Analyzing the Top 10 Lineups From 2020 DraftKings NFL Millionaire Makers
It’s that time of year again! For the past few seasons I have been doing a thorough review of what the Top 10 lineups look like for the largest field DraftKings Millionaire Maker. Sometimes you have to beat a few hundred thousand opponents, and sometimes you have to beat over 1 MILLION other lineups to come out on top. It takes a good strategy and some luck to end up walking away the winner each week, but there are some trends that have emerged over the past few seasons that can help you make lineups that look like previous lineups that have won or competed at the highest levels in these extremely large field tournaments.
After you have read what happened in the 2020 season in this article, I encourage you to also go back and read last year’s article if you have never read my summary pieces before. There are plenty of lineup techniques to learn and it always helps to have as much information as possible. You can find the 2019 piece here.
Don’t forget that while this study looks at the very specific Millionaire Maker tournaments, the data within can be applied to many of the large field tournaments around the industry where you are attempting to beat 100,000+ entries.
I have a full set of 2020 data from Weeks 1-17 that I looked at for this review, so let’s get started trying to dissect what it takes to put yourself in contention to win these types of contests!
Stacking
I think most of you know how important the correlation between a QB and his pass-catching options are, but if you are new to DFS or this article this is what we refer to as stacking. While Patrick Mahomes and Travis Kelce can each have big games on their own, it’s reasonably likely if one has a big game, the other will as well. There is also a difference between using players from the same team (Team Stack), and using players from the same game (Game Stack). Game stacks have become the preferred way to build lineups over the years as not only does a QB correlate with his own pass catchers, he also correlates very well with players from the other team in the same game. Here are how many teams and game stacks were in each of the 170 lineups that graced the Top 10 last season.
Some Notes
157 of 170 lineups used some form of a stack. Of those 157 stacks, 146 or 93% of them came from four specific stack types.
— 33 from 2-man Team Stacks
— 23 from 3-man Team Stacks
— 45 from 3-man Game Stacks
— 45 from 4-Man Game Stacks
Straying too far from these types of stacks seems to be a losing proposition.
Game stack usage in Top 10 lineups has increased from 34% in 2018, to 58% in 2020. I’m not sure there is a ton more room for it to go up, but I generally don’t expect that number to move down.
We make it very easy for you to stack however you choose in LineupHQ here at RotoGrinders. You can easily make any stack you want in the Stacks tab when making lineups no matter how many you wish to create.
Here is the lineup that won the Millionaire Maker Week 1 last season, it utilized a 4-man game stack paring Aaron Rodgers with two pass-catchers and also using Adam Thielen from the same game from the Viking in hopes of a shootout.
Flex Usage
With so many options available at the Flex, deciding what position to use can often be troublesome. Here are what the Top 10 lineups did at the flex spot in 2020.
— 34% used an RB in the Flex
— 54% used a WR in the Flex
— 12% used a TE in the Flex
Here is what the field did as a whole at the Flex position:
— 43% used an RB in the Flex
— 46% used a WR in the Flex
— 11% used a TE in the Flex
With game stacks tending to be the preferred option of Top 10 lineups, it makes sense that we see WRs being the top option in the flex. If you are going to game stack pass-catching options from both teams, you simply have to start using WR more often in the flex. WRs also tend to have the highest ceiling of any position. Combine those two points and it’s easy to see why WR was the top option among top lineups.
The field is also starting to understand this as WR in the flex usage has increased from 27% in 2018 to 54% in 2020 in Top 10 lineups.
Total Lineup Ownership
Before we talk about ownership on a player level, let’s take a look at it from an entire lineup perspective first. Many casual players fail to take into account that ownership is one of the most important aspects to consider when making lineups in these large-field tournaments. I will call the Total ownership a lineup uses (Average Ownership) or (AO) from here on out.
Here is the AO of the Top 10 lineups each week for the 2020 season.
As you can see there are some peaks and valleys, with AO typically being lower when there are no bye weeks and having some spike weeks as ownership gets condensed during the bye weeks as there are fewer teams playing. There are some outliers but this trend generally holds true year after year. Another trend over the years is the start of the season also tends to have a lower ownership total as we have the least amount of information of what each team is trying to do on offense and which players are seeing increased roles. Be very willing to think outside the box for the first few weeks.
Now that I have three years of data, another interesting trend is emerging. Here is the AO of every Top 10 lineup for each of the past three seasons.
