ALICE.report
Start the Solo Probe

The method

Alice does not judge.She counts.

A neuro-quest provokes behavior. The Alice Assessor records it: who spoke, who interrupted, how long a decision took, who handed over the captaincy and who never did. This page explains how 120-180 minutes in the Room become 92 indicators, 25 competencies and 7 team-dynamics dimensions, which eight models the scoring draws on, what keeps the count honest, and what we refuse to measure.

Open the sample report Talk to the Chief Methodologist
Indicator Stream · sample Venice 1498 · minute 47
  1. 00:47:12 Five More Minutes used by P4 (Venice, mission 4) card
  2. 00:47:40 Quiet Counsel: P8 with P3 (captain), 30 s card
  3. 00:48:02 Decision committed by captain after 212 s timing
  4. 00:49:10 Plan adopted: authored by P8. Indicator 17, influence without authority speech

Audio only. Four lines from one minute of one session; the full stream becomes the Reel in the Alice Report.

01 · The left column

The left column has a ceiling

A questionnaire measures what a person is willing to say about themselves on a Tuesday afternoon. That is useful, and it is limited in three specific ways. It is self-report: the respondent is the only witness. It is static: one answer, no clock, no consequence. And it is rehearsed: anyone who has sat through a 360 knows the vocabulary. The same three limits apply to interviews, and to most assessment centers, where one or two observers write down what they remember.

A simulation removes all three. The witness is the record, not the respondent. The Clock makes every answer cost something. And the plot cannot be rehearsed, because Alice changes it in response to what the team does. Nobody argues with a clock.

The objection we hear most is that people will perform for the camera. They do, for about fifteen minutes. Then the Clock drops to sixty seconds, the plan breaks and the captaincy changes hands. Performance does not survive minute 47. The Pressure dimension measures exactly that delta, and the Heat Map shows who kept their quality when the Clock turned.

What each instrument can witness
InstrumentWho is the witnessClockCan it be rehearsedWhat you receive
Engagement surveyThe respondentNoneYesA score per question
360 feedbackColleagues, from memoryNoneYesRatings and comments
Structured interviewOne or two interviewersNoneYesNotes and a rating
Assessment centerTwo to four observers, by handPer exercisePartlyObserver ratings, days later
AliceThe timestamped recordFive minutes per mission, sixty seconds at the peaksNo: the plot branches on the team’s decisions92 indicators, 25 competencies, 7 dimensions, the Alice Report 15 minutes after the last decision

Simulation-based learning retention: 75% versus 5% for lectures. Industry research; a directional figure, shown here in small type for that reason.

02 · The instruments

The plot is the instrument

The method borrows game mechanics (a rotating captaincy, cards, a clock) because they provoke behavior a questionnaire cannot. Each of the three exists for a diagnostic reason, and each one feeds named indicators.

01 · Captain Rotation

Nobody may be captain more than twice,and never twice in a row.

The rule removes the title from the room. Whoever holds the seat has to brief, decide and yield within one era, and whoever does not hold it has to influence without authority. The Captain Log in every report shows who took the seat, who handed it back, who never asked for it, and how decision latency and voices heard changed under each captain.

Indicators 39-46, delegation and adaptive leadership, read mostly from this one rule.

02 · The Pressure Clock

Five minutes per mission. Sixty seconds at the peaks. Dynamic windows of three, five and ten minutes that Alice sets from the team's own pace.

03 · Opportunity Cards

Eight ways to intervene, each with a diagnostic cost. A card that is never used says as much as one that is.

