Back by popular demand (since I got a TON of requests for this) is the Expected Fantasy Points. This does take quite a bit of work, so my passion and desire for doing this will be directly related to the feedback and demand for this type of thing.

​

My Method:

We compare a players current “fantasy opportunities” to the points they have actually scored.

I looked at the last 3 year average of Points per Pass Attempts, Rush Attempts, Targets per position. I included both Red Zone stats, and non Red Zone stats, to creates formulas to determine the expected points each person should have scored, based on their opportunities (touches).

I also included position rank given that better players produce more per touch than average players.

Back-Testing Model 2022:

Evaluating this model, I compared the correlations of Actual Points vs my “Expected Points” for 2022 and got these results:

  • QB: .96 correlation
  • RB: .98 correlation
  • WR: .92 correlation
  • TE: .79 correlation (.91 when you remove Taysom Hill lol!)

2022 Full Season: 2022 Expected Fantasy Points (Full Season)

Simplified Example:

Over the last 3 years, the WR1 averaged:

  • .53 points per rush attempt (non Red-Zone).
  • 1.47 points per target (non Red-Zone).
  • .59 points per Red Zone Rush Attempt
  • 2.71 points per Red Zone Target.

Through weeks 5-7 Ja’Marr Chase (Expected WR1) has:

  • 0 non Red Zone Rush Attempts,
  • 26 non Red Zone Targets,
  • 0 Red Zone Rush Attempts,
  • 6 Red Zone Targets.

Using this data, we would expect his fantasy points to be 54.4 Points. He has ACTUALLY scored 55.7 Points. Which is about 102% as expected. Which is about exactly as we expect.

How to use this:

You can view this report like a “Weighted Touches” report. If a player is averaging 15 touches a game, but they are all LOW quality touches (Non Red Zone), this will *LIKELY* result in poor fantasy expected points. This helps you filter out “noise” if one of those touches happens to rip off for 50 yards and a score.

Hopefully this will help identify players who are over producing their touches, and are LIKELY to see negative regression (Sell-High) and identify players who are getting the opportunities which SHOULD lead to higher fantasy production (Buy-Low) if their current level of opportunities continues.

Why Trailing 3 Weeks?

We want to have a large enough body of data (touches) for the law of averages to pan out, but also want to be available to spot trends (shifts in opportunities). If someone had monster targets in week 1, that is less relevant towards our ROS outlook than what they did last week. I felt doing a “rolling 3 weeks” allows us the best of both worlds.

​

SHEETS:

EXPECTED FANTASY POINTS (WEEKS 6-8)

EXPECTED FANTASY POINTS (FULL 2023)