How the Leak Detector Found My Biggest Bleed (And Fixed It in 3 Weeks)

I’d been breakeven at $25NL for 6 months. Grinding 3k hands a week, reviewing maybe 10 spots, convinced my river play was the issue.

The leak detector found the real problem in 90 seconds.

What it flagged

After importing 15,000 hands from PokerStars, the report grouped findings by severity:

High (sample > 500 hands):

• C-bet too high IP: 94% c-bet frequency in position, vs GTO ~65-70%
• No 3-bet from CO/BTN at 12bb stack: 0% 3-bet from CO with 10-14bb in MTT spots

Medium (sample 200-500 hands):

• Over-defend BB vs SB open: defending 72% of hands vs SB (GTO: ~55%)

The high-frequency c-betting was bleeding money. Villain adjusts to your stat — if you’re c-betting 94%, good players start floating every flop and raising high.

The fix

I spent 2 weeks drilling BTN c-bet decision making in the ICM Trainer, specifically choosing “check back” instead of “bet” with medium-strength hands on connected boards. The habit change took about 1,200 practice hands.

Week 3 results: c-bet frequency dropped to 68%. Win rate moved from -1bb/100 (breakeven minus rake) to +4.2bb/100.

Why this matters

Manual review catches the hands you notice. The leak detector catches the patterns you don’t notice — because they happen uniformly across your range and feel normal.

It’s the difference between a coach watching your session and a coach with a database of every decision you’ve made for 6 months.

Start your leak report from your next imported session. The top finding is usually the most surprising one.