Cloak and Data

Celebrities UK - Season 2 · through episode 10

Hannah Fry's method

Professor Hannah Fry and Matt Parker asked how Alan Turing would have played The Traitors: score every vote as a clue, add the clues up, and cross off the Traitor line-ups the votes rule out. These graphics run their method on this season's votes and update after every episode.

Turing points

Turing points, to Ep 10 T4Every clue in a player's votes adds or takes away points: right leans Traitor, left Faithful. Every clue in a player’s votes adds or takes away points.Turing points← FaithfulTraitor →Sebastianlate pile-on ×3, named by a Traitor ×2, stayed out−0.63Romeshburn ×2, late pile-on ×2, named by a Traitor, stayed out ×2−2.03WHAT EACH CLUE IS WORTHBurn vote: a vote nobody else cast (first three tables)+0.45Burn vote at a later table−0.17Late pile-on: first vote for someone already under fire+0.43Stayed out: voted with others, not for who went−1.06Named by a banished Traitor−0.43Method: Hannah Fry & Matt Parker, “How would Alan Turing have played the Traitors?”@cloakanddata | cloakanddata.co.uk
Updated after episode 10, round table 4

Universes left

Every possible set of three Traitors among the players still in, crossed off as the votes rule them out.

Universes left, to Ep 10 T4Every possible line-up of three Traitors among the players still in, crossed off as the castle learns more. Every possible trio of Traitors, crossed off as we learn more.Universes left0line-ups still possibleout of 1,330 at the start01,000StartRichard banishedBella murderedRob murderedJoe banishedJerry banishedJoanne banishedJames A. banishedMaya banishedKenny banishedJames B. banishedOnly crossing off players who have leftAlso using the votes: Traitors don’t vote for each otherunless one is going home, so a voter and a target who survivedcan’t both be TraitorsMethod: Hannah Fry & Matt Parker, “How would Alan Turing have played the Traitors?”@cloakanddata | cloakanddata.co.uk
Updated after episode 10, round table 4

Every universe left

Each remaining set of three Traitors, most likely first. Darker tiles are more likely, based on Turing points.

Every universe left, to Ep 9All 17 possible line-ups of three Traitors, most likely first, weighted by Turing points.Every universe leftNames in red: Traitors known to viewersJames B.MichaelSebastian32%James B.MayaMichael27%James B.Leigh-AnneMichael7%James B.Leigh-AnneSebastian7%James B.Leigh-AnneMaya6%Leigh-AnneMichaelSebastian5%Leigh-AnneMayaMichael4%Leigh-AnneRomeshSebastian3%Leigh-AnneMayaRomesh3%James B.KennyMichael2%James B.KennyMaya2%KennyMayaMichael1%KennyMayaRomesh1%James B.KennyLeigh-Anne<1%KennyLeigh-AnneMichael<1%KennyLeigh-AnneMaya<1%KennyLeigh-AnneRomesh<1%Darker = more likely, based on Turing pointsJames B., Michael, Sebastian is the most likely line-up (32…Method: Hannah Fry & Matt Parker, “How would Alan Turing have played the Traitors?”@cloakanddata | cloakanddata.co.uk
Updated after episode 9

Who could be Traitors together?

How many of the remaining sets of three Traitors include each pair of players. A cross means the votes rule the pair out.

Who could be Traitors together?, to Ep 9For each pair of players, how many of the 17 possible Traitor line-ups include both. A cross means the votes rule the pair out. 17 line-ups left.Who could be Traitors together?Names in red: Traitors known to viewersJames B.MichaelSebastianMayaLeigh-AnneKennyJames B.in 8Michael4in 8Sebastian22in 4Maya33in 8Leigh-Anne4434in 11Kenny3344in 8Romesh1232in 4Each square: line-ups that include both playersA cross: the votes rule the pair out. “in N”: line-ups that player is in6 pairs tie for most common, each in 4 line-upsMethod: Hannah Fry & Matt Parker, “How would Alan Turing have played the Traitors?”@cloakanddata | cloakanddata.co.uk
Updated after episode 9

Who's in the most universes?

James B. is in the most universes: a 83% chance of being a Traitor.

Who's in the most universes?, to Ep 9Each player's chance of being a Traitor: their share of the possible line-ups, weighted by Turing points. Chance of being a Traitor, weighted by Turing points.Who’s in the most universes?James B.83%Michael77%Sebastian47%Maya43%Leigh-Anne36%Kenny7%Romesh7%Before weighting: share of the possible line-upsNames in red: Traitors known to viewersMethod: Hannah Fry & Matt Parker, “How would Alan Turing have played the Traitors?”@cloakanddata | cloakanddata.co.uk
Updated after episode 9

Would Hannah have caught them?

Hannah Fry's method picked a Traitor at 5 of 11 round tables.

Would Hannah have caught them?, to Ep 10 T4At each round table: who the castle banished, and who Hannah Fry's method would have picked from the votes before it. Who went, and the method’s pick from the votes before.Would Hannah have caught them?THE CASTLETURING POINTS PICKEp 4✗JulieFaithful✓James A., James B.joint topEp 5✓HannahTraitor✓James A., James B., Hannahjoint topEp 6✗JoeFaithful✓James A.+0.88 pointsEp 7✗JerryFaithful✓James A.+0.71 pointsEp 8✗JoanneFaithful✓James A.+1.14 pointsEp 9✓James A.Traitor✗James B.−0.18 pointsEp 10 T1✓MayaTraitor✗James B.−0.18 pointsEp 10 T2✗Leigh-AnneFaithful✗James B.−0.18 pointsEp 10 T3✓KennyTraitor✗James B.+0.25 pointsEp 10 T4✗James B.Faithful✗James B.−0.18 points4 of 10banishments caught a Traitor5 of 10picks included a TraitorA random pick finds a Traitor about 23% of the time.Names in red were Traitors at that table, as viewers know now.Method: Hannah Fry & Matt Parker, “How would Alan Turing have played the Traitors?”@cloakanddata | cloakanddata.co.uk
Updated after episode 10, round table 4

Method from Hannah Fry and Matt Parker's video, with clue weights from TraitorBase's re-count. Not affiliated with either.

Instagram TikTok