clock menu more-arrow no yes

Filed under:

You tend not to matter: An excavation of the crap-heap that was our 2015 fantasy football draft

New, comments

This year's fantasy football season has seemed like an especially unknowable hallucination. A lot of us drafted our way straight into the cellar. Let's see if we can fix some of your awful fantasy teams by drafting them completely at random.

Your life is a protracted brokering session between you and the world, which is sitting on the business side of the desk, hands folded, sometimes pointing out its thumbs and leaning forward a bit in emphasis, as it neatly translates the legalese of its handbook into the words you know: you can't do that. You aren't authorized to operate that machinery. It's not a matter of permission, but of capability. The mobile hung over your crib was not ornamental, but instructive: it wobbles and beeps at you, entirely out of your reach, and none of it was your idea. In the ensuing years, the world will pull you into its office, again and again, and explain to you in as many different ways as it knows how that your life is less an arena in which you have agency, and more a very long parade of things happening to you.

An obscure file, lost in the cabinets, details your failure of a fantasy football season. It was two degrees of separation removed from what is real. In the first place, your team played a sport, which is performed within the artifices of rules and painted boundaries, and is thereby non-real in at least a few senses of the term. Further, your team is imaginary. Not only were you the leader of a mirage of people, those people within that mirage were doing an artificial thing. A dream within a dream. You still lost.

The same number of fantasy players experience this year in and year out, of course, but this season felt especially arbitrary. These, according to average draft position (ADP), were our first 10 picks.


The data represented here is based on projections I made. I'll drop the process behind them at the end of this post.

We knew better than to believe that, say, the top overall pick was guaranteed to be the best fantasy football player. Anyone who's been around a while knows to pursue approximations rather than chasing certainties. This year, we aimed low, and we missed almost entirely.

Adrian Peterson, while not the very best running back so far this season, is a perfectly fine top overall pick. Antonio Brown, who is only a couple points shy of this season's best wideout, is a good get at No. 5. At No. 8, Matt Forte is a little disappointing, but not disastrous. The rest are disastrous, whether by injury (Bell, Charles, Bryant), disappointing performances (Lynch, Anderson), or some combination of the two (Lacy, Luck). This might have been the worst first round of fantasy football drafting we've ever seen. A lot of that was beyond the control of the folks we drafted, much less ours. That is the point.

We fell so hard that I'm not sure whether to laugh at how wrong we were, or marvel that we didn't fall harder.


That green trendline represents how these players would line up if sorted from best fantasy performance to worst. Each trendline comes with its own mathematical designator -- the coefficient of determination, or "R-squared." The R-squared number signifies how well a trendline fits the data, and it ranges between 0 and 1. If it's 1, that means that the line perfectly fits the data; if it's 0, the trendline doesn't fit the data at all.

I bring this up because the green line's R-squared was north of 0.9 -- a very good fit. The gray line was what happened when I asked Google Sheets to draw up a trendline to represent the value of the players as we all drafted them. That trendline was around 0.02. There was only the most vague shadow of a trend, if even that. The computer did the best it could do, which was not very good. It took a look at our dots and said, "hell, man, I don't know. I don't know what y'all did."

There has been uncertainty at every major position. Andrew Luck, who lots of us heralded as the de facto best fantasy quarterback last summer, has fallen hard. Peyton Manning, who was on average the third quarterback selected, has been historically bad. Dez Bryant's injury has been devastating to teams who spent a first-round pick on him, and Jimmy Graham -- one of the surest bets in fantasy football -- was dealt to a team that can't figure out how to use him.

But the most crucial fantasy position, running back, has produced more chaos than any other.


Same story with that gray trendline: it's not a good statistical fit. More compelling to me is that the traces of a trend the computer did see is almost horizontal. It's almost flatlining. It looks almost as flat as it would look if every running back in football were exactly the same. When we drafted these people, we did not understand a damn thing.

Stop on that point for a second: how many of us really knew what we were doing? I mean, I certainly didn't. My team is trash. Would those of us in the cellar had drafted better if we had just blindly thrown darts at a wall? The tried-and-true "I don't give a shit" maneuver is to simply auto-draft a team by ADP, but clearly, there's no reason to believe it would have served us well.

What if we gave even less of a shit than that, and just drafted our teams entirely at random? Well, we're gonna find out.

Earlier this week, I asked y'all to fill out a form telling me a little about your bad fantasy team. I asked you to tell me the quarterback, running backs, wideouts, and tight end you drafted, and I also asked you to write a short poem about your crappy team.

