Back in December of 2020, I was listening to a DFS Army podcast and Geek beautifully articulated the comparison between The Queen’s Gambit and DFS Strategy. He said something like, “Ya’ll need to be like the girl from Queen’s Gambit and stay up at night replaying old matchups in your head.. Maybe not replaying old matchups on the ceiling because you have to be on drugs to do that but this should keep you up at night! You are not going to get better until you put in this kind of effort to be successful.”
Also known as the Backtest.
That really stuck with me and I decided to dedicate part of the offseason looking for past trends that will make me a more profitable DFS player in 2021. As the old saying goes, History repeats itself.
So why the Milli Maker? Because if you can identify trends that winners use to take down the hardest contest to win in DFS, there is a lot to be learned there to work your way back to smaller GPP contests. This is just part 1 of evaluating GPP’s.
DFS, more so than any other form of Fantasy Football, is a game of statistics. Predicting what is the most likely scenario to happen doesn’t help you in a field of hundreds of thousands of DFS players all aiming for that same prize pool. Therefore, the only way to take down a big tournament is to differentiate your lineup & we must dissect the winning lineups to find trends that we can use going forward.
Stacking trends
One of the most important aspects - but easily the trickiest part - of winning DFS Tournaments is stacking. Both Team Stacks (A QB and 1-2 of his pass catchers) and game stacks (Stacking players from one team and also players from the opposing team in that same game.) First I’m trying to identify the proper game environment for these kinds of stacks to have the highest probability of the most points scored, and secondly how players are stacked within those games.
One thing sharp DFS players look for when building a lineup for tournaments is the Vegas line totals to identify these game environments that are likely to produce shootouts and thus fantasy points, so we’ll start there.
Just looking at the Milli Maker winners from the first month of the season, you can definitely identify a trend of lineups featuring team stacks and game stacks from those highest projected game totals.
For example, in week 2 of the 2020 NFL season, you had
ATL/DAL at 53 points
BAL/HOU at 52 points
GB/DET at 48.5 points
TB/CAR at 48.5 points
KC/LAC at 48 points.
It was evident early on in the 2020 Season that both Dallas and Atlanta could score points, but also had poor Defenses that surrendered a lot of points. This is the type of game that sets up to be a shootout & the kind of game you want to stack. In fact, the week 2 Milli maker winning lineup had Dak at QB, Cooper at WR, Lamb at WR, and Calvin Ridley as the bring back WR option. This team stack consisted of 3 players Dak, Cooper, Lamb. The game stack used Ridley as the bring back.
Furthermore, Dak, Cooper, & Lamb were in the Milli Maker winning lineup again just two weeks later in a similar situation where they faced Cleveland in a game total that was projected at 55.5 points. This time the bring back option was OBJ.
It doesn’t end there as Dallas played Seattle in Week 3. Early on in the season the narrative was, “Let Russ Cook.” and he was cooking while the Defense had their struggles stopping the other QB from cooking. This game featured two high powered offenses with struggling defenses. This was another type of game you would target and the Milli Maker winner had a Game stack of Wilson, Metcalf, Lockett, and Gallup.
(One thing to add here is that the vegas lines are just a guide that help us zero in on a group of games that we want to stack but picking the right stack that actually exceeds the high projected score is the first big step in constructing the winning lineup.)
As you would imagine, the first four weeks of the season had Milli Maker winners with exactly 4 players from the same high scoring game, as well as a combination of other smaller team/game stacks. However, from there the formula for building a winning Milli maker lineup changed drastically.
The ingredient you will notice the most is flexibility. If the game script projects a high probability of a shootout, stack 3-4 players in that game. If it doesn’t, don’t force it. You have to be willing to zig when others zag.
Through 16 weeks of the 2020 NFL Season:
There was a stack with 2 or more players from the same game in 100% of the winning milli maker lineups.
There was a stack with 3 or more players from the same game in 81% of the winning milli maker lineups.
There was a stack with 4 or more players from the same game in 43% of the winning milli maker lineups.
In short, if your milli maker lineups didn’t correlate with at least one stack, you might as well have just burned a $20 bill for every entry you placed in the milli maker. Some would argue that’s the case in the Milli Maker anyways, but hey it sure is fun to dream a bit before lock about taking down the biggest pinata in DFS.
Roster ownership projections
One of the biggest factors to consider in building a lineup for a massive GPP such as the Milli Maker is total ownership of your lineup. You’re looking to beat out 400,000 to 500,000+ lineups so you have got to be different in your constructions. One measurement for that is projected ownership. In 16 weeks the lowest ownership was week 15 with 66% total ownership... which is crazy low. The highest was in week 12 with 173% ownership. Normally that is way too high of a target ownership when building your lineup but sometimes you’ve got to get silly for the milli and this silly lineup brought it home in week 12. Taking out the highest and lowest, only changed the average by 2 points. The average ownership over 16 weeks was 108%. There were 7 weeks under 100% so I wouldn’t shy away from building a line up under 100% ownership in the least bit, but I would definitely keep it under 130% as there were 7 weeks with ownership between 100-129%. If you’re keeping score, that’s 14 out of 16 weeks with ownership below 130%.
