down, I counted the number of 50 or higher-graded prospects (i.e. the sort which provide surplus value) in each system. The results follows.
| Team | 70 | 65 | 60 | 55 | 50 | Surplus WAR | Surplus WAR Rank | BA Rank |
| Boston | 2 | 2 | 1 | 2 | 13 | 42.5 | 1 | 4 |
| Los Angles Dodgers | 2 | 3 | 3 | 11 | 37.0 | 2 | 1 | |
| Texas | 1 | 1 | 5 | 1 | 8 | 35.5 | 3 | 7 |
| Houston | 2 | 3 | 6 | 9 | 34.5 | 4 | 2 | |
| Atlanta | 1 | 5 | 5 | 11 | 34.0 | 5 | 3 | |
| Colorado | 1 | 5 | 4 | 13 | 33.5 | 6 | 6 | |
| Washington | 1 | 1 | 3 | 2 | 13 | 33.5 | 6 | 5 |
| Minnesota | 1 | 4 | 3 | 10 | 30.5 | 8 | 10 | |
| Milwaukee | 1 | 2 | 8 | 10 | 29.0 | 9 | 9 | |
| New York Mets | 1 | 1 | 8 | 12 | 27.0 | 10 | 15 | |
| Pittsburgh | 1 | 3 | 4 | 11 | 26.5 | 11 | 11 | |
| St. Louis | 1 | 7 | 13 | 26.0 | 12 | 14 | ||
| Cincinnati | 1 | 3 | 2 | 15 | 25.5 | 13 | 12 | |
| Philadelphia | 1 | 2 | 6 | 9 | 25.5 | 13 | 8 | |
| Tampa Bay | 1 | 2 | 5 | 12 | 25.5 | 13 | 13 | |
| Cleveland | 1 | 3 | 2 | 14 | 25.0 | 16 | 17 | |
| Kansas City | 1 | 1 | 6 | 14 | 25.0 | 16 | 21 | |
| New York Yankees | 5 | 2 | 11 | 23.5 | 18 | 16 | ||
| Chicago Cubs | 1 | 9 | 11 | 22.0 | 19 | 20 | ||
| Toronto | 2 | 5 | 10 | 18.5 | 20 | 24 | ||
| San Francisco | 1 | 5 | 13 | 16.5 | 21 | 19 | ||
| Oakland | 1 | 6 | 9 | 17.0 | 21 | 18 | ||
| Seattle | 1 | 3 | 16 | 15.5 | 23 | 28 | ||
| Chicago White Sox | 2 | 2 | 11 | 14.5 | 24 | 23 | ||
| Arizona | 4 | 14 | 13.0 | 25 | 22 | |||
| San Diego | 7 | 5 | 13.0 | 25 | 25 | |||
| Detroit | 1 | 4 | 7 | 12.5 | 27 | 26 | ||
| Baltimore | 2 | 2 | 6 | 12.0 | 28 | 27 | ||
| Miami | 1 | 2 | 10 | 11.0 | 29 | 29 | ||
| Los Angles Angels | 1 | 8 | 5.5 | 30 | 30 | |||
| Total | 8 | 16 | 63 | 126 | 329 | 710.5 |
My thoughts:
- Boston has the best system by this measure, with some heavy hitters at the top: both Yoan Moncada and Anderson Espinoza feature 70 grades, while Rafael Devers and Andrew Benintendi earn 65s. The second-ranked Dodgers were the only other team with two 70-grade prospects.
- Three distinct tiers exist: one ranging from 33.5 WAR to 37.0 WAR (six teams) and another from 25 WAR to 28.5 WAR (nine teams). The final one covers 11.0 to 15.5 WAR (seven teams). These three tiers make up 22 of the teams with the other teams sprinkled in-between.
- I had to check the Angels system a couple times to see if it is this bad. They only have one player graded at 55. The Cubs, by comparison, have nine.
- It was necessary this year for an organization to employ at least one 65 or higher-graded player to have a top-half farm system. The top 17 ranked systems had one such player, the bottom 13 didn’t.
- Some preseason contenders — the Red Sox, Dodgers, Rangers, Astros, and Nationals — have the system depth to go get a decent player at the trade deadline. Other teams like the Angels, Tigers, Diamondbacks, and White Sox just don’t have as many pieces to trade.
Well, I have the farm systems ranked using one publication. I plan on doing the same with MLB.com’s prospect rankings and our own Dan Farnsworth’s once each on is complete. Until then, it looks like the Boston Red Sox can put a claim on the best farm system in the majors.
Jeff, one of the authors of the fantasy baseball guide,The Process, writes for RotoGraphs, The Hardball Times, Rotowire, Baseball America, and BaseballHQ. He has been nominated for two SABR Analytics Research Award for Contemporary Analysis and won it in 2013 in tandem with Bill Petti. He has won four FSWA Awards including on for his Mining the News series. He's won Tout Wars three times, LABR twice, and got his first NFBC Main Event win in 2021. Follow him on Twitter @jeffwzimmerman.
