The 2023 NFL Draft is still weeks away, and one major Underdog best ball fantasy football tournament has already filled — the Big Board. To follow up the Big Board, Underdog has, naturally, launched a new Big Board tournament… but with a major twist. This Big Board is a SUPERFLEX tournament. In essence, a Superflex format encourages fantasy drafters and fantasy managers to start two quarterbacks, with the second eligible to start in a flex spot alongside the traditional RB/WR/TE flex positions. As you might imagine, this boosts the values of QBs immensely.
There has been extensive data analysis work done on the pick-by-pick data for each of Underdog’s Best Ball Mania tournaments from the past three years, but there is no pick-by-pick data publicly available for any Superflex tournaments they have run. So I figured… why not simulate my own?
NOTE: If you’re looking to enter a Big Board tournament or plan on drafting later in the year, we’re constantly updating our Best Ball rankings to give you the biggest edge with our friends over at Underdog Fantasy (Promo code: FANTASYPTS).
New signups to Underdog get both a deposit match of up to $100 and a Fantasy Points Standard subscription for just $5!
You’ll also want to subscribe to the Fantasy Points YouTube Channel, where we have been doing Underdog best ball drafts every Wednesday night at 7:30 Eastern, including Big Board Superflex drafts.
Underdog Fantasy Superflex Simulation
Using the ADP from last year’s Big Board Superflex, I was able to simulate 5,000 drafts to give me simulated pick-by-pick data for 60,000 teams. From that point, I just went week-by-week to figure out the weekly scores for each team based on their roster of players to find out their regular-season total score.
It is important to point out that the simulated data is likely not a perfect representation of human drafting — the simulation did not try to intentionally stack players from the same team, for example. The only restriction I put on the simulation was limiting rosters to a maximum of four QBs, nine RBs, 10 WRs, and three TEs. Other than that, I let the simulation run to get a wide variety of possible team combinations based on ADP.
Underdog Fantasy Superflex Roster Structure
There is a major difference in the roster requirements for this tournament compared to non-Superflex tournaments in best ball. Instead of starting three WRs each week, one of those WR spots is now a Superflex spot. This means that instead of having to play three WRs, you can play an additional QB, but also another RB or TE as well. Ideally, you will be starting two QBs each week, as they are the highest-scoring players in fantasy football.
Below is a table showing how teams scored based on the number of QBs on the roster:
Unsurprisingly, more QBs led to higher-scoring teams. Three QBs stand out as the most optimal roster construction, with the best average regular season score, the highest floor, the second highest ceiling, and the highest average regular season score among teams with a score in the top 10% percentile. Going with two QBs can work, but you really need to nail those two. There is nothing wrong with going with four QBs either (Geno Smith was basically free last year and was a great fourth QB pick.)
It is a bit surprising to see more RBs lead to higher-scoring teams. My main takeaway was that people were drafting so many WRs before RBs that it led to great RB values in a format where you have to start the same number of WRs as RBs (non-Superflex Underdog draft rooms tend to skew heavily to WRs, so there seemed to be some carryover despite the roster format changing). We do see around seven RBs to be the sweet spot for average scores and ceiling.
Since we saw more RBs lead to higher scores, it makes sense that we see more WRs lead to lower scores. One fewer WR spot really does make it so that you don’t have to load up on WRs if you can nail elite WRs earlier. The elite WRs like Justin Jefferson and Tyreek Hill were much more consistent last year than a lot of RBs last year, and you can find high-level RBs late that earn unexpected playing time due to injuries to starters. That dynamic doesn’t exactly exist for late-round WRs.
The ideal number of tight ends is kind of all over the place. There isn’t really a clear takeaway, with many of the results being similar. I would say you want to structure your teams based on where you draft TEs. If you draft Travis Kelce, you likely don’t need three TEs, and may be able to get away with only drafting one. You can also pass on an elite TE and grab three late and hope you nail a high-ceiling outcome like Evan Engram was last year.
Below is a table showing the top 15 most common roster constructions the simulation produced.
We can clearly see that having a similar number of RBs and WRs should be your goal when drafting Superflex teams on Underdog. Having only four or five RBs did not lead to better teams in any metric. This is possibly the biggest edge the simulation produced! Draft rooms have leaned extremely WR heavy, both this year and last year, with drafters loading up at the position. In the Superflex format, that doesn’t hold up as an optimal strategy due to the starting roster requirements.
It’s also important to note that drafting four QBs performs really well. Remember QBs are the highest-scoring position in Underdog’s half-PPR scoring system. Making sure you are maximizing two QB spots each week is very important.
Early-Round Underdog Superflex Draft Strategy
Now that we know generally how we want to structure our teams, let's take a look at how we should attack the draft.
The table above shows the average scores of how teams attacked the first two rounds and the first three rounds based on the position they selected. First off, you definitely want to draft at least one QB in the first two rounds, and possibly two in the first three rounds. If you wait to grab your first or second QB until the fourth round this year, that would put you with a QB like Russell Wilson, Derek Carr, or Kyler Murray as your first or second QB.
