Script Refinement Quality Evaluation on Dramatron-generated Fantasy Scripts
88Character Development Win %Dramaturge
Evaluation Results
| Method | Links | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| DramaturgeComparison Baseline=Original Scripts2025.10 | 88 | 8 | 4 | 86 | 10 | 4 | 90 | 7 | 3 | 92 | 6 | 2 | |
| DramaturgeComparison Baseline=GPT-4o2025.10 | 76 | 18 | 6 | 74 | 20 | 6 | 78 | 17 | 5 | 80 | 16 | 4 | |
| DramaturgeComparison Baseline=Qwen-max2025.10 | 74 | 20 | 6 | 73 | 21 | 6 | 76 | 19 | 5 | 78 | 18 | 4 | |
| DramaturgeComparison Baseline=GPT-4.1-mini2025.10 | 71 | 21 | 8 | 69 | 23 | 8 | 73 | 20 | 7 | 75 | 20 | 5 | |
| DramaturgeComparison Baseline=Deepseek-r12025.10 | 68 | 24 | 8 | 66 | 26 | 8 | 70 | 23 | 7 | 72 | 22 | 6 | |
| DramaturgeComparison Baseline=GPT-4.12025.10 | 64 | 26 | 10 | 61 | 28 | 11 | 66 | 25 | 9 | 68 | 24 | 8 | |
| DramaturgeComparison Baseline=Gemini-2.5-pro2025.10 | 60 | 30 | 10 | 58 | 31 | 11 | 63 | 28 | 9 | 66 | 26 | 8 |