The dupe tax: why your "unique" lineup has twelve copies

Everyone is drinking from the same well

Every entrant in a large-field GPP starts from roughly the same place: the same projections, the same salary cap, the same handful of "obvious" value plays that unlock the slate. Run those inputs through an optimizer — or through the human instinct to not leave points on the bench — and the astronomically large space of possible lineups collapses into a few heavily-trafficked constructions.

That's duplication. It isn't bad luck, and it isn't other people "stealing" your build. It's the predictable result of thousands of people solving the same puzzle with the same pieces.

Joint ownership is the tell

You can estimate how crowded a lineup is before the contest locks. The signal is joint ownership: how popular all nine of your players are, together. Each high-ownership player you add multiplies the odds that someone else lands on your exact roster. One 30% chalk piece is fine. Six of them, glued together by the same two value plays everyone found, is a convention, not a lineup.

Those numbers come from the duplication model inside SimSlate's contest simulator, which measures how often its own ownership-faithful field process reproduces each construction. You can try it on your lineup right now with the free Dupe Tax Calculator — enter nine ownership percentages, get expected copies in any contest size.

What splitting first place actually costs

Here's why this matters more than almost any projection argument. GPP payouts are savagely top-heavy: the top 1% of finishers take most of the pool, and first place alone can be 15–20% of it. Duplicated lineups tie. Tied lineups split the prizes across their rank band.

If twelve copies of your lineup hit the perfect ceiling game and "win" a million-dollar tournament, you don't win a million dollars. You win roughly the average of the top twelve prizes — often a fifth or less of the headline number. Two lineups with identical scores do not pay the same. The chalkier one always pays less. That gap is the dupe tax.

Uniqueness earns nothing — it just stops the bleeding

The flip side matters too: there is no bonus for being different. A one-of-a-kind lineup keeps 100% of whatever it wins — that's all. Forcing weird, low-probability builds just to be unique swaps a small tax for a large one (a lineup that never wins keeps 100% of nothing).

The target is leverage: builds with real, correlated paths to the top of the scoreboard that the field underweights. Leverage keeps the winning odds and dodges the tax.

How SimSlate prices it in

Most tools ignore duplication entirely — every lineup is scored as if it keeps the full prize. SimSlate's simulator gives each lineup a crowd-width based on how often the field process reproduces it, and splits prizes across that band in all 50,000 simulated contests. The ROI you see on the board already has the dupe tax taken out. Chalk doesn't need to be banned; it needs to be priced.

The free plan runs a real 1,000-lineup field simulation — no card required — and the methodology page walks through the full model if you want the receipts.