Semi-professional is an awkward label. It sounds like a hedge, a way of describing someone who didn’t commit. In practice it describes something more specific: a player who takes poker seriously enough to track it properly and to expect it to be profitable, but who doesn’t need it to cover the rent.
We covered the reasoning behind that choice in how to be a professional poker player without losing at life. This piece is the operational side: how many hours, at what stakes, with what bankroll, and how to tell whether it is actually working.
Know your real win rate first
Before any of the structural decisions matter, you need to know whether you’re a winning player, and most people who believe they are haven’t established it to any useful standard.
Live poker produces roughly 25 to 30 hands per hour. Online, multi-tabling, you might see fifteen times that. This matters because win rates only become meaningful across sample sizes that live players accumulate very slowly. A live player putting in ten hours a week sees somewhere around 15,000 hands a year, which is well short of what is needed to tell a real edge from a run of good cards.
The practical response is to keep measuring but hold the conclusions loosely. Track every session: date, room, stakes, hours, result. After a year you’ll have something worth looking at, and more importantly you will have the habit. Players who don’t track reliably remember their results as better than they were.
Bankroll when poker isn’t the rent
Standard bankroll guidance assumes poker is your income and a downswing threatens your ability to keep playing. For a semi-pro with outside income, the calculation really does change, because you can reload from earnings and you’re never forced to move down at the worst moment.
That doesn’t mean bankroll rules stop applying. It means the number can be lower than the full-time equivalent while still being large enough that no single session affects how you play. The threshold that matters is psychological rather than mathematical: your bankroll is adequate when losing a buy-in doesn’t change your decisions for the rest of the session.
Two rules make this work in practice. Keep poker money in a separate account, so the balance is a real number instead of a vague sense of how you’re doing. And decide your stakes at the start of the month, away from the table, where the decision gets made by whatever just happened.
Choosing when to play
This is the single largest advantage of not playing full-time, and it’s routinely wasted.
A full-time player has to take the games available during their working hours. You don’t. Weekend evenings in Texas club games, late nights, and the hours after a big local tournament busts are consistently softer than a Tuesday afternoon. If you can only play ten hours a week, those ten hours can be drawn almost entirely from the best available windows.
The corollary is that you should be willing to leave. Driving to a card room and finding a bad lineup isn’t a reason to sit anyway. The drive is a sunk cost, and playing a game you don’t want because you already made the trip is how a structural edge gets given back. Our guide to the best poker rooms in Texas covers which rooms tend to run which games.
Session length and the concentration ceiling
Most players have a window inside which they play their best game, and it is shorter than they think. For a lot of people it’s somewhere between three and five hours, after which decisions get faster, looser and more automatic.
Find your number by tracking results by session length. If your long sessions are noticeably worse, your concentration ceiling is showing up in the data. Then set a session length slightly inside it and treat it as fixed, with one exception: if the game is truly exceptional, staying is correct. Just be honest about whether the game is good or whether you are stuck and want to get even.
Study time is where part-time players fall behind
The real cost of playing part-time is improvement rate. Someone studying solver outputs for twenty hours a week will pull ahead of you technically, and there is no clever structure that avoids this.
What you can do is spend limited study time well. Reviewing your own marked hands from recent sessions beats working through material unrelated to spots you actually face. Studying the specific formats you play beats general theory. And a small amount of consistent weekly study beats occasional long sessions, for the same reason spaced repetition beats cramming.
If you use training software, pick one tool and use it properly instead of subscribing to three. We have a discount on GTO Wizard if you want a solver-based option.
Taxes and record-keeping
Poker winnings are taxable income in the United States, including from live club games where nobody hands you a form. The session log that tells you whether you’re winning is the same log that supports your filing position, which is a reasonable second argument for keeping one.
The treatment of poker income differs depending on whether it is a hobby or a business, and that distinction has real consequences for deductions. Talk to an accountant who has dealt with gambling income before instead of guessing from forum posts.
Knowing whether it is working
Set a review point a year out and decide in advance what you’re measuring. Hourly rate is the obvious metric, but it isn’t the only one that matters. Whether you still enjoy playing, whether sessions are costing you sleep, and whether the money is going anywhere other than back into play are all reasonable things to check.
If poker is profitable and you still want to play next month, the structure is working. If it is profitable and you’ve started to dread it, take that seriously, because the entire advantage of not needing the income is that you are free to act on exactly that signal.