If you’re anything like me, you hate wasting an early pick on a onesie position*. What drives many a fantasy football novice to jump early on that shiny QB or TE is largely the perceived advantage of having a “set it and forget it” player that will start for you week in and week out at that position. For those brave enough to eschew the temptation of convenience, the world of streaming awaits. If you aren’t familiar with the concept of streaming, it is predicated on the belief that you can compensate for not having a star player at a certain position by swapping out mediocre players each week based on the opposition matchup. I’d be remiss if I didn’t mention at this point that my old editor at numberFire, JJ Zachariason (@LateRoundQB), and coiffure artist extraordinaire, CD Carter (@CDCarter13) host a great podcast examining streaming opportunities called Living the Stream that I highly recommend for more in depth discussions on the topic.
So, the question for those willing to give streaming a try becomes what exactly constitutes an ideal matchup? It’s easy enough to look up how defenses are performing against a particular position, but that really isn’t as helpful as it seems. For example, if you discover that the Giants are giving up an average of 19 fantasy points to opposing QBs per game, that says nothing for how many points you can expect to get out of the next QB on their schedule. That average of 19 points could have come from facing Aaron Rodgers/Drew Brees caliber QBs, in which case the defense is quite stout, or it could have come from facing Captain Butt Fumble & Friends, in which case 19 points is a huge number and a sign of defensive weakness.
In trying to untangle this mess, I decided to partner with my good friend excel and try to make sense of the data. I started with a primary assumption: every quarterback has a base expected score that will go up or down depending on the opponent. To get this base score, I took the average points scored by a QB, excluding games that they did not play/finish (simple enough, right?) To get the defensive adjustment, which I referred to as the “handicap”, I first determined how many points a defense gave up to a QB each week, and then found the difference between that number and the average number of points scored by that QB. These differences were then averaged to get the average number of points you could expect above or below the base QB points when playing a particular defense. All points are in reference to ESPN standard scoring.
Without further ado, here’s how the numbers look:
The first thing that stood out to me was the fact that I expected the matchups to have much more of an effect on the QB performance. Instead, these numbers suggest that an optimal matchup will give a QB an extra 6 points on top of their base score, while the worst possible matchup will drop them down 6 points below their base score. The second thing that I found interesting was how evenly spread and balanced the handicap numbers were. Take from that what you will.
Some take home messages for me as I looked at these numbers were:
In the near future, I'll take a look at the numbers for TEs to see if there is a similar trend. Ultimately, I think I've decided that I don't feel super comfortable taking QBs that I don't like in the hopes of swapping them out each week based on perceived matchup advantages. Of course, averages are not destiny and there is always the chance that Eli Manning goes off for 40 points at Dallas this year, so remember to take these numbers with a grain of salt as you decide on your own drafting strategy this year.
*A onesie position, a term coined by CD Carter, is a position for which there is only one in your starting lineup each week, which means that supply is often greater than demand in any given league.