Projecting
Fantasy Football Statistics
During the decade that we've been
forecasting fantasy statistics, fantasy football is, by far, the
toughest sport to forecast of the 4 major U.S. sports.
Possible explanations abound, but logic dictates the most likely
explanations for this phenomena are:
-
The limited size of the sample
in any given year (only 16 games!) ,
-
The rather high occurrence
rate of injuries, and
-
The dependence on other position
players & the coaches playbook.
The Limited sample size:
To produce any meaningful forecasting data, professional
statisticians consider 100 observations to be the minimum
acceptable size of a sample. As you can imagine, 16 regular season
games falls well beneath that threshold. However the cumulative number of
observations within a game can constitute meaningful data. An
example of this is the average running back who attains 300+ carries in
a season. Although the results of these numbers are meaningful,
they are skewed by the only 16 (or less) different opponents that
a particular running back faces during the season.
High Occurrence of Injuries:
For anyone who has not played or watched football....football is a brutal full
contact sport....with lots of injuries! And more importantly no
one can predict injuries. Thus relying on a drafting model
(cheat sheet) which is purely based on individual performances is missing the bigger picture.
Dependence on other Position
Players & Playbook: In baseball a
batter comes up to the plate and takes a full hack 95% of the
time, conversely the opposing pitcher is looking to record an out
almost 100% of the time. In football however, almost every offensive position relies on other
factors outside of their immediate control.
Examples include: a running back who relies
on his line to find an opening, relies on his coach to call the
play....a quarterback who relies on his offensive line to give him time, his
wide receiver to find an opening, his coach to call the pass play....a
receiver relies on the quarterback to find the time, his
quarterback to find his hands, and finally the coach to call the
play (not in this order). As you can see there are a lot of variables that need to
work in sync for our fantasy players to perform adequately. Thus,
because of this dependence
factor, predicting individual performances in fantasy football is
much less predictable as many will have you believe.
The
key to being competitive, is not knowing who
to draft, but rather how to draft. Our
VAM drafting model puts you in a position to be competitive
without the improbable accuracy of the projected player stats.
Sure we spend months working on our individual player projections, and
our projections model is 35%+ more effective than using the prior
year stats, but it's the drafting order we predicate which
will make you competitive. Bottom Line: We do the tedious work,
so you don't have to. Whether you are a new or experienced fantasy
football player, Fantistics utilizes the probability assumptions
contained within the laws of mathematics to give you an edge.
Below is a quick
look at some of the forecasting research we've done based on
prior year statistics on the major fantasy football positions. It's just part of the what we use in our
projections model: |