AO has gown down from 118% in 2018 to 110% in 2020. Just like with some of the other topics above, it seems like the field is starting to understand on some level that factoring ownership into your lineups in these extremely large field tournaments is important.
Low-Owned Players
In order to beat over One Million other lineups in Week 1 this year, and well into the hundreds of thousands of lineups most other weeks, you generally need a player or two in your lineup that perform at a high level that most others do not even consider. For my articles here on RotoGrinders, I consider that threshold to be a player that is 5% owned or less.
In 2020 the average Top 10 lineup used 1.9 players that were 5% or less owned in the Millionaire Maker. One interesting note over the years is that while Total Lineup Ownership has gone down as shown in the chart above, the number of low-owned plays per lineup has actually gown down over the years as well. Here are the number of low-owned plays per lineup over the past three years
2018 – 2.4
2019 – 2.2
2020 – 1.9
You would think that with AO going down over the past three seasons that the number of low-owned plays per lineup would be rising, but this is not the case. I believe this is due to the increase of game stacking and the use of correlated plays in lineups proving to be the better tactic overall when trying to take down these large-field tournaments. However, even though it’s much easier to nail a particular game stack than an individual low-owned play, with nearly two of the nine roster spots of the average lineup being filled by a 5% or less owned play, it is still something you should be paying attention to. Here is a positional chart of where every low-owned play came from during the 2020 season.
One position that really stands out after looking at three years of data, is that last year we were VERY good at selecting the correct QBs in our lineups in 2020. Here are the number of times low-owned QBs have appeared in Top 10 lineups over the past three seasons.
2018 – 67 (39.4% of lineups)
2019 – 66 (44% of lineups) Only 15 weeks of data used in 2019
2020 – 48 (28.2% of lineups)
Last season we saw the “good” QB plays come through for us nearly all the time. We will talk more in-depth about bust rates later, but of the Top 3 highest owned QBs each week, only 25% of them “busted” last season. This compared to a 58% bust rate in 2019 and a 44% bust rate in 2018. There could be many reasons for this. Are we seeing the elite QBs simply be that much better than the mediocre ones? Did having fewer fans in the crowd during Covid-19 help offense? Did fewer holding penalties help?
To me all of these factors helped contribute to the highest-scoring season in NFL history, with over 49 points being scored per game. To put that into perspective in 2019 we saw just under 46 points per game being scored.
If we see another big spike in scoring, we might just see the best and highest owned QBs continually put up the best numbers week in and week out like we did last year. But if some of these factors even out a little, we may see more lower-owned QBs show up in Top 10 lineups this season than last. Just a thought to keep in your head as we start the 2021 season.
Low-Owned Running Backs
Running Backs that are 5% owned or less have held reasonably steady over the years in terms of lineup usage and appeared 54 times or in 31.7% of Top 10 lineups. The question we need to answer is what does a 5% owned RB look like that can appear in a Top 10 lineup. Here is an example from last season and a short explanation on what happened that week.
Here we have the Week 4 winning lineup from RoyalPain21. We have Joe Mixon coming in at 3.8% against the Jaguars, COME ON! The same Jaguars that were so bad they got the #1 pick in the draft. How does this happen? One of the key things to remember in NFL DFS is that recency bias is VERY strong with the general public. If a player did badly the week before, or has a bad run for two weeks, that player is often left for dead by the DFS community. In the previous two weeks, Joe Mixon had failed to rush for more than 50 yards or score a touchdown. WHAT A BUM! We failed to realize just how bad the Jaguars were in Week 4 even though they gave up 138 yards and two TDs rushing to the Dolphins the week before. Looking back it seems ludicrous that as a collective whole we allowed Joe Mixon to be 3.8% owned. Keep that in mind when searching for low-owned plays at the position. In general however, outside of finding a player like Mixon at low ownership in a obvious situation, RB is one of the least favored spots to go searching for a low-owned player that can get you to the top.
Low-Owned Wide Receivers
The usage of low-owned WRs in Top 10 lineups dipped last season which correlates with the drop in low-owned QBs showing up in the Top 10. If we are using the best QBs with their high-owned receivers and they generally do well, there isn’t much room for the low-owned WR to shine. With three WR spots and a flex available, Top 10 lineups had 680 chances to roster a low-owned WR. Of those 680 chances, a low-owned WR was used just 127 times or 18.6%. This number was 25.2% in 2019, and 24.4% in 2018. With WRs being used in the flex more then ever last season, this drop in low-owned WRs is even more pronounced.
