Who Wants to Be a Millionaire?
We analyzed 80 DraftKings Millionaire Maker winning lineups from 2021 through 2025 to see what the million-dollar rosters actually had in common. The answer was not “fade all the popular plays” or “always bring it back.” The winners were usually correlated, usually had a couple low-owned leverage pieces, and very often included at least one player who completely broke the slate.
Correlation + selective leverage + one enormous value hit.
Across the full five-year construction sample, 86.2% of winners stacked their QB, the median winner used two players below 5% ownership, and among the 29 lineups with complete player-level salary data, 86.2% had at least one player deliver 6x salary value.
Five seasons. Eighty winners.
We built a standardized dataset of DraftKings NFL Millionaire Maker winning lineups from 2021 through 2025.
The construction sample contains:
That distinction matters. We use all 80 winners for stacking, ownership and FLEX analysis. Salary and points-per-dollar findings are restricted to the 29 full-lineup contests where those fields are actually available.
This is a study of winning lineups, not all submitted lineups. It tells us what Millionaire Maker winners looked like. It does not prove any one construction choice independently causes a lineup to win.
What did the typical winner look like?
69 of 80 winners paired their quarterback with at least one same-team pass catcher.
32 of 80 winners paired their quarterback with at least two same-team pass catchers.
35 of 80 winners used at least one player from the opposing side of their QB's game.
The median winner used two players below 5% ownership.
The median winner still used one player at 20% ownership or higher.
Among full lineups, 25 of 29 contained at least one player who returned 6x salary value.
Yes, Millionaire Maker winners usually stacked their QB.
The strongest construction signal in the entire dataset was quarterback correlation.
69 of the 80 winners — 86.2% — used at least one same-team pass catcher with their quarterback.
Only 11 winners used a naked QB.
That last number is important.
The common DFS rule of thumb is often presented as:
But fewer than half of the winning lineups actually used the opposing-side bring-back.
The data supports the value of QB correlation much more strongly than it supports the idea that you must always force an opponent into the same game stack.
Stack your quarterback aggressively. Treat the bring-back as a tool, not a requirement.
How the winning stacks were actually built.
The most common winning construction was actually the simplest correlated build:
Double stacks were also extremely viable. Forty percent of all winners used at least two same-team pass catchers with their QB.
That is a useful reminder for large-field tournaments: concentrating more of your lineup around one passing offense is not automatically “too much correlation.”
Million-dollar winners were not nine contrarian darts.
The median winning lineup used:
Cumulative ownership simply means adding the ownership percentage of all nine players in a lineup. It is not the percentage of the field that owned that exact lineup.
More important than the 111.7% number itself is what sits underneath it:
That is very different from intentionally making every roster spot uncomfortable just for the sake of being unique.
You do not need nine low-owned players. You need enough leverage to separate from the field without giving away too much projection.
Most winners found at least one massive salary-adjusted score.
In the 29 contests where we have complete player-level salary and scoring data:
The last number may be the most important.
A 6x player means a $5,000 player scoring at least 30 DraftKings points, a $4,000 player reaching 24, and so on.
In 25 of 29 full winning lineups, at least one roster spot generated that kind of salary-breaking performance.
That is why pure projection optimization is not enough. You are not just looking for nine players likely to hit median outcomes. You need access to performances that can break the slate relative to salary and ownership.
The FLEX was basically a coin flip between RB and WR.
Across all 80 winners:
There is no strong five-year case here for automatically forcing either RB or WR into the FLEX.
The interesting part is how much the profile moved by year.
| Season | QB Stack | Double Stack | Bring-Back | RB FLEX | WR FLEX | TE FLEX |
|---|---|---|---|---|---|---|
| 2021 | 100.0% | 27.8% | 50.0% | 38.9% | 50.0% | 11.1% |
| 2022 | 88.9% | 50.0% | 55.6% | 33.3% | 61.1% | 5.6% |
| 2023 | 83.3% | 38.9% | 16.7% | 33.3% | 55.6% | 11.1% |
| 2024 | 80.0% | 40.0% | 46.7% | 53.3% | 33.3% | 13.3% |
| 2025* | 72.7% | 45.5% | 54.5% | 72.7% | 18.2% | 9.1% |
*2025 contains 11 qualifying contests in the source archive.