In your research did you see if a 70 grade hitter is as valuable as a 70 grade pitcher?? Wondering if setting different values for hitters and pitchers would make sense.
Maybe, but it wouldn’t have worked with my methodology from the book. Without rewriting the an entire article, it just wasn’t possible.
Does this factor in risk? For example, BA has Devers one spot above Espinoza in their rankings, but I assume here that Espinoza is worth 3.0 WAR more than Devers. (Not to mention Seager, who is 17 spots above both of them but worth no more WAR than Espinoza here.) Thanks.
When I looked at the BA grades originally, I never took risk into account. If the player was a 70, I fill the risk is baked in. Looking at the Red Sox in 2012, first year of full team grades, BA had:
1. Middlebrooks (60)
2. Bogaerts (65)
3. Swihart (65)
Middlebrooks may have looked safer, but the upside of the higher grades is more important. In the long run, the 65 grades will on average give more surplus value than 60 grades.
I guess it comes down to knowing you will have known mediocre talent or the chance of a star. I will take the stars.
The risk isn’t baked in. A 60/extreme isn’t the same as a 60/low. Generally though, you can slide a half-step down the grade scale to lower the risk (ie, a 50/medium could be given a 45/low – of course going in the opposite direction is not always true if the player just doesn’t have that kind of ceiling).
If a player has the star potential they generally have a pretty high floor as eventually they will be given every opportunity to succeed by one organization or another. A 70 grade prolly has a 45 floor almost no matter what just based off talent alone they will make rosters.
I love the principle behind this methodology. I’m more glad that you’ll apply it to other ranking systems. There are usually huge jumps between the same player between, say, Baseball Prospectus and MLB.
Getting this into a more aggregate sphere would be great.
Is it fair to give Yoan Moncada all that surplus value, when you consider he’s already reportedly cost the red Sox 100+ million.
Yoan Moncada has received a 31.5 Million dollar signing bonus and the Red Sox had to pay a 100% tax on that signing bonus of an additional 31.5 Million dollars. Other than that 63 Million dollars, they have not spent anything more on Yoan than any other prospect.
I am not sure where you are getting the idea that the Red Sox spent an additional 37+ Million dollars on Yoan.
Bad reporting
Only $63 million for a guy who played in A ball last year?
Sigh. That’s how the other half lives, I guess.
Moncada is already paid for. Going forward, he costs what any other prospect costs. If a team where to trade for him, they wouldn’t owe him any of the 62mil. The Redsox already paid that.
Does the surplus value of a player like Moncada include the $60M it cost the Red Sox to sign him?
No, not in this case.
You have 542 prospects with 50 (Average) or above future grades. That seems like a lot given there are 750 active major league players. I’m not saying it is wrong, but my gut tells me that is optimistic. There were 227 MLB debuts last year, and 24 of those managed 1.5 bWAR in a historic class. I’m curious everyone’s guess how many 50 grade or higher prospects there should be in all the farm systems at one time.
I agree the number may be high, I am just going off BA grades. It was the most available set of information. MLB.com is get to the point where they may be usable. Here is last year’s spread with Kiley here at FanGraphs:
http://cdn.fangraphs.com/blogs/wp-content/uploads/2015/02/Top-200-Prospects-Distribution.png
This should be something we can calculate. I’m too lazy to actually do it, but I would guess it should be closer to 300. Something like graduating 1/4 of those 300 a year, every year. Figuring a 10 year career or something. They may not produce the first year, but over a 10 year career you need 75 ‘average’ new players a year or something. Obviously the number of ‘average’ players you need to replace the pool highly depends on how long the average career lasts.
My rough guess was similar, and I arrived at it in a similar fashion. The fact that only 24 players put up 1.5 bWAR in their debut (albeit in partial seasons) made me thing that it was a bit smaller: 200-250. I’ve heard many say that the current top 100 list is weak. Maybe I’ll go with 200. Can someone keep track and let me know in 15-20 years?
Just to respond to a few very reasonable comments, I don’t think people should get too caught up in the nuances between prospects grades and such. The reason being that to worry that much about it is to assign FAR too much precision to this exercise.
One person rightly pointed out that many given overall grades could be taken +/-5 points, and I think that’s on point. It’s extremely unlikely that EVERY listed prospect for one team would have his grade inflated or deflated to the full 5 points of that range, so the variance likely cancels out. Even if it doesn’t, again there should be an understood lack of precision here.
So, the value of the exercise is in seeing where teams generally rank. And it shouldn’t surprise that they all end up in the vicinity that we would expect.