Fair ranking philosophy

iFixedX is an algorithm R&D vehicle: learn how to improve discovery at X-scale without shipping a separate “boost slot” product.

Principles (product north star)

  1. No artificial boost lanes — do not reserve feed slots that only small accounts can win; that feels rigged.

  2. Keep stochastic surfacing — the feed stays partly “lotto-like”; better candidate generation and scoring improve odds for merit, not guaranteed impressions.

  3. Merit over pity — smaller accounts should win when posts are genuinely strong, not because of diversity injection.

  4. Richer signals (directional):

    • Semantic quality / substance (Grok seasoning, bounded cost)

    • Originality vs noise

    • Conversation quality

    • Author consistency over time

    • Who engaged (network quality), not raw follower count alone

“Exp sharing” (design space)

Ideas under discussion (not all implemented):

  • Reputation-weighted engagement

  • Discovery credit when quality accounts amplify smaller ones

  • Author-level features accumulated in Pipetrix mixerMetaJson and optionally Account rollup (seenPostCount, lastSeenAt)

Implementations should extend existing types (Tweet, canonical ingest, search.ts) — not a parallel social graph DB unless necessary. Replay raw Grok batches via Grok ingests when re-scoring without new xAI calls.

Two rankers today

Mode

Location

Character

Default (engagement-style)

platform_style

Closer to velocity / engagement heuristics

iFixedX rank

refined

Text match + recency + anti-spam sliders

Grok supplies candidates and context (mixerNote, topic batches); local code owns ordering.

Evaluation

Use Ranking lab sliders; ingest snapshots rankingLab into canonical mixerMetaJson for offline comparison.

Grok collaboration

When Grok on the box proposes scoring changes, start from:

  • src/mixer/merge.ts

  • src/mixer/buildFeedPosts.ts

  • src/feedtrix/canonicalIngest.ts

  • server/forYouCorpus.ts

  • server/search.ts

See Grok agent handoff (on-box).