Running QBs - A Safer Bet in Fantasy Football?

I've always felt like QBs that can run the ball have a higher floor than QBs who remind me of... well, me in any kind of athletic competition.

Author's note for some context: I tried out for the golf team (certainly a top 30 or 40 sport in terms of athletic demand) in high school and shot a clutch 63 on the final day of tryouts……on 9 holes….so I am prettayy good.

Now that I've established my athletic golf-chops, let's talk about a totally different sport.

Hypothesis: QBs that are generally regarded as dual-threat QBs will have a higher floor than the their slower colleagues.

I pulled the top 10 fantasy QBs for the last 10 years to see if this intuition above was correct. Here we can see the top 10 QBs from each of the last 10 seasons. 100 dots total if you didn't want to count them.

Dad - here is a helper on how to read the above:

• Each dot represents a QB performance for a season.

• The further right the dots, the more points a QB scored that season.

• The higher up the dots, the bigger % of a QBs points came from running the ball (yards & TDs).

No major trend I am seeing above. So we’ll continue.

First matter at hand was figuring out how to divide QBs performance each year into groups - 1 for those more reliant on the ground & 1 for those more reliant on the air. Luckily for me I took a CodeCademy Data Science course about 7 years ago and remember about as much of that as I do senior year of college (for the record I had a total of 7 silver bullets in those 5 years - I wasn’t “cool”).

What I remember from that nerd course was basically go down the list of top rushing seasons until you run into Blake Bortles. When you hit Bah-lah-kayyy, you know you have gone too far. Below is the same chart as above, but color coded with the statistically legit "Bortles Line". Above the Bortles line are the "capable with legs" group & below the Bortles line are the "distributors" cohort.

Need to do a quick check to make sure the QBs in each bucket pass the sniff test. I can't have my first published scientific model say that Joe Flacco is an elite runner.

Here are the QBs who probably excelled at the shuttle run in 7th grade (capable with legs).

And the QBs that have at one time thought to put a chair halfway up the stairs (distributors):

There are a few surprises here and frankly some names I didn't think I'd see, but enough about Andy Dalton. It appears the Bortles-line (Bortles Barrier? Bortles Border?) did it's job - let's move on!

And a quick look at the size of each bucket by season:

Now, we will pull the individual game stats for these players & the seasons where they made the top-10 list. Once we have that, we can plot every single game broken out between our 2 buckets and see if there is a meaningful difference in performance between the groups. I'll filter out games where these QBs didn't play more than 5 snaps to focus on the games they actually suited up.

Alright so it does seem that running QBs have a slightly better floor (+0.6 PPG), a better median game (+0.6 PPG) and better breakout games (+1.4 PPG). I was expecting more of a gap, especially in the floor but it’s not huge. The higher ceilings for the runners will certainly help me overcome some of the other boneheaded decisions I make on game day.

Is any of this statistically significant? I don't know - ask a statistician. While this analysis wasn't the sweeping validation that my intuition was wildly correct, I'll safely ignore the takeaways and continue to institute my "no distributors" draft strategy, even if that means taking Justin Fields 1st overall.

Since we have the data, we might as well take a look the top 10 games from all those dots above:

Peyton absolutely slinging at the ripe age of 37. Dude out here giving "banana with a lot of brown spots on it" energy and dropping SEVEN TDs! And if you guessed The Sheriff (Peyton) as the only QB to make the top 10 list with negative rushing points, you are correct. He wasn't my guess as I have always thought of him as an elite rusher: