Updated: Mar 31
Whilst best ball exploded in popularity in 2021 with Underdog’s presence growing larger and Sleeper adding best ball functionality, the mainstream providers were still predominantly rolling out single quarterback leagues. FFPC have run Superflex leagues for several years but they are far less popular than the rest of their products. My main ADP provider is Establish The Run, who collate ADP daily from all of the big sites. Last April when I enquired if they could add Superflex data they found there had only been two Superflex drafts in the previous 14 days. Meanwhile, on Sleeper many of us were getting busy. Sleeper’s introduction of best ball scoring allowed many people who hadn’t tried the format previously to jump in, whether that was a paid entry or completely free. Massive contests with top-heavy payouts can be daunting to the uninitiated and season-long leagues provide an easier gateway to the best ball world. ** UPDATE ** A matter of hours after posting this Underdog launched their first Superflex contest. Think of the takeaways in this as possibly applicable. The league settings I used will be different to Underdog's half PPR settings. With Underdog having just one flex spot in addition to the Superflex, and only two WR spots, as opposed to their normal three, I would expect the importance of QB's to dictate success rate even more. Whilst I have made conclusions about builds and draft capitals below, it's worth keeping in mind these might not translate directly.
As many of you who know me will know, I hosted a lot of best ball leagues on Sleeper last year, and will do so again this year if Draftkings best ball continues to disappoint and Underdog remains inaccessible to the U.K. Half of the leagues I ran were Superflex and eliminating the free ones that were aimed at the first time best ball drafters, I finished the year with a sample size of 17 12-team leagues to mull over the data contained within. It’s worth pointing out that 17 leagues are by no means a large sample size and an NFL season as a whole is not a large sample either. It’s also worth mentioning upfront that when we look at best ball data we are often trying to translate that looking forward and it can often go wrong. The hot topic of this early best ball season seems to be structural drafting and when picking certain players makes the most sense. I would strongly urge you to read this article by Hayden Winks on the right time to draft a Quarterback, or more to the point, your second one.
I can’t give you all the answers here for how to draft a perfect team, but parsing through the drafts of 204 teams there did start to become noticeable trends. This will be the first of several articles searching for Superflex answers.
One last thing before we jump in, these leagues were based on similar setups, but with slight differences and all cost between $10-20 entry. All leagues were full PPR with starting rosters typically including 1 QB, 2 RB's, 2 or 3 WR's, 1 TE, 1 Superflex and 2 Flex spots.
A Leagues (8) - Standard Superflex 12 team league 20-24 roster spots.
B Leagues (4) - Similar but with bonus points for big plays.
F Leagues (5) - Standard Superflex 12 team 50/50 payout structure.
All drafts were slow drafts ranging from March to September 2021. Drafters ranged in familiarity with best ball from seasoned veterans to first-time players, who typically were still used to Superflex formats.
I’m a big believer in the idea that roster construction matters. Merely picking the best available player at every pick will not give you the balanced team you need to win a league. After all, if we drafted the players who score the most points with every pick, projections would tell you to take mainly quarterbacks. In regular best ball, plenty has been written about the more optimal ways to construct your roster. Ideas like taking only two QB’s if you’ve selected one early are no longer well-kept secrets but common-place in drafts. However, in best ball should we be aiming for four QB’s or can we get by with three?
If you’ve drafted Superflex before you’ll be well aware of the run on Quarterbacks that can come when people start to panic that they don’t have their second or third Quarterback. The Quarterbacks dictate the whole draft and your ability to resist being caught up in a run can change a lot. Across 17 leagues I found nine teams that drafted three quarterbacks and went on to win their league, as well as eight teams that drafted four. Meanwhile, teams who finished last drafted three QB’s 11 times, and four QB’s six times. If we expand this out across the rest of the 204 teams we can see that the differences in the percentage of roster spots allocated to the QB position were marginally different. In a 12-team league, it’s quite hard to take more than four Quarterbacks and it’s rare managers risk having just two. In all 17 leagues the managers who did risk this finished an average of 8.5 out of 12 players. Whilst I sadly don’t have any definitive answers here, keep reading for more nuggets about the right times to draft our Quarterbacks in the section below.
