ROI in poker: how to calculate it, what’s a good ROI and when it means something

ROI in poker (return on investment) is your tournament profit divided by everything you paid to enter, shown as a percentage. Pay $550 to play fifty $10 + $1 tournaments, cash for $660 in total, and your profit is $110: an ROI of 20%. Put simply, every dollar you put into buy-ins came back as $1.20.

The formula takes a minute to learn. The harder questions are what to include in it, what a good number looks like, and how many tournaments you need before your ROI says anything about your skill. We answer the last one with a simulation of three formats, and the answer is more tournaments than most players expect.

How to calculate ROI

The formula:

ROI = (total winnings − total invested) ÷ total invested × 100%

Using the example above: ($660 − $550) ÷ $550 × 100% = 20%. A negative result means you lost money: winnings of $440 on the same $550 give an ROI of −20%.

ROI is a tournament stat. Each entry has a fixed cost, so it’s easy to measure what came back per dollar. Cash game players use a winrate in big blinds per 100 hands instead, covered in our guide to poker winrate.

What counts as your investment

Everything you paid to play goes on the cost side, not only the first buy-in:

• The fee (the “+$1” in $10 + $1). Leaving it out makes every result look about 10% better than it is. How fees work is in our guide to rake in poker.
• Re-entries and rebuys. If you fire three bullets into a $22 tournament, your investment is $66. See our guide to rebuys.
• Add-ons, the extra chips bought at the end of a rebuy period. Our guide to the add-on explains when they’re worth it.

What counts as winnings

Everything you won goes on the other side: prize money, bounties in knockout events, and tickets won in satellites, at the value of the tournament they get you into. If you play satellites and track only cash prizes, your satellite ROI looks like a loss and your main-event ROI looks too good.

Freerolls have no cost, so they don’t fit an ROI calculation; track them separately, as profit, as our freerolls guide suggests.

ROI across different buy-ins

If you play several buy-in levels, calculate your overall ROI from the totals, not by averaging the percentages. Here’s why, with one player’s year:

The simple average of +30% and −10% is +10%. The real figure is +1.4%, because most of the money went into the $55 games where the player lost. Overall ROI is weighted by dollars, not by tournaments. When you see a strong ROI from a player who mixes stakes, ask how much of their volume sits at the level with the best number.

Why 0% ROI already beats the field

Every tournament keeps its fee. If all the players in a $10 + $1 tournament were equally skilled, each would get back $10 for every $11 paid, on average. Their ROI would be −9.1%.

So an ROI of 0% isn’t break-even skill: it means you beat the field by enough to cover the fee. A player at +5% in $10 + $1 games is about 14 points better than an average entrant. The higher the fee as a share of the buy-in, the more skill it takes to reach zero, which is why cheap tournaments with 10% or higher fees are harder to beat than they look.

The same logic applies to lottery sit-and-gos. In GG’s Spin & Gold, the fee is about 7%, so equal players expect about −7%. Our Spin & Gold guide shows where that number comes from.

What’s a good ROI in poker?

There’s no single good number, because ROI depends on the format, the field size, the buy-in and the fee. Instead of a chart of targets, it’s more useful to know which way each factor pushes:

• Field size. Bigger fields pay a few players a lot, so a strong player can reach a much higher ROI in a 1,000-player tournament than in a 9-player sit-and-go, with much bigger swings on the way.
• Buy-in. Higher buy-ins draw stronger players, so the same player usually shows a lower ROI as they move up, while earning more per tournament.
• Speed. Turbo and hyper-turbo structures leave less room for skill, which compresses ROI toward the fee.
• Fee. A 10% fee costs 9.1 points of ROI before skill; a 5% fee costs about 4.8.

To give a sense of scale, here’s what our simulation needed to produce a 20% ROI in a 1,000-player field paying the top 15%: the player had to reach the money 19.4% of the time instead of 15%, and finish higher once there. That’s a clear edge, well below anything that would feel like “always cashing”. Small differences in finishing position make large differences in ROI.

Compare your ROI with your own past results in the same format and buy-in, not with numbers quoted for other games. Our guide to multi-table tournaments covers how formats differ.

How many tournaments before ROI means anything

Tournament results swing hard, and ROI swings with them. To measure how hard, we simulated three players with a known true ROI and checked what they’d see on their own results after different numbers of tournaments:

• a sit-and-go player in 9-player games paying 50/30/20%, true ROI 10%;
• an MTT player in 100-player fields, true ROI 20%;
• an MTT player in 1,000-player fields, true ROI 20%.