What does this mean for our lineups this season? If scoring jumps 3+ points per game like last season, I believe we will see a similar result as last season for low-owned WRs. This would again lead to many lineups in the Top 10 not needing those sub 5% owned gems that post big games. If we see scoring stay the same or decrease, it’s very likely that more high-owned QBs and their correlated high owned WRs end up faltering. This would lead to more low-owned WRs hitting the Top 10 lineups.
At the start of the season I see no reason not to embrace the uncertainty and will plan to include a fair share of low-owned WRs in my extremely large field tournament lineups.
Low-Owned Tight Ends
Low-owned TEs saw their usage spike last season as 51 low-owned TEs made it into the 170 Top 10 lineups overall. This compares to a lower percentage in both 2018 and 2019. After looking at the Top 10 lineups over and over, it really just came down to us needing some salary savings at the position. In Week 1 last year we had 9 of the 51 low-owned TEs show up in lineups, led by Dallas Goedert at $4,100 who came in at 3.6% and scored 27 FPTS.
There are also plenty of low-owned TEs showing up with rather mediocre performances of even getting ZERO fantasy points. They tend to show up in Top 10 lineups in this fashion when they are very cheap since using a low salary at TE allowed you to spend up on the correct stack of the week.
If you do go cheap and low-owned at TE, try to get players that have some touchdown equity like a Dallas Goedert from last season, a Logan Thomas from last season who was very cheap for awhile, and if you don’t go for the cheap TEs, try to use the stud players like a Travis Kelce or Darren Waller when the rest of the field is on a cheap TE for the week and the studs at the position stay under 10% owned.
Low-Owned Defense/Special Teams
Low-Owned D/STs appearing in Top 10 lineups dipped in a big way last season. In 2018 we saw 34% of lineups use a 5% or less owned D/ST. In 2019 that stayed at 34%. Last season we saw that number dip to 23.5%. Another interesting fact about the D/ST spot is that in five different weeks the actual winning lineup scored nine or less points from the D/ST position.
What really happened last season both at the D/ST position and across many of the low-owned factors, was really due to the increase in scoring, which in turn led to game stacking becoming even more important than ever. If we continue to see a scoring spike, or even hold steady at the pace of last season, I believe many of the low-owned statistics I look at will hold steady or continue to trend downward in 2021. If scoring dips, which is certainly possible, then we are likely to see the need for low-owned players in lineups rise to levels from 2019 or 2018. Remember when building lineups to start the season that last year was a very big outlier in terms of the scoring per game increase and that any dip in scoring per game, or another large increase, could drastically change what top lineups look like in 2021 in terms of low-owned players.
Bust Rates
Just like with the low-owned players above, bust rates saw some very large discrepancies in a few positions compared to previous seasons. First let’s take a look at what the bust rates were in 2018 and 2019 among the three highest owned players at each position.
Both graphs look rather similar from 2018 and 2019. Now let’s compare that to 2020.
Here we have a graph that looks nothing like what we saw in the previous two seasons. At the QB position the bust rate was cut from 57% to 25% from 2019 to 2020. Since QB bust rates fell so dramatically, it also makes sense that WRs saw their bust rates drop to a three year low as they fell from a 64% bust rate in 2019 to just 43% in 2020. While I do think we will always get slightly better as a community in selecting the “best” plays, this is yet another statistical oddity that we can chalk up to the increased scoring last season in my opinion. If scoring dips this season, I would expect to see some regression to the 2018-2019 bust rates at the QB and WR positions.
The other three positions of RB, TE, and DST all remained relatively close to their historical bust rates last season. This was very impressive at the RB position since we only had Christian McCaffrey and his high floor/ceiling combo for three games last season. As a fun note CMC was under 7% owned in two of the three games he played last year and had two touchdowns in all three games. I actually think the return of CMC will have a big impact on DFS tournaments this season if he can regain his dominant form.
Tight Ends have had at at least 49% of the Top 3 highest owned players bust each season, and reached an all-time high last year when 58% of the Top 3 highest owned at the position failed to produce anything meaningful. It remains one of the best positions to swerve off the highest owned players week in week out. When players like Hayden Hurst, Drew Sample, and Chris Herndon project for high ownership on any given week, I would suggest looking elsewhere as over the years these fringe TE talents that pop in ownership end up holding your lineups back in a significant way as you are much better off taking a different option that is much lower owned in these large field GPPs. TE is also a position that you can totally miss on and still have a Top 10 lineup. There are quite a few instances of TEs scoring less that 5 FPTS in Top 10 lineups and even some absolute ZERO’s every year. This just further shows that it’s a position that you should be willing to move away from the highest owned players.
