Quorum's verdict isn't a majority vote. Each of the 10 seats sketches a little bell of how it reads the stock — where it lands and how confident it is — and the stack of all 10 bells is the verdict. The peak of the stack is the chip. The width of that peak is how much the council actually agrees with itself.
Picture each seat's view as a little hill on a left-to-right scale: bearish on the left, bullish on the right. Confident seats draw tall narrow hills — they know exactly where they stand. Uncertain seats draw wide low hills — they're hedging. Put all 10 hills on the same scale, add them up, and that mountain range is the verdict.
The chip we surface is wherever the mountain's tallest point lands. The shape around that peak is the part most people miss:
No Conclusion.That's it. Ten seats, ten hills, one mountain range, one chip. Sounds neat. The boundary case is where naïve implementations break.
Picture our bearish seat again: very confident, very bearish, pinned to the far-left edge of the scale. A naïve "just add up the hills" implementation lets that hill collapse into a needle-thin spike crammed against the wall. The mountain's tallest point is still the bull peak on the right, so the signal reads Upside Signal cleanly. Tidy. And it just laundered away the most cited dissent in the room.
Our math is engineered against exactly that. Without the parameters: each seat's hill is shaped so it tapers into the wall instead of slamming into it, the tail can't escape the scale, and no single seat is allowed to draw a spike narrow enough to hide. The effect: a 4-1 outcome with one well-cited dissenter actually shifts the mountain. The signal might still say Upside Signal, but the curve visibly widens on the bear side, and the dissent text sits under the chip with its named falsifier intact.
Each seat is required to issue a named falsifier — the specific event that would force it to flip. "If Q3 ARR misses by >5% I drop BUY → HOLD." That's the actionable part: not the chip, but the named condition that flips it.
The 5-0 unanimous verdict is the one you should read most carefully — it usually means the seats are pattern-matching the same training data, not converging on fundamentals. A 4-1 with a load-bearing dissent is a verdict the system argued itself into. The bell-curve shape lets you see which one you're looking at without reading the full debate.