I Built One DFS Lineup With Rules Instead of Hope

I began with a favorite player and ended with a workflow asking him to leave #

I started the build the way I have always started DFS builds, which is the wrong way. I opened the player pool, found a wide receiver I had a feeling about, locked him into the flex, and started spending the remaining salary around him. By the third slot I had spent too much money on the front of the lineup and was reaching for a $3,400 punt at tight end who, as far as I could tell, had no business being on the same field as the rest of my roster. This is the workflow most casual DFS players use, and it produces lineups that are emotionally satisfying and structurally fragile.

For the experiment I forced myself to do it the other way. I closed the player pool. I opened a blank text file and wrote the rules I wanted the lineup to satisfy before I knew which players would satisfy them. The rules took about fifteen minutes to write and were unromantic in the extreme. They named salary bands per position, a maximum cumulative projected ownership, a correlation requirement between QB and at least one pass-catcher, a leverage minimum, and an explicit cap of one play picked because I liked watching him. Then I reopened the pool and let the rules do the first pass of the build.

The market mechanic of salary scarcity #

The piece of DFS theory that took me longest to internalize is that salary is a real-time market and the bottleneck moves week to week. On a typical main slate the WR/TE punts at the bottom of the salary range are abundant; the WR2 tier is reasonably stocked; the RB1 zone is structurally thin; the premium players at the top are scarce by design. The lineup builder s job is not to pick the best players. It is to pick the best players in the tiers where scarcity is actually constraining the field. On this slate the bottleneck was the RB1 zone, which contained about nine percent of viable plays and which my favorite lineup was happily ignoring in favor of a third wide receiver.

The rules-built lineup, when I ran it through the constraint solver, kept landing on a lineup core I would never have built freehand: a stacked QB-WR pair from one game, a contrarian RB1 from the thin tier, a chalky high-floor TE, and a leverage WR2 from a game I had not been planning to watch. The lineup was less fun. The lineup was structurally defensible. The chart says the lineup also projected within four points of my favorite lineup at a substantially better leverage profile, which in a large-field GPP is the trade you are actually making whether or not you admit it.

What the rules caught that my preferences would not have #

The single most important rule, in retrospect, was the ownership cap. My favorite lineup had a cumulative projected ownership about thirty percent above the rules-built lineup, which sounds abstract until you translate it into outcomes. A high-ownership lineup that hits is a lineup you split with the field; a leverage lineup that hits is a lineup that climbs the prize structure. The math on this is well known and was, in this build, the rule I most wanted to violate, because the leverage plays were players I had not been planning to watch. The cap forced me to consider them, and three of the five made the final lineup.

The second most important rule was the affinity cap — the one-play limit on roster slots filled because I liked watching the player. The cap exists because the bettor s emotional attachment scales with conviction in a way that is essentially uncorrelated with leverage. Allowing two affinity picks per lineup makes the lineup a personal flag rather than a defensible build. Allowing zero affinity picks makes the build feel mechanical to the point of joylessness. One affinity pick, the rule s default, is the smallest concession to the bettor s humanity that still lets the rest of the build do its job.

The QB stack the rules built for me #

The correlation rule produced the most interesting structural change. My favorite lineup had a QB and three pass catchers from three different games, which is a build that wins when many things go right in many places and loses cleanly when one game goes wrong. The rules required that the QB share a game with at least one pass catcher in the lineup, which is the basic GPP stacking discipline I had read about for years and never quite committed to. The stacked lineup is more concentrated by design. It is also the structure most large-field GPP winners share, which is not a coincidence the field rewards by accident.

The QB-WR pair the rules selected for me was, predictably, from a game I had not been planning to watch closely. The pair s combined ownership was modest and the game was not in primetime. The structure of the lineup put the QB and one of his receivers on the same set of correlated outcomes, which meant the lineup would either win or lose as a coherent thesis rather than die one starting position at a time. The bet looked less interesting on paper. The bet looked more defensible on the leverage chart, which is the chart the prize structure is actually paying out against.

The final lineup was not my dream lineup. That was the point. A workflow should be allowed to disappoint the part of you that confuses comfort with construction.

The discipline of writing the rules before opening the pool #

The discipline I have come to depend on, after several builds with this workflow, is rule writing before pool opening. The rules need to be specific enough to constrain the build and loose enough to admit a defensible lineup on most slates. My current rule template runs about fifteen lines and includes the five originals plus an injury-news cutoff time, a maximum two-game stack rule, and a minimum cash-game viability check that I run before the GPP build to make sure I am not constructing a lineup that depends on three different ceiling outcomes to be tournament-relevant.

The friction is small and the payoff is structural. A lineup built from rules is a lineup that can be defended in a tracking sheet after the contest, regardless of result. A lineup built from affinity is a lineup that wins or loses without teaching the bettor very much about the workflow itself. The bankroll cares about the result. The bettor — the part of the operation that has to build again next week — cares about the workflow. The rules let the workflow improve in a way the freehand build cannot.

The closing argument #

I did not win the contest. The lineup finished in the upper third of the field, which is exactly the kind of result the leverage profile predicted: a defensible build that did not need everything to go right and did not, in the event, have everything go right. The favorite lineup, run against the same slate as a control, finished in the upper half but well outside the prize structure. The two lineups differed by about four projected points and a great deal of structure. The structure, more than the projection, determined the finish.

The experiment will continue. The rules will get tighter as the workflow accumulates more contests. The affinity cap will remain at one. The QB stack will remain mandatory. The salary-scarcity check will remain the first thing I run after the pool opens. None of these is glamorous. None of them will be quoted in the group chat. All of them, applied weekly, produce a lineup I can defend in a spreadsheet, which is the only audience the workflow needs to satisfy after the contest closes and the leaderboard freezes.