ShowOn

Splitting the Losses Feed Into Wins Only Cuts Tilt Quits 19%

A six-week poker experiment found that logging losses separately from wins cut tilt quits by 19%, suggesting loss framing matters as much as size

Splitting the Losses Feed Into Wins Only Cuts Tilt Quits 19%

A small experiment run through a mid-stakes online poker room's hand-history tool found that players who routed half of every losing session into a separate "wins-only" ledger — tracking wins in one column, losses in another, and refusing to net them — quit 19% less often after a downswing than a control group that saw a single combined balance. The finding, drawn from 1,240 sessions logged over six weeks in early 2024, suggests the framing of a loss matters as much as its size when it comes to the tilt quit.

What the split actually changes

The intervention is almost embarrassingly simple. Instead of one running total, players kept two: a wins feed that only ever ticked upward, and a losses feed that sat apart. A $200 down day never subtracted from the visible win count. The psychological mechanism isn't mysterious — it's a form of mental accounting that keeps a bad hour from erasing the memory of a good one.

The control group saw the standard net figure every session. Their quit rate after a losing stretch of three or more buy-ins was 31%. The split-feed group came in at 25% — the source of the 19% relative reduction.

Why the effect showed up after losses, not before

Session length before the first loss was nearly identical between groups, within 4 minutes. The divergence appeared only in the 90 minutes following a drawdown. That timing matters: it points to the split feed acting as a brake on the post-loss spiral rather than as a general engagement booster.

The numbers behind the 19%

The raw quit counts were 96 of 310 control sessions versus 78 of 312 split-feed sessions that ended within 90 minutes of a three-buy-in drawdown. The 19% figure is the relative drop; in absolute terms it's about 6 percentage points. That's a real but modest effect, and the sample is small enough that a skeptic could reasonably ask for a replication before trusting it.

The study also didn't control for stakes precisely. Median buy-in was $22 in both arms, but the split-feed group skews slightly toward players with fewer than 500 lifetime hands logged — a group that may quit more for reasons unrelated to tilt.

Where this brushes against how the industry frames losses

Operators have long known that surfacing net losses prominently can accelerate churn, which is why some dashboards bury the running total. The split-feed approach goes further by structurally preventing losses from touching the wins view. Whether that's a genuine player-protection tool or a subtle way to keep people in seats longer is the uncomfortable question. A 19% reduction in tilt quits is good for the player who avoids chasing; it's also good for the room's session-length metrics.

Regulators in states like New Jersey and Pennsylvania have pushed for clearer loss disclosure, not less. A tool that hides net position from the player runs against that grain, even if the intent is to reduce impulsive quitting.

The open question

What the experiment can't tell us is whether the split-feed players were making better decisions or simply staying longer in a bad game. Quitting after a drawdown isn't always tilt — sometimes it's the smartest move at the table. The next useful test would track win rate, not just quit rate, among players who stay. If the 19% who stuck around kept bleeding, the intervention is a churn fix dressed as a tilt fix. If their results held steady, it's something more interesting. Until then, the number is worth noting and not much more.