September
2008 | Player Projection
accuracy is Vital on Draft Day article by Anthony A. Perri
(Fantistics Founder and
Statistician)
Bigger doesn't necessarily mean better
when it comes to fantasy football..... and this expression
holds true when considering accuracy in player projections and
year end results.
If you are new to Fantistics,
please allow me to indulge you for a few minutes.
We differentiate our service by
basing our player prognostication on fundamental analysis. Our pre-season
projections formula takes over 15 indicators into account to
produce a base projection for each player stat category.
It has allowed us to enjoy an successful prognostication track record which has translated into results
for our
loyal community. What we do here is not by chance or gut
feel. I'm a numbers guy who relies on statistical probability
to make most of the decisions in my life. My eyes see a world
created using a perfect algorithm, an
algorithm that defines both
nature and our universe. In my universe, The Fantasy Universe,
I sheepishly try to apply the same logic.
We fancy ourselves by providing
our subscribers with proven strategies, including the successful
VAM draft strategy.
A strategy that is time tested, not only through our 10 years of
service on the internet, but also through a historical simulation
model. We first introduced Value Above Mean (VAM or position
scarcity) to the draft day regiment in 1999, and have never
deviated from this model. Why should we! It's proven to be 20-30%
more effective than the typical drafting patterns. And this year
for the first time, subscribers can see the results for
themselves. The new Mock Draft Simulator
allows fantasy GMs to see the results of using our VAM
strategy verses computer opponents which utilize Average Draft
Position (ADP).
By season's end we usually have a good handle on how we performed with our fantasy player
projections. Through our year end survey we poll Subscribers
for their year end league placement. Using this, we can gauge how well our projections performed.
Another gauge we utilize is our success rate in correctly
forecasting player
production direction (progression or
digression).
I am frequently asked how our
projections perform against other services. As a consumer it's
only natural to inquire about this. I believe that draft day
accounts for at least two-thirds of your year end outcome. The
other 1/3 is season management, assuming that luck (in either
direction) plays little or no factor. My intention when I
first began to openly post both our success and failures on a
player by player basis (back in 2000), was to have the rest of the
industry follow suit. A way for the consumer to judge the best
product in the market place. It really hasn't happened that way,
and unfortunately all we have out in cyberspace presently are
small studies.
In early August, subscriber Stephen
Cohen sent us an analysis comparing our 2007 football player
rankings against two of the bigger names in the Fantasy Sports
industry (ESPN and
CBS Sportsline). Cohen
took the top 120 players, ranked by ADP (outside of Kickers and
Defenses), heading into the 2007 fantasy season. To compare how
each of the 3 services performed, the prognosticator who came closest to
the actual season
ending player rank received 1 point, a tie would split that point. The results
(details):
Comparison Ranking Top 120
Offensive Players - 2007:
-
Fantistics
42.5 Points 39.0%
-
ESPN
35.0 Points 32.1%
-
Sportsline
31.5 Points 28.9%
