| As much as we try to count on the same defenses year after year, the
formula never works straight up. Why do team defenses change with the frequency that they
do, and how can we find a more reliable predictor of future success? Why do the Top Fantasy Defenses Change Yearly?
I believe the primary reasons for this phenomena are:
It only takes one weak link in a defense to
cause an exploitation by opposing offenses. An example of this is an offensive teams'
ability to exploit a weak cornerback, while the opposite side cornerback who may have the
ability to not only defend against points scored, yardage earned, but also accumulate
those all important fantasy defensive points, will see little or no action the whole game.
Many leagues highly reward defensive
touchdowns, which encompass more of a "luck" than "skill" factor.
Don't get me wrong, certain defensive players have more ability to get in the position to
intercept errand throws, or scoop up fumbles, but it frequently also takes a number of
missed tackles after the turnover for this feat to be accomplished.
Teams do not play the same schedule every
season and the outcomes are never the same even if they did ("That's why they
play the games"). Game plans really have a huge impact in this game, and usually
dictate how many chances a defense will get at scoring points.
16 games is not a significantly relevant
statistical sample size, many within the mathematical field consider a sample size of 40
to be a minimum borderline figure.
There are probably another half dozen sound reasons why the
same defenses do not continue their dominance or failure season after season. As a head
coach in the NFL it's their responsibility to address the weak links in the defense and
their game plans during the during the off-season. As a fantasy head coach it's our
responsibility to address their changes and to access these changes over the entire
schedule, game by game.
Fantasy Football is a Statistics Game
If you do not already know it, fantasy football is a
"number's game". If you put in the statistical work required (or join us, which
you already have!), you will finish ahead of where you would have without understanding
the impact of VALID projections and a sound draft strategy.
As a statistician, it's my obligation to report that the correlation coefficient (r)
between prior season's fantasy defenses and the following season's fantasy defenses is
extremely poor. For those of you who forgot the definition of correlation (from a
statistical perspective I don't blame you), correlation coefficient measures the
relativity between a series or set of numbers. In this case we are measuring how much a
prior year's fantasy numbers influenced the outcome of the following season. If we take
this a step further and square the R, we come up with the coefficient of determination (r
squared). The values for r2 range from 0 to 1. The closer to 1 the higher the confidence
we have in the correlation between the numbers.
Here's the revelation, over a 10 year history we found that
the correlation of determination (r2) for fantasy points from prior years
to the following season for the top 10 defenses was .13....ouch that's
pretty bad!
Wait! All is not lost, we can improve these numbers
significantly by addressing and accessing the current situations. What we did in our
analysis of Defenses, was take each game played and assess the fantasy potential of each
defense, based on a number of criteria:
The Offensive Criteria:
Project how many Sacks each
offensive line will yield
How many Interceptions will
each offense yield
How many fumbles will each
offense yield (another difficult number to forecast)
Project how many points
will each offensive unit score this season
Project how many total yards
will each offensive unit total this season
The Defensive Criteria:
How many Sacks would each
defensive unit accumulate
How many forced fumbles
would each defensive unit accumulate
How many Interceptions
would each defensive unit accumulate
How many touchdowns would
each defensive unit accumulate (a difficult number to predict)
How many points would each
defensive unit yield this season (used to assess dominance)
How many total yards would
each defensive unit yield this season (used to assess dominance)
The Fantistics Matrix System
The only way to truly & accurately access Defensive
Team rankings, is to look at each team's opponent throughout the season....it's pain
staking work (but that's the difference between our service and a free service). We
created three grids, and proceeded to weigh each of these criteria based on a simple
defensive scoring system (2 points for Interceptions/2 points for recovered fumbles/6
points for defensive touchdowns/1 point for each Sack). One grid measured the
offensive criteria and the other grid measured that of the defensive criteria.
Finally, a third grid was created that matched up each team's
schedule and produced a value for each game played (see the grid below).
Fantistics 2002 Team Rankings Grid (by week)
Rank |
Team |
1 |
2 |
3 |
4 |
5 |
6 |
7 |
8 |
9 |
10 |
11 |
12 |
13 |
14 |
15 |
16 |
17 |
Sum |
VAM |
| 1 |
PIT |
7.0 |
6.7 |
0.0 |
8.2 |
7.6 |
8.1 |
6.6 |
7.2 |
8.2 |
7.5 |
6.2 |
8.1 |
7.4 |
8.2 |
7.7 |
7.1 |
7.2 |
119.0 |
21.1 |
| 2 |
CHI |
7.6 |
7.7 |
7.8 |
7.3 |
6.4 |
0.0 |
7.6 |
7.6 |
7.4 |
7.2 |
6.9 |
7.6 |
6.4 |
7.5 |
6.6 |
8.0 |
7.3 |
116.8 |
18.9 |
| 3 |
TBB |
7.5 |
7.2 |
6.7 |
8.1 |
7.5 |
8.2 |
7.1 |
7.7 |
7.4 |
0.0 |
7.7 |
6.2 |
7.5 |
7.5 |
7.4 |
5.9 |
6.5 |
115.8 |
17.9 |
| 4 |
SDC |
8.1 |
8.2 |
6.1 |
6.9 |
7.3 |
7.0 |
6.7 |
0.0 |
6.4 |
6.7 |
5.8 |
7.2 |
7.3 |
6.7 |
7.1 |
7.0 |
7.5 |
111.9 |
14.0 |
| 5 |
CIN |
6.0 |
8.0 |
7.3 |
6.9 |
6.4 |
5.7 |
0.0 |
6.0 |
8.0 |
7.0 |
8.0 |
5.7 |
7.0 |
7.5 |
7.2 |
7.4 |
6.9 |
111.0 |
13.1 |
By no means will this Matrix grid accurately predict the
final fantasy rankings at the end of the upcoming season, however via backtesting we are
able to increase our confidence level by 37%! That means you can expect
to be 37% closer to accuracy in selecting your teams, and probably 100% better than most
of the guesses we see posted on the internet.
In our draft strategy (VAM), we also discount the value
(add a level of probability) to these results so that we can ascertain a more accurate
value oriented draft order. The second to last column on the grid below is the
accumulation of these weekly matchups, and the final column is a measure of VAM....viola the 2002 Defensive Team Rankings!
Fantistics
2002 Team Rankings Grading Grid (by week)
|
