Simplified trees are reduced versions of the complete poker game tree, used by solvers because the full tree is too large to compute.
Simplification makes solving possible, but introduces approximation
| Full Tree |
Problem |
| Every possible bet size |
Infinite options |
| Every chip amount |
Billions of nodes |
| Every possible runout |
Exponential growth |
Reality: No computer can solve the complete poker game tree.
| Full Game |
Simplified |
| Bet any amount |
33%, 50%, 75%, 100%, 150% pot |
| Infinite options |
4-6 options |
| Full Game |
Simplified |
| Check, bet, raise |
Check, bet (limited sizes) |
| Multiple raise sizes |
1-2 raise options |
| Full Game |
Simplified |
| All possible runouts |
Grouped similar runouts |
| K♠ vs K♥ |
Often treated same |
| Fewer Options |
More Options |
| Faster solves |
Slower solves |
| Less accurate |
More accurate |
| Miss some strategies |
Capture more nuance |
| Street |
Common Sizes |
| Flop |
33%, 75% |
| Turn |
50%, 75%, 100% |
| River |
50%, 75%, 100%, 150%+ |
| Situation |
Consider Adding |
| Deep stacks |
Larger sizings |
| Studying overbets |
150%+ pot options |
| Specific spots |
Sizes villain uses |
| Street |
Sizes |
| Flop |
50% |
| Turn |
66% |
| River |
66%, all-in |
Good for: Quick study, general patterns
| Street |
Sizes |
| Flop |
33%, 75% |
| Turn |
50%, 75% |
| River |
50%, 100% |
Good for: Most studying
| Street |
Sizes |
| Flop |
25%, 50%, 75%, 125% |
| Turn |
33%, 66%, 100%, 150% |
| River |
50%, 75%, 100%, 150%, 200% |
Good for: Detailed analysis, specific spots
| Priority |
Choose |
| Speed |
Fewer sizes |
| Accuracy |
More sizes |
| General learning |
Simple trees |
| Specific spots |
Complex trees |
| If Solver Uses |
It Might Mean |
| Largest available size |
Would use bigger if available |
| Smallest available size |
Wants to bet small |
| Mix of sizes |
Sizing matters for that hand |
Simplified trees make solver work possible but approximate reality. Include bet sizes that match real games, balance speed vs accuracy, and understand that missing sizes affect results. For learning patterns, simpler trees work fine; for specific spots, add more options.