At the D/ST position we have seen at least a 50% bust rate for the top three highest owned teams in Top 10 lineups each of the past three seasons. From my observance of these lineups over the years I have noticed a few trends. If you use a higher-priced D/ST in the $3200+ range you generally need to hit the nuts to compete for a top lineup. If you go with a cheaper defense, you really give yourself some outs by being able to spend more money on the higher end position players that have huge games or create the correct stack of the week. There are plenty of $2500ish or less D/ST’s in Top 10 lineups that score less than seven points, and even some that score negative points. With three straight seasons of a 50%+ bust rate on the highest owned plays, don’t be afraid going against the grain at the position, especially if the teams you use are on the cheap end.
Lineup Examples
In this section I’m going to highlight a few of my favorite lineups from the Top 10 in 2020 while attempting to breakdown what may or may not have been thought of by the maker of that lineup.
Let’s start with the Week 6 winning lineup from WackyUncle
This lineup starts off with a 4-man game stack of the ATL at MIN game with only one player in that stack over 10% owned. Paired with that we have one of the more powerful “secondary correlations” of RB/OppWR which used Ronald Jones from TB and Marques Valdes-Scantling from GB from another game in an attempt to add even more correlation to the lineup. A low-owned DeAndre Swift against the Jaguars and a low-owned D/ST round out the lineup. A work of art that took home a million dollars.
Here is a lineup from randers that took 2nd place in Week 10
In this lineup we have a shootout potential game in a dome with a BUF@ARi 4-man game stack which was generally a great way to start lineups last season. Added to this was the “secondary correlation” of RB/Opposing Pass catcher of DeAndre Swift and Logan Thomas from the WSH@DET game. An RB that gets plenty of work in games that might score plenty of points correlates very well with opposing pass-catchers in the same game, which makes this addition a very powerful one.
Below is the winning lineup from Week 14 from Fortyfortner, which utilized a different form of a stack.
In this lineup, we have a 2-man team stack of Derek Carr and Nelson Aghohlor at extremely low ownership. Added to this we see a Bears stack as they faced the Texans at home and were in line to put up plenty of points. David Montgomery was tearing it up to close the season out and Allen Robinson is always a threat to post a big game. Adding two players from the same team as an additional stack can be very viable when they are in a spot like the Bears were in Week 14. Not only is that “secondary correlation” used, but we also see a third correlation used with Derrick Henry + AJ Brown and Tyler Eifert from the TEN@JAX game. By getting just three games correct in how they would play out, this player was able to take up seven of the nine lineup spots with highly correlated plays. One of the best ways to beat 200,000+ entries is by having as much correlation in your lineup as possible whether it be on a singular 4-man game stack, or a few different correlations from separate games.
You don’t always need secondary correlations, in fact with the uptick in scoring and 4-man game stacks last season in Top 10 lineups, secondary correlations actually saw a reasonable decrease in usage in Top 10 lineups. The main point to take away is that you typically need a large amount of correlation in your lineups to compete at the top. That can be in the form of a singular 4-man game stack, or it can be a combination of correlations shown above. If you need more information on how to use LineupHQ to make secondary correlations, here is a quick tutorial video on some basics.
Summary
I mentioned a few times that the increase in scoring really changed the way the game was played season. Here is a chart from wizardofodds that demonstrates just how big the jump was:
Image Credit: wizardofodds
The 2020 NFL season saw the biggest jump in points per game scored in recent history. After seeing scoring around 46 points per game for both the 2018 and 2019 season, the jump to over 49.5 points per game in the 2020 season had a very big impact on what strategies you needed to use to compete at the top of these large field GPPs. Was this increase in scoring due to having less fans in the stands, was it due to seeing more than one less offensive holding penalty per game called, was it due to no preseason? These are all questions we need to ask ourselves as we make lineups to start the 2021 season. If scoring remains the same or goes up, we are going to see even more 4-man game stacks with secondary correlations at the top of tournaments. If scoring goes back to the 2018-2019 era, lineup construction and the statistics shown throughout this article will likely regress back towards those years where scoring remained lower.
That’s it for the 2020 review, make sure to come find the bi-weekly 2021 articles that will be feature here at RotoGrinders starting after Week 2 of the 2021 season as we try to dissect what will be winning this season!