RUN 20260722-224837 (design)

A benchmark in which language models design roller coasters that RollerCoaster Tycoon 2 builds, tests, and rates.

Standings

design mode — models build from scratch

modelscoreintensitynausea similaritybesttokens / cost
1st claude-opus-4-6 no successful coaster —
2nd claude-sonnet-5 no successful coaster —

score = excitement, scaled down when similarity to a stock library design exceeds 0.5 (an exact or mirrored copy scores 0)

claude-opus-4-6full detail →

Round 1 build failed

build failed (57/57 pieces placed): set_status(testing): Track is not a complete circuit

track program (57 pieces)
{
  "pieces": [
    "begin_station",
    "middle_station",
    "middle_station",
    "middle_station",
    "end_station",
    "flat_to_up_25",
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    "up_25_to_flat",
    "flat_to_right_bank",
    "banked_right_turn_3",
    "right_bank_to_flat",
    "flat_to_down_25",
    "down_25",
    "down_25",
    "down_25",
    "down_25",
    "down_25",
    "down_25",
    "down_25_to_flat",
    "flat_to_right_bank",
    "banked_right_turn_3",
    "right_bank_to_flat",
    "flat_to_up_25",
    "up_25",
    "up_25",
    "up_25",
    "up_25",
    "up_25_to_flat",
    "flat_to_right_bank",
    "banked_right_turn_3",
    "right_bank_to_flat",
    "flat_to_down_25",
    "down_25",
    "down_25",
    "down_25",
    "down_25",
    "down_25",
    "down_25",
    "down_25",
    "down_25_to_flat",
    "flat",
    "flat_to_right_bank",
    "banked_right_turn_3",
    "right_bank_to_flat",
    "flat",
    "flat",
    "flat"
  ],
  "ride_type": 52,
  "start": {
    "dir": 2,
    "x": 50,
    "y": 60
  }
}
park screenshot, claude-opus-4-6 round 1 view 1
claude-sonnet-5full detail →

Round 1 build failed

build failed at piece 10 (10/42 pieces placed): rejected at cursor (x=50, y=60, z=216, dir=0): Too steep for lift hill

track program (42 pieces)
{
  "pieces": [
    {
      "t": "begin_station"
    },
    {
      "t": "middle_station"
    },
    {
      "t": "end_station"
    },
    {
      "t": "flat_to_up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25"
    },
    {
      "chain": true,
      "t": "up_25_to_up_60"
    },
    {
      "chain": true,
      "t": "up_60"
    },
    {
      "chain": true,
      "t": "up_60"
    },
    {
      "t": "up_60_to_flat"
    },
    {
      "t": "left_turn_3"
    },
    {
      "t": "down_25"
    },
    {
      "t": "down_25"
    },
    {
      "t": "down_25_to_down_60"
    },
    {
      "t": "down_60"
    },
    {
      "t": "down_60_to_flat"
    },
    {
      "t": "left_turn_3"
    },
    {
      "t": "flat_to_up_25"
    },
    {
      "t": "up_25"
    },
    {
      "t": "up_25_to_left_bank"
    },
    {
      "t": "left_bank"
    },
    {
      "t": "left_bank"
    },
    {
      "t": "left_bank_to_up_25"
    },
    {
      "t": "up_25_to_flat"
    },
    {
      "t": "left_turn_5"
    },
    {
      "t": "flat_to_down_25"
    },
    {
      "t": "down_25"
    },
    {
      "t": "down_25_to_flat"
    },
    {
      "t": "left_turn_3"
    },
    {
      "t": "flat_to_left_bank"
    },
    {
      "t": "banked_left_turn_5"
    },
    {
      "t": "left_bank_to_flat"
    },
    {
      "t": "s_bend_left"
    },
    {
      "t": "left_turn_3"
    },
    {
      "t": "flat"
    },
    {
      "t": "flat"
    },
    {
      "t": "brakes"
    },
    {
      "t": "flat"
    }
  ],
  "ride_type": 52,
  "start": {
    "dir": 0,
    "x": 60,
    "y": 60
  }
}
park screenshot, claude-sonnet-5 round 1 view 1

← all runs