Is Ignition Poker Beatable? Pool Leak vs Rake, Measured

“Is Ignition still beatable?” gets asked every month on 2+2 and Reddit, and every answer is an anecdote. We can do better than anecdotes, because Ignition’s hand histories show every dealt seat’s hole cards. That lets us price every preflop decision the pool makes against a solver chart and add up what the pool gives away. Then we can compare that number with what the rake takes.

What the rake costs at a table of perfect players

Ignition takes 5% of the pot, no flop no drop, capped by the number of players dealt: $1 heads-up, $2 three-handed, $3 at four or five, $4 at six or more. At $1/$2 the six-handed cap is 2 bb.

If six players all followed the same equilibrium chart, each seat would lose 3.9 bb/100 to the rake at NL200. That is the chart’s own expectation when it plays itself through the rake model. At the tables in our sample the rake actually taken per seat was higher, 5.4 to 5.8 bb/100, because the pool plays more pots and bigger ones than equilibrium does.

So the bar is about 4 to 5.5 bb/100. To win, the pool has to hand you more than that.

How much the pool hands over preflop

For every pool decision on an NL200 ring table we looked up the chart’s EV for the action taken and the chart’s best action, and took the difference. Summed per player, the pool loses:

• +1.70 ± 0.19 bb/100 per player on nodes the solver prices reliably (33,719 decisions).
• +4.14 ± 0.52 bb/100 when every priced node is counted (36,448 of 36,535 decisions).

The gap between the two is decisions on nodes the solver barely visited: lines after a limp, after an off-chart 3-bet size or a jam. The solver’s EVs there are noise (its tree has AA “losing 95 bb” by calling a 3-bet in one limp line), so we report them but don’t lean on them.

A chart-follower collects roughly one pool player’s average leak: each opponent’s mistake is shared among the other seats, five opponents, a fifth each.

Where the pool’s preflop money goes

The pattern matters more than the total. The pool’s opening ranges and its folds are close to the charts. What it gets wrong is calling: flatting opens it should fold or 3-bet, calling 3-bets with hands that should fold, and calling 4-bets. Open-limping, the leak everyone talks about, is the smallest line on the chart; we measured how rare it is in the limping post.

Other stakes and Zone

Two things stand out. The NL100 pool leaks about 60% more than NL200, which is the ordinary stake effect. And NL200 Zone leaks more than NL200 ring, even though it is the same stake and largely the same players. Fast-fold play seems to make people call wider. There is more on that in Zone vs ring.

The rake moves the other way at the micro end: NL5 pays about 10.4 bb/100 in rake against 5.4 at NL200, so a softer pool does not mean a bigger margin.

What we actually won

Measuring the pool’s leak says how much is on the table. It doesn’t say how much anyone keeps, so we also ran our own hands through a luck adjustment (AIVAT-style: the cards dealt, the seat, and each board card are replaced by their expectation, which removes noise without changing the mean).

Error bars are one standard error. The adjustment cuts the per-hand standard deviation from 9.1 to 6.9 bb, which makes each hand worth about 1.5 to 2 raw hands. It is not the 10× you get heads-up, because at a 6-max table most of the remaining noise is the opponents’ cards and choices, which no adjustment can remove.

Those ring hands were played preflop within 0.2 to 1.3 bb/100 of the same charts we priced the pool against. With preflop worth +2 to +4, most of that +12.5 was won postflop.

So: beatable?

Yes, at NL100 and NL200 ring, for a player who follows the charts preflop and plays reasonably postflop. The pool is not soft in the way 2010 pools were. Its opens and folds are nearly correct. It leaks by calling, and the leak preflop is smaller than the rake. The money is in the postflop decisions those loose calls set up: capped, wide ranges out of position in single-raised and 3-bet pots.

That is also why “GTO preflop charts” alone disappoint people on this site. The charts are necessary. They are not where the win rate is.