What does Community’s winning-session percentage actually measure?
Read Toolchest’s all-time Community summaries without confusing winning sessions, recorded plays, total profit or your next bet’s probability.
A strategy with a high Winning sessions percentage has a particular record among the completed sessions shared with Community. That percentage does not tell you the chance of winning your next spin, hand or roll. It also does not tell you whether the combined sessions made money.
This guide describes the simplified Community interface in Tink’s Toolchest 2.4.1 on Apple platforms, publicly released November 12. Android availability of this version has not been independently confirmed; check your installed version before expecting the same layout. Earlier releases can show different Community controls.
Start with the scope above the numbers
Open Community and choose a game or strategy from the leaderboards. A game summary says All strategies · All time; a strategy summary identifies its game and says All time. Those labels matter: a game total combines the shared sessions in that game, whereas a strategy detail narrows the selection.
The summary is not limited to the first visible list of recent sessions. The app requests the available pages to calculate the all-time summary. Recent Sessions is a way to inspect individual records, not a label that changes the summary into today’s results. Pull to refresh when checking for changes; a connection error is not evidence that there are no sessions.
These are shared records, not a census of every person who played a casino game. Sharing is optional, and records can be corrected or removed. Different people can also choose different session lengths and stopping points. Keep that selection in mind before comparing two strategies.
Understand the denominator
For these summaries, a session with recorded net profit above zero is a Win, below zero is a Loss, and zero is Even. Winning-session percentage divides winning sessions by all sessions in the selected scope. Even sessions remain in the denominator.
Here is a hypothetical example, not a real Community result:
| Recorded outcome | Sessions |
|---|---|
| Win | 6 |
| Loss | 3 |
| Even | 1 |
| Total | 10 |
The winning-session figure is 6 ÷ 10 = 60%, not 6 ÷ 9. Wins and losses alone add to nine because the even session is neither. Read the total before treating a gap between Wins plus Losses and Sessions as an error.
A winning majority can still lose money
Suppose those six winning sessions each gained $10, the three losing sessions each lost $30, and the even session broke even. The combined result is $60 − $90 = a $30 loss, despite 60% winning sessions. Those invented amounts illustrate arithmetic; they are not claimed Community records or suggested stakes.
Version 2.4.1’s recent-session cards show Win, Loss or Even, rather than publishing dollar amounts on those cards. A leaderboard percentage therefore cannot answer “Which strategy earned the most?” Nor does it establish comparable risk: a short session and a long session each count as one session.
Recorded plays are a separate measure
Depending on the game, a detail summary labels recorded activity as spins, hands or rolls. The count adds play counts from records that contain that information. A session without a recorded play count is still a session; it does not establish that zero plays occurred. When none of the selected records contain play counts, the summary displays a dash.
Do not divide total wins by recorded plays and call the result a per-play win rate. The numerator counts sessions; the denominator would count individual plays, and some sessions may have no play count. They describe different units.
Use the numbers for recordkeeping, not a forecast
A useful sequence is: read the game or strategy label, check Sessions, then inspect Recent Sessions for dates, duration and outcomes. Compare like scopes and avoid treating a small set of shared results as proof that a strategy is superior.
Casino rules and payouts determine the mathematical house edge. A high historical session percentage does not remove that edge or guarantee future profit. For independent random outcomes, previous results do not make the next result due. Keep time and spending limits separate from what appears on a leaderboard.
The Community sharing guide explains which history becomes public and what happens when sharing is turned off. The observed spins guide explains why recording a result is different from placing a wager. Tink’s Toolchest can help organize records; those records are not predictions.
Sources and verification
- Tink’s Toolchest Apple listing and version information
- Gambling Commission: return to player, randomness and house edge
- Toolchest privacy and Community controls
Product labels, summary pagination, outcome classification and play-count handling were checked against the publicly released Apple 2.4.1 implementation. The hypothetical percentage and profit examples were independently recalculated. No real customer sessions were created, altered or published for this guide.