But first, I needed to make sure to filter out all the mediocre-to-great teams, which I had no use for, because I'm only trying to fix the bad teams. As any statistician or journalist knows, the best means of verifying data is to stop and ask, "oh but for real?"


Thank you all for your honesty. This allowed me to filter out a full third of the teams in question, leaving me with a sample of about 400.

I decided to select five teams at random. I would look at the players they drafted, see when they were drafted, and randomly draft a different player who was lower on the ADP list -- in other words, a player who was likely to still be available. For example, if the quarterback you drafted was Drew Brees, who was the fifth quarterback was taken by ADP, you had zero chance of randomly drafting Andrew Luck or Aaron Rodgers or anyone else above him.

So clearly, the deck is stacked against us here. And if we can make at least three of these five teams better by throwing darts at a wall like this, I will take this as a complete repudiation of ADP and our agency within the gelatinous, fundamentally un-sturdy realm of fantasy football. I will declare all numbers imaginary, live the rest of my days in the forest, and make the remainder of my marks upon this world with nothing but my bare feet, my teeth, and a jagged rock.

First up: let's hear from "Anonymous." Most of us declined to give a name.


This is such a brutally bad team. You can retrace this individual's steps: Bryant (injured) was selected first, followed by Demaryius Thomas (bad, on account of Peyton Manning being his quarterback). They doubled their misery with the selection of Manning, who has played very badly while hurt. That is far worse a fate than that of a quarterback who is benched with injury; at least that way you know to replace him. If you drafted Peyton Manning, you were faced with the decision of whether to bench the most prolific point-scorer in the history of fantasy football. Only now, halfway into the season, has that decision been made for you.

The randomly re-drafted team is atypical of how we're going to do, believe me. We hit the jackpot. Cam Newton has been great, and Devonta Freeman -- on average, the 40th running back selected, and the 119th player selected -- has been absolutely incredible. Score one for the arbitrary nature of existence.



We've crashed back to Earth, thanks in large part to tybg's selection of Colin Kaepernick, a quarterback with such a late ADP (he was the 18th quarterback selected) that he fell beyond the reach of the data set I'd prepared. What stuns me about this team is that tybg was able to draft both Jamaal Charles (3.3 ADP) and C.J. Anderson (11.3) ADP. There could not possibly have been very many teams in this league. And yet, tybg reached all the way down to 18 to draft Kaepernick. That is bad drafting, but it's also on-purpose as shit, and on-purpose is what we're after.

Since Kaepernick fell outside my data set, I wasn't sure what to do. Ultimately, I decided to let the random number generator select literally any quarterback with a later ADP, from Jay Cutler to Dan Orlovsky.

It went all the way to the end of the list, No. 66. It selected Dan Orlovsky. It would seem that the fates also have a purpose in mind. That selection single-handedly torpedoed our random team's chances.



We're back! The segment of a team we randomly drafted here is not great, but it's good enough to beat Anonymous' team. This is a team so bad that were it any worse, I wouldn't believe it was drafted in good faith. It's just full of piss-poor luck: Romo's been hurt, Lacy's been bad and hurt, Anderson's been bad, Bryant's been hurt, and Sanders' production has been capped on account of his quarterback situation.



Just a little bit short. This was heartbreaking. Tight end Jordan Reed had such a late ADP -- he was the 26th tight end drafted -- that he fell outside of my data set. I just did not account for y'all drafting quite like this. But in Anonymous' case, it paid off, because Reed has been decent.

Nonetheless, I faced another dilemma, and once again, I decided to err on the side opposite my bias. I let the number generator pick any tight end drafted after Reed. It selected Virgil Green, who is on pace to score 32 fantasy points this season.

Minus a couple unlucky breaks, we could very well be sitting at 4-0 right now. Instead, it's 2-2. The fifth randomly-selected entry in the spreadsheet will decide everything.



uh hey



An explanation of my projections:

These projections represent an estimation of a player's total points scored in 2015. Bye weeks were accounted for. Specific match-ups were not. I simply calculated a player's average number of points per game, multiplied that by the number of games remaining for that player's team, and added it to the player's present season totals (through Week 10). However, I did exclude all players on injured reserve from this process. For example, Jamaal Charles' standard-scoring fantasy value is not a projection, but his real-life season total. All projections are based on non-PPR standard scoring.

* * *

Pretty Good video: Lonnie Smith fought the 1980s and beat its ass down