Islands in the Stream: Deconstructing Fantasy Team Defense

Chris Morgan
July 29, 2014
@FantasySophist

It happened last year and this year, it’s happening again.

Look at any set of ADPs and you’ll likely find the Seattle Seahawks defense drafted smack in the middle of the draft. The fact that you could have (and probably did) pluck the Kansas City defense off the waiver wire in 2013, may be the single strongest argument in favor of streaming team defenses in standard leagues. But, despite evidence to the contrary, some people still value a top fantasy defense over potentially game changing players in the middle rounds.

Don’t think a mid round pick is valuable? Here’s a name with a similar 2013 ADP to the Seattle defense: Josh Gordon.

That’s quite an opportunity cost.

Of course, that’s an extreme example. But, that’s the kind of talent you should be looking for in the middle rounds, i.e. a player with the potential to significantly outperform their ADP. Even Golden Tate or Fred Jackson, both whom also had similar ADPs to Seattle in 2013, would have paid off more handsomely than a team defense.

Defenders of drafting highly rated units like the Seattle Seahawks commonly justify the pick with an argument like this: “Well, I know the Seahawks are good, therefore they must be the safest defensive pick”. That argument commits two fantasy football sins: First, it equates a good NFL defense with being a good fantasy defense, which is not necessarily the case. In fact, in many ways it’s like comparing apples to hand grenades. Second, that argument significantly overvalues the impact of so called ‘top team defenses’.

First, let’s look at how standard team defenses are awarded fantasy points. Depending on your league’s scoring rules, these settings will probably look familiar. I’ll call this the standard model of defensive scoring:

Standard Model of Defensive Scoring

  • 1 point for sacks
  • 2 points for interceptions
  • 2 points for fumbles recovered
  • 2 points for safeties
  • 6 points for defensive touchdowns
  • 6 points for special teams touchdowns
  • A range of -4 – 10 points for points allowed

There is some variation, for example, some leagues penalize for yardage surrendered, award points for special team yardage, or modify the point values for these categories. But, this is generally the most common model of team defensive scoring.

Sacks, interceptions, fumble recoveries, and safeties are all weighted about evenly, though these categories are clearly not equitable. Consider safeties: According to pro-football-reference.com, in 2013 there were only 20 safeties recorded and only two teams recorded more than 1. In 2012, there were 13 safeties. (In 2011, there were 21. In 2010, there were 13. In 2009, there were 14. I think you get the picture.) By contrast, for the whole league in 2013, about 500 interceptions and 300 fumble recoveries were recorded. That may seem like a lot more (and it is) but that’s still only an average rate of about one of each per game league wide.

Defensive and special teams touchdowns are weighted the heaviest, but these are also relatively abnormal. For example, in 2013  the barnstorming Kansas City Chiefs and their #1 ranked fantasy defense  returned four kicks (two punts and two kick-off returns), five interceptions, and two fumbles for touchdowns. Conversely, the #2 ranked Seattle Seahawks defense returned a grand total of zero kicks, two interceptions, and one fumble for touchdowns. The upshot here: these big plays aren’t normal occurrences. They are, by definition, extraordinary events for which fantasy points are the reward. A good NFL defense simply does more than the standard model scores.

So, what does it mean to be a ‘top fantasy defense’? Favorable match-ups and strength of schedule. Look at the dark horse Kansas City Chiefs of 2013. According to Jon Oliver of SBNation the Kansas City Chiefs’ preseason Strength of Schedule ranked 26nd overall, but actually proved to be 32nd  overall after the final standings. Their first nine games were played against teams who finished the season with a combined record of 52-80, and when the Chiefs faced bottom tier competition, their fantasy numbers soared. Generally speaking, it’s easy to beat up on the little guys. During their first nine games, the Chiefs blew the doors off of everyone to the outrageous tune of 16 points per game and were a formidable island in the stream. But, they took a nose dive on the back six and averaged 9 points per game. What happened? Peyton Manning happened twice, and in those games the Chiefs only scored 8 total points. Then they fell flat against the Chargers and the Colts to the tune of -2 total points. Ouch.

However, the Chiefs still managed very a  respectable average of about 13 points per game and they were largely undrafted in the preseason. But, even the mighty Chiefs were a tale of two defenses and were top performers (i.e. Top 6 in the weekly standings) no better than 50% of the time. The Seahawks were Top 6 performers only about 43% of the time. Yikes!

But, there are at least three problems with that assessment of ‘top performance’ . First, and most obvious, there’s no way to know how difficult a team’s strength of schedule is until about halfway through the season. But, that breaks in favor of streaming. If the Seahawks finish outside the Top 10 in fantasy defense this year because they have the hardest strength of schedule (they did win the Super Bowl, after all), they could still be a good NFL defense while being an average fantasy defense.

The second problem is that my assumption of ‘top perfomance’ could be misleading. There may have been a few weeks, possibly many weeks, when the Seahawks defense scored more than your opponent’s defense or at least did not ‘damage’ your chances of winning that particular game. In other words, the Seahawks my not have needed to be a Top 6 defense to have outscored an opponent’s defense. But, that volatility works against drafting team defenses early. One of the key differences between the team defensive position and every other position in fantasy football is the small pool of ‘players’. There are only 32 defenses and there can only be 32 defenses, or however many teams that happen to be in the league at a time. In other words, Jamaal Charles may not finish every week at the top ranked rusher, but he’ll likely finish ranked higher than 80-90% of his peers. That isn’t necessarily the case with team defenses where the standard model produces positional ties at the highest rate of all positions, with the possible exception of Kickers (but no one is trying to draft Matt Prater in the 8th round). Therefore, titles like ‘Top 6’ and ‘Top 12’ are of little help in assessing how effective a top defense actually is.

The third problem is related to the nature of the standard model of defensive scoring: because of favorable match-ups due to relative strength of schedule, mediocre real world defenses can often post stud fantasy numbers. The best example of this from 2013 is nearly any defense playing against the New York Giants. Let’s assume you streamed every defense you could have against the Giants (not likely but you could have gotten most of them off the waiver wire). The Giants surrendered 194 points to opposing defenses in 2013, which would have been good enough to rival the Chiefs in terms of overall points. In fact, if you streamed against the Giants last year (12.93 points per game), you were about as consistent as the Chiefs (13.27 points per game). And the list of teams the Giants faced wasn’t exactly a roll call of powerhouse defensive talent in 2013: Dallas twice, Philadelphia, Oakland, and  Minnesota were among the defenses out of which you could have gotten respectable performances. And, on average you would have performed better than the #2 ranked Seahawks.

Using San Francisco as a baseline of ‘worst’ top fantasy defense, the below chart compares three top fantasy defenses to teams which could have been streamed against the Giants in 2013:

average team defensive perfomances 2013

(Chart excludes bye weeks and week 17)

There is little evidence to suggest that drafting a defense early pays off. There are much cheaper ways to come by a defensive positional advantage. Only three of the Top 6 team defenses drafted before the season finished in the Top 6 last year and two were either undrafted or underdrafted. And yes, while the Seahawks did finish #2 overall, that’s still underperforming expectations.  The truth is, there are no safe picks at defense for the value of a high pick, unless you define safe as possibly 50-50.

 

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: