Vitya Largus: Why Online Poker Regulars Break Even for Months

Hi! I want to talk about rake, game selection, bankroll management, and other topics that matter to anyone starting out on a professional poker career.

To keep the discussion substantive, I brought together experts from the apps and the European rooms. The goal was to paint a broad picture to show what's happening in the industry, how to make the professional journey as painless as possible, and how to bring reality as close to expectations as we can.

So here's our set of hands pulled from various online poker rooms. I filtered the sample by VPIP and PFR, which left me with a global population of regulars we know nothing else about (except that, judging by their stats, they came to the tables to make money).

Their average win rate before rakeback was negative: -0.79 EV bb/100

The single most important figure is how much rake the average player pays, and it's striking: 9.6 bb/100. Rake varies by platform. The club apps charge more, while the European poker rooms charge less.

CoinPoker's line shows 4.2 bb/100, but unfortunately that number no longer holds: the room raised its rake a few months ago. I deliberately left it in the results to show the direction the industry is heading.

What we have, then, is a pool of NL50 regulars playing at a slight loss. The sample is clearly mixed: some grind purely for rakeback, others select their games carefully, and there are both experienced players and newcomers.

Now I'll hit the buttons and let the program run the numbers and compile them into a table. It's kept as general as possible. There's no goal of getting into the fine detail; the main thing is to set the direction of my thinking and paint the big picture.

On average, five regulars pay 9.6 bb each, or 48 bb/100 combined. Next we add an amateur. A lot depends on their VPIP: at a VPIP of 45 the player pays about 20 bb/100, while at a VPIP of 65-70 it's roughly 10 bb more.

An amateur like this pulls the regulars into more pots and opens up more action, so the total rake at the table rises as well.

Rather than getting bogged down in the finer points, let's look at some rough but realistic numbers. The table pays 73 bb/100, and even under that rake load it can turn a long-term profit without rakeback. I want to stress that we're talking about the average, faceless pro (although there are players who confidently beat the game before rakeback).

Now consider a situation where this population of regulars is up against an amateur with a VPIP of 0-44. On average the sample runs at -4.8 bb/100, and winning is only possible with good rakeback, because too much is lost to the rake. These figures won't necessarily match your own results: if you're a winning player, you'll do better.

The next tier is VPIP 45-64, and things look better here. The regulars start paying more rake, but their win rates climb too. The numbers are still very low (0.64 EV bb/100) but again, these are averages. A competent regular will post a higher win rate.

Moving on to VPIP 65-100. Sitting at a table with an amateur like this is almost always profitable. There are exceptions. An opponent with a VPIP of 70 who, rake included, loses only 40 bb/100. If that amateur simply halved their VPIP, they could actually beat the game. Conversely, some play a VPIP of 40 and still lose at -100 bb/100. But on average, against opponents like these (VPIP 65+), competent regulars are all but guaranteed a win rate of 11 bb/100.

It all comes down to the rake, which you have to factor in.

These figures are approximate rather than exact, but I'd still suggest keeping them in mind. It will help you properly plan your climb through the limits and the growth of your bankroll.

Let's talk about the expected value of the average regular. With an amateur whose VPIP is 0-44, expectation is roughly zero: the average win rate runs from -2 to +2 depending on skill. There's no profit before rakeback.

And yes, these amateurs are easy to find. Just sit down at any NL50 table and, more often than not, at least one such enthusiast will be there. This is what a zero-selection game looks like.

The key is that the share of your games played against these opponents doesn't cross a certain threshold. If half your games are against a single amateur with a VPIP under 44 and you don't have strong rakeback, you'll end up with the classic rake dump.

In the 45-64 VPIP sample you can expect 7-10 bb/100, depending on the rake. Those who play a lower rake and understand how to play against amateurs well will get 10 bb. Most of the field is fairly weak, though, so 7 bb is already a good result.

VPIP 65+. A good win rate here is 20 bb/100, and in exceptional cases strong players can reach 30 bb/100. Some players stay tight and cautious against other regulars but perform very well against amateurs and have phenomenal table selection.

Factoring in the rake at the lower stakes, your win rate here shouldn't fall below 15 bb/100.