CardWhat it does in the RoomWhat it revealsIndicators it feeds
Cancel the OrderOverturns the captain's last decision.The team's appetite for constructive opposition, and how the captain takes it.30, interruptions during a dispute · 45, accepts Cancel the Order without retaliation
ReplayReturns the team to the previous mission for a different choice.Admitting error; who argues for the retry.50, uses Replay to correct, not to avoid · 84, changes approach after one failed attempt, not three
Five More MinutesExtends the mission Clock.When the team buys time versus commits.47, decision latency under the Clock · 62, uses Five More Minutes on the mission that needs it most
Quiet Counsel30 seconds in private with any one participant.Informal channels and alliances; who gets chosen.17, influence without authority · 26, uses Quiet Counsel to include, not to conspire
Hand Over the CaptaincyPasses the captaincy immediately.Hidden leaders and willingness to yield.41, yields the lead when asked · 46, steps back cleanly when no longer captain
Ask a LocalAlice voices a hint through a character; costs Autonomy points.Dependence on outside help.9, asks a Local with a precise question, not a plea · 10, counts the Autonomy cost before asking for help
Personal MissionA private objective slightly at odds with the team's.Priorities and ethics.16, reaches agreement between conflicting Personal Missions · 32, declines a Personal Mission that would hurt the team
Silent MinuteOne participant is silent for one minute; Alice mutes the mic.Alternative channels and who notices.8, checks in with the silenced teammate after Silent Minute · 75, re-engages after Silent Minute
“Pick a captain. Sixty seconds. Nobody keeps the seat twice in a row.”
Alice, in the Briefing

03 · The pipeline

From a timestamp to a percentile

The pipeline, in order: Signals → 92 Indicators → 25 Competencies → 7 Dimensions → the Cohort → the Alice Report. Each step is a count, a latency or a ratio. None of them is an opinion.

Signals

4

kinds, from the audio and the Room's own event log

  • Speech

    What is said, by whom, when, and how: turn-taking, interruptions, pace.

    Interruption: P3 over P7 (3rd this era)

  • Decision

    The choice made at a fork and its outcome.

    Mission 2 committed: seal the east wing. 2 of 3 scrolls saved

  • Timing

    Latency, recovery time and pacing against the Clock.

    Decision committed by captain after 212 s

  • Card

    Which Opportunity Card is used, when, and on whom.

    Silent Minute: used by P5 at 01:51:05 on P2

The Room emits the decision, timing and card signals itself: it knows when the Clock started, who holds the captaincy and which card was used. Speech is transcribed and attributed to a participant tile. Nothing on this list requires a camera.

Indicators

92

events with a timestamp, each with a number, a name and a rule for when it fires

Competencies

25

individual-level, five families of five; each rests on three or four indicators, none on one

Communication 1-19
  • Communication clarity: indicators 1, 2, 3, 4
  • Empathy: indicators 5, 6, 7, 8
  • Stakeholder management: indicators 9, 10, 11, 12
  • Cultural competence: indicators 13, 14, 15
  • Negotiation: indicators 16, 17, 18, 19
Collaboration 20-38
  • Collaboration: indicators 20, 21, 22, 23
  • Trust and psychological safety: indicators 24, 25, 26, 27
  • Conflict resolution: indicators 28, 29, 30, 31
  • Team identity: indicators 32, 33, 34
  • Accountability: indicators 35, 36, 37, 38
Leadership 39-57
  • Delegation: indicators 39, 40, 41, 42
  • Adaptive leadership: indicators 43, 44, 45, 46
  • Decision-making and execution: indicators 47, 48, 49, 50
  • Motivation and initiative: indicators 51, 52, 53
  • Strategic thinking: indicators 54, 55, 56, 57
Execution 58-75
  • Project and organizational skill: indicators 58, 59, 60, 61
  • Resource management: indicators 62, 63, 64, 65
  • Risk management: indicators 66, 67, 68, 69
  • Analytical ability: indicators 70, 71, 72
  • Persistence: indicators 73, 74, 75
Adaptability 76-92
  • Creativity and innovation: indicators 76, 77, 78, 79
  • Flexibility and change: indicators 80, 81, 82
  • Learning agility: indicators 83, 84, 85
  • Stress tolerance: indicators 86, 87, 88, 89
  • Emotional self-regulation: indicators 90, 91, 92

Hover a cell for its number and name. Most indicators are counts or latencies, normalized for team size and for the participant's own baseline. A few are judgments on the transcript, and each of those is quoted in the report as a Moment.

