Of his 2018 predictions at 80%, Scott Alexander got 10 right and 3 wrong, a score of 77%.

True if Scott Alexander's 2018 predictions and his January 2019 grading post show 10 correct and 3 incorrect at the 80% confidence level (10/13 ≈ 77%). Mostly true if the post reports these figures but a recount of the original list differs by one prediction, or if the percentage is only slightly mis-rounded. Mixed if the counts differ by two. Mostly false if the post reports a different tally at 80%, or the numbers apply to a different confidence level. False if the counts are clearly different (e.g. off by more than two). Any later correction he published governs.

Set before the research began.

1. E1 Predictions For 2018 Slate Star Codex · 6 February 2018 Own record

   “2. Democrats take control of the House in midterms: 80%”

2. E2 Predictions For 2018 Slate Star Codex · 6 February 2018 Own record

   “34. No co-bloggers make 3 or more SSC posts this year: 80%”

3. E3 Predictions For 2018 Slate Star Codex · 6 February 2018 Own record

   “Chance of Trump’s approval not breaking 50% probably closer to 95% than 90%.”

E1 and E2 quote two separate predictions at the 80% level from Scott Alexander's 2018 predictions post, and E3 is a note about a different confidence level. None of the evidence gives the full list of 80% predictions, how each one turned out, or the tally in the January 2019 grading post. Without those, the claimed 10 right and 3 wrong (77%) cannot be checked or recounted.