16 Common Poker Leaks: What Winning Players Do Differently
A leak is a mistake you repeat without noticing. The 16 leak rules PokerInk checks automatically, with the benchmark ranges winning players sit inside.
Aug 17, 2026 · 9 min read · PokerInk Team
A leak is a mistake you make repeatedly without noticing.
That last part is the problem: by definition, you cannot feel your own leaks. The hands where you limped a suited ace, called a river a little light, or gave up on a turn all felt reasonable at the time. Each one, individually, might even have been fine. A leak is not one bad decision. It is a frequency that is wrong across hundreds of decisions, and frequencies are invisible without data.
Below are the 16 leak rules PokerInk checks automatically against your logged hands, with the benchmark ranges the rules engine uses. The ranges describe disciplined, winning-style play in typical No-Limit Hold'em games; your exact game may shift them a little, but being far outside one across a real sample is worth your attention.
Preflop leaks
| Leak | What is measured | Benchmark |
|---|---|---|
| Playing too many hands | VPIP: how often you voluntarily put money in preflop | 21-27% |
| Playing too few hands | The same VPIP, from the other side | 21-27% |
| Not raising enough preflop | PFR: how often your preflop entry is a raise | 18-23% |
| Limping too often | The gap between the hands you play and the hands you raise (money in without a raise, limps and cold-calls alike) | 2-4 points |
Preflop is where most recreational players lose the most, because preflop mistakes repeat every orbit. A VPIP of 35% does not feel loose from the inside; it feels like "I had playable cards." The two numbers to internalize together are VPIP and PFR: winning players keep them close, because entering the pot with a raise keeps the initiative and the limp is almost always the worst of the three options.
Flop leaks
| Leak | What is measured | Benchmark |
|---|---|---|
| Continuation betting too often | Flop c-bet frequency after raising preflop | 55-70% |
| Not continuation betting enough | The same frequency, from below | 55-70% |
| Folding to continuation bets too often | How often you fold when facing a flop c-bet | 30-50% |
| Weak flop play | Average AI coaching score for your flop decisions | 6-8 out of 10 |
The c-bet band is wide because texture matters, but the failure modes at the edges are real: c-betting every flop turns your range face-up and burns money on boards that smash the caller, while c-betting rarely lets opponents realize equity for free. Folding to more than half of all c-bets makes you an ATM for anyone paying attention, since c-bets are often made with bluffs or weak hands.
Turn and river leaks
| Leak | What is measured | Benchmark |
|---|---|---|
| Giving up on the turn after a flop c-bet | How often you continue betting the turn after c-betting | 45-55% |
| Weak turn play | Average AI coaching score for turn decisions | 6-8 out of 10 |
| Missing river value bets | How often you bet or raise on the river | 25-40% |
| Calling too much on the river | How often you call river bets | 25-40% |
| Weak river play | Average AI coaching score for river decisions | 6-8 out of 10 |
The turn give-up is one of the most expensive patterns in live poker: fire the flop, get one call, shut down. Opponents who float the flop specifically to take the pot away on the turn feast on it. On the river, the two leaks mirror each other: missed value bets are money left on the table with the best hand, and light calls are money paid off with the worst one. Light river calls are the most expensive mistake per occurrence, because they happen when the pot is biggest.
Position and session leaks
| Leak | What is measured |
|---|---|
| Not adjusting play by position | Whether your play differs between early and late position; winning players play meaningfully more hands, and more aggressively, in position |
| Tilt detected across sessions | Decision scores swinging widely across multiple sessions |
| Performance drops late in sessions | Decision quality declining as a session gets long |
These three do not live in a single hand at all, which is why no amount of hand-by-hand review finds them. Playing the same range from under the gun and the button is invisible until someone lines your hands up by seat. Tilt and fatigue are patterns across sessions and hours, visible only in a journal that spans them.
How detection actually works
There is no magic here, and that is the point: every rule above is a frequency compared against a range, computed from hands you recorded. PokerInk shows each triggered leak with your current value, the benchmark, and a severity level, and the trainer then drills the specific skill behind it.
The one thing the rules cannot do is run on hands that were never recorded, which is why the whole system starts with a journal built for the table. If you want a head start on the numbers themselves, the free preflop range charts and outs and equity tables cover the preflop and math sides with no account needed.
Run the 16 leak rules on your own game
PokerInk is free to start on iPhone, Android, and the web. See pricing.
Frequently asked questions
What is a leak in poker?
A leak is a systematic, repeated mistake in your game: playing too many hands, limping instead of raising, giving up on turns, or calling rivers too light. Leaks are invisible from the inside, which is why they persist. They show up only in data collected across many sessions.
What are the most common poker leaks?
Preflop leaks come first for most players: playing too many hands (a VPIP far above the low-to-mid twenties) and open-limping instead of raising. Postflop, the classic pattern is continuation betting the flop and then giving up on the turn. Each of these is one of the 16 automated rules PokerInk checks.
How do I find my own poker leaks?
Record your hands, then compare your frequencies against benchmark ranges rather than judging individual hands. One bad call is noise; a river call rate far above the benchmark across dozens of hands is a leak. PokerInk runs 16 named rules across your journal automatically and shows your value against each benchmark.