Loading up on early-round WRs seems like a bad strategy. It is important to note that last season, Round 1 through 3 WRs included Cooper Kupp and Ja’Marr Chase, who both missed significant time, so that could be skewing these numbers a bit. That said, I do think you want one of your first three picks to be a WR, given that having an early-round WR on your team led to the second-highest top 10% team final score averages.
Drafting early-round RBs seems to raise your floor, but does not necessarily raise your top outcomes. I don’t think you have to prioritize getting an RB early, but I’d lean toward wanting one if the draft falls my way. You can make up for not getting one by loading up later and having more RBs than most teams.
Travis Kelce and Mark Andrews made up most of the early-round TE picks based on ADP. Kelce, unsurprisingly, rose the floors and ceilings of teams he was drafted on, but you definitely don’t want to take more than one early-round TE, and you don’t have to take one at all.
Mid-Round Underdog Superflex Draft Strategy
Your QB strategy in Rounds 4 through 7 is not too important, the above results would seem to argue. Drafting three QBs in this range does show a high top 10% score average, but that was somewhat of a small sample, with only 837 teams utilizing that strategy in the simulations. In general, I likely already have at least one QB on my team by Round 4, so I’m not looking to load up here and likely will try to draft no more than one QB in this range based on how the draft is going.
We really see this as “The Sweet Spot” for drafting running backs (the Superflex format pushes “The Dead Zone” further back). This was the range Josh Jacobs, Travis Etienne, and Saquon Barkley were drafted in last year. Wide receivers were not as strong in this range last year, with guys like Courtland Sutton, Jerry Jeudy, Michael Pittman, Elijah Moore, and Michael Thomas coming off the board. I don’t know if we see such a difference in WR and RB performance from this range in 2023, but it does seem like a good idea to target RB or WR positions in this range over QBs and TEs — the key is not to draft way more WRs than RBs on your team as a whole, as has been established.
Avoiding tight ends in this range makes sense based on how the second tier of elite TEs like George Kittle, Kyle Pitts, and Darren Waller performed last year. Current ADP for this year would include Kittle, Pitts, T.J. Hockenson, and Andrews in this range. I am generally fading this group unless they fall past ADP.
Late-Round Underdog Superflex Draft Strategy
Round 8 and beyond will provide a lot of flexibility for how you fill out your team. How you approach the first seven rounds will really drive how you should attack the late rounds. If you took two early-round QBs, you probably don’t need more than one late-round QB. On the flip side, if you took only one QB in the first seven rounds, you might want to take three late-round QBs.
Geno Smith was one of the last QBs drafted last year, and is likely skewing these results to some degree. I don’t know if we will ever see a high-level QB go at the back of drafts again. We also saw Jacoby Brissett go late and provide good QB weeks throughout the season. Don’t be afraid to take shots on guys like Brissett, Mac Jones, or even guys like Gardner Minshew and Baker Mayfield, who both could start a significant number of games for their teams this year and have good supporting casts.
Again we see drafting a larger than usual number of RBs as beneficial. However, your ceiling does seem to become capped if seven or more of your RBs come from the late rounds. Doing an “Anchor RB” strategy — taking one higher-level RB and then six late round RBs — seems very viable here.
Having more elite WRs and fewer late-round fliers was very beneficial based on the simulations. Only taking one or two WRs earlier and then taking seven or eight late-round WRs performed much worse than structuring your team to have around seven or eight WRs, but having five or fewer of them from the late rounds, looks like the optimal strategy. The importance of having one fewer WR position to fill can’t be understated. You don’t need nine or more WRs to give you strong WRs scores each week.
Tight end strategy seems very flexible once again. If you miss one of the elite TEs of your choosing, going with three late-round tight ends was very viable, posting the highest average score of this sample. Pairing an elite TE with one or two late-round tight ends also looks good. The average scores are all relatively close, though, so just structure your tight end position based on how your draft has gone so far and take value when it falls to you.
You definitely want to have 1 QB on your team by the end of the 3rd round, and sometimes 2 QBs before the end of round 3.
Plan to finish every draft with 3-5 QBs depending on the draft capital you allocate to the position. Late-round options like Mayfield and Minshew can give you much-needed spike weeks in your Superflex spot over skill position players.
Don’t draft so many WRs. It is okay to have a similar number of RBs and WRs on your team, if not sometimes more RBs than WRs.
Late-round RBs are much more valuable to a team than late-round WRs, especially this time of year when so many backfields will shift before the start of the season for various reasons.
High-end tight end production gets a slight boost due to needing a tight end to fill only one of seven starting skill player spots vs. one of eight in non-Superflex drafts.
It is important to remember while there were over 50,000 teams analyzed in the simulation results, this still is only one season of data. Every season can be different, and an injury to any player can greatly skew results. However, with no data analysis for this Underdog Fantasy Superflex format out there, this analysis can still help us build better-constructed teams than our opponents and gain an edge.
The next steps will be to analyze how these simulated teams would have performed in the single-week rounds of the tournament. (Maybe having more WRs was beneficial in the playoff rounds, but not the regular season. Maybe having more QBs becomes even more important when you want ceiling outcomes in two spots each single-week round, etc.)