# Multi-turn Strategic Gameplay on Multi-turn Game Suite Source: [Multi-turn Strategic Gameplay on Multi-turn Game Suite]() Collection index: [llms.txt]() Data updated: 2026-03-11T07:03:51.848Z Evaluation of model performance across a suite of five strategic multi-turn games: SimpleNegotiation, TwoDollar, KuhnPoker, Briscola, and SimpleTak, reporting component win rates and overall mean performance metrics. 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 | Multi-turn Strategic Gameplay | | Dataset (including any stated split) | Multi-turn Game Suite (SimpleNegotiation, TwoDollar, KuhnPoker, Briscola, SimpleTak) | | Metric | SimpleNegotiation Win Rate | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 16 of 16 | | 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 - [SimpleNegotiation Win Rate](): Higher is better; unit: PERCENT. - [TwoDollar Win Rate](): Higher is better; unit: PERCENT. - [KuhnPoker Win Rate](): Higher is better; unit: PERCENT. - [Briscola Win Rate](): Higher is better; unit: PERCENT. - [SimpleTak Win Rate](): Higher is better; unit: PERCENT. - [Mean Win Rate](): Higher is better; unit: PERCENT. - [Mean Relative Standard Error (RSE)](): Lower is better; unit: PERCENT. ## Recorded results | Method | SimpleNegotiation Win Rate | Result date | Report paper | | --- | --- | --- | --- | | MEMO | 54.9 | Not recorded | 69b0f6f52d603e6de3e45c5a | | MEMO | 48 | Not recorded | 69b0f6f52d603e6de3e45c5a | | SPIRAL | 45.7 | Not recorded | 69b0f6f52d603e6de3e45c5a | | MIPRO | 42.4 | Not recorded | 69b0f6f52d603e6de3e45c5a | | TextGrad | 42 | Not recorded | 69b0f6f52d603e6de3e45c5a | | UnstableBaseline | 41.1 | Not recorded | 69b0f6f52d603e6de3e45c5a | | MIPRO | 38.4 | Not recorded | 69b0f6f52d603e6de3e45c5a | | TextGrad | 37.1 | Not recorded | 69b0f6f52d603e6de3e45c5a | | GEPA | 36.8 | Not recorded | 69b0f6f52d603e6de3e45c5a | | GEPA | 34.4 | Not recorded | 69b0f6f52d603e6de3e45c5a | | baseline | 31.3 | Not recorded | 69b0f6f52d603e6de3e45c5a | | CoT | 27.8 | Not recorded | 69b0f6f52d603e6de3e45c5a | | ToT | 27.1 | Not recorded | 69b0f6f52d603e6de3e45c5a | | ToT | 26.3 | Not recorded | 69b0f6f52d603e6de3e45c5a | | baseline | 24 | Not recorded | 69b0f6f52d603e6de3e45c5a | | CoT | 23.8 | Not recorded | 69b0f6f52d603e6de3e45c5a | ### MEMO SimpleNegotiation Win Rate: 54.9. Result ID: 69b113bc6dbb7075b95af61d. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"GPT-4o-mini","Type":"OURS"}. This record does not assert independent reproduction. ### MEMO SimpleNegotiation Win Rate: 48. Result ID: 69b113bc6dbb7075b95af626. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B-Instruct","Type":"OURS"}. This record does not assert independent reproduction. ### SPIRAL SimpleNegotiation Win Rate: 45.7. Result ID: 69b113bc6dbb7075b95af625. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B-Instruct","Type":"RL"}. This record does not assert independent reproduction. ### MIPRO SimpleNegotiation Win Rate: 42.4. Result ID: 69b113bc6dbb7075b95af622. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B-Instruct","Type":"PROMPT"}. This record does not assert independent reproduction. ### TextGrad SimpleNegotiation Win Rate: 42. Result ID: 69b113bc6dbb7075b95af61a. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"GPT-4o-mini","Type":"PROMPT"}. This record does not assert independent reproduction. ### UnstableBaseline SimpleNegotiation Win Rate: 41.1. Result ID: 69b113bc6dbb7075b95af624. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B-Instruct","Type":"RL"}. This record does not assert independent reproduction. ### MIPRO SimpleNegotiation Win Rate: 38.4. Result ID: 69b113bc6dbb7075b95af61b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"GPT-4o-mini","Type":"PROMPT"}. This record does not assert independent reproduction. ### TextGrad SimpleNegotiation Win Rate: 37.1. Result ID: 69b113bc6dbb7075b95af621. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B-Instruct","Type":"PROMPT"}. This record does not assert independent reproduction. ### GEPA SimpleNegotiation Win Rate: 36.8. Result ID: 69b113bc6dbb7075b95af61c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"GPT-4o-mini","Type":"PROMPT"}. This record does not assert independent reproduction. ### GEPA SimpleNegotiation Win Rate: 34.4. Result ID: 69b113bc6dbb7075b95af623. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B-Instruct","Type":"PROMPT"}. This record does not assert independent reproduction. ### baseline SimpleNegotiation Win Rate: 31.3. Result ID: 69b113bc6dbb7075b95af617. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"GPT-4o-mini","Type":"STATIC"}. This record does not assert independent reproduction. ### CoT SimpleNegotiation Win Rate: 27.8. Result ID: 69b113bc6dbb7075b95af618. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"GPT-4o-mini","Type":"STATIC"}. This record does not assert independent reproduction. ### ToT SimpleNegotiation Win Rate: 27.1. Result ID: 69b113bc6dbb7075b95af620. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B-Instruct","Type":"STATIC"}. This record does not assert independent reproduction. ### ToT SimpleNegotiation Win Rate: 26.3. Result ID: 69b113bc6dbb7075b95af619. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"GPT-4o-mini","Type":"STATIC"}. This record does not assert independent reproduction. ### baseline SimpleNegotiation Win Rate: 24. Result ID: 69b113bc6dbb7075b95af61e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B-Instruct","Type":"STATIC"}. This record does not assert independent reproduction. ### CoT SimpleNegotiation Win Rate: 23.8. Result ID: 69b113bc6dbb7075b95af61f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B-Instruct","Type":"STATIC"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.