KATOH’s Organizational Rankings
Last week, Kiley McDaniel published his write-up of the Milwaukee Brewers farm system, marking the 25th installment of his “Evaluating the Prospects” series. Once he makes it through the five teams left on the table, he’ll be publishing his ranking of all thirty teams’ farm systems. In anticipation of this release, I thought it would be interesting to create an organizational ranking based exclusively on KATOH — my prospect projection system. The methodology for making this list seems pretty obvious: Simply add up all of the KATOH projections for the players on each team. In practice, however, this isn’t nearly as straightforward as it sounds.
First, I had to decide what to do with players who appeared in only a small number of games. For example, Carlos Rodon — the White Sox top prospect — receives a projection of 7.9 WAR through age-28. That projection seems pretty reasonable for Rodon, but it’s only based on 95 batters faced over 21 innings. That sample’s not nearly large enough to mean much of anything.
To address this problem, I decided to only consider players who logged at least 200 plate appearances last season (or 200 batters faced for pitchers). This inevitably means that some big-name prospects (like Rodon) wind up being excluded, but there’s really no other way to go about it. KATOH projections are based on 2014 stats, and for guys who only appeared in a few games, the 2014 stats are almost meaningless.
There’s also the issue of knowing the affiliations for thousands of minor league players. My database of 2014 stats includes each player’s organization from the 2014 season, but a lot has changed since then. It would be easy enough to account for the players who were traded, but what about the hundreds of nondescript minor league free agents who changed organizations over the winter? It would be futile to try to figure out where each one ended up.
Due to this restriction, I only included KATOH’s top 400 players in this analysis. I tracked down and confirmed the whereabouts for all 400 of these players, but after seeing some of the names that ranked in the high 300’s, I decided it would be foolish to go any deeper than 400. So prospects outside the top 400 — those projected for less than 1.8 WAR through age-28 — were not considered.
I’m sure you didn’t click on this article to read about methodologies and the caveats, so I’ll cut to the chase. Below, you’ll find the organizational farm system rankings according to KATOH.
| Team | Sum of WAR thru 28 |
| Texas Rangers | 101.0 |
| New York Yankees | 89.2 |
| New York Mets | 88.2 |
| Chicago Cubs | 83.9 |
| Los Angeles Dodgers | 80.4 |
| Houston Astros | 63.8 |
| Boston Red Sox | 62.4 |
| Pittsburgh Pirates | 55.9 |
| Arizona Diamondbacks | 50.9 |
| Tampa Bay Rays | 49.6 |
| Kansas City Royals | 47.8 |
| Seattle Mariners | 46.3 |
| Atlanta Braves | 46.0 |
| St. Louis Cardinals | 45.2 |
| Cleveland Indians | 38.3 |
| Detroit Tigers | 37.4 |
| Toronto Blue Jays | 37.3 |
| Minnesota Twins | 36.8 |
| Chicago White Sox | 36.3 |
| Oakland Athletics | 33.6 |
| Baltimore Orioles | 33.4 |
| Washington Nationals | 33.4 |
| Colorado Rockies | 33.0 |
| Philadelphia Phillies | 32.9 |
| San Francisco Giants | 32.7 |
| San Diego Padres | 30.1 |
| Cincinnati Reds | 28.2 |
| Milwaukee Brewers | 20.9 |
| Los Aneles Angels of Anahiem | 17.2 |
| Miami Marlins | 16.6 |
And here it is in the form of an infographic.

Somewhat surprisingly, its not the Chicago Cubs, but the Texas Rangers who take the top spot. Outside of Joey Gallo, the Rangers don’t have any elite prospects according to KATOH, but they have a dozen guys who project for between 3 and 10 WAR through age-28. The average team only placed five players in this range. This glut of merely good prospects was enough to vault them to the top of the list.
Texas’ depth goes further than that, however. All told, the Rangers have 24 guys in the top 400, which puts them in a tie for the second most. Here’s a look at this breakdown for all 30 teams.
The moral of the story is that depth — rather than high-impact talent — wins out in KATOH’s eyes. Although the Cubs have several guys with excellent projections, their lack of depth (according to KATOH) knocks them down to fourth overall. Meanwhile, the Rangers, Yankees and Mets overtake them due to their sheer quantity of decent prospects. This doesn’t necessarily prove that current evaluations place too much value elite prospects. It very well could be that KATOH’s just overly conservative when it comes to crème de la crème prospects. However, I also wouldn’t rule out the possibility that KATOH might be on to something here.