Dimensions

7

team-level, each computed from its own signals across the whole team and the whole session

  1. 01 Trust admissions of error, defensive language
  2. 02 Conflict interruptions in disputes, Cancel the Order
  3. 03 Voice airtime share, who is asked
  4. 04 Velocity decision latency, commits per mission
  5. 05 Roles handovers, captaincy pattern
  6. 06 Pressure quality at the 60-second peak
  7. 07 Learning Replay, reuse of what worked

Not averages of the 25 competencies. A team can hold Signature individuals and still score low on Voice, because Voice is who was heard, not who was able.

The Cohort

n

one anonymized set: every team that has run a session, in any world. Its size is never stated

Every dimension and every competency is reported as a cohort percentile. It holds across worlds because the cards, the Captain Rotation and the Clock are identical in every world, and every indicator is normalized before it is compared, so a Blackout team and a Timefold team land on the same scale. The Cohort grows with every session, and so does the resolution of the percentile. We do not publish its size.

The Alice Report

15min

after the last decision; it opens the debrief

Every score in the report links to at least one Moment: a timestamped transcript excerpt that evidences it. Text only; raw audio is deleted within 24 hours unless the client opts in.

The last page, Method and Confidence, states the confidence level for each dimension and lists what Alice could not observe.

Open the sample report

Five of the 92

An indicator is an event with a timestamp. It has a number, a name, a signal type and a rule for when it fires. The report refers to indicators by number and name, so a score can always be traced to the minute that produced it.

IndicatorNameSignalFires when
17Influence without authoritySpeechA plan authored by a non-captain is adopted by the captain
30Interruptions during a disputeSpeechOne participant starts speaking over another while a disagreement is open
41Yields the lead when askedCardThe captain hands over after a request from the team, measured by latency from request to handover
47Decision latency under the ClockTimingSeconds from the mission brief to the captain’s committed decision
86Decision quality in the last 60 seconds of a missionTimingOutcome of decisions committed inside a peak, against decisions committed outside one

Five levels

Each competency is reported on five levels. The levels are defined by how often and how reliably the competency's indicators fired, not by comparison with colleagues.

LevelDefinition
AbsentNone of the competency’s indicators fired in a situation that called for them
EmergingFired once, or only after prompting from Alice or a teammate
WorkingFired in most situations that called for it, with at least one miss
StrongFired in every situation that called for it, in at least two eras
SignatureFired in every situation, in at least three eras, and at least once inside a 60-second peak

Why four eras and not one

Every world loads every family, but each era leans on three competencies by design. A competency that only fires once is Emerging, whatever it looked like.

  • Alexandria 48 BCanalytical ability · stress tolerance · collaboration
  • Venice 1498communication clarity · adaptive leadership · empathy
  • Amsterdam 1702project and organizational skill · creativity and innovation · resource management
  • Orbit 2100creativity and innovation · delegation · learning agility

Seven dimensions

Dimensions are team-level. They are not averages of the 25 competencies; each is computed from its own signals, across the whole team, across the whole session.

DimensionSignals it is computed fromWhat a low score looks like in the Room
1TrustOpenness of information, defensive language, admissions of error.Nobody says “I got that wrong” until Alice shows them
2ConflictDisagreements raised, resolved or suppressed; tone under disagreement; Cancel the Order usage.Fourteen interruptions and zero uses of Cancel the Order
3VoiceAirtime distribution, interruptions, who is asked and who is not; response to Silent Minute.Two voices hold 44% of the airtime; one participant is asked for a view twice in 171 minutes
4VelocityDecision latency under the Clock; commits per mission; Five More Minutes usage.A median of 104 seconds to commit; a 212-second peak at minute 47
5RolesHandover behavior, delegation, role drift across eras; the captaincy pattern.Two handovers in four eras, both by card; three participants never captain
6PressurePerformance delta between normal windows and 60-second peaks.Decision quality halves when the Clock turns
7LearningAdaptation after a failed mission; Replay behavior; reuse of what worked.The same plan fails three times before anyone argues for a Replay

04 · The scoring key

Eight models, one scoring key

The 92 indicators were written by a Chief Methodologist with 20+ years in leadership development, mediation and facilitation, and they were scored by hand, by two human assessors, in sessions run in 2024-2025. The indicator set synthesizes eight established models. Each one contributes a specific lens, and each lens shows up in a specific place in the report.

