Framler has scored Tempus AI daily since Jul 1 — 37 published scores, of which 31 are old enough to be graded against what the stock actually did next. Both are below, unfiltered.
The purple line is the 0–100 score we published for TEM each day. In the green zone (above 60) the model leans bullish; in the red zone (below 40) it leans bearish; in between it isn't making a strong call. 26 of these 37 days sat in the middle zone — the model has not made a strong call on this stock recently. Every point is committed to a public, Bitcoin-anchored ledger — we can't quietly rewrite a bad call.
Markets are in an expansionary regime. Momentum and growth factors tend to perform well.
Most forecast sites will print an exact price for TEM a year out. We won't: a point target is a guess dressed up as precision. What we publish instead is a 0–100 ranking score with an honest uncertainty range, a daily record of every score we've ever published (anchored in Bitcoin, so it can't be edited after the fact), and a pre-registered public bet on whether the score ranks stocks better than chance. If the model is wrong, this page will say so — that's the product.
Each bar is what TEM actually did over the 7 days after one of our published scores — up bars, the stock rose; down bars, it fell. A “call” is a day the score left the middle zone (above 60 = we said up, below 40 = we said down), and it's counted right only if the stock then moved our way. Faint bars are days we made no call — saying “we don't know” is allowed here. These are raw counts on one stock, far too few to prove skill either way — the serious, pre-registered test across all 1,001 stocks is the public bet.