Circa Took In $38.5 Million This Month and Kept Roughly None of It

Editor's note: TrueEdge builds odds tools and earns affiliate commissions from licensed sportsbooks, including DraftKings and FanDuel, whose economics this piece treats unkindly. Circa pays us nothing. Weigh the conflict anyway.

Registration closed at 2 p.m. Pacific on Saturday. When Circa Sports counted, it had 30,884 entries across three pro football contests and a combined prize pool of $38.7 million — the largest in the history of legal American betting contests, and about $8.7 million past the guarantee the company published on August 3.

Now do the arithmetic Circa does not do for you.

Survivor costs $1,000 an entry and drew 25,017 of them: $25,017,000 collected, $25 million posted as the prize pool. Grandissimo costs $100,000 and drew 77: $7.7 million collected, $7.7 million posted. Subtract those two fields from the total and 5,790 entries are left for Circa Million VIII, also $1,000 apiece, which brings in $5,790,000 against a $6 million guarantee.

Circa is short. Not by much — roughly $210,000, the gap between what the Million took in and what Circa promised to pay out — but short, deliberately, on a guarantee it invented and could have set lower. The house's cut of $38.5 million in entry fees is a negative number.

That is not a loss leader in the ordinary sense. A loss leader is priced below cost. This is priced below zero and then rounded further down.

The sentence on Circa's own website that explains it

"All contest signups must be completed in person at one of our Nevada locations."

Thirty thousand eight hundred and eighty-four entries. Every one of them required a human being to stand at a counter in downtown Las Vegas with a photo ID. Out-of-state entrants can use Nevada proxy services to submit weekly picks, but the act of entering is physical, and the deadline is a Saturday afternoon in September.

Read that as a customer-acquisition line item and the numbers reorganise themselves. On August 24, Circa had 12,827 entries. By Saturday it had 30,884. Roughly eighteen thousand entries — call it several thousand distinct people, given the ten-entry cap and the rebuy rule — were sold in the final nineteen days, into a building that also contains a hotel, a casino floor, three sportsbook counters and a sports bar with a screen the size of a tennis court.

Circa does not have to disclose what a September weekend of that traffic is worth. Nobody does. But the alternative model is public and audited, and it is grim.

What the other model costs

DraftKings reported second-quarter 2026 revenue of $1.443 billion, down 5% year over year. Monthly unique payers rose 9% to 3.6 million. Average revenue per payer fell 13% to $132. Adjusted EBITDA was $115 million against a planned 10% increase in acquisition spend, and management's own explanation for the revenue decline named "increased promotional reinvestment associated with new customer acquisition."

That is the promo treadmill described in its own accounting language: more customers, each worth less, bought with money that reduces the revenue line they were bought to raise. It works — DraftKings is enormous and growing its payer base — but the unit economics deteriorate as the field expands, because a bonus bet is a discount and a discount is permanent once the market expects it.

Circa's contests do the opposite thing. The prize pool is funded entirely by the participants. Circa's marginal cost per entrant is a clerk's time, a printed ticket and whatever the risk room spends grading picks. The company's advertising is the size of the number, and the number is generated by the customers.

There is a second-order effect that is easy to miss. Because the pool is a pass-through, Circa has no incentive to cap it, and because it has no incentive to cap it, the guarantee can be set aggressively enough to be newsworthy. The $20 million Survivor guarantee was a marketing claim that the customers then made true — and overshot by $5 million. DraftKings cannot run that play. A national book's promotional guarantee is an expense; Circa's is a forecast.

The strongest case against reading it this way

Here is the objection, and it is a real one.

Circa is not acquiring bettors. It is acquiring the wrong bettors, on purpose, and calling it a strategy. Circa's entire public identity is that it takes all comers and does not limit winners. The population that will fly to Las Vegas, hand over $1,000 and grind twenty weeks of survivor picks is the sharpest, most price-sensitive, least profitable cohort in the sport. DraftKings spends $132 a year acquiring somebody who parlays four legs on a Sunday and loses. Circa spends $210,000 acquiring thirty thousand people who will beat its closing numbers. On that reading, the contest is not a funnel; it is a very expensive way to fill a building with people who are bad for the book.

That objection is strong enough that we should concede most of it. Circa's contest entrants almost certainly hold worse than a national book's promo cohort. If Circa's model depended on those 30,884 people losing at market rates, it would be a poor model.

It does not depend on that. Circa operates in six states — Nevada, Colorado, Illinois, Iowa, Kentucky and Missouri — with an Arizona licence making seven, against DraftKings' twenty-plus. It is not running a national customer-count race and cannot win one. What it is running is a Las Vegas property whose sportsbook is the reason people choose that property, and the contests are the mechanism that puts a date on the calendar. A hotel night, a casino floor and a September Saturday's foot traffic are the revenue. The pick 'em is the reason for the trip.

The falsifiable version: if this is acquisition rather than theatre, Circa's entry counts should keep compounding faster than its state footprint grows, and its guarantees should keep being overshot rather than missed. Survivor went from 18,718 entries to 25,017 in one year, a 34% increase, while the state count did not move. Grandissimo went from 69 entries to 77 at $100,000 a seat. Both guarantees were beaten. Watch next August: if Circa raises the Survivor guarantee again and the field does not clear it, the flywheel has stalled and this reading was wrong.

What we do not know

We do not know Circa's handle, its hold, or what fraction of contest entrants bet anything else while they were in the building. Circa is privately held and discloses none of it. We do not know how many of the 30,884 entries are distinct people rather than the same syndicate maxing ten and rebuying ten. We do not know whether the $25 million Survivor figure is exact or rounded from $25,017,000, which is why the $210,000 shortfall above is stated as roughly.

What is not in doubt is the shape. Derek Stevens, crediting his staff on Saturday, thanked "the people in the risk room, the customer support, the writers" — the three groups whose labour is the actual cost of a contest that charges no rake. Last season's five surviving entries split $18,718,000, $3,743,600 each, and Circa kept nothing from that either.

A book in six states just posted the biggest prize pool in the industry by declining to take a cut of it. The books in twenty-five states are explaining to investors why their revenue per customer fell 13%. Only one of those is a pricing decision the other one could copy, and it is not the expensive one.