Here's a table of the threshold values for each sample. If your win rate is worse than the figures above even under a very high rake load, then something is wrong with your strategy against amateurs, and with your game in general.

Now to selection. If a player is set on prioritizing it rather than grinding everything in sight, then at NL50 the share of games against opponents with a VPIP of 65+ should ideally reach 25%. The 45-64 VPIP range accounts for about 50%, and the 0-44 range for the remaining 25%.

In practice it's messier: you can't put in a real volume and hold your selection at the same time, especially playing in a single room. More often than not the 65+ VPIP share "bleeds over" into the lower brackets, and overall win rates drop.

If you select at NL50 at the rates above and play well, you can expect 10 bb/100. But at those numbers it makes sense to move up in stakes. There are fewer amateurs, but the rake is also lower. I'm not trying to talk you into or out of selecting; just keep it in mind.

Let's take the average monthly volume of a player with normal selection who doesn't take many days off and plays six tables. He's not a pure grinder, but he's fairly dedicated. Into the variance calculator we'll enter a standard deviation of 81.7 bb/100 (based on the H2N4 results), a sample of 40,000 hands, and a win rate of 2.5 bb/100.

Click "Calculate."

The numbers show that without good selection there's a 70% chance your monthly result lands somewhere between losing six and winning twenty-six buy-ins. The probability of finishing 40,000 hands in the red is 27%, meaning we'd end up in the red more often than one month in four!

Let's look more closely at the factors you need to weigh so these months don't become a problem: selection, rakeback, your own skill level, and bankroll depth.

If we're going to sacrifice selection, we should at least pay attention to position. Having direct position on an opponent produces results 2-3 times better than being one seat out of position (-1) relative to the amateur.

It's important to keep rakeback in mind. When you play with good selection you'll have fewer tables, but a consistently high win rate. If you go the rakeback route instead, you need to work out the effective rake in advance and make sure it's worth the effort.

Back to the calculator: let's enter 7.5 bb/100. This is the win rate of a player with good selection who plays 25% of their volume against amateurs with a VPIP of 65+.

Now let's shorten the sample to 25,000 hands. The win rate has tripled compared with the first case, and the probability of finishing the month in the red is only 7%.

A high win rate delivers stable results. The swings in the graph become far smaller than they are for a player focused on rakeback.

That said, neither approach is inherently better or worse because different styles suit different people. If your selection is poor, the swings will be frequent: some months a year you'll lose, and a few others you'll break even. With good selection you'll only break even for a month or two, with maybe a slight loss, but otherwise life becomes much clearer and more enjoyable.

None of this comes for free, though: you'll have to put in extra hours scouting the lobby, join more rooms, and so on. Sometimes that's more tilting than simply sitting down and playing in peace. Some players opt for Zoom and, thanks to rakeback, come out ahead even when they'd otherwise be playing at a loss. The important thing is to be honest with yourself and do the simple math on what you can realistically expect.

Another factor is your relative skill level, meaning, how confident you feel in a tough field. If you often don't understand what's happening at the tables and aren't sure what to do, don't overrate yourself while underrating your opponents. Better to underrate yourself a little and approach your win rate with cautious expectations.

Finally, you have to account for bankroll depth. Ask yourself: are you prepared to play 160,000 hands at breakeven or a slight loss before rakeback? Would one or two months without profit force you to drop down in stakes because of withdrawals? The math here is simple. Compare your real win rate against the database, look at your results against amateurs, and take the largest hand sample you have. That's your benchmark.

Don't fixate on your most recent hot stretch. That's a dangerous path, and it's easy to lose your way. We never know our true win rate in the present moment because it's always a reflection of the past. But from the past we can draw assumptions that tend to hold up.

Better to underrate yourself than to end up in a spot where, after a withdrawal, you don't have enough bankroll to put in the volume you need at your limit (for example, to keep your rakeback status).

There are different approaches, but one thing always holds: everyone who sits down at the table has to beat the rake, whether directly or through rakeback. Ignore that and the journey only gets harder. Best to play it safe from the very start. Good luck!

Original YouTube Recording