DIME
Get the picksJoin

FEEDBACK · findings note

pinnacle-default

Pinnacle is the default reference book for CLV / sharp-line analytics. Don't ship code that defaults to FanDuel/DraftKings as "primary" without first checking that Pinnacle is in the data — it usually is, and it's the right reference.

← All findingssource · memory/feedback_pinnacle_default.md

User caught it 2026-05-24: shipped CLV bookPriority as [FanDuel, DraftKings, BetMGM, ...] without Pinnacle on the list at all.

What was wrong: Pinnacle is the sharpest book in the world; its closing line is the standard CLV reference. Measuring CLV against FanDuel/DraftKings overstates apparent edge because those books are slower and looser. Pinnacle is in odds-service (under regions=eu, scraped for 581/2480 prop snapshots on the day of this discovery) — I just hadn't checked.

Effect of the fix: NBA mean CLV across all stat buckets dropped 0.5-0.6 percentage points uniformly when switching FanDuel → Pinnacle as reference. That's the textbook signature of using the right book. The yesterday-quoted +1.5-1.9% numbers were soft-book-inflated; the +1.0-1.4% post-fix numbers are the realistic measure.

Rule for next time:

  • Before shipping any CLV / line-comparison / sharp-line code, query the odds source first to see which books are present. curl odds-service ... is 5 seconds. Don't guess.
  • Default reference for CLV: Pinnacle if available, then FanDuel, then DraftKings. Anything else is unjustified.
  • When odds-service adds a sharper book in the future (e.g. Bookmaker, BetISN, CRIS), it goes ahead of FanDuel too.
  • The principle: sharpness > recognizability. The book with the most recreational marketing is rarely the book with the sharpest line.

Same discipline pattern as feedback-verify-before-extend — pre-shipping verification of an assumption I held with too much confidence.