There have been some interesting advanced stats released to the public in recent seasons that allow us to delve more deeply into player performance. For example, BABIP allowed us to look not only at a player's batting average overall, but also how well he did specifically on balls in play. We could identify whether that player had an inordinately high or low percentage of hits after making contact and whether his results were skewed by luck or largely a product of skill.

Now we can go beyond basic BABIP with "line drive percentage," "pull percentage" and "hard-hit percentage." A player who didn't strike out a ton and hit the ball hard more often than average should have a good batting average. If he didn't, we can reasonably conclude he was unlucky, and we should therefore project him to do better going forward.

The problem, however, with BABIP and even these deeper, more precise indicators is they're all backward looking. The question you're answering with them is: "How good or lucky was the player in the past?" We usually have a decent idea about how good a player was with his basic stats, and these new ones merely allow us to get a little more precise.

This is helpful to an extent. I'd like to know whether a .300 hitter really had the skills of a .285 or .315 one. But is this really what wins fantasy leagues for us? I would argue it's not.

The biggest factor in winning fantasy leagues is finding players whose values greatly exceed their costs, i.e., nailing the breakouts (as well as avoiding the reverse.) While some players might regress positively in BABIP or HR/FB to your benefit, breakouts are typically players who perform at a level in Year X+1 that vastly exceeds anything they did in Year X. I'm looking at Bryce Harper, Manny Machado, Jake Arrieta and Dallas Keuchel from 2015. Are those more a matter of their bad luck regressing back toward the mean or the light bulb going on, and their career trajectories taking on an altogether new arc? If you look at the Steamer projections for any of those players for 2015, you can see they were not even close to their actual output. And Steamer is all about regressing players with outlier luck stats like BABIP. That tells me the positive regression a system like Steamer prices in cannot possibly account for their breakouts. (I use Steamer not only because they price in the luck-based regression, but also because they're the most accurate system available.)

Here's a comparison below of the 2015 Steamer projections and actual seasons for these four players:

Hitters:

Pitchers:

As you can see, the projections which regressed Keuchel's supposed good BABIP luck from 2014 (.296 despite a 3.6:1 GB/FB ratio) did so in the wrong direction (Keuchel had a .277 BABIP on 3.71 GB/FB.) Moreover, Steamer had no way of foreseeing his strikeout rate jump from 6.6 to 8.4 per nine innings. In short, Keuchel was a huge value and on so many league-winning teams because (1) he showed what looked like luck might in fact be skill; and (2) his skills improved.

A similar story could be told for the other players - it's not merely that they stayed healthy all year, but also their rate stats greatly exceeded what they produced in 2014 and did so even if you priced in expected age-related growth.

This is not to say certain advanced stats couldn't potentially be used as indicators of breakouts - a young player who hits the ball especially hard might in fact be the type that hits 40 home runs two years hence. But those breakouts happen because of skills growth above and beyond the prior performance and not due to a normalization of luck. For that reason, no backward looking stat, no matter how advanced, can account for a player's future growth. The best it can do is point to the possibility.