You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Bin valuation: contaminants paid at their own price distort mixed buckets #35
Found while building the hammermill-only start (#3): Sim.binStats pays every material in a bucket its own price times the bucket's grade. A 90 percent ferrous "sinks" bucket is therefore still paid for the copper, brass and zinc buried in it, which makes mixed buckets worth more than they should be and distorts the NEXT PURCHASE ranking (a sink-float tank on raw shred shows +$238/t, ahead of the magnet).
Work:
Price a contaminant at no more than the price of the material the bucket is sold as (the dominant group), or at a scrap-dealer discount; a buyer of ferrous shred does not pay copper prices for stray copper.
Give mixed non-ferrous (zorba) and mixed heavies (zebra) their own tradeable prices so a honest mixed bucket is still worth something.
Re-measure the hammermill-only margin on cars after the change; the start may need the ELV feed price or START_BANK retuned (the worker measured revenue dropping from $138/t to $65/t under the strict rule).
Re-check the NEXT PURCHASE order: with honest valuation it becomes magnet, sink-float, eddy, air, which matches real practice.
Keep tests/lines.js, tests/contracts.js and tests/progression.js passing; add a valuation test.
Activity
jackgary86-dev commentedon Oct 3, 2026
Taking this in an automated run; a pull request will follow.
Generated by Claude Code
jackgary86-dev commentedon Oct 3, 2026
Pull request: #36 (branch ticket/35-bin-valuation)
Generated by Claude Code
Bin valuation: contaminants paid at the dominant group's price (#35)