Diving Head First into Minor League Averages

As a projectionist, one of the most difficult players to forecast are rookies with no previous MLB experience. While there have been many attempts at translating minor league performance into Major League equivalents, we will never get the conversion perfectly right. It’s hard enough to project established veterans, so with such varying competition, a more limited body of work with which to analyze, and wildly fluctuating league and park factors, minor leaguers are a real challenge.

In the last couple of years, FanGraphs has added minor league stats to player pages, including advanced metrics such as HR/FB rate and SwStk%. This is fantastic. However, what’s still missing is context. We only know that Aaron Judge’s 2017 HR/FB rate of 35.6% was outrageously good because it was significantly above the 13.7% MLB average. Without that context, we would be judging Judge blind.

So if context is what you demand, then context I will provide. I painstakingly downloaded all our minor league data, grouped by level (AAA: All Leagues, AA: All Leagues, etc), and then calculated a myriad of metrics. Unfortunately, it would have been too time consuming to download the data grouped by each league at each level, though I acknowledge those averages would be even more valuable.

The resulting comparison will allow us to compare performance by level and help us convert minor league rates to the Major Leagues. After each table of averages, I will present the level factor using MLB as the base. Essentially, you would multiply the level factor by the player’s rate at that level to get the MLB equivalent. Caveat: this is not the same as calculating a Major League Equivalency for a rookie or creating a projection, as there is far more that would go into such a calculation. This is simply a comparison of how each league as a whole compares to the rest and MLB, not how a minor league hitter would perform in the Majors.

First, we’ll start with the triple slash, which won’t have a league factor table following.

Triple Slash Line by Level
Level AVG OBP SLG ISO OPS
MLB 0.255 0.324 0.426 0.171 0.750
AAA 0.266 0.334 0.415 0.149 0.748
AA 0.254 0.326 0.384 0.130 0.710
A+ 0.255 0.324 0.381 0.126 0.705
A 0.250 0.321 0.375 0.124 0.695
A- 0.245 0.320 0.351 0.106 0.670
R 0.263 0.344 0.389 0.126 0.733

Interestingly, it’s the Rookie league that gets on base at the highest clip. After that, the league OBP collapses at Low-A and gradually rises, before plummeting back again at the Major League level. But what MLB has that the other leagues are first starting to develop is power. The leaguewide power spike is quite evident here, as ISO is significantly higher than anywhere else. We know that power takes time to develop, and we could see that as it rockets up from Low-A to Single-A, before surging again at Triple-A. This suggests that if your favorite top prospect has yet to display the immense power he was projected for through the Double-A level, don’t fret just yet. That power spike will hopefully occur at Triple-A.

Now let’s check in on the league average batted ball metrics, along with a couple of other random rates.

Batted Ball Metrics by Level
Level LD% GB% FB% IFFB% BABIP HR/FB AB/HR PA/SB
MLB 20.3% 44.2% 35.5% 9.6% 0.300 13.7% 27.1 73.3
AAA 20.4% 43.7% 35.8% 20.8% 0.317 9.7% 36.5 57.8
AA 20.2% 44.3% 35.5% 21.6% 0.306 8.3% 43.3 58.2
A+ 19.4% 44.8% 35.7% 21.0% 0.316 7.4% 49.2 46.3
A 18.3% 45.4% 36.4% 21.4% 0.311 7.1% 50.3 43.0
A- 20.0% 47.0% 33.0% 23.5% 0.313 5.8% 69.1 41.4
R 19.3% 46.0% 34.7% 21.7% 0.327 6.8% 54.9 38.7

Now I present to you a table of level factors, calculated as a percentage of that level the MLB mark represents. So the formula is simply MLB/Minor league level. That way you could treat each mark as a multiplier for the level to get the equivalent MLB mark.

Batted Ball Metric Factors by Level
Level LD% GB% FB% IFFB% BABIP HR/FB AB/HR PA/SB
AAA 0.995 1.011 0.992 0.462 0.946 1.412 0.742 1.268
AA 1.005 0.998 1.000 0.444 0.980 1.651 0.626 1.259
A+ 1.046 0.987 0.994 0.457 0.949 1.851 0.551 1.583
A 1.109 0.974 0.975 0.449 0.965 1.930 0.539 1.705
A- 1.015 0.940 1.076 0.409 0.958 2.362 0.392 1.771
R 1.052 0.961 1.023 0.442 0.917 2.015 0.494 1.894

Phew, that’s a lot of data. Let’s start with the batted ball type distribution, which is composed of the first four columns after the level indication. These include the line drive rate (LD%), ground ball rate (GB%), fly ball rate (FB%), and infield fly ball rate (IFFB%). We can’t be sure if differences here are due to stringer classification bias, hitter skill, or a combination of both.

If we assume skill, then we find that before reaching Double-A, hitters haven’t yet mastered their line drive stroke. That LD% mark jumps in Double-A and essentially remains flat through the Majors. On the other hand, when prospects debut in professional baseball, they kill worms a bit more often, pounding the ball into the ground with slightly greater frequency. Thankfully, that worm killing tendency erodes as prospects climb the ladder and eventually stabilizes beginning at Double-A.

Fly ball rate is interesting, because at the lowest two levels, prospects hit flies less than 35% of the time. But after that, aside from a spike above 36% at Single-A, the tendency flattens out. I always held the belief that young hitters learn to lift the ball more as they reach their peak power years, but this trend indicates no such movement.

