Built for serious live cash players

The live-cash reality

Live cash gives you nothing for free. No hand histories, no automatic import, the same rooms and the same faces, and a sample that is always smaller than it feels. The practical problem is familiar: players often cannot tell running bad from playing bad, and a buy-in and cash-out log leaves out the costs and context needed to answer it.

Overshove is shaped around exactly that reality. Logging happens at the buy-in and takes seconds, so it survives contact with an actual session. Reads are filed by venue, because your edge lives in a specific pool, not an abstract one. And the variance layer works on the sample you actually have, not the one an online tracker assumes you can generate.

A week on the grind

Tuesday, 6pm, 2/5 at your regular room. You open a session at the buy-in: venue, stake and buy-in prefilled from last time, clock started. That is the whole at-the-table interaction. A rebuy is two taps. Otherwise the tool disappears while the cards are in the air, which is the point.

1am, car park, cool-down. Cash-out and tips go in, and you mark the two hands that mattered while they are fresh: the 200 BB three-bet pot where you folded the river against the tightest reg in the room, and a flatted spot you are less sure about. One line on state: last hour was C-game, left on time anyway. Ninety seconds, engine on, home.

Thursday repeats the loop at a different room, and the notebook already holds your reads on that pool, filed by venue. Sunday morning is the desk block: thirty minutes, the week's marked hands in one list. The river fold goes through your solver and comes back fine. The flat does not, so it gets a leak tag and a study log entry against it, which means in a month you can check whether it actually stopped appearing. Last stop is the dashboard: BB/hr by venue, the confidence band on your sample, and a sober look at whether the planned 5/10 shot is bankroll-approved or just a good week talking.

What 40 hours actually earned

Take a month most players would call solid: 40 hours at 2/5, up £1,800 at the cage. A minimal buy-in and cash-out log records £45 an hour and stops there. Add what the month actually cost, £240 in dealer and service tips at roughly £6 an hour plus £160 in travel and food, and the true result is £1,400, or £35 an hour. That is 7 BB/hr, not 9. Same cards, same decisions, and the gross number overstates the hourly by 29%.

The sample cuts the other way too. Forty hours at roughly 28 hands an hour is about 1,120 hands, which is closer to noise than signal. One month tells you almost nothing on its own; it only becomes information when it lands in a running sample alongside everything else you have logged.

Variance, quantified on your sample

Now the harder number. Say you have logged 300 hours at 2/5 and the sample shows 5 BB/hr, which at 28 hands an hour is about 18 BB/100. With a live full-ring standard deviation around 100 BB/100, those 300 hours are roughly 8,400 hands, and the 95% confidence interval on that win rate runs about plus or minus 21 BB/100. The band still straddles zero. That is not a flaw in the method, it is the maths of live volume, and it is why judging your decisions rather than your monthly results is the only honest option.

It is also where records are most likely to break. Logging often stops during a downswing, which is precisely when the record earns its keep. Overshove puts the band on the chart, shows whether the current stretch sits inside normal for your win rate and standard deviation, and keeps the sample growing through the exact period where a spreadsheet gets quietly abandoned.