There was a question about defenses in the Friday edition of the Mailbag, and I’m afraid I made a mess of it. Paul Bakalars of Minnesota wrote in to ask about luck and defenses, and the numbers I created were poorly structured and full of errors.
So I will take another stab at it.
As best I can tell, there is a substantial amount of luck/chance with defenses. When you start looking at the numbers – the correct numbers – the logical conclusion is that only if you think a defense is really special should you be selecting it before the last couple of rounds.
For this study (and the numbers are definitely right this time around) I used the fantasy scoring system of 6 points for touchdowns (including kick returns), 2 for takeaways and 1 for each sack.
Everyone’s rankings of the defenses are different, but most of those lists are heavily influenced by how the defense performed the previous year. We tend to see the top defenses from 2013 being selected earliest in 2014 drafts. So I used that as my starting point. How much better can we expect a defense to be when it ranked higher than another defense in the previous season?
For the first cut, I broke the 32 teams into four quarters. The top 8 teams (from the previous year) went into the first group, followed by the teams that ranked 9th to 16th, then followed by the teams ranked 17th to 24th. The bottom 8 teams from the previous year (rankings 25 thru 32) were in the final group.
That isn’t how teams were exactly drafted, of course, but it’s a reasonable approximation. And using that as a template, we can look back and see how they performed. Using 2002 as a starting point (that’s when the league moved to 32 teams) it turned out this way:
| AVERAGE DEFENSIVE POINTS SINCE 2003 | ||||
|---|---|---|---|---|
| Year | 1-8 | 9-16 | 17-24 | 25-32 |
| 2003 | 7.2 | 8.5 | 6.1 | 6.5 |
| 2004 | 7.2 | 7.6 | 7.4 | 7.1 |
| 2005 | 7.4 | 7.1 | 6.9 | 6.6 |
| 2006 | 7.5 | 6.0 | 7.8 | 7.1 |
| 2007 | 8.3 | 7.5 | 6.4 | 7.3 |
| 2008 | 6.7 | 6.9 | 6.5 | 6.3 |
| 2009 | 7.7 | 6.6 | 7.0 | 6.2 |
| 2010 | 7.3 | 6.3 | 7.3 | 7.1 |
| 2011 | 7.4 | 7.5 | 5.9 | 6.6 |
| 2012 | 7.6 | 7.1 | 6.7 | 6.6 |
| 2013 | 7.2 | 7.4 | 7.0 | 7.3 |
| Avg | 6.8 | 6.5 | 6.3 | 6.2 |
Notice that in five of the 11 seasons (almost half) the teams that had been ranked 9th to 16th the previous year actually outperformed the top 8. Overall, there isn’t much spread between the two groups. The top 8 teams (88 teams in the 11 years) averaged 6.8 fantasy points per game, just .3 ahead of the 88 teams in the second group. There’s a half-point drop down the third group (6.3), and then the bottom group is at the rear (6.2).
In one of the 11 years (last season) the bottom 8 teams actually outperformed the top 8 teams. That was fueled by Kansas City going worst to first last year.
Let’s focus it further, though. Most fantasy leagues have only 12 teams, so it’s only about half of the defenses that are relevant. So rather than breaking them into groups of eight, let’s group them in fours – top 4, followed by 5th to 8th, followed by 9th to 12th. And 13th-16th is the last group, working under the assumption the final 16 teams won’t even be drafted.
Using that model, the scores work out this way.
| AVERAGE DEFENSIVE POINTS SINCE 2003 | ||||
|---|---|---|---|---|
| Year | 1-4 | 5-8 | 9-12 | 13-16 |
| 2003 | 7.2 | 7.3 | 9.9 | 7.1 |
| 2004 | 7.7 | 6.7 | 7.7 | 7.4 |
| 2005 | 6.5 | 8.2 | 6.8 | 7.4 |
| 2006 | 8.3 | 6.7 | 6.6 | 5.4 |
| 2007 | 8.5 | 8.1 | 7.8 | 7.3 |
| 2008 | 6.8 | 6.5 | 7.0 | 6.9 |
| 2009 | 8.5 | 7.0 | 6.5 | 6.8 |
| 2010 | 7.3 | 7.2 | 6.5 | 6.1 |
| 2011 | 6.6 | 8.1 | 7.9 | 7.1 |
| 2012 | 7.6 | 7.7 | 7.1 | 7.2 |
| 2013 | 7.0 | 7.3 | 7.9 | 6.9 |
| Avg | 6.8 | 6.7 | 6.8 | 6.3 |
Broken out that way, there seems to be little sense – no sense, really – in using resources to capture a defense because it happened to rank a little higher the previous year. The teams ranked 1st thru 4th were only slightly better than the next two groups of four. Both the 5th-8th group and the 9th-12th group finished with a higher average in five of the 11 years. You start to see a more meaningful drop outside the top 12.
And if you break it all the way down and look at exact draft positions, it starts to get really crazy. If, over the last 11 years, you had drafted the No. 1 defense every year (based on how they performed the previous year) you would have only the 9th-most points at that position.
The No. 2 spot ended up with the highest-scoring defenses over the last 11 years, so that makes sense, but No. 3 finished in 7th place, No. 4 finished in 22nd and No. 5 ranked 19th. Crazy.
The four highest-scoring positions (again, just going off the previous year) belonged to slot No. 2 (1st), slot No. 6 (2nd), slot No. 10 (3rd) and slot No. 23 (4th).
Two of the four worst slots belong to teams that ranked 30th and 31st the previous year, so that makes sense. But the worst defenses of all were those who ranked 17th the previous year. And the No. 13 spot came in 29th.
For me, it underscores that defense is a fluid position. It’s not enough to simply trot out last year’s numbers. When drafting a defense, it’s helpful to have seen some production the previous year, but you should also be giving plenty of weight to scheme and personnel. Are the key defensive guys back? And is it a system where they’re trying to get after the quarterback.
—Ian Allan