Using Sabermetrics For Fantasy Baseball Part 15 - Minor League Stats

Once you've grown accustomed to having advanced tools to help make fantasy decisions, it can feel disorientating to be without them. Prospects are increasingly becoming a focal point in both real and fantasy baseball, but the minors simply do not have all of the data available for MLB players. For example, advanced plate discipline stats, Pitch Info, and anything Statcast-related are all currently unavailable for minor league campaigns.

Does this mean we go back to looking at ERA and batting average as the only indicators of future performance? Of course not! Instead, we do our best to work with what we have. The process begins by looking at the environment. Higher levels of competition result in more accurate data, so you should start by excluding anything lower than Double-A if a player's track record allows it.

Here's how to effectively use this data to give you an edge in your fantasy baseball league throughout the season.

How to Interpret Minor League Stats

The first point to remember is that the baseline for certain predictive metrics is different on the farm. Mike Podhorzer of FanGraphs.com has an excellent article detailing the specifics. For example, Double-A hitters collectively posted a .306 BABIP last year, while their Triple-A counterparts managed a .317 figure. Both marks are significantly higher than MLB's .300 BABIP, making a performance that looks fluky actually league-average.

Another common sticking point is IFFB%. Double-A batters posted a ludicrous 21.6% IFFB% on their fly balls last year, while their Triple-A counterparts were only slightly better (20.8%). This leads many fantasy owners to conclude that EVERY minor league prospect has a massive pop-up problem, but this is not the case. The stat is calculated differently on the farm, and you need to halve it to get something approaching an MLB projection.

Like MLB, each minor league and ballpark also has its own unique quirks and tendencies. For example, the Pacific Coast League is a Triple-A league notorious for inflating offensive statistics. Imagine if an entire league played in Coors Field every game. That's basically the PCL.

For PCL players, a batting line may look good at first glance, but really represent only an average performance. Likewise, pitchers may put up dreadful numbers even after they are ready for the Show. For instance, a certain PCL pitcher put up a 9-7 record with a 4.60 ERA in 133 IP in 2014. His K% was a robust 24.9%, but none of his other stats screamed MLB ready.

However, some fantasy owners noticed that his BABIP against was a ludicrous .378, a number that would almost certainly regress in a different environment. The pitcher never ran a BABIP that high in any other minor league stop. His LOB% of 67.2% would likely climb as the BABIP dropped. We have FIP for minor leaguers, and this pitcher's was 3.70--still not great, but much better than his ERA.

Despite ugly Triple-A results in 2014, this pitcher pitched in the majors for 150 innings in 2015. His 9-7 record repeated itself, but his ERA fell to 3.24, right in line with a FIP of 3.25. The K% he flashed in the PCL translated to the majors, where he posted a strong 27.5% rate. His name is Noah Syndergaard, and he definitely had owners kicking themselves by the end of 2015 for trusting minor league surface stats. Nothing changed in 2016, as Syndergaard went 14-9 with a 2.60 ERA and 29.3% K%. Injuries limited him last year, but he was still elite in his 30 1/3 IP (2.97 ERA, 1.31 FIP, 27.4% K%).

If memorizing each league's tendencies is too overwhelming for you, you can look at Weighted Runs Created Plus (wRC+) as a shortcut. This metric sets 100 as the league's average offensive output, with each number higher or lower representing a one percent difference in either direction. This means that a wRC+ of 95 is five percent worse than league average, while a mark of 110 is 10 percent better. While the formula does not directly translate to fantasy value, park and league adjustments are already included in the calculation.

Another common problem with minor league statistics is sample size. It is easier to run an unsustainable BABIP or HR/FB in a small sample than a larger one. The minor leagues compound this problem by allowing a healthy player to be called up or demoted multiple times in one season, leaving us with two or more partial season samples instead of one full season of statistics. Astros shortstop Carlos Correa illustrates this, as he had a grant total of 246 PAs at Double-A and Triple-A combined before his MLB call up in 2015.

Due to the small sample, Correa's BABIP was unreliable. In this situation, I like to examine the player's plate discipline numbers because they stabilize (or become predictive) more quickly. At Double-A, Correa had an 11.3% BB% against an 18.8% K%, indicating a strong knowledge of the zone. Triple-A saw his BB% drop slightly to 10.6%, but a drop in K% to 12.4% made his overall plate discipline profile stronger.

Correa posted a 9.3% BB% and 18.1% K% en route to his Rookie of the Year award in 2015. Correa was even more willing to walk in 2016 (11.4% BB%), but struck out a little more often as the league adjusted to him (21.1% K%). These trends held steady last season, as Correa posted a 11% BB% and 19.1% K%.

Plate discipline is harder in the majors than the minors, and we don't have the additional information provided by metrics such as O-Swing%. Still, Correa seemed to possess strong discipline in the minors and managed to take it with him as soon as he was called up to the bigs. In general, a player won't be completely overmatched in the majors if he had strong plate discipline numbers in the minors.

The examples above were chosen because they now have more than one season of MLB data confirming their minor league trends, but this methodology could have helped you in 2017. For example, Rhys Hoskins combined stellar BB% marks (13.5% at Triple-A last year, 12.1% at Double-A in 2016) with sky high FB% (48.6%, 51.6%) and HR/FB (18.2%, 19.9%) rates to profile as an impact power bat with enough plate discipline to avoid hurting your batting average. Owners who took a chance on him got a .259/.396/.618 line with 18 HR in 212 PAs.

By contrast, blindly believing minor league surface stats could have pointed you in Dominic Smith's direction. He slashed .330/.386/.519 with 16 HR at Triple-A Las Vegas before his MLB debut. However, Las Vegas is the Coors Field of the PCL, helping him compile a 28.3% LD% and .380 BABIP nobody could sustain in New York. He was also allergic to fly balls (26.2% FB%), making power difficult to project. He ended up slashing .198/.262/.395 with nine dingers, burning owners who counted on him for the stretch.

Stealing bases is easier in the minors, but elite success rates are still something to look for when projecting fast players. Age is also a factor for minor leaguers, as a 28-year-old dominating a bunch of teenagers at Rookie ball isn't really that impressive.

Conclusion

To conclude, the fact that we do not know a minor leaguer's average airborne exit velocity or BABIP on ground balls does not prevent us from analyzing minor league players for fantasy purposes. We have tools such as BABIP and BB% for hitters and FIP and LOB% for pitchers. We can still place these numbers into context by examining any given league's tendencies. Finding rookie breakouts before they happen is still challenging, but that's what makes it a worthy endeavor.

More Sabermetrics & Fantasy Baseball Strategy