The model uses a 10% fee and pays the top 15% in MTTs, with first place taking about 25% of the pool in a 100-player field and 14% in a 1,000-player field. Skill shifts the player’s finishing positions upward; each result comes from thousands of simulated careers.

The table shows the share of these winning players who’d still be looking at a negative ROI:

After 100 tournaments in big fields, a genuine 20% winner is about as likely to be down as up. After 1,000, one in seven is still showing a loss. And the spread works both ways: after 500 big-field tournaments, the same player’s measured ROI can be anywhere from about −18% to +69%.

What it means in practice: in sit-and-gos, a few hundred games start to tell you something; in large-field MTTs, you need thousands. Until then, judge your play by the decisions you made rather than the bottom line. Our guide to variance covers the bankroll side of these swings.

ROI vs ITM, ABI and hourly rate

ROI is the headline, but three related numbers explain it.

ITM

ITM (in the money) is the share of tournaments where you cashed. It’s steadier than ROI and useful for spotting problems, but it isn’t profit. A player who cashes often but rarely goes deep can have a high ITM and a negative ROI, because min-cashes barely cover the buy-ins. Around the bubble, playing to cash instead of to win raises ITM and lowers ROI.

ABI

ABI (average buy-in) is the total invested divided by the number of tournaments. It tells you the stake behind the ROI: a 20% ROI with an ABI of $5 and one with an ABI of $100 are very different results.

Profit per tournament and per hour

A higher ROI doesn’t always mean more money. At 30% on $5.50 tournaments you make $1.65 per tournament; at 8% on $55 tournaments you make $4.40. If both take the same time, the lower ROI earns more than twice as much. Hourly rate adds speed and the number of tables: profit per tournament × tournaments per hour. When deciding whether to move up, compare dollars per hour, not ROI.

ROI vs winrate and EV

Cash games don’t have a fixed entry fee, so players there measure winrate: big blinds won per 100 hands. A winrate and an ROI can’t be compared directly, and neither can be turned into the other without knowing volume and stakes.

EV (expected value) is the average result of a decision, not of a session. Good decisions have positive EV; ROI is what they add up to over time, plus luck. In the short run, track whether your decisions were +EV; in the long run, ROI shows whether you were right about them. Our guide to EV in poker explains the difference with examples.

How to track your ROI

A spreadsheet is enough. Log one row per tournament with these columns: date, site, format, buy-in, fee, re-entries and add-ons, prize (including bounties and ticket value), finishing position and field size. Total invested, total won and ROI then take three formulas, and you can filter by any column.

Two other options fill the gaps. SharkScope collects tournament results across many sites and shows ROI, ABI and graphs, including for your own screen name. On GGPoker, the built-in PokerCraft tracker shows your tournament results without any outside software; our PokerCraft guide explains what it shows.

GGPoker, CoinPoker and WPT Global ban third-party trackers while you play. A spreadsheet you update after the session, or the site’s own tools, keep you within the rules.

Using ROI to find leaks

Your overall ROI hides patterns. Split it by one factor at a time and leaks start to show:

• Format: regular speed against turbo, bounty against freezeout. Many players are strong in one and losing in another. See our guide to freezeout poker.
• Buy-in level: a winner at $5 can be a loser at $50; the mixed buy-in table above shows how one level can hide another.
• Field size: big-field results need far more volume to mean anything, so judge them last.
• Time and session length: late-night results and the last hours of long sessions often trail the rest.

Look for segments with a few hundred tournaments behind them before you act on them. A split with 40 tournaments in it is noise, however striking the number.

ROI mistakes

Most errors with ROI make the number look better than it is, or lead to decisions the number can’t support.

Trusting a small sample. Fifty tournaments at +80% says almost nothing. The table above shows how often a real winner looks like a loser, and the reverse is just as common.

Forgetting re-entries and fees. Counting only the first buy-in in a re-entry tournament can turn a losing format into an apparent winner.

Comparing ROI across formats. A 10% ROI in 9-player sit-and-gos and a 10% ROI in 2,000-player MTTs describe very different edges.

Moving up after one big score. A single deep run can lift ROI by dozens of points. Move up when your bankroll and a long record support it, and treat a sudden drop afterwards as variance, not proof, which is the moment tilt tends to start.

ROI in poker FAQ