Methodology

Each platform receives its own score from 0 to 100. A score near 50 means the verified conversation is balanced or uncertain; higher scores indicate a more positive mood and lower scores indicate a more negative mood. It is not a market-share percentage, sales estimate, popularity ranking, community-vote result, or probability of winning.

The calculation begins at neutral and moves according to the difference between positive and negative conversation. Evidence confidence pulls weakly supported results back toward 50. Recent momentum is converted to the same 0–100 scale, then the final score combines 80% confidence-adjusted sentiment with 20% momentum and is bounded between 0 and 100.

Because evidence is deliberately conservative and confidence-adjusted, scores normally cluster around the middle instead of jumping to artificial extremes. Scores can be compared with a platform's prior snapshots, but they do not need to add up to 100 across platforms.

The homepage change is the current Vibe Score minus the same platform's score in the immediately preceding successfully published research snapshot. It is a difference in score points, not a percentage. Failed or unpublished runs are not baselines; actual intervals can vary.

Detailed analytics also retain an approximate 24-hour comparison: the newest published snapshot at least 20 hours earlier, preserving tolerance for scheduler drift. Delayed jobs can make that baseline older than 24 hours; its actual elapsed time is displayed. We do not interpolate or invent historical scores.

If an earlier snapshot or platform score is missing, the comparison is unavailable, not zero. The public JSON exposes both deltas with each baseline's timestamp, score and elapsed hours. The existing deltaPoints field remains an alias for the previous-snapshot delta. Current scores and the seven-day rolling leader are unchanged.