In this series, RotoViz writers will be using the Draft Dashboard, the flagship tool of the FFDRAFTPREP Package, to practice approaching 2018 drafts. Twelve authors will be participating and each will be mock-drafting, from a different starting position, using the tool. The aim of this series is to outline the strategies employed by our writers, the information they are using to inform their decisions, and to obtain an understanding of how draft slot plays a role in the development of their approaches. At the end of each mock, the author will compare their team to those drafted by the computer and self-assess whether or not they “won” the draft.
I learned a great deal about fantasy football, drafting, logic, and problem solving during the literally thousands of hours I spent building this tool. Programming logic into it that could appropriately respond to a number of settings, varying league sizes, roster configurations, league drafting habits, and ADP sources, to build out realistic fantasy rosters was quite the task.
I’ve spent countless hours watching the AI driving the computer-drafted teams duke it out while testing for reasonableness, tweaking calculations, and searching for bugs. Every once in a while, the computer would put together a team that I absolutely loved. Sometimes, it would do so by approaching the draft in ways that my human mind would not.
Man vs. Machine
With the fear of the machines becoming sentient, taking over, and erasing human civilization becoming more real each day, it’s time that I take on my creation and show it its maker.
RotoViz subscribers, consider yourself lucky because this is a historic event. A Tyson-Ali title fight, a Brady-Rodgers Super Bowl, a Lebron-Jordan finals — they have nothing on this matchup. It’s like Bobby Fischer inventing a version of Deep Blue or Komodo and then facing it in a winner-take-all death match.1 This is like Chuck Norris challenging himself to a staring contest, John Madden bad mouthing Brett Favre, or relationship George walking through that door — worlds are going to collide!
Can I out draft the fruit of my blood, sweat, and tears, or will it scrape together a superior roster?
Fear not humans, I’m going to take my laptop out to the woodshed, smack down the caps lock, and bitch slap this thing into the third row like I’m Dikembe Mutombo and it’s Muggsy Bogues.
All writers will be drafting in 12-team PPR leagues and using “Average PPR” as the ADP source in their mocks. Rosters will consist of 1QB, 2RB, 3WR, 1TE, 1Flex (RB/WR/TE), 1K, and 1DST. The computer teams will be set to “Mixed”, meaning that they will build out their rosters using a mixture of approaches. Some will go running back or wide receiver heavy, others will emphasize positions equally, and one or two may go Zero-RB or Zero-WR.
For my draft, I’ll be picking 10th.
In PPR leagues, I always aim to play a WR in the flex. The reason for this is simple:I like my chances of landing four weekly contributors at WR more so than landing three weekly contributors at RB. Further, in general, a WR can fail to live up to expectations and still provide some level of production. This is much harder for an RB to achieve, especially when drafted with an early pick. Given that the league employs PPR scoring, a flex-level WR is likely to have a higher weekly floor than a flex-level RB.
In the mocks I participated in earlier in the summer, I found myself less pleased with teams in which I accumulated three strong RBs than those that were built around three strong WRs. Granted, I could strike a balance, but I like to facilitate in-season management by building a core strength into my roster — one that will not need to be addressed as the season progresses. As I’ll be drafting from pick 10, there will be stretches of 18 picks that pass between mine. I’ll need to keep this in mind, especially when making my third and fourth picks.
Notes: Round 1 – 6
- WRs went off of the board faster than I would have liked. At my first pick, only two Tier 1 WRs were available.
- Fortunately, I got lucky and Team 12 reached on Mike Evans. This allowed me to select Julio Jones.
- I got lucky again as Doug Baldwin, the final Tier 2 WR, remained available at my third pick.
- Through the first six rounds of the mock, I owned two Tier 1 WRs, one Tier 2 WR, one Tier 3 WR, and two Tier 3 RBs.
- The other teams in the mock had constructed their rosters in a variety of ways.