The order of these rankings may seem a bit odd in some cases, but they appear to be at least somewhat grounded in reality. Last month, using Kiley’s FV grades, Dave Cameron put together a rough estimate of what each team’s farm system is worth. Unsurprisingly, KATOH’s WAR values correlate respectably with Cameron’s dollar values.
KATOH’s organizational rankings are interesting to look at, but I’d caution against taking them too seriously. Not only is the KATOH model far from perfect, but my cutoffs for this list were completely arbitrary. If I had instead chosen to consider, say, the top 500 players with at least 300 plate appearances, there’s a good chance things would move around a bit. By no means are these rankings gospel, but at the very least, they’re something to think about while we wait for meaningful baseball to start up again.
Chris works in economic development by day, but spends most of his nights thinking about baseball. He writes for Pinstripe Pundits, FanGraphs and The Hardball Times. He's also on the twitter machine: @_chris_mitchell None of the views expressed in his articles reflect those of his daytime employer.


“First, I had to decide what to do with players who appeared in only a small number of games. For example, Carlos Rodon — the White Sox top prospect — receives a projection of 7.9 WAR through age-28. That projection seems pretty reasonable for Rodon, but it’s only based on 95 batters faced over 21 innings. That sample’s not nearly large enough to mean much of anything.”
Holy methodological dumpster fire, Batman!
It seems to me that simply ignoring projections for top prospects with limited playing time is maybe the absolute worst way to go about it.
I don’t think that’s a terrible decision given the model, but it implies that KATOH has no regression component based on sample size. Is that really true? If a player puts up a 10% K rate and a .200 ISO over a month’s worth a plate appearances he gets the same projection as someone who does that over a full season? The latter is clearly a much better prospect, statistically. That seems like an area where KATOH has a lot of room to improve.
You are correct. As it stands, KATOH does not regress for sample size. I’m working on this now, and will definitely incorporated it into the next iteration of KATOH.
regression will help, but also more data, such as the 346 innings that Rodon threw in college.
Thanks. College stats are also in the works. In the process of gathering and cleaning the data.
“Holy methodological dumpster fire, Batman!”
Congratulations, this is the most FanGraphs sentence ever uttered
It has become increasingly clear over the last few years that what Sandy Alderson and his team have done well is accumulate and develop young talent. I would put an emphasis on the develop–it seems like they have had an inordinate percentage and number of prospects and young players take big positive leaps over the last few years (Harvey, Lagares, Duda, deGrom, Syndergaard, Herrera, Matz) versus the busted prospects or stagnant major leaguers (Tejada, Davis, Tapia). This has made the Mets one of the deepest teams in MLB, and enabled them to literally shrug off the spring training loss of a young power pitcher worth 2.5 fWAR last year (Wheeler). They are not the best team, but I would be surprised if injuries are the reason they miss the playoffs, unless they have significant injuries at SS or bullpen.
How much worse can they really get with a significant injury to SS?
http://www.fangraphs.com/blogs/maybe-that-mets-shortstop-situation-isnt-such-a-disaster/
Flores could easily be an above average shortstop. If he is hurt, Tejada should be OK, but was sub-replacement level in 2013.
What stands out to me is the huge gulf between the top and bottom systems. Is the top farm system in the game likely to out-produce the worst by a more than factor of 6? I can’t tell if that’s too conservative or too generous, but its cool to see a number associated with it, and maybe something to calculate using the systems from 10 years ago. Thanks!
It wouldn’t surprise me in the slightest that the best teams outpace the worst in terms of WAR over a 6 year period by a factor of 6.
Wait until the Red Sox loonies see this list….”there’s no way we’re only 7th…Swayhart is ta greatast evah!”
Meanwhile Amaro will be adding this to his CV….”Look, we’re 7th from the bottom!”
So good (so good)
Yes, those loonies. Not the loonies who create strawmen in anticipation of said loonies…
http://www.baseballamerica.com/teams/?t=1003
Sorry, but in the time that you wrote this all of the Rays (past, present and future) got hurt so you can probably move them to the bottom.
To find out where all the minor league free agents signed, write a script to download major and minor league team rosters from mlb.com
Good idea. Unfortunately that’s a little above my pay grade right now, but I’ll keep it in mind.
actually not that difficult, I have it run automatically every night, with Python and MySQL. If you email me I can send details
Brian. Brian? Will I appear here at Fangraphs again this year? I was so smart about Mookie Betts last year, when no one else was, I think I deserve another go.
I am a system of projection, not a robot. Promise.
Your “infographic” is a bar chart lol
On the subject of small sample sizes, can we get KATOH projections for Chris Sale and Clayton Kershaw based on their minor league rehab appearances last year? This would be completely pointless, but possibly enjoyable.