A 60-session review of hand-history data from low- and mid-stakes no-limit hold'em tables found that flagged "rabbit-hunting" behavior — a player asking the dealer to expose what would have come next after folding — fell 12% at tables where floor staff had been briefed to log and verbalize the request. The finding, drawn from 41,600 logged requests across 18 rooms between January and June 2025, suggests that the act of reporting, rather than penalizing, is doing most of the work. The 12% figure is modest, but it is the first session-indexed number of its kind circulated among US card-room supervisors, and it complicates the assumption that rabbit-hunting is either harmless theater or an integrity problem requiring rule changes.
What Rabbit-Hunting Actually Is, and Why Rooms Track It
Rabbit-hunting is the practice of asking the dealer to run out the board after a hand ends without showdown, usually after a player folds and wants to know whether the fold was correct. It is not cheating. It is not collusion. It is, depending on whom you ask, either a courtesy the dealer extends to keep the table happy or a small leak in the game's information economy.
The case against it is procedural. Exposing undealt cards at a live table creates a record — sometimes photographed, sometimes relayed to a neighbor — of cards that were never in play. In a room running a bad-beat jackpot, that record has no cash value, but in a room running promotions tied to specific board textures, it can. More practically, rabbit-hunting slows the game: a dealer who runs out three cards adds roughly 8 to 12 seconds to a hand, and at a table averaging 28 hands per hour, that compounds.
The case for it is social. Players who fold a big draw want to know. Denying them creates friction with the dealer, who did not write the rule. Most US rooms have settled on an informal middle ground: the dealer runs it out if the table is not busy, and declines if the floor is watching. That inconsistency is exactly what the session-60 data tried to measure.
The Session-60 Dataset
The review covered 60 sessions per room, with sessions defined as four-hour blocks on Friday and Saturday evenings, when request volume peaks. Observers logged every rabbit-hunt request, whether it was granted, and whether the requesting player re-entered the table within the same session after busting.
The re-entry figure is the one that moved. At control tables — rooms where staff continued business as usual — 31.4% of players who requested a rabbit-hunt and subsequently busted re-entered within the same session. At intervention tables, where dealers were instructed to say a short scripted line ("I'll log that request for the floor") before running out the board, the re-entry rate fell to 27.6%. That is the 12% relative decline in the headline.
The absolute numbers matter more than the percentage. A 3.8-point drop in re-entry, applied to a table with 9 seats and an average of 2.1 bust-outs per session, is roughly one fewer re-entry every 12 to 13 sessions per table. That is not a transformation. It is a detectable signal.
Why Re-Entry, and Not Revenue
The choice of re-entry as the dependent variable was deliberate. Revenue per session is noisy — a single whale distorts a table's numbers for a month. Re-entry is a cleaner proxy for the behavior rooms actually care about: the player who busts, feels a small sting, and sits back down to chase. Rabbit-hunting sits adjacent to that sting. The player who folds a flush draw, watches the dealer turn the flush card, and then busts 20 minutes later is a player with a live grievance. Logging the request appears to interrupt the loop between grievance and re-buy.
The mechanism is not fully understood. The working hypothesis among the supervisors who reviewed the data is that verbalizing the request — making it a transaction with the floor rather than a private moment with the dealer — adds just enough friction to break the automaticity of the re-buy. It is the same logic behind deposit-limit prompts that ask a player to confirm a number rather than click through it.
The Counterargument: You Are Measuring the Wrong Thing
Not everyone who reviewed the data accepts the interpretation. Two objections recur.
The first is selection. Intervention tables were not randomized. Rooms volunteered, and rooms that volunteer for a rabbit-hunting study are plausibly rooms already concerned about table tempo or player behavior. The 12% decline may reflect a room-level shift in attention — floor staff walking the section more often because they were briefed — rather than anything specific to rabbit-hunting.
The second is displacement. A player who does not re-enter the same table may simply re-enter a different one, or a different room, or an online site. The dataset tracked same-session, same-room re-entries only. If the 12% decline is entirely displacement, the operational value is close to zero, and the intervention is just moving players around the floor.
Both objections are fair, and the review's authors concede them in the working paper's limitations section. The honest read is that the 12% is a hypothesis-generating result, not a policy recommendation.
What Rooms Are Doing With It
Despite the caveats, several rooms have adopted a lighter version of the intervention: no script, no log, just a standing instruction that dealers pause for two seconds before running out a board. The pause is the whole intervention. Early anecdotal reports from three rooms suggest request volume drops when the pause is consistent, though no session-indexed data has been published.
The more interesting development is on the player side. A handful of regulars at mid-stakes tables have started asking dealers not to rabbit-hunt on their behalf, citing tempo. That is a cultural shift, not a regulatory one, and it is the kind of change that tends to outlast floor directives.
For US rooms operating under state compacts that say nothing about undealt-card exposure, the practical guidance is narrow: rabbit-hunting is not a rules problem, so do not write a rule. If a room wants to reduce re-entry churn, the data suggests that making the request visible — to the floor, to the player, to the table — is a cheaper lever than changing the game. Whether that visibility is a genuine harm-reduction tool or just a well-dressed placebo is the open question the next round of session data needs to answer.