Leverage
Apart from Total Ownership, one of the biggest factors you need to consider when building a lineup for a massive contest like the milli maker, is what are your opponents doing? You need to zig when they zag. Especially in this contest. In GPP’s, Fade the chalk as they say. For example when you’re looking to use Seattle in a stack and most of the field is stacking Wilson/Metcalf, you can probably get leverage on the field by using a lower owned Wilson/Lockett stack or maybe stacking the other side of the game and getting off of Wilson/Metcalf entirely and just using Lockett as the bring back option. In the week 3 of the milli maker, for example, two game stacks were used for a total of 6 players, and 2 of those players were Greg Olsen and Lazard at 3.4% and 2.9% ownership, respectively. CHumleigh54 got leverage on the field with 2 players at very low ownership.
Overthinking the Flex Spot
Overthinking that flex spot. Who hasn’t done it? You have your game stack that you want, you have identified your studs as one offs, but you have no idea whether to put at RB or WR in the flex. Or will it be a tight end week? Do you start Kelce or Waller and then play some stone cold minimum Tight end at Flex to squeeze in the studs at WR?
In 16 weeks of the Milli Maker contest:
19% of the winning lineups had a Tight End in the flex spot.
25% of the winning lineups had a RB in the flex spot.
56% of the winning lineups had a WR in the flex spot.
When in doubt, the answer is WR. It’s like picking C in a multiple choice test when you don’t know the answer.
Let’s explore this further:
Week 1 it was WR Ridley at $6100 in a game stack with Olsen.
Week 2 it was TE Reed at $2600 as a one off. Kelce was in the TE spot.
Week 3 it was RB Henry at $7600 as a one off. He paired with 2 $4000 RB’s.
Week 4 it was RB Mike Davis at $5700 as a one off.
Week 5 it was WR Anderson at $5900 in a game stack with Gurley.
Week 6 it was RB Swift at $4500 as a one off.
Week 7 it was WR Hopkins at $8200 in a game stack with Wilson/Lockett.
Week 8 it was WR Bourne at $3500 in a game stack with Metcalf.
Week 9 it was WR Davis at $3200 as a one off.
Week 10 it was WR Reynolds at $3500 as a one off.
Week 11 it was WR Lamb at $5000 in a game stack with Cook/Thielen
Week 12 it was RB Ekeler at $6100 in a game stack with Gabriel Davis.
Week 13 it was TE Waller at $6100 in a game stack with Carr.
Week 14 it was TE Eifert at $3100 in a game stack with Henry/Brown
Week 15 it was WR Mooney at $3900 in a team stack with Monty.
Week 16 it was WR Crowder at $4500 as a one off.
Probability wise, you’re best putting a WR in the flex in some kind of stack. The exception to this seems to be when they’re low priced as they don’t have to do much to hit value. If you do put a RB in the flex, you’re best using them as a one off. The only RB who was stacked was Ekeler who is more of a receiving back so it makes sense he was stacked with the cheap Davis. If going tight end in the flex, it makes sense to pair a tight end close to minimum pricing with a stud like Waller/Kelce who produce more like elite wide receivers.
Positional Pricing
Over 16 weeks, I broke down the pricing of each position to see if I could spot some trends. Weeks 1-8, the average QB price in the winning milli maker contest was $7250, weeks 9-16, it dropped to $6412. Why the difference? In all 16 weeks, the Salary Cap is $50,000, so obviously a shift away from spending money at one position, would lead to more money elsewhere - As no milli maker winner spent less than $49,900 when constructing their lineup.
The Average money spent at the RB position in weeks 1-8 was $6793. In fact, without the week 8 outlier, it was $5500. (Henry and Cook) From weeks 9-16, the average jumped up to $7100.
Comparatively, the average amount spent at the WR position weeks 1-8 was $5900, and $5978 weeks 1-16 so that position held steady for the entirety of the season.
The average price of the tight end position, however, declined with the average price of the QB as it was $5100 in the first 8 weeks and then averaged $3900 the rest of the season.
My guess was that with such high priced RB studs like Henry, Cook, and Kamara, it made more sense to find a pay down option at QB/TE to have one of these studs in your lineup because they have the kind of ceiling you need to bring home the milli maker. In any week, any 3 of those RB’s can go off for 40-50+ points.
How would one do this but still have adequate production at the QB Position? One option is finding a backup QB with running ability who gets thrust into the starting lineup. The Week 11 winner had Taysom Hill at only $4800 which allows for the payup of Cook at $9,000. Also the week 15 winner had Hurts at $5900 who produced 40.82 points to the tune of 6.9X! Squeezing in Monty and Ridley at an average price of $7600 is much easier to do when your QB is under 6K. Cheat code engaged.
Even a non running QB like Carr gave you the perfect salary relief in back to back weeks to bring home the milli maker in weeks 13 & 14. With game totals of 59 points, and 71 points, Carr averaged 32 points and was never more than 6K. He also brought Waller with him with 48 points in week 13, and Agholor with 24 points in week 14.