Most Challenging Pick
At my fourth pick, I was faced with the hardest decision of the draft. I had an incredible stable of WRs assembled in Odell Beckham Jr., Jones, and Baldwin. However, only four Tier 3 RBs remained, compared to eight WRs. Given that 18 picks would pass before I’d have the option to draft one the backs, it seemed unlikely that they’d still be available. In most drafts, I’d prioritize Tier 3 WRs over Tier 3 RBs. In this case, I liked my WRs so much that mixing in an RB and then pivoting back to WR seemed like a reasonable option. To add some context, here are the available player listings showing the RBs and WRs I was choosing between:
I decided to attempt to have my cake and eat too, assuming that there would be at least one, maybe two, of the Tier 3 WRs available at my fifth-round picks. I selected Derrius Guice.
In a frustrating turn of events, only one Tier 3 WR remained when the snake came back around, but two Tier 3 RBs were still available. So instead of being able to grab two Tier 3 WRs and one Tier 3 RB in Rounds 4-6, I’d have to get two Tier 3 RBs and 1 Tier 3 WR or one Tier 3 WR, one Tier 3 RB, and one Tier 4 WR. While this certainly wouldn’t play out this way every time, it does highlight the mistake that I made. Rather than taking the most appealing concrete option, I opened myself up to a situation in which there was a very real possibility I’d have to settle on a less desirable configuration.
Most Surprising Pick
With my seventh pick, I went tight end and selected Delanie Walker. Outside of leagues that implement TE premium scoring, I rarely draft TEs before the later rounds of drafts. However, when making this pick, Walker was the only Tier 2 TE remaining. I had no interest in the available RBs, and I believed that Walker has the potential to outscore the available Tier 4 WRs. I’d be back up in five picks and could grab one of these WRs then.
Overall, I was pleased with my team. It has a great group of WRs in Beckham, Jones, Baldwin, and Golden Tate. I would prefer to have one more Tier 3 WR in that mix, but I compensated for this by drafting just three RBs. This gave me more darts to throw on WRs that have the potential to finish as Tier 3 options. I also selected Jordan Reed as a way to capture upside. If he hits, he could essentially be a Tier 3 WR. Guice, Lamar Miller, and C.J. Anderson will be able to get me through the first couple weeks of the season. I can address the position at this time when I have better information. The other backs that I could have added would have little utility until deeper in the season, and there’s a good chance I’d end up dropping them anyway. I selected the Lions as my DST. The team will be facing the Jets to open the season which looks like one of the more favorable matchups of the Week 1 slate.
Favorite Computer Team
I really like what the computer did with Team 12. It went WR heavy, opening the draft by selecting four WRs. Jamison Crowder, Corey Coleman, and Mohamed Sanu provide depth to a WR corps that has significant potential. While the roster doesn’t have a clear-cut second option in the running game, the mixture of Marshawn Lynch and Doug Martin should be able to bridge the team through the first couple of weeks. If Nick Chubb can win Cleveland’s RB1 role, Team 12’s starting roster would be more than formidable. Nyheim Hines adds an element of upside to the team and Michael Thomas, Mike Evans, Adam Thielen, and Josh Gordon could all conceivably finish as WR1s.
Despite being impressed by the squad that Team 12 assembled, I can genuinely say that I like my team better and believe it was the strongest one assembled in the mock. I prefer my QB and our TEs possess similar ranges of outcomes. While I consider Gordon to be a Tier 3 WR, and Thomas a Tier 1, his group of receivers has trouble stacking up to mine. It’s hard to assemble a group of starters that average a Tier of 1.75 and even when optimizing the Tiers assigned to Team 12s WRs, they average out to a Tier 2.25 WR.
Overall, I believe that my team has a better mix of starters, a deeper bench, and at least equal, if not more upside.
The humans win.
Draft Board and Team Rosters
- I don’t really know anything about Chess, so forgive me if I blew this analogy. (back)