Performance of Top Running Backs — 2019 Update!

I have a lot of questions about life; this post does not cover any of them.

I’ve always wondered what makes a fantasy player elite within their position group? Outside skill conversations, are there other factors about a player’s situation you should take into account when it comes time to draft or pickup? I’m going to attempt to answer this question.

First in a series

This is the first post in a series on fantasy performance by position — running backs, wide receivers, quarterbacks, and tight ends. My goal is to look at each position independently and then knit together a cohesive drafting and pickup strategy. Check back for updates!

Team Wins

The figure below shows the density of final running back rank against team wins. There is a strong trend of higher RB performance on teams with more wins. This could be the result of good teams having good players or some halo effect players get when they are on a good team. Just for the sake of it, let’s look at the correlation coefficient between these two (correlation != causation so take it with a grain of salt).

The correlation between total team wins, and fantasy points are 0.15, meaning there is a slight positive correlation. This makes sense — teams that win tend to score more points. This suggests that you should draft and pick up players who are on good teams. This is a decent data signal so that I will dig deeper into this in future posts in this series.

Correlation Wins vs. Fantasy Points0.15

Running Back Tiers

When I say “tiers,” I’m talking about breaking position groups into general categories — elite, good, average, bad, etc. Tiers are essential to fantasy sports, especially when drafting a team. Tiers will change year-to-year as players cycle through the league and teams alter their offensive/defensive strategies, but they are generally stable enough that past performance can accurately predict them year-to-year. The thing that makes tiers useful when drafting is that you can come up with a strategy that doesn’t depend on when individual players are selected. You can say, go to your draft with the #6 pick knowing that there are five elite RBs. If one of them makes it to you, you take them. Otherwise, you fall back on a different plan.

All the analysis in this section operates under the assumption that we know who will end the season at each rank. It is a hypothetical exercise if we could go back in time to last season and redo our draft(s).

Some positions groups have a lot of elite players; others have drop-offs after just a handful of the best players. The image below shows some interesting declines in points. The cool things I see:

• The top 2 running backs are significantly better than the rest. Getting one of these two guys is a top priority if possible given your draft position.
• RB and QB 1–5 are significantly higher scoring than the rest
• After the first two RBs taken, pivoting to WR appears to be the best play if your goal is to maximize points.
• If you don’t have the top TE, it’s a black hole. What’s interesting here, if you can get a top 10 RB or top 20 WR, it probably doesn’t make sense to reach the top few TEs, assuming your goal is to maximize your points scored. A top TE gives you an advantage over other players, but it might cost you an edge at WR or RB. Another strategy with TE, grab the top TE because there always seems to be a guy or two off waivers you can pick up at RB or WR that end up being a top talent.

A good follow up analysis would be to look at how often the top players turn over each year. This will be something I attempt to investigate in the next few articles, or at the bottom of this one if I don’t get lazy.

Point Fluctuations for Top Running Backs

If we zoom into just RB, understanding how each final position ends with total score can be significant. I find it useful to understand bands of potential scoring so I can evaluate potential risk.

Yearly Data

We’ll start by analyzing the scoring bands at the end of the year.

The boxplots are just a redrawing of the RB violin plot in the Potential Running Tiers section. This shows some interesting points; the top running back will have the largest possible score relative to the next position because there is no upper bound limit to the points for their rank. This will be interesting to compare against the other positions to see if it’s something we can exploit.

Full disclosure, I am going to do a full write up on WR next, In the interim, here is the same plot, only with WR. There is a much less pronounced drop off from first to the second position, and really, from position to position. This is probably a function of more WR being relevant or something along those lines.

When it comes to drafting, if I am confident the RB I am drafting will be the best overall, I will likely be taking that person, over the number one WR.

Weekly Changes in Points

The yearly analysis is fun and matters for the long game. Unfortunately, this isn’t baseball where you have 22 weeks of the regular season for players to regress to their mean performance, unlike baseball where you can reasonably rely on this.

The first thing that I noticed out is the ability for any RB to score over 20 points in a given week. I’m not surprised; it’s more of a confirmation. The biggest issue, as you work your way to lower ranks is the total downside risk that is present. This is also reflected in their final position. I want to find a guy who will not kill me each week. Generally speaking, the Median points scored is not very different from rank 15–32; the gap in the outcome is a function of the weekly variance in points scored.

2019 Recap

• Last year shows some pretty large variances week to week with ~1/3 of all the top 32 players capable of scoring 30 points in a given week.
• The top 11 all have scores that span the 20 point margin within their IQR (IQR is the difference between the 25th and 75th percentile, meaning half of all values fall between the upper and the lower boundaries of the box.
• Christian McCaffrey / Alvin Kamara are interesting cases because their IQR is large. It’s probably no coincidence they are receiver heavy RBs.
• The RBs outside of the top 10 had lots of downside risk with weekly scoring, meaning, they can kill you in any given week. This fact is true about anyone in any position. What’s stark about this group is what ranges the box spans, which are where 50% of the points scored fall into.

Conclusion

Everyone has their draft strategy going into the season. The person in my 14 team league who has won 4 of the ten years, it’s been around, swears by going all in on wide receivers and punting everything else. There might be something to that strategy, and hopefully, we can address that strategy in the wide receiver post. Based on looking at running backs, I will focus on drafting someone at the top and cascade my decisions on other positions from there and originally

published at gridironai.com.