# Multi-agent reinforcement learning on SMAC 5v5 v2

Source: [Multi-agent reinforcement learning on SMAC 5v5 v2](<https://www.sota2.com/research/sota/multi-agent-reinforcement-learning-on-smac-5v5-v2>)

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

Data updated: 2026-06-30T08:02:53.022Z

Evaluation of Multi-Agent Reinforcement Learning performance on the SMAC 5v5 v2 scenario, reporting faction-specific rewards and resource costs (KB) for Protoss, Terran, and Zerg.

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-agent reinforcement learning |
| Dataset (including any stated split) | SMAC 5v5 v2 |
| Metric | Protoss Reward |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 10 of 10 |
| 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

- [Protoss Reward](<https://www.sota2.com/research/sota/multi-agent-reinforcement-learning-on-smac-5v5-v2.md?metric=Protoss%20Reward>): Higher is better; unit: SCALAR.

- [Protoss Cost (KB)](<https://www.sota2.com/research/sota/multi-agent-reinforcement-learning-on-smac-5v5-v2.md?metric=Protoss%20Cost%20(KB)>): Lower is better; unit: SIZE.

- [Terran Reward](<https://www.sota2.com/research/sota/multi-agent-reinforcement-learning-on-smac-5v5-v2.md?metric=Terran%20Reward>): Higher is better; unit: SCALAR.

- [Terran Cost (KB)](<https://www.sota2.com/research/sota/multi-agent-reinforcement-learning-on-smac-5v5-v2.md?metric=Terran%20Cost%20(KB)>): Lower is better; unit: SIZE.

- [Zerg Reward](<https://www.sota2.com/research/sota/multi-agent-reinforcement-learning-on-smac-5v5-v2.md?metric=Zerg%20Reward>): Higher is better; unit: SCALAR.

- [Zerg Cost (KB)](<https://www.sota2.com/research/sota/multi-agent-reinforcement-learning-on-smac-5v5-v2.md?metric=Zerg%20Cost%20(KB)>): Lower is better; unit: SIZE.

## Recorded results

| Method | Protoss Reward | Result date | Report paper |
| --- | --- | --- | --- |
| FULLOBS | 17.4 | Not recorded | 6a434d8eba73165dbb743ae2 |
| FULLOBS | 17.1 | Not recorded | 6a434d8eba73165dbb743ae2 |
| HICOMM | 16.7 | Not recorded | 6a434d8eba73165dbb743ae2 |
| HICOMM | 16.3 | Not recorded | 6a434d8eba73165dbb743ae2 |
| CACOM | 16 | Not recorded | 6a434d8eba73165dbb743ae2 |
| CACOM | 15.6 | Not recorded | 6a434d8eba73165dbb743ae2 |
| T2MAC | 15.6 | Not recorded | 6a434d8eba73165dbb743ae2 |
| T2MAC | 15.2 | Not recorded | 6a434d8eba73165dbb743ae2 |
| PARTIALOBS | 14.8 | Not recorded | 6a434d8eba73165dbb743ae2 |
| PARTIALOBS | 14.5 | Not recorded | 6a434d8eba73165dbb743ae2 |

### FULLOBS

Protoss Reward: 17.4. Result ID: 6a437813fbc373592c94207f.

[Reporting paper](<https://www.sota2.com/research/paper/6a434d8eba73165dbb743ae2.md>)

Recorded evaluation configuration: {"Algorithm":"MAPPO"}. This record does not assert independent reproduction.

### FULLOBS

Protoss Reward: 17.1. Result ID: 6a437813fbc373592c94207a.

[Reporting paper](<https://www.sota2.com/research/paper/6a434d8eba73165dbb743ae2.md>)

Recorded evaluation configuration: {"Algorithm":"IPPO"}. This record does not assert independent reproduction.

### HICOMM

Protoss Reward: 16.7. Result ID: 6a437813fbc373592c942083.

[Reporting paper](<https://www.sota2.com/research/paper/6a434d8eba73165dbb743ae2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a434d8eba73165dbb743ae2.md>)

Recorded evaluation configuration: {"Algorithm":"MAPPO"}. This record does not assert independent reproduction.

### HICOMM

Protoss Reward: 16.3. Result ID: 6a437813fbc373592c94207e.

[Reporting paper](<https://www.sota2.com/research/paper/6a434d8eba73165dbb743ae2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a434d8eba73165dbb743ae2.md>)

Recorded evaluation configuration: {"Algorithm":"IPPO"}. This record does not assert independent reproduction.

### CACOM

Protoss Reward: 16. Result ID: 6a437813fbc373592c942081.

[Reporting paper](<https://www.sota2.com/research/paper/6a434d8eba73165dbb743ae2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74211e209131cea5fd89.md>)

Recorded evaluation configuration: {"Algorithm":"MAPPO"}. This record does not assert independent reproduction.

### CACOM

Protoss Reward: 15.6. Result ID: 6a437813fbc373592c94207c.

[Reporting paper](<https://www.sota2.com/research/paper/6a434d8eba73165dbb743ae2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74211e209131cea5fd89.md>)

Recorded evaluation configuration: {"Algorithm":"IPPO"}. This record does not assert independent reproduction.

### T2MAC

Protoss Reward: 15.6. Result ID: 6a437813fbc373592c942082.

[Reporting paper](<https://www.sota2.com/research/paper/6a434d8eba73165dbb743ae2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71f51e209131cea53c7a.md>)

Recorded evaluation configuration: {"Algorithm":"MAPPO"}. This record does not assert independent reproduction.

### T2MAC

Protoss Reward: 15.2. Result ID: 6a437813fbc373592c94207d.

[Reporting paper](<https://www.sota2.com/research/paper/6a434d8eba73165dbb743ae2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71f51e209131cea53c7a.md>)

Recorded evaluation configuration: {"Algorithm":"IPPO"}. This record does not assert independent reproduction.

### PARTIALOBS

Protoss Reward: 14.8. Result ID: 6a437813fbc373592c942080.

[Reporting paper](<https://www.sota2.com/research/paper/6a434d8eba73165dbb743ae2.md>)

Recorded evaluation configuration: {"Algorithm":"MAPPO"}. This record does not assert independent reproduction.

### PARTIALOBS

Protoss Reward: 14.5. Result ID: 6a437813fbc373592c94207b.

[Reporting paper](<https://www.sota2.com/research/paper/6a434d8eba73165dbb743ae2.md>)

Recorded evaluation configuration: {"Algorithm":"IPPO"}. This record does not assert independent reproduction.

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