Following feedback raised yesterday by several members on UKMuscle, i took the points made seriously and reviewed the discussion in detail.

This morning we carried out a full, manual re-check of the poll data with those concerns in mind. The intention wasn’t to protect any outcome, but to make sure the results reflect genuine, individual participation as accurately as possible. we do appreciate the constructive input and have acted directly on it.

What we did to tighten and clean the voting

• Enforced strict one-vote-per-person rules
  Multiple submissions originating from the same IP address were collapsed into a single valid vote, regardless of browser, device, or submission name used.
• Applied advanced device fingerprint consistency checks
  We did not rely on OS or browser alone. Votes were reviewed using combined technical markers (IP behaviour, user-agent strings, screen characteristics, timing signatures, and request patterns). When multiple submissions shared the same composite fingerprint within short intervals, this was treated as repeat participation and reduced to one valid vote.
• Removed coordinated burst voting
  Rapid clusters of votes arriving within the same 90-second window, particularly when traced back to the same referral source (e.g. Telegram, Reddit, Facebook, etc.), were identified as coordinated activity. This does not reflect natural, independent voting behaviour and was corrected accordingly.
• Filtered VPN and proxy reuse
  Submissions originating from shared VPN or proxy networks were limited to a single vote to prevent artificial inflation through reused exit points.
• Merged obvious name variants
  Alternate spellings, shorthand, or rebranded entries referring to the same lab were combined into a single entry to prevent padding through naming variation.
• Manually reviewed vote timing patterns
  Organic voting shows irregular spacing and variation. Where submissions instead displayed repeated timing signatures or tightly grouped intervals, those entries were corrected.

The guiding principle throughout was consistency and fairness. Votes that appeared genuine and independent were preserved; votes that showed clear signs of coordination or repeat participation were corrected.

We’re committed to keeping the poll as transparent and legitimate as possible.

Technical footnote

Device checks rely on correlated signals rather than any single identifier. No personal data is collected or stored beyond what is required to prevent duplicate submissions. The process is designed to identify repeat participation patterns, not individual users.