Example Leak Audit - NL5 (Rush & Cash + Regular)

EXAMPLE REPORT

50,348 hands at $0.02/$0.05 on GG Poker — Rush & Cash and regular 6-max ring, combined.

Every number in this report is measured against the alternative you had, not against zero. Every rate carries a reference. Nothing in here is here to flatter you.

This is an example of what a Leak Audit looks like, published so you can see the work before you commission it. It is not based on a client's data — no client's hands, results or details appear anywhere in this report, and no client report is ever published. The database behind it is a real one and the analysis is the real analysis: the numbers, the findings and the conclusions are exactly as they came out. Hand identifiers and dates have been removed.

---

50,348 hands at $0.02/$0.05 on GG Poker — 40,141 hands of Rush & Cash and 10,207 of regular 6-max ring, including 1,015 hands on full-ring Ante tables.

Result: -$228.45, which is -9.07 bb/100 — big blinds lost per 100 hands, the standard way results are compared across stakes.

Straight answer up front. You're losing at this stake, over enough hands to be sure of it — the give-or-take range runs from -15.3 to -2.8 bb/100 and doesn't touch zero. And measured properly against equity, almost none of it is bad luck: about 1 bb/100, which is a rounding error next to the hole. The single biggest cause is that you fold your big blind 84% of the time and your small blind 86%, while the hands you do play from those seats are among the most profitable in your database. Fixing everything on this list closes roughly half to three-quarters of the gap to break-even. That's the honest arithmetic, and the rest of this document is the evidence for it.

How to read this report. Wherever you see "usually around X%," that's a widely-used rule of thumb for a solid 6-max player — not a number measured from this pool. Treat it as a ballpark, not a target. And every leak cost is measured against what the alternative would have done, never against zero: calling a 3-bet and losing 1bb isn't a 1bb mistake if folding would have lost 2bb.

Where the money actually went

Every hand falls into exactly one row, and the rows add up to your result:

The largest single line is the blinds you post and fold. That's structural — every player pays it, and you earn it back by winning pots — but it belongs at the top, because a red winrate invites the assumption that it's all mistakes, and it isn't.

Why the leak list below doesn't add up to -9.07. The leaks are counterfactuals, not slices of your loss. Money you never collected — the steals you didn't attempt, the blinds you folded — never appears in a results table at all. Money you did lose on a call only counts as a mistake for the part beyond what folding would have cost. This table says where the chips went; the next one says what to change. Different questions, and they don't reconcile.

Results by seat

(The rare full-ring seats — UTG1, MP1 and HJ, 50 to 107 hands each — are omitted; their ranges are far too wide to read.) The usual shape is button best, descending to under the gun, both blinds negative with the big blind worst. Yours follows that except the button looking slightly worse than the cutoff. The usual cause is opening too tight on the button — but that isn't it here; you open slightly more from the button than the cutoff. Both ranges include zero, so at this sample size it's most likely nothing.

The red line and the blue line

The red line is not your problem, and it would be easy to tell you it was. The right yardstick is what you're forced to put in: you post 24.1 bb/100 in blinds and antes. Recovering all but 7.5 of that without ever showing a hand is fine. Chasing it — bluffing more, barrelling more — would most likely make things worse rather than better.

The blue line is the one that should bother you. You play only 15.7% of hands, which is very tight. A range that tight should be winning showdowns comfortably, because you're only there with strong hands. Instead you're losing at showdown. That's the thread tying several findings below together: you arrive at showdown slightly too often, with hands slightly too weak.

What we count as a cooler, a bad beat, and variance

These words get used loosely at the table. Here they have fixed meanings, and the rules are set deliberately strict — every hand wrongly filed as "bad luck" is a hand you're told you could do nothing about, when it may be exactly the one you need to play differently.

Bad beat (suckout). Once the flop came down you had a made hand — a pair or better — and your opponent did not: they were drawing, held nothing, or held a pocket pair below the board that needed to hit a set. They got there anyway.

Cooler. Once the flop came down, both of you had a genuine hand — an overpair, a set, two pair using both of your own cards, or better — and one of you held the bigger one. Nobody misplayed it. KK into AA is a cooler.

Not a cooler, not a bad beat — just second best:

• A pocket pair below the board that never improved. QQ on a king-high board losing to a king is having the second-best hand.
• "Two pair" made with one of your cards plus a pair already on the board. A-Q on A-K-4-4 reads as "two pair, aces and fours." The fours belong to everyone. You have one pair — and the paired board is exactly what lets your opponent hold a full house. (A genuine pocket pair on a paired board, AA on 9-K-3-7-K, is different and does count.)
• Coinflips. A-K against QQ all-in preflop is a flip, whichever way it lands.
• Money that went in wrong preflop. If you called an all-in that was a 5-bet or higher holding QQ or worse — or a 4-bet-or-higher shove holding JJ or worse — that is not a cooler whatever it ran into. At this stake a 5-bet shove is aces or kings almost without exception. The hand was decided by the call.

Variance is only ever measured as your actual result minus what your equity said to expect, across all such spots, wins included. A list of losses is not a measure of luck.

What's counted: hands that reached a showdown heads-up with a flop dealt and both hands shown. Multiway showdowns and the rare no-flop all-in aren't classified, so the true totals are somewhat larger. The bias is deliberately toward undercounting.

How much of this was actually bad luck?

About 1 bb/100. Not half, and not a quarter.

The tables below list every bad beat and cooler in both directions. They're real hands and worth reading. But adding them up and calling the total "luck" is wrong, and here's the trap: you're a tight player, so when you reach a showdown you're usually the one holding the made hand and your opponent is usually the one drawing. In this sample you were the made hand against a draw 352 times and the drawer 225 times — a ratio of about 1.6 to 1. You lost 98 of those 352 as the favourite (28%) and won 69 of the 225 as the underdog (31%). Both are almost exactly what the odds say should happen.

So the reason "bad beats taken" outnumbers "bad beats given" isn't that the deck ran against you. It's that you were on the favourite side more often — and favourites lose about three times in ten. The 254 hands where your made hand simply held aren't in these tables; they're just wins. That's what makes a list of the losses look like misfortune.

Measured properly — actual result against what equity said to expect, across all 577 of these spots, wins included — you ran about 1 bb/100 unlucky. Slightly unlucky. Not a story.

What that means for the rest of this report: of your -9.07 bb/100, roughly one point is luck. The other eight are decisions and structure. That's a harder message than a big variance number would have sent, and it's the true one.

The hand lists below are kept for a different reason: they show where you're getting money in as a favourite and losing it. How often you end up in those spots is partly a choice.

The worst coolers you took

The best coolers that went your way

The worst bad beats you took

The best bad beats you gave

Money that went in wrong preflop — not coolers

Four hands that look like coolers are excluded, because you called a 5-bet-or-higher shove with a hand that range has crushed: QQ into AA (Hand 17, -$5.86), QQ into KK (Hand 18, -$5.00), TT into KK (Hand 19, -$5.00) and 99 into QQ (Hand 20, -$5.09). Together -$21, about -0.83 bb/100. A 5-bet shove at NL5 is aces or kings. Queens and below fold — and the deeper fix is not to 4-bet a medium pair into a 3-bet in the first place, which is what creates the spot.

What to change

Be clear about what this adds up to. You're losing at -9.07 bb/100, with about one point of that being luck. The claimable edge above is roughly 3 to 8 bb/100, and the wide range is honest: it depends entirely on how well the hands you add to your blind-defence and stealing ranges perform. The pessimistic end closes about half the gap; the optimistic end closes it and a little more. Either way, the blinds are worth more than everything else on the list combined, and that's where to start.

The leaks, one at a time

1. You fold your blinds more than four times in five — by far the biggest leak

What's happening: facing a single raise you defend the big blind about 21-23%, and across all big blinds including limped and multiway pots it's 16%. In the small blind it's 14.5%. Both are far too tight; a reasonable target in each is somewhere around 30%.

Here's the number that should settle it: the hands you choose to play from these seats make money — and the small blind figure, +107 bb/100, is the highest of any seat in your database. You are already good at this. You're doing it one hand in six.

Why it matters: the blinds are the two seats where you already have money in the pot before you act. That's a discount nobody else at the table is getting — you only have to put in the rest, and the pot is already laying you a price. Folding throws the discount away. It shows in the seat results: the big blind cost you -52.2 bb/100 across this sample, and folding every hand would have cost -100. You're recovering roughly half of what's recoverable, in the seat where the most is available.

What it's worth: taking each to 30% is +1.1 to +3.3 bb/100 in the big blind and +1.4 to +4.1 in the small blind. The low ends assume the added hands earn only a third of what your current defends earn — deliberately pessimistic — and even those are the two largest items here.

But widen with judgment, not indiscriminately. Defend more against late-position opens — the button and small blind open wide, so your equity against them is far better than against an under-the-gun raise, where your current tightness is closer to correct. Prefer suited and connected hands over offsuit broadways, because out of position you want hands that flop draws and make easy decisions rather than hands that flop one pair and a guess. From the small blind, lean toward raising rather than calling — you'll be out of position for the rest of the hand, and taking it down immediately is worth a lot. And expect your red line to get slightly worse as you do this: you'll see more flops and fold more of them. That's the correct trade, not a warning sign.

2. You're not stealing enough from late position — +0.5 to +1.0 bb/100 available

Why it matters: when everyone in front of you has folded, the pot is essentially orphaned — only the blinds remain, and they'll be out of position against you for the rest of the hand. It's the cheapest, safest money in the game. Every one of these opens you're already making shows a profit, which is the tell that there's room: if the current, tighter range makes money, the next tier of hands very likely does too.

What it's worth: bringing each seat up to the reference rate at your current profit per open would have been about +$24. But widening means adding weaker hands, so your profit per open will drop as you do it — a more honest range is +0.5 to +1.0 bb/100, or $12 to $24. That dip is fine. The goal is more total profit, not a higher average.

3. You check too passively on the flop — real, but not costable

What's happening: when you check and get bet into, you raise back 4.5% of the time. Usually around 8-12% is reasonable.

Why it matters: a check that never turns into a raise is fully transparent. Opponents learn that betting into your check is free — the worst outcome for them is a call, and often you fold. That lets them value-bet thinner and bluff more freely. A believable check-raise range — some strong hands plus a few good semi-bluffs — changes the price of betting into you.

What it's worth: we can't honestly put a number on it. Your 46 flop check-raises averaged +$0.24 and your check-calls -$0.05, but that comparison is meaningless — you check-raise your best hands, so of course they do better. The value of check-raising more isn't in those hands; it's in the bets your opponents stop making, and that never appears in a hand history. Real leak, unmeasurable from this data.

But don't overcorrect into check-raising only monsters. A range that's exclusively two pair and better is nearly as readable as never raising at all, just in the other direction. You want a few hands in there that can raise without a made hand — which is what the next section is about.

4. Semi-bluffing with draws — passive, but at least not readable

Continuing with a draw is only half of playing it. The other half is whether the draw ever gets played aggressively.

Why it matters: a flush draw or open-ender is the ideal hand to raise with, because it has two ways to win — they fold now, or you hit later. Only ever calling gives up the first one entirely.

But here's the saving grace, and it's genuine: your aggression with draws is identical to your aggression with made hands — 10% raise both, 7% check-raise both. So a raise from you doesn't announce a made hand. Opponents can't read you; you're passive, not transparent. That's a smaller and much more fixable problem, and draws are the cheapest place to start fixing it: raise a draw, get called, and you still have all your outs.

5. Calling 5-bet shoves with QQ or worse — -0.83 bb/100

Four hands, all lost, listed above. A 5-bet all-in at this stake is aces or kings. But the real fix is one step earlier: don't 4-bet TT, JJ or QQ into a 3-bet, because that's what builds a pot big enough to price you into calling off with the worst of it. Take those hands to the flop instead, or fold them.

6. Calling 3-bets with suited aces — -0.34 bb/100, watch rather than act

When you open and get 3-bet, you fold 36% of the time (322 of 885). Usually around 55-65%. At first glance that looks like a serious leak — but the alternative to calling isn't breaking even, it's folding, which costs you the open (about 2.2bb). Measured against that, your 447 calls came out +567bb ahead, +1.27bb each. Calling has been the better choice overall. Broken down by holding:

Only suited aces trail folding: about -$8. But that's 33 hands and a few big pots, so it doesn't survive the outlier test — a direction, not a finding. It's plausible on its face: a small suited ace against a 3-bet is mostly hoping to flop a flush draw while being dominated by the A-K/A-Q-heavy ranges people 3-bet with. Lean toward folding suited aces below A-T; leave the rest of the calling range alone, because it's working.

7. Double gutshots — a counting habit worth fixing permanently

A double gutshot — two separate inside draws at once, say 9-8 on K-J-7 where both a ten and a six make your straight — has eight outs, exactly as many as an open-ender. You fold open-enders 8% of the time and double gutshots 38%. You're treating them like gutshots, which you fold 47% of the time and correctly so, because those have four. Your hand is twice as strong as you're playing it.

But this is 16 hands, so the money involved is small. It's here because it's a cheap, permanent fix: when you count outs, count both gaps.

What we checked and it's fine

• Your flop folding is disciplined. Facing a c-bet you fold 54.7% overall (usually 40-55%), and the shape is right: nothing folds 98% of the time, one overcard 95%, backdoor draws 83-91%, while flush draws continue 97% and open-enders 92%. If folding feels excessive, the fix is upstream in which hands you play preflop, not in calling more on the flop.
• Paying off big river bets — checked directly, because it's a common leak. Across 34 calls of a river bet of 15bb or more holding one pair or worse, you won 10, and against folding you came out +$8.61 ahead. Not a leak here.
• Cold-calling raises — button 10.4%, cutoff 6.1%, middle position 5.0%, none losing money.
• Squeezing — 4.9% of the 1,967 chances, +$0.06 per squeeze. Fine.
• Barrelling the turn — 55% of the time after a called c-bet (usually 50-60%), and profitable; better than checking the turn in the same spots.
• A-K and A-Q. A-K offsuit (+147 bb/100 over 426 hands) and suited (+277 over 177) are both robust winners. A-Q offsuit +33 and suited -77 are both within noise — the suited figure turns positive once the biggest pots are set aside. You also fold missed A-K/A-Q about 80% of the time facing a bet against 93% for random missed hands, which is the correct shape: two overcards is a live hand, not air. Nothing to change.
• Open sizes — 2.0bb from every seat outside the blinds and 3.0bb from the small blind, both standard. You mix in 2.5bb and 3bb, and the bigger sizes fold the blinds slightly more often — but that's a format difference, not a sizing effect: your 2bb opens are 97% Rush & Cash and your 3bb opens 90% regular ring. Nothing to change.
• A3s and T9s, which show large losses (-$13 and -$11) — read hand by hand, two or three enormous coolers account for most of it in each case. Don't change how you play them. The one hand worth your own look is Hand 21, where you raised a flop bet with just a straight draw and jammed over the re-raise into a set.

What we didn't get to

• Turn and river play in full detail — sizing street by street, when to fire a second or third barrel versus give up, value-betting thresholds. The barrelling and river-call numbers above are a good start, not a full review.
• Session and tilt patterns — would need session timestamps, which weren't part of this pass.

Neither is "checked and clean" — they're simply unexamined, which is different.

---

The technical appendix

"Give or take X to Y" means: if you ran this exact stretch of luck over and over, 95 times out of 100 the real figure would land somewhere in that range. The narrower the range, the more you can trust the number.

Overall

50,348 hands, -9.07 bb/100, give or take -15.3 to -2.8. Standard deviation 7.15bb per hand.

Preflop

Tighter than the usual band with a healthy small gap — a solid, slightly tight shape overall.

Showdown

Slightly too often, slightly too rarely successful — the blue-line story in two numbers.

Blind defence, by who opened

Defence barely moves with the opener's seat, when it should widen substantially against the button and small blind. Three-bets from the big blind: 305 of 5,861 spots.

Stealing

Fold-to-3-bet detail

885 spots: folded 322 (36.4%), called 447, 4-bet 116. Calls measured against folding: +1.27bb each, +567bb total. By class, per call: suited connectors +4.80 (43 calls), big pairs +4.23 (43), suited broadways +2.31 (96), small pairs +1.25 (91), mid pairs +0.23 (77), offsuit broadways -0.07 (64), suited aces -5.15 (33).

Check-raising, by street

Flop c-betting

Facing a c-bet, folds by what you actually held

Overall 54.7% across 1,446 chances (usually 40-55%). Across every category, only eight individual hands in the whole sample were folds where the price was clearly good enough — the discipline is genuinely good, the double-gutshot row aside.

Barrelling after a flop c-bet

Checking the turn instead averaged -$0.10.

Cold-calling an open (blinds excluded — those are blind defence)

Squeeze spots

Faced an open plus a caller 1,967 times: squeezed 97 (4.9%, usually 5%+), flat-called 266, folded the rest. Average on the squeezes: +$0.06.

Open sizes, by seat

Median 2.0bb from UTG, MP, CO and BTN (usually 2-2.5bb); 3.0bb from the small blind (usually 2.5-3.5bb). Sizes are only ever compared within a seat, since across seats size is confounded with how many players are left to act.

Where the money goes, by exit point

The field

Opponents play 23.3% of hands, raise 16.8%, and fold to a c-bet 51.7%, across 249,465 player-hands.

Site fees, for the record

43.1% of hands raked, roughly 6-8 bb/100 depending on how it's attributed. Identical for every player at the table and not something you can influence — listed for completeness, not as a finding.