Last of the batted ball metrics is perhaps the most fascinating — IFFB%, or popup percentage. Remember how earlier I wondered whether differences related to classification bias or batter skill? I’m fairly sure that for IFFB%, it’s the former. Minor leaguers were generally consistent with their popup rates around the 21% to 22% range. But in the Majors, that rate is halved! Do minor league hitters have difficulty squaring up the ball, making them more prone to the dreaded popout? Or are minor league stringers classifying significantly more fly balls as popups than MLB stringers are? Given the relative consistency of the other batted ball rates, I would bet its the stringer causing issues here.

This is an important finding — no longer do you have to panic because your favorite minor leaguer (I’m talking to you Willie Calhoun) just posted a crazy high IFFB% (Calhoun posted an elevated 30.3% IFFB% at Triple-A last season while still with the Dodgers organization, and then an absurd 37% mark after being traded into the Rangers organization). Okay okay, so even if we halve Calhoun’s IFBB%, it’s still too high, but it’s not as outrageously so as the pre-adjusted rate initially led us to believe.

Moving past batted ball type distribution, we land on one of our favorite advanced metrics, BABIP. I have written many times in past articles that minor league BABIP is dramatically higher than MLB BABIP. Now I have proof. So when you see that a hitter at Triple-A just posted a .320 BABIP and think he’ll perform similarly at the Major League level, you may be severely disappointed. This metric would have seriously benefited from being broken down by league within each level, but that’s a project for a different day.

The HR/FB and AB/HR (lower is better) columns confirm what we already know and match the first table’s ISO trend. Though you figure that a prospect’s HR/FB rate is going to decline once hitting the Majors given the higher quality of pitching he will face, perhaps this is no longer the case. Or, this is just another artifact of selection bias, as only the best of the best ever make it to the Majors. Probably a combination of multiple factors. The bottom line is that I need to consider the possibility that a top prospect holds his minor league HR/FB rates, or even increases it in his first year.

Geez, does MLB have the need for speed or what?! The PA/SB (lower is better) mark nearly increases in lockstep, taking only a brief pause when moving from Double-A to Triple-A. This is the effect of aging, as speed is a skill of the young. While there are probably examples of some players legitimately getting faster or drastically improving their stolen base skills, the vast majority slow down each year and will decide to run less frequently.

Last, let’s dive into batted ball direction data and plate discipline metrics, followed by a table of level factors.

Batted Ball Direction & Plate Discipline Metrics by Level
Level Pull% Cent% Oppo% SwStr% K% BB%
MLB 39.8% 34.9% 25.3% 10.5% 21.6% 8.5%
AAA 40.4% 27.5% 32.0% 10.3% 20.5% 8.5%
AA 40.4% 26.6% 33.0% 10.5% 20.4% 8.7%
A+ 44.4% 24.3% 31.3% 12.1% 22.0% 8.3%
A 44.5% 24.3% 31.3% 11.9% 22.1% 8.2%
A- 44.9% 22.9% 32.2% 12.0% 23.0% 8.6%
R 44.8% 23.9% 31.3% 18.3% 21.3% 9.5%

Batted Ball Direction & Plate Discipline Metric Factors by Level
Level Pull% Cent% Oppo% SwStr% K% BB%
AAA 0.984 1.268 0.789 1.020 1.056 0.998
AA 0.985 1.314 0.766 1.001 1.058 0.980
A+ 0.896 1.435 0.809 0.864 0.984 1.024
A 0.895 1.438 0.809 0.885 0.976 1.041
A- 0.886 1.526 0.785 0.873 0.941 0.990
R 0.889 1.460 0.807 0.574 1.016 0.898

The Pull% trend is one of the most surprising observations here. I was under the impression that as hitters grow into their power, they learn to pull the ball. The data tells the opposite story, with batters going up the middle more and more frequently, and Major League batters giving up on opposite field shots. Was this lack of interest in going the opposite way a side effect of the leaguewide power surge? Nope, as that mark has remained between 25% and 26% since 2009.

Check out that Rookie league SwStk%! Either hitters getting their first taste of professional action are flailing away or there is some weirdness going on in the calculation. The mark remains slightly elevated through High-A, then declines and flattens from Double-A on. So don’t be alarmed if your favorite prospect’s SwStk% is a bit higher than you would like at the low levels.

Finally, we end with hitter strikeout and walk rates. Despite a strikeout epidemic occurring at the Major League level, it’s no higher than minor league levels. After starting at just above 21% in the Rookie league, strikeout rates spike before gradually improving through Triple-A. Once in the Majors, however, hitters have been willing to trade strikeouts for power.

In another surprise, walk rate remains quite stable from Low-A and up. Only the Rookie league has the world’s most patient hitters, but that patience erodes immediately after. The walk rate aging curve tells us that young hitters improve their plate patience in the early years, but that seemingly occurs while already in the Majors, not while still in the minors.

There is a whole lot more to analyze in the wonderful world of minor league baseball, but this introduction should provide the context necessary to improve upon the accuracy of minor league hitter evaluations.

Mike Podhorzer is the 2015 Fantasy Sports Writers Association Baseball Writer of the Year. He produces player projections using his own forecasting system and is the author of the eBook Projecting X 2.0: How to Forecast Baseball Player Performance, which teaches you how to project players yourself. His projections helped him win the inaugural 2013 Tout Wars mixed draft league. Follow Mike on Twitter @MikePodhorzer and contact him via email.