The skill positions remain as important in Superflex best ball as they are in Single QB best ball. Running backs were frequently under or over-drafted by last-place teams, whilst winners stayed close to the overall averages of the league. This isn’t a remark on whether zero or hero RB works, more a reflection that taking a lot of running backs, or very few, rarely worked out for drafters in these leagues. After all, we can hardly know what average to aim for mid-draft. Winners allocated an average of 33.24% of their roster spots to running backs, and whilst losers and the overall average was very similar there are more details to explore. Seven of the seventeen last-placed teams allocated significantly more or less roster spots to running backs, ranging as high as an incredible 70% and as low as 20%. In a 20 round draft that would represent 14 running backs for the high figure and four for the low. I’m not averse to four running back builds in a regular best ball but in Superflex, it seems very hard to hit those numbers at the same time as getting QB’s at the right times. Winners allocated between 20-45% of their picks on running backs and 30% was the most common. If we’re thinking in 20 rounds, that would translate to six running backs. More on where to select them down below.
In a normal best ball league I, like many others, lean heavily on wide receivers in my builds and my most successful team last year reached the Draftkings final with 10 receivers carrying the load. In Superflex, I’ve found myself struggling to get quite so receiver heavy at times and the winners of my leagues trended similarly. The average across all teams was 45.34% of your roster being allocated to receivers, meanwhile, winners dropped by 1.5% to 43.82%. The most common allocation for winners was 45% (9/20), and whilst similar for losers again we saw losers stray above and below this number heavily. Losing teams allocated as much as 60% (12/20 spots) and as little as 30% (6/20).
I’ve been a believer that if I draft an elite tight end early, such as Travis Kelce, I wouldn’t use more than one other pick on a tight end, so I was surprised to see that no teams won their leagues with only two tight ends. Instead, 11 teams won with three and six won with four. Two team builds were more common in the losers section though, where teams allocated between 10% and 25% of their picks on tight ends. Even in leagues with additional flex spots, it seems that taking more than three tight ends did very little for your chances. It’s rare that tight ends outside of the elite range provide the type of spike weeks receivers are able to.
From this very small sample size based upon 2021’s data, it becomes obvious that taking a stand at wide receiver or running back more often than not did not pay off. The winning teams stuck close to what felt like a natural build of 3-6-8-3 if we round the numbers off and typically they had 5-7 running backs or 8-9 wide receivers. This doesn’t mean that strategies like Robust RB, Zero or Hero RB couldn’t work, but when players employed them they needed to not do so at the expense of overall construction. The question remains though when did taking certain positions pay off?
Now that we know edges were small in roster construction it makes sense to jump into round by round picks and see if edges were gained more routinely in that area. The average pick numbers in this section are taken from averaging each winner/loser's average round they picked the relevant position in.
As I mentioned earlier there’s nothing quite like the run on second and third quarterbacks to start a panic for drafters. People start grabbing their third and you still have just one and now you’re faced with Sam Darnold and Taysom Hill as your only options. Superflex drafts are all about finding the right time to pick the right QB’s but also not panicking straight away if you miss out. Last year we saw players such as Tom Brady and Matthew Stafford return an excellent value outside of the first couple of rounds of Superflex drafts and they should be a reminder for this upcoming year that there are still options as the rounds go on.
The average of all 204 teams saw teams select their first Quarterbacks in the second round (2.33). Winners waited slightly later on average at 2.40 and last-placed losing teams waited even later till 2.73. More interestingly the second quarterbacks were selected by winners in 6.13 (round 6) whereas losers made sure they had their second QB by 5.78.
It’s worth noting that in 17 leagues no team who drafted QB-QB back to back in rounds one and two went on to enjoy any form of success, their average finish was 7.7.
Losing teams took their third QB on average at 10.07, which was just ahead of the total average of 10.24, but winners held out a full round and a half later on average at 11.76. There were plenty of pitfalls to avoid in the later rounds, but our winners were rewarded for not rounding out their position as early as others.
In regular best ball, you’ll often see sharp drafters aim to have four wide receivers by round seven at the latest, and it’s not unusual to see good drafts have five. In Superflex, this becomes tricky as the Quarterbacks have been pushed up high and we know we’ll need to get one if not two before the tiers drop too dramatically.
The total averages saw a team's first receiver be drafted in round 3 (3.74) and winners were close to this with an average of 3.94. Losers however waited too long and averaged a significantly later 4.50 before grabbing a receiver. This theme continues for teams wide receivers two through four, with winners being significantly ahead of losing teams and nearly always ahead of the league average.
