For online players (coming later)

Why Overshove shipped live first

Online poker has had capable trackers for two decades. Live cash never got the same treatment. Online, your database can separate results, sample size and review context automatically. Live, a basic buy-in and cash-out log leaves players guessing. That is where we built first.

The honest version of the roadmap: live cash is fully supported today, online is planned, and no date gets announced until it is real. But most of the system already translates. The review loop, the study log, leak tagging, bankroll rules and the variance engine care about decisions and samples, not about whether a person or a server dealt the cards. The missing piece is the import pipeline that parses online hand histories, and that is the piece still to come.

What online support will look like

Hand-history import is the core of it: point Overshove at a hand history folder and get sessions, stakes and results parsed with no manual entry. BB/100 will sit alongside BB/hr, because online samples are measured in hands, not hours, and the variance maths will be tuned to each. Per-stake and per-format splits follow from there.

What it will not be: a HUD. Overshove is a session tracker and study system, and that holds online. If you want live stats on opponents while you play, other tools already do that well. If you want to know whether your win rate is real and whether the leak you worked on last month is actually closing, that is the lane we stay in.

Beta players shape the build order. Overshove's core tracker is free, and roadmap priority follows the people already using it. If online import is the thing you need, join and say so.

A week on the hybrid grind

Plenty of online regs already play live once or twice a week, and that is where Overshove earns its place today. Tuesday night you sit 1/3 at the local room: buy-in, stake and venue logged in about ten seconds before the first hand. Mid-session you flag two hands during a dealer change, a few taps each, no typing at the table, no phone out mid-hand.

The cool-down happens in the car park: two minutes filling in the flagged hands while the action is fresh. Positions, sizings, the river decision you were not sure about. That habit is the live equivalent of the hand history you take for granted online, and it only works if capture is faster than the memory fades.

Wednesday, before the online grind starts, the live hands get the same treatment your database hands get: run the flagged river spot through your solver, tag the leak, log fifteen minutes in the study log against those exact hands. Friday, second live session, same loop. Sunday review: BB/hr net of rake and tips by venue, plus the variance read on your live sample, which behaves nothing like your online one. Which brings us to the numbers.

Same 40 hours, a tenth of the sample

The instinct that trips up online players moving into live is sample size. Forty hours of live full ring at 28 hands per hour is 1,120 hands. Forty hours four-tabling online at 75 hands per hour per table is 12,000 hands. Same seat time, more than ten times the sample.

Run that through the standard error maths, using 100 BB/100 as the standard deviation for both columns (live full ring typically runs 80 to 120 BB/100). Standard error is the SD divided by the square root of the sample in hundreds of hands. Live: 100 divided by the square root of 11.2 is roughly 30 BB/100. Online: 100 divided by the square root of 120 is roughly 9 BB/100. After the same month of volume, your live win rate estimate is over three times noisier.

That is why live tracking leans on things online tracking never needed: the true hourly after rake and tips, venue and game-selection data, session conditions, a stop-loss you actually honour. The sample will never bail you out live, so the surrounding data has to. When online import lands, both halves of your game sit in one place, with the maths handled correctly for each.