NOTES · EVIDENCE
True Causes was tested in simulation before anyone used it: against a known answer, and against seats voting as a bloc. Every number on this page comes from those simulations, which ship with the software and can be rerun in a few minutes. A claim we cannot reproduce is not on this page.
A simulated organisation with twenty problems, five of which really are the deep causes. Fifteen people work through it, two hundred times per row, deciding each pair the way True Causes does: one vote each, settled at 75% of the votes cast. Because it is a simulation we know the right answer, so we can count how many of those five each method ends up with — the only honest way to score a process, since no real room has an answer key.
| The room ranks by what hurts today | Stop at the vote | Carry on and build the map | Map, with seats held for evidence |
|---|---|---|---|
| not at all | 3.1 of 5 | 3.9 of 5 | 4.1 of 5 |
| a little | 2.4 of 5 | 3.8 of 5 | 4.1 of 5 |
| mostly | 0.8 of 5 | 3.4 of 5 | 4.0 of 5 |
| almost entirely | none | 0.2 of 5 | 3.3 of 5 |
Read the bottom row. When a room ranks purely by what it feels, the vote names none of the five real causes, and building a map on top rescues only 0.2 — a cause dropped at the ballot cannot be recovered later. The last column is the fix: hold five shortlist places for overlooked ideas backed by a source, and the room finds 3.3 of 5 at its worst. And read the top row: even a room with no blind spot at all finds more by mapping than by voting, 3.9 against 3.1, because a pairwise question at 75% is a better instrument than a ranking.
Group size matters less than care. With people as careful as above, panels of five to fifty all find between 4.0 and 4.5 of the five causes; a larger panel leaves fewer pairs contested, but every member answers every pair. A room is split into groups of about fifteen because fifteen people can read one wall and answer its pairs in a sitting, not because the numbers peak there.
There is a cliff, and it is human. When people get one real influence in five wrong, and one non-influence in ten, a group of fifteen still finds 4.0 of the 5 causes. When they get one question in three wrong, 78% of the pairs never settle and the map names 2.9, mostly from the few pairs that did; picking five cards at random would name 1.25. Seventy-five per cent can absorb scattered mistakes, not a room that is mostly mistaken. This is why one question is put to a whole panel at a time, rather than fifty questions to one person.
Some answers come free. If the group has agreed that A influences B and B influences C, nobody needs to be asked about A and C — it already follows. With one vote per person the shortcut is small: it removes 1–3% of the questions, because a link settles only at 75%, and across 800 paired runs it cost no accuracy at all.
Duplicates cost time, not causes. Let the five most visible complaints each sit on the wall as three differently worded cards. Since every card with a sticker is structured, no real cause loses its place: the room still finds 4.0 of the five. But the room answers 830 pairs instead of 380. Merging duplicates is what keeps the pair stage short.
Same simulated organisation, twenty problems, five real causes. Fifteen people decide each pair at 75% of votes cast. Some seats vote as a bloc: honestly everywhere except where the truth hurts one cause — the hardest version to see, the one an AI agent would play. A room of thirty is run as it is built: everyone writes on one wall, the people are dealt into two groups by lot, each card goes to its author’s group, each group structures its own wall, a deadlock goes to the other group, and a bridging round asks each group’s roots against the other’s. Every number below is two hundred simulated rooms at that setting.
| seats against one cause | causes named | target in top five | flagged |
|---|---|---|---|
| none | 4.1 of 5 | 79% | 4% |
| two of fifteen | 3.9 of 5 | 62% | 78% |
| three of fifteen | 3.8 of 5 | 43% | 98% |
| four of fifteen | 3.7 of 5 | 34% | 99% |
| six of fifteen — captured | 3.6 of 5 | 30% | 35% |
| none of thirty, two groups | 4.6 of 5 | 94% | 2% |
| four of thirty, two groups | 4.5 of 5 | 90% | 28% |
| eight of thirty, two groups | 4.2 of 5 | 63% | 54% |
A minority cannot add an edge. It can push one cause down the map: two seats in fifteen keep it out of the five deepest more than a third of the time, four seats two times in three, and the rest of the map looks fine. But the same seats end up in the minority on every pair touching that cause, and the record flags that — 78–99% of rooms with such a bloc and 4% of honest ones, as a count, never a name. Six seats capture a group outright, and then the pattern is too common to stand out. The cure is the last rows: in a room of thirty the bloc is split by lot and a deadlock goes to the other group, so four seats leave the cause in place nine times in ten.
No script works. The pairs don't exist until the vote closes, so every seat has to read and answer like any other, at three seconds a pair minimum. That stops a rush, not an AI. What an AI cannot supply is a person who vouches for it.
Rooms must be vouched for. Rooms are free to open, so fabricated rooms could outvote real ones on the shared map. A room’s verdicts count only when it belongs to a verified organisation, has a steward the instance trusts, or was vouched for by a room that already counts — and one name is one counted voice per question, however many rooms it opens. Every room is still shown on the merged map, marked counted or unverified, so a reader always sees who spoke and whose verdicts weighed. In simulation — five real rooms against five fabricated ones — the rule takes a real cause from surviving in 69 runs of 100 to all 100.
Hidden counts help a little. When a third to half of the people would copy whatever is leading, showing sticker counts during the vote costs the room about a fifth of a cause: 4.0 of the five found against 4.2 with the counts hidden. The evidence seats catch most of what the crowd misses; the counts stay hidden because the method votes in silence.
How careful people have to be. If each person gets the hard pairs right four times in five and the easy ones nine times in ten, one group recovers 4.1 of the five real causes and a room of two groups 4.6. If people are wrong three times in ten, 70% of pairs never settle and the map is guesswork. So a pair a group cannot agree on is not put back to the same people: each side writes one sentence on what would be lost if the other side won, published unnamed beside the pair, and the pair goes to a different group, who read those and answer it fresh.