Multi-agent Reinforcement Learning on StarCraft Multi-Agent Challenge (SMAC) benchmark leaderboard · SOTA2 Research

Multi-agent Reinforcement Learning on StarCraft Multi-Agent Challenge (SMAC)

1001c3s5z Win RateMAPPO

Evaluation Results

Method	Links
MAPPO2021.03		100	100	100	100	100	100	100	100	100	96.9	100	100	100	96.9	100	89.1	96.9	96.9	90.6	84.4	93.8	88.3	100
IPPO2021.03		100	100	100	100	100	100	99.2	100	100	96.9	100	100	98.4	96.9	100	87.5	96.9	93	86.7	82.8	69.5	84.4	98.4
RODE2021.03		100	—	—	100	100	—	—	—	—	—	100	78.9	100	—	—	71.1	93.8	95.3	89.8	96.8	96.8	78.1	65.6
MAPPO2021.03		100	100	100	100	96.9	100	100	100	100	93.8	100	98.4	96.9	84.4	96.9	65.6	71.9	81.2	51.6	75	93.8	78.1	93.8
MAPPO2021.03		96.9	100	100	100	100	100	98.4	100	100	93.8	100	99.2	100	96.9	100	88.3	96.9	96.9	87.5	63.3	85.9	85.9	98.4
MAPPO2021.03		96.9	100	100	100	96.9	100	100	96.9	100	96.9	100	100	95.3	87.5	93.8	68.8	53.1	89.1	28.1	18.8	89.1	81.2	93.8
QMix2021.03		96.1	95.3	96.9	96.9	95.3	96.9	97.7	96.9	97.7	95.3	100	98.4	92.2	92.2	85.9	75.8	88.3	95.3	87.5	82.8	39.1	9.4	84.4
LBI2024.10		94.59	—	—	—	—	—	—	—	—	—	—	92.82	71.44	—	—	87.98	—	90.45	95.96	83.34	—	66.61	87.45
MAPT2024.10		74.77	—	—	—	—	—	—	—	—	—	—	72.33	58.28	—	—	56.01	—	66.85	19.21	34.9	—	37.16	45.91
MA-TREX2024.10		64.76	—	—	—	—	—	—	—	—	—	—	55.38	22.45	—	—	50.01	—	48.78	21.52	36.16	—	28.38	30.59
MADT2024.10		61.35	—	—	—	—	—	—	—	—	—	—	80.9	61.9	—	—	79.78	—	82.76	54.34	27.44	—	30.94	69.85
BC2024.10		16.44	—	—	—	—	—	—	—	—	—	—	0	17.37	—	—	13.78	—	26.19	0	0	—	9.28	0
MA-AIRL2024.10		7.88	—	—	—	—	—	—	—	—	—	—	0.05	24.75	—	—	11.59	—	41.69	0	0	—	16.47	0.76