The year-to-year movement is a good reason not to overfit one season of DFS results into a permanent rule. Winning construction evolves with pricing, player archetypes, field behavior and the NFL environment.
Stacking stayed dominant, but it became less absolute.
The QB stack rate declined each season in our sample:
That does not mean stacking stopped working.
Even in 2025, nearly three quarters of the winners still stacked their QB. But the field has become more sophisticated about correlation, and naked-QB builds have shown they can still win when the underlying player projections and ownership create the right path.
The trend is interesting. The five-year aggregate is more important: QB stacking remained the dominant construction by a wide margin.
If I were building three lineups, here’s where I’d start.
Eighty winning lineups do not give us a magic formula. But they do give us enough information to build a repeatable process.
If I were entering three Millionaire Maker lineups this week, this would be my starting point.
Identify three quarterbacks with legitimate slate-winning ceilings and build one lineup around each. Pair each QB with one or two of his pass catchers depending on the offense, pricing and how concentrated the targets are.
For each QB stack, identify the opposing player most capable of benefiting if the game turns into a shootout. Use him when the role, price and ceiling make sense. Only 43.8% of winners used a bring-back, so correlation should not come at the expense of a clearly better play.
This is where I would spend the most research time. Look for a player whose opportunity may be greater than the field realizes: an injury-created role, a backup becoming a starter, increasing usage, a breakout candidate, or a favorable depth-chart or schematic change.
Once the stack and primary value thesis are established, fill the remaining positions with high-upside plays. Do not avoid a strong player simply because he is highly owned. The lineup only needs enough differentiation to separate from the field while preserving upside.
The median winner used two players below 5% ownership, and 86.2% of the full winning lineups contained at least one 6x salary-value performance. The goal is not merely to find a cheap player — it is to find a cheap player with a plausible path to a score his salary was not designed to accommodate.
Optional bring-back → value thesis → fill with ceiling.
Optional bring-back → value thesis → fill with ceiling.
Optional bring-back → value thesis → fill with ceiling.
Injuries, promotions, increasing usage and breakout indicators are where low-cost upside can appear before ownership catches up.
There is no single millionaire template. There is a millionaire profile.
The 80 winning lineups do not give us one rigid formula.
They give us a set of tendencies:
86.2% of winners did. This is the strongest construction tendency in the dataset.
Only 43.8% used one. The QB stack mattered more than automatically including the opponent.
The median winner had two sub-5% players and one 20%+ player.
86.2% of full winning lineups contained at least one 6x salary-value performance.
Winning lineups generally used almost the entire cap.
Across five years, FLEX usage was almost evenly split between WR and RB.
How we built the dataset.
This analysis uses DraftKings NFL Millionaire Maker winning-lineup data from 2021 through 2025.
The five-year construction sample contains 80 winning lineups: 18 from 2021, 18 from 2022, 18 from 2023, 15 from 2024 and 11 from 2025.
Historical 2021–2023 construction and ownership data was normalized from published Millionaire Maker winner tables. The 2024–2025 lineups were reconstructed at full player level and validated against weekly NFL statistics. Three late-2023 contests also contain full player-level data.
We therefore separate the analysis into two levels:
Full-lineup sample: 29 winners. Used for winning score, salary usage and points-per-dollar / 6x value analysis.
QB stack: quarterback plus at least one same-team WR or TE.
Double QB stack: quarterback plus at least two same-team WR/TE pass catchers.
Bring-back: at least one player from the opposing team in the quarterback's game.
6x value: DraftKings fantasy points at least six times a player's salary in thousands. Example: 30 DK points from a $5,000 player.
Cumulative lineup ownership: the sum of the ownership percentages of all nine roster spots. It is not the percentage of entries that owned the exact lineup.
2025 note: the source archive contains 11 qualifying contests for 2025 in this study.
The study is descriptive. It identifies the characteristics of winning rosters but does not establish the independent causal value of any one lineup-building decision.
← BACK TO INSIGHTS