The key takeaway here is being flexible and taking what the schedule gives you. In the first month of the season, you’re building game stacks with QB’s at the average price of nearly 7K + 2 pass catchers from the same team + an opposing team pass catcher. In the last 6 weeks of the season, we saw 4 weeks with a sub 6K QB in the winning lineup. Not only was the pricing much lower, but in two of those matchups, the QB’s were ran naked (without a pass catcher aka stack.) Those Qb’s were Hurts/Hill. The idea in the construction of those lineups is that Hurts/Hill don’t throw enough to bring a WR with them to have the kind of ceiling you need to take down the milli maker, so it’s best to have them unstacked.
Another point on matchups is that it’s often believed that Defense is one of the most volatile positions and that it’s best to find a pay down option. I would agree with this to an extent, but this is less true in the milli maker. The average price of the Defense in the Milli Maker was $3,000. 11 out of the 16 winning Milli Maker lineups had a top 10 ranked defense. The other winners were usually sub par defenses in great matchups like Dallas vs the burrowless Bengals and Cleveland vs the Eagles led by turnover prone Wentz.
Again, the idea is play the matchups but don’t be afraid to pay up for a top 10 Defense that is priced in the $3K-$3500 range. At the end of the day, you generally need 245 points or more to win the milli maker, and having a defense that can score gives you a much better chance to get there. In fact, the average points for Defense in the winning lineups totalled 16.5 points. If you’re keeping count, that is still over 5X.
Player ceiling
The last thing I wanted to identify is the ceiling. In 16 weeks of the Milli Maker contest, the low winning score was 217 points and the high was 279 points. There were multiple weeks with the winning lineup scoring over 250 and the average for 16 weeks was 243 points. If we’re shooting
at a target to win the Milli Maker, I’m setting it at about 250. Divided by 9 starting positions, that equals 27.77 or about 28 points per starting position.
Theoretically, if you were building a lineup of 9 players and expecting them to average 27.77 points, you’d need some core players with a high floor & ceiling, and you need to pick the right game stack where a shootout can take some players with a lot of variance to their ceiling, or close enough to their ceiling to keep the player average high enough to hit that 250 point target.
This is where constructing a lineup that can take down a big tournament eliminates the bigtime players from the JAGS. You can have a correlated lineup, but if you don’t have the horses capable of producing 250 points, you’ll never reach that many points in a game.
This is where you want to have the DK’s, Tyreek’s, Davante’s, Cooper’s of the world as the core of your lineup, but also some low priced, high variance players in that stack that can vastly outperform their salary.
https://hellorookie.com/nfl/2020-draftkings-millionaire-maker-winners/
For example, look at the week 1 milli maker winning lineup. The target was a GB/Min game stack consisting of Rodgers, Adams, Thielen, and Lazard. Now it’s clear that the 3 players of Rodgers, Adams, and Thielen have the kind of floor and ceiling to combine for 83 points for that 27.7 average that we need. In fact, their combined average was 37.45 points. This average and game script was enough to bring a low floor player like Lazard to almost his ceiling with 18 points - which brought the 4 player average to 32 points, well ahead of the 27.7 needed for a 250 point lineup.
So the correlated game stack hit and this game stack of 4 players produced more than half of the points needed for a 250 point total. Chumleigh54 didn’t only have 1 stack in his lineup. He also chose to feature a smaller game stack between Seattle & Atlanta which ended up being a high scoring game as well. In this game stack he chose the high floor/high ceiling player of Ridley who came through with 36.90 points and brought Greg Olsen with him for 12.40 points. Low scoring but still 3X for his $4,000 Price tag and still cumulatively ahead of the 27.7 point average needed by each player. Between 2 game stacks and a total of 6 players, you have 179.86 points, which is about a 30 point average from 6 roster spots.
The final 3 spots:
The easiest part about selecting Running Backs is generally they shouldn’t be stacked with other players from the same team as they don’t have a positive correlation with pass catchers or the quarterback from the same team. However, sometimes they can be used as bring back options in a game stack featuring a game that shoots out. In this instance, Mostert and Jacobs went unstacked and combined for 61 points. Mostert and Jacobs are talented backs with median prices and favorable week 1 matchups so were solid plays with the high priced QB/WR’s combo of Rodgers, Adams, Thielen, & Ridley.
Lastly, The Washington Football Team’s top 10 Defense in 2020 got the best of Wentz week 1 and scored 15 points. At only $2,000, produced 7.5X!
This roster was beautifully correlated & constructed which allowed it to produce the 255 points required to make $1,025,000 in week 1 of the Milli Maker.
Whether you’re playing in The Milli Maker or other large field GPP’s, replaying old matchups is the equivalent of film study for DFS players & something i’ll be regularly evaluating in 2021 and beyond. Subscribe for weekly reviews of contest winners In both large field GPP’S and smaller tournaments such as the 3 max.
Join us on Friday nights Weekly this season as we debut a new Podcast ‘Makin a Milli’ starting 09/17.
Find the Fantasy Fling Podcast on Itunes, Spotify, or wherever you listen!
Comments