# Heads-Up No-Limit Hold'em (HUNL) on GTOWizard benchmark

Source: [Heads-Up No-Limit Hold'em (HUNL) on GTOWizard benchmark](<https://www.sota2.com/research/sota/heads-up-no-limit-hold-em-hunl-on-gtowizard-benchmark>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-05-29T10:55:50.555Z

Evaluation of agent performance in Heads-Up No-Limit Texas Hold'em poker using the GTOWizard benchmark environment, measured by metrics like Big Blinds won per hand (mbb/h).

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Heads-Up No-Limit Hold'em (HUNL) |
| Dataset (including any stated split) | GTOWizard benchmark |
| Metric | mbb/h |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 4 of 4 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [mbb/h](<https://www.sota2.com/research/sota/heads-up-no-limit-hold-em-hunl-on-gtowizard-benchmark.md?metric=mbb%2Fh>): Higher is better; unit: SCALAR.

## Recorded results

| Method | mbb/h | Result date | Report paper |
| --- | --- | --- | --- |
| Slumbot (CFR) | 194 | Not recorded | 6a191de3dbd7150506f82c06 |
| GPT-5.5 (zero-shot) | 132 | Not recorded | 6a191de3dbd7150506f82c06 |
| PokerSkill | 61 | Not recorded | 6a191de3dbd7150506f82c06 |
| Loki (rule-based) | -200 | Not recorded | 6a191de3dbd7150506f82c06 |

### Slumbot (CFR)

mbb/h: 194. Result ID: 6a1970affbc373592c1496eb.

[Reporting paper](<https://www.sota2.com/research/paper/6a191de3dbd7150506f82c06.md>)

Recorded evaluation configuration: {"Training":"Millions CPU-hrs","Interpretable":"No","Scales with LLM":"No"}. This record does not assert independent reproduction.

### GPT-5.5 (zero-shot)

mbb/h: 132. Result ID: 6a1970affbc373592c1496ec.

[Reporting paper](<https://www.sota2.com/research/paper/6a191de3dbd7150506f82c06.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69c4bd6b0636e5bf4afc7494.md>)

Recorded evaluation configuration: {"Training":"None","Interpretable":"Partial","Scales with LLM":"Yes"}. This record does not assert independent reproduction.

### PokerSkill

mbb/h: 61. Result ID: 6a1970affbc373592c1496ed.

[Reporting paper](<https://www.sota2.com/research/paper/6a191de3dbd7150506f82c06.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a191de3dbd7150506f82c06.md>)

Recorded evaluation configuration: {"Training":"None","Interpretable":"Yes","Scales with LLM":"Yes"}. This record does not assert independent reproduction.

### Loki (rule-based)

mbb/h: -200. Result ID: 6a1970affbc373592c1496ea.

[Reporting paper](<https://www.sota2.com/research/paper/6a191de3dbd7150506f82c06.md>)

Recorded evaluation configuration: {"Training":"None","Interpretable":"Yes","Scales with LLM":"N/A"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.