Predictability of Game Scores and Fantasy Positions (during five years: 2017-2021)

The following shows how predictable/random the last years have been for "Vegas" (betting lines), in comparison to 2021. For each year, I show the correlation coefficient between lines and results:

The season predictability (from betting lines) has been trending to a level almost as low as 2017. The spread has been most difficult, in relative terms.

The blue line (implied game score) is trending towards 0.3 (the level of '17), which indicates the difficulty of making good picks based on match-up-- especially D/ST. So let's check the streaming positions (K,D/ST,QB) and see that D/ST and QB do actually follow a similar trend:

I am also indicating the predictability in the first half of each season, for a more direct comparison against the current year.

A few takeaways:

  • The QB and D/ST trends roughly match the trend of Vegas-implied team scores (shown in the first chart above).

  • D/ST plunged again in 2021 due to several weeks of mid-season betting line upsets.

  • The first half of each year has actually been more predictable for D/ST (and QB), compared to the second half of the season.

  • Kicker has apparently gotten more difficult the last couple years-- roughly following the same trend as betting line spread predictabilities.

 

Finally, we might as well also show the Flex position predictabilities for good measure:

The "first-half-season trend" is shown separately only for TE, because it is the only position with notable difference between the start and end of each season (getting more predictable by season-end).

  • By eyeball, all the RB/WR/TE trends follow a trend like the O/U line of the first plot.

  • TE tends to be most predictable, but only comes through in the second half of each season.

  • RB (again: total team-RB, not individual RB1s) is least predictable, most seasons.

  • Team-WR is having a great season so far, despite increased randomness for all other positions.

 

TL;DR-- Fantasy rankings have (all) encountered a slump due to increased levels of upsets (upsets relative to betting lines). There are not many ways to beat betting line consensus, and D/ST and QB have taken the biggest hits.