Whilst with Quarterbacks it was tough for teams to overcome a QB-QB start, it was equally hard for teams who took two receivers in the first three rounds. Without diving deeper I would imagine this forced them to miss out on either a quarterback or running back and tilted their team off balance.
Running back strategies can be a contentious subject in best ball and whilst I enjoy using Zero RB and Hero RB in regular best ball those strategies can be somewhat harder to pull off in Superflex leagues.
Our 204 teams averaged their first running back by 1.98, just squeezing into a first-round designation. Both winners and losers weren’t overly close to this figure with losers at 1.54 and winners even earlier at 1.41.
The winners and losers separated much more as they picked their RB2’s with winners waiting till 5.03, later than the league average of 4.70 and much later than the losers who pulled the trigger at 3.56 (Round 3). RB3 selections saw the winners creep back to the average with 8.33 being almost identical to the 8.28 average, but losers had already secured that spot almost a full round before at 7.48. By the time teams selected their RB4’s every one was fairly in line with no more than 0.27 separating all groups in the twelfth round.
Whilst for other position groups going heavy on them in the first couple of rounds often harmed a team's chances, at running back teams regularly saw both good and bad finishes if they took two RB’s in the first three rounds and averaged a 7.2 finish. So taking two RB’s early wasn’t problematic, but when teams did that they needed to move off the running backs for a while before selecting their third back, or their chances of winning dropped.
Lastly, we come to the tight end position, where we have already established that there was no edge in taking less than three in this format.
Winning teams selected their TE1 at 6.14, almost a round and a half ahead of the 204 team average of 7.59 and even further ahead of the losing teams who waited till 8.62 to select their first tight end.
No position is the difference between good and bad more obvious than at tight end and whilst many argue to go early or late and heavy for tight end, this pushes me much more towards ‘make sure you go early’. Ten times the winner had selected their TE1 by round six, and only three times out of seventeen a team went on to win their league if they waited till the double-digit rounds to select a tight end.
Whilst we found few edges in roster construction, it’s become clear that there are edges in taking certain types of players at certain times, as long as you stay balanced in the process. Many people have written about the running back dead zone and nobody covers it better than Jack Miller of Establish The Run. Without deeper analysis I would guess that typical dead zone running backs are being pushed into the areas where losing teams typically took their RB2 and RB3.
The ideal draft saw teams take their players in this method;
QB1 - earlier than average
QB2 - later than average
QB3 - later than average
WR1 - earlier than average
WR2 - earlier than average
WR3 - earlier than average
WR4 - earlier than average
RB1 - earlier than average
RB2 - later than average
RB3 - later than average
RB4 - earlier than average
TE1 - earlier than average
Whilst many factors will go into a successful draft, including at the very heart of things often picking the true league winners will do it, it’s completely possible to construct a team that can overcome not having those players. To win a league we need to be cognisant of the players we pick and how many of them we choose to pick. Looking at a relatively small sample of teams we can conclude that for the 2021 season a roster build of 3-6-8-3 was most successful and typically we wanted to wait later than most for our QB2 and QB3, hit WR heavy early on but not before we got our first RB, and in amongst all of that make sure we got a solid tight end as well.
There’s a lot to balance there and balance really does seem to be the name of the game. Don’t draft a position heavily if it means you’ll neglect another as a byproduct of it. When the giants of best ball start featuring Superflex heavier we’ll start to get a better handle on year to year data over a greater sample but for now in these infant days of best ball where the on-trend strategies change year to year, it would be wise to keep the points made above in mind.
As always, I love hearing your thoughts on the content I'm writing about. If you've enjoyed this, or have differing thoughts I'd love to hear them. The full version of the data I scraped is available to download below, with team names removed for their privacy. More articles will be coming over the next few weeks on this subject and if you'd like to get involved with Superflex Best Ball, give me a nudge on Twitter and in the coming weeks, I'll be sure to let you know what leagues I'm planning on running this year. All my leagues are on Sleeper with money securely held on Teamstake.com.
If you enjoyed this article and want to support my work and this website then please Like or Retweet the Twitter posts, it really helps me grow my audience and continue to provide content. Alternatively, you can buy me a coffee - https://buymeacoffee.com/tstrack
Thanks again for reading. Happy drafting.