Questioning Question's Questions

Riddle me this: Why did you take a defense in the eighth round?

A wise man once said, “Are you mad cause I’m asking you 21 questions? Are you my soul mate? Cause if so girl, you a blessin’.” He went on to say something like, “$#%@#^%” and “&^^@#$.” The modern day poetry of Curtis James Jackson III was probably not the inspiration for the Questions section on numberFire. It was also not the inspiration for this column. But it was good background listening while I slogged through 120 mock drafts for the following missive.

One of my favorite parts of our little slice of the Internet is the quality of answers given to the questions on the aforementioned section of our site. I really enjoy participating and seeing the varying responses and reasoning behind them. So as a love letter to the numberFire Questions community, I decided to try and make the entire thing an obsolete wasteland by answering the most common queries in one neatly packaged bundle of affection.

Why do people draft defenses in the eighth round?

Seeing people draft their defense before the second to last round is more annoying than chewing on tinfoil while listening to Phish pump out a 24 minute jam. It’s also just as awful.

When I look at “Rate My Lineup” posts on numberFire and see the Seahawks defense, I know for a fact that they were drafted in the 11th round or before. I also know that expert mocks continually have defenses coming off the board as early as the eighth round. To get an idea of how pervasive this is, check out the following chart with data culled from 10 expert mocks I randomly pulled from a handful of mainstream sites (including numberFire).

Number of Defenses Drafted Before The 14th Round

RoundD/ST Drafted

There are more than six defenses per mock coming off the board by the 13th round. Players like T.Y. Hilton, Shane Vereen, Daryl Richardson, Mike Williams, Chris Givens, and Josh Gordon are all going in that area. We are talking about a mix of real life, high-upside NFL starters and super solid producers. But instead, people are picking a defense?

Curious why this is happening, I went through the last four seasons of D/ST ADP and stats and came up with some compelling data:

- In three of the last four seasons, the ADP of the number one defense was over 160. Allow me to repeat that: In standard leagues, the number one defense was undrafted three of the last four years.

- The only season since 2008 where the number one defense was actually drafted was 2012, when the Chicago Bears were seventh among D/ST’s in ADP.

- After the top three each year, defenses 4-10 were separated by an average total of about 1.5 points per game. This is to say, they were more or less the same.

- Only three times (out of 12) had a top-three defense been in the ADP top five.

- The average finish of a top three drafted defense was 10.2. For a top-five drafted D, it is 11.6.

- 45% (9 of 20) of defenses drafted in the top five finish out of the top 10.

- 21 of the 40 top-10 defenses were undrafted.

2008-2012 D/ST Average Finish By ADP

D/ST Rank According to ADPAverage Final Rank

While there is certainly an advantage to drafting one of the top two or three defenses, it is very difficult to predict who that will be. You are welcome to try, but the numbers show it is a fool’s errand.

Why does everybody put so much emphasis on the first round?

One of the most asked questions on numberFire goes something like this, “I have the X pick, who should I take? I really like LeSean McCoy, but Arian Foster will be there and I know I am supposed to take him. What would you do?”

My answer is always the same: Take the guy you like best.

But what about rankings and projections?

Hooey, I say!

The top 10 is very close. Debating C.J. Spiller against Doug Martin against Jamaal Charles is pointless. They all have pros and cons, and in the end, measure very equally against each other. Take the guy you want.

In the last five years, covering 50 top-10 picks, only eight ended up being major busts*, and seven of those eight missed significant time with injuries. If you remove the other 10 players who missed two or more games due to injury, we end up with 21 of 28 position players finishing in their positional top 10.

The moral of the story is that if your guy doesn’t get hurt, he will be fine. Save for a crystal ball, we won’t be predicting injuries, so all I can say is draft whoever you think will have the biggest year and don’t look back.

*defined as WR/RB finishing out of the top 20 and QB’s out of the top 10

Why do owners try and shoehorn themselves into a specific strategy?

Another commonly asked question: “I have pick X, should I go RB-RB-WR or WR-RB-RB or RB-RB-RB?”

My advice is always the same: As long as you get two running backs in your first four picks, it doesn’t really matter.

To help make my point, I spent a good amount of time (and glass or two of bourbon) in front of a mock draft simulator. I simulated the following six combinations of picks five times each from three different draft slots (2, 5, 8):


Sticking to the basic rule of drafting at least two running backs and two wide receivers, I then picked out of the same three slots ten times each using my normal draft method (RB in the first round, play it by ear thereafter).

120 mocks later, I had to be brought back from the brink of insanity. Once stable, I used numberFire’s projections to total up the points for each five player lineup. Please note that I did not mock any of the first-round wide receiver combinations for the second overall pick, as Calvin Johnson is almost never being drafted that early. The Results:

Mock Results

Structure2nd Pick5th Pick8th PickAverage
No Structure929912918920

As you can see, the only method that delivered consistently better results is when I didn't follow a predetermined pattern. With the rest of the projections there isn't much consistency or any real discernible pattern. When you look at a graph of the raw data, you will notice the same.

Here is the point where I admit that simulators are not the real thing. I will also cede that five drafts of each type at each pick slot isn’t exactly a massive sample size. All that said, I do feel very confident that even with an acceptable margin of error factored in, my results are statistically valid.

Other Takeaways From My Experiment

- In the mocks where I grabbed my second running back in the fourth round, I was stuck with the likes of Chris Ivory, Darren McFadden, or Lamar Miller. While a case can be made for each of them as an RB2 this year, I would feel a lot more comfortable playing them at flex. If it were a real draft, I’d have to adjust my strategy and make drafting steadier, lower risk back up running backs a significant priority. You simply can’t lack depth at a position when you are relying on the likes of Run DND.

- Between these simulated mocks and the live ones I’ve done this preseason, I think I am going to end up with a higher than normal proportion of teams with a wide receiver at flex. The early run on running backs leaves the pool very thin as soon as round three. I find myself continually taking, for example, Reggie Wayne over the Miller's of the world to slot in at the flex position. I would still advocate targeting a running back as your flex, but we are drafting more of them in the first two rounds than at any point since 2007, so it is getting thinner faster than at any time in recent history.

YearRB's Taken in the First Two Rounds

Some of you may be wondering why I didn't include tight ends and quarterbacks in my mocks. The issue with their inclusion is that I would have had to go at least 10 rounds per mock. Due to the nature of draft simulators, once you get past the early rounds, I don't think the results would have been statistically reliable enough to present to our readership.

Why do you need a backup quarterback?

Q. What do an extra thumb, an evil twin, NASCAR jackets and backup quarterbacks in 10-team leagues all have in common?

A. You shouldn't want any of them.

Nearly every “Rate My Lineup” post on our Questions section has two quarterbacks. Why are you backing up Tom Brady? Doesn’t your league have a waiver wire? Stop wasting valuable bench real estate on a replacement level player who will likely see your lineup one time the entire season.

Why are you so anxious to make trades?

Slow down! You just drafted! We have two weeks to go before the season starts. So just relax, take stock of your team, check out the other rosters, and let trades develop organically. Don’t try to force something to fill a perceived hole when we haven’t seen a single regular season snap.

Should I drop X For Y?

If you are asking this, we are probably talking about your last bench spot. Generally speaking, unless you are chock full of them already, that slot should be reserved for a high-upside sort of guy. Considering the only way they will see my lineup is if they perform above expectation, I would much rather have Rueben Randle than Malcom Floyd sitting on the end of my bench.

If train A leaves the station going 60 miles per hour and train B leaves one hour later going 85 miles per hour, how long will it take train B to catch up with train A?