Understanding the Numbers Behind Fantasy Football Matchup Rankings

a matrix of numbers

Not all things are created or reported equally. This fact came to light when I wrote the article, “This Guy or That Guy? Fantasy Football Week 12″. In analyzing which offensive players to sit or start, we use a data metric for fantasy points allowed to the position by opposing defenses. If an opposing defense is stout and allows the fewest fantasy points in the league to running backs, try to avoid flexing your RB against them. But as we discussed, your studs like McCaffrey, are matchup proof. You roll with them every week. At the same time, this elite run defense could be last in the league against wide receivers. That is one hair to split and not just assume that a team’s defense is solid across the board. This is not, however, the main point I wanted to make.

Raw points v. Schedule Adjusted Fantasy Points Allowed (aFPA)

I looked at the fantasy matchup for New Orlean Saints running back Tony Jones against the Buffalo Bills on Thanksgiving night and noticed a huge discrepancy between ESPN’s matchup ranking and 4for4’s. ESPN had the Bills as #9 against running backs while 4for4 was at 25. Big difference, please explain. In 10 games, the Bills defense gave up an average of 19 points a game to running backs (half ppr). In raw scoring that places them 9th on ESPN’s matchup rankings. By comparison, the Jets are dead last on everyone’s list with over 30 points a game to running backs. The key phrase here is “raw points”. The 19 points a game allowed to the Bills’ running back opponents does not factor in their strength of schedule, or put another way, the quality of running backs they faced. 4for4, however, has a unique, maybe proprietary way, to handle rankings that eliminate the schedule bias. It is called Schedule Adjusted Fantasy Points Allowed (aFPA).

Schedule Adjusted Fantasy Points Allowed (aFPA) explained

According to the 4for4 website, “To understand how aFPA works take two teams from 2015, the Dallas Cowboys and the Kansas City Chiefs. After Week 15, Dallas and KC ranked first and second, respectively, in aFPA versus opposing quarterbacks. In the 10 weeks leading up to Week 16, Dallas had allowed just 11.6 raw fantasy points per game to opposing quarterbacks, while KC had given up 14.3, but raw fantasy points allowed doesn’t remove schedule bias. The Schedule Adjusted Fantasy Points allowed metric recognized that in that 10 week period, the group of QBs that Dallas faced averaged 14.4 FP/g, while KC played against QBs that posted 19.1 FP/g. The Chiefs’ ability to hold that group of signal callers to just over 14 points FP/g is clearly an impressive feat, and aFPA rewarded KC with a high ranking.”

An illustration to make it real

Pretend you and your buddy make a bet about who can date the most women over a 30 day period. Your buddy kills it and takes out 20 different women while you manage only dates with 5 women. He wins, right? Yes, based on the raw numbers he does. But let’s say that all 20 of his lady friends were dog-ugly, meth-head skanks, with rotten teeth and who were overjoyed by the prospects of a free meal. On the other hand, all 5 of your dates were Sports Illustrated swimsuit models. Does it feel right when he asks you to Venmo him the prize money? Probably not. That’s how 4for4 felt about using raw numbers to measure the strength of defenses against every offensive skill position. The rankings should be weighted by the quality of the competition, as you might agree with your dating bet.

Why this long article to explain 4for4’s Schedule Adjusted Fantasy Points Allowed (aFPA) ranking metric? Because in my articles it’s the ranking system I’ll use when analyzing players. If I recommend New Orlean Saints running back Tony Jones against the Buffalo Bills because the Bills are 25th against the run and your fantasy football app shows the Bills ranked #9, you’ll understand the reason for the discrepancy.

Sorry if I let you down on the Tony Jones pick, but Tony Pollard more than made up for it. I’m 1-1 on my Tony’s.

 

.

Follow me!

Leave a Reply

%d bloggers like this: