Fantasy Baseball Applications of zxwOBAcon Analysis

In part one of this series, I established that zxwOBAcon is the primary driver of season-to-season changes in xwOBAcon and that the biggest risers in zxwOBAcon each season tend to be rebound candidates. Furthermore, rebound candidates recover all of their zxwOBAcon losses more often than not and frequently outperform their season x zxwOBAcon in season x+2 despite declines in season x+1.

In part two, I applied those findings to 2020 hitters and identified potential draft values for 2021. That process involved developing low-end, high-end, and middle xwOBAcon projections for 2021 rebound candidates and using those projections to establish draft pick targets for the rebound candidates.

This article is the third and final part of the series, in which I’ll outline future applications of the findings discussed in part one. I’ll also review some potential opportunities for improvements and areas for future research.

• Fantasy baseball draft kit
• Fantasy baseball rankings
• Team Sync platform and Draft Assistant
• Fantasy baseball mock draft simulator
• Fantasy baseball draft cheat sheets
• Fantasy baseball closer depth charts
• Fantasy baseball prospects

Effects of Drops in zxwOBAcon on ADP

Although several factors can influence a hitter’s xwOBA each season (many of which have nothing at all to do with contact quality), drops in zxwOBAcon are fairly highly correlated to drops in xwOBA. To that end, season-to-season changes in zxwOBAcon explain about 60% of the season-to-season changes in xwOBA. 

xwOBA is not comprehensive when it comes to fantasy value, but the metric represents a solid proxy for a hitter’s fantasy value. Understandably, then, fantasy managers have been reluctant to draft zxwOABcon rebound candidates compared to their non-rebound candidate counterparts. 

In drafts leading up to the 2020 season, the average rebound candidate saw their ADP increase around 73 picks from drafts leading up to the 2019 season compared to a one-pick increase for non-rebound candidates. Additionally, rebound candidates were more than 1.5 times more likely to see their ADPs rise in 2020 compared to non-rebound candidates. The draft price drop that rebound candidates experience is apparent in the graph below.

Given that zxwOBAcon fallers recover all of their losses (and then some) in the season following those losses more often than not, fantasy managers can and should capitalize on the significant draft-day price reduction that rebound candidates experience. 

Setting ADP Expectations For Rebound Candidates

In part two of this series, I outlined a methodology to set high, low, and middle xwOBAcon projections for rebound candidates. Fantasy managers can integrate those xwOBAcon projections with projections for hitter strikeout and walk rates, as well as any other relevant projections (stolen bases, playing time, etc.) to develop draft price targets for each player.

Of course, fantasy managers would need to choose one of the three xwOBAcon projections for effective integration into draft price target exercises. The middle projection should be used in almost all cases since rebound candidates in general are most likely to post an xwOBAcon near their middle projections, but there are cases where it may make sense to go with the high or low projection. For example, a high projection might make sense for a hitter whose zxwOBAcon drop was fueled by an injury that they have since recovered from, and a low projection might make sense for a hitter who has seen their zxwOBAcon fall in several consecutive seasons.

Overall, the results of this research are likely best used as a guideline for fantasy managers, and the two primary takeaways should be the following:

1. There’s no reason (generally speaking) to expect a hitter whose fantasy value declined based primarily on contact quality to maintain that diminished fantasy value in the next season.
2. Rebound candidates tend to be undervalued in drafts, and fantasy managers should capitalize on that inefficiency.

Future Research

Arguably the most significant issue with the research I’ve outlined in this series is the possibility of survivorship bias. It’s possible that the sampled rebound candidates tend to bounce back so strongly simply because they are higher quality hitters, and that lower quality hitters who see their zxwOBAcon fall significantly in a season fail to meet the batted-ball qualifications for the sample in consecutive seasons. Beyond dealing with the possibility of survivorship bias, there are a few other areas of research that would be interesting follow-ups to this series.

The first area that comes to mind is understanding what separates rebound candidates who successfully rebound from those who continue to see their zxwOBAcon fall. None of the size of a rebound candidate’s zxwOBAcon decline in season x+1 nor changes in average exit velocity or launch angle from season to x to season x+1 are predictive for a hitter’s zxwOBAcon in season x+2, so what is? Possibilities might include age and injury-related factors, among others.

Another area for potential research is an examination of how hitter zxwOBAcon changes in season. Do hitters who start hot in terms of zxwOBAcon tend to maintain that hot start through the entire season? It would also be interesting to see the stabilization point for zxwOBAcon; at what point does a sample of a hitter’s zxwOBAcon include more signal than noise? These are all questions that I’ll leave open for now and look into over the rest of the offseason.

Notes:

1. ADP data is tough to find this time of year. For this analysis, I used ADPs from Smada’s 2 Early Mocks, found here.
2. Rebound candidates are hitters whose zxwOBAcon declined from season x to season x+1.

More 2020 Fantasy Baseball Advice