  1. 01

    MBTI

    A vocabulary for how individuals take in information and reach decisions; used to read decision style under the Clock.

    In the reportGap Cards: the decision-style line

  2. 02

    Process Communication Model (PCM)

    How communication style shifts under distress, and what the shift looks like in speech.

    In the reportPressure dimension; communication clarity and emotional self-regulation

  3. 03

    Belbin Team Roles

    Role balance: which roles a team needs and who is covering them.

    In the reportRoles dimension; the pairing recommendation on each Gap Card

  4. 04

    Drexler-Sibbet Team Performance Model

    The stages a team moves through from orientation to renewal, and where it stalls.

    In the reportLearning dimension, read across four eras

  5. 05

    Situational Leadership

    Leadership behavior matched to the readiness of the people being led.

    In the reportCaptain Log; adaptive leadership and delegation

  6. 06

    Lencioni's Five Dysfunctions of a Team

    The ladder from trust through conflict and commitment to accountability and results.

    In the reportTrust and Conflict dimensions

  7. 07

    Lencioni's Six Types of Working Genius

    Where each person’s energy goes in a cycle of work, and where it drains.

    In the reportTeam-role tendency; the owner-per-step logic of the 90-Day Playbook

  8. 08

    Hawkins' High-Performance Team model

    Trust, shared learning and results orientation treated as one system.

    In the reportThe backbone of the team-level checklist used in 2024-2025; the structural source of the 7 dimensions

Names are trademarks of their respective owners. Alice synthesizes through these models and is not affiliated with or certified by them.

The models are the scoring key, not the output. Alice does not type anyone. A Gap Card never says what someone is; it says what they did, at which minute, and how often.

05 · Bias controls

What keeps the count honest

Validity in a behavioral instrument comes from mechanisms, not from assurances. These are ours.

  1. 01

    Provenance.

    Before Alice listened, two humans did. The 92 indicators were written and scored by hand in sessions run in 2024-2025, with a human director and two human assessors in the room. Alice automates the observation; it did not invent the checklist.

  2. 02

    Observable events only.

    An indicator fires on something that happened at a timestamp: a sentence, a handover, a card, a committed decision. The Assessor does not infer intent, mood or personality. Where the model has to judge a sentence (Indicator 28, separates the problem from the person), the sentence is quoted in the report as a Moment.

  3. 03

    Titles are joined after scoring.

    In the Room, the Assessor scores P1 through P12. It does not know who the CEO is. Names and roles are attached to the report after the scores are final, so the COO’s interruption counts the same as anyone else’s.

  4. 04

    Personal baselines.

    Speech rate, volume and turn length are compared with the same participant’s own first fifteen minutes, not with a population norm. A fast talker is not scored as stressed for being a fast talker.

  5. 05

    Normalization for team size.

    Airtime share, interruptions and voices heard are expressed relative to the team’s own band, so a team of 5 and a team of 12 land on the same scale.

  6. 06

    Speaker attribution you can check.

    Each participant joins from their own device with their own tile, and the Room Check confirms the microphone before the Room opens. Headphones are recommended because echo confuses the Assessor. Unattributed utterances are logged as unattributed, and the Method and Confidence page reports how many there were.

  7. 07

    Accented English.

    Sessions run in English today, and speech processing handles accented English. Other languages are on the roadmap and we say so.

  8. 08

    One scale, every world.

    The cards, the Captain Rotation and the Clock are identical in every world, and airtime, latency and interruptions are normalized before they are compared, so a Blackout team and a Timefold team are scored on the same scale, against the same Cohort. The report names the world beside every percentile.

  9. 09

    Four eras, not one.

    A competency that fires once is Emerging, whatever it looked like. Signature requires three eras and a 60-second peak. The session is long because a short one would flatter everyone.

  10. 10

    Confidence stated per dimension.

    The last page of every report lists the confidence level for each of the 7 dimensions and what Alice could not observe: muted microphones, dropped connections, a participant who left the Room. A dropped connection pauses the Clock and lowers the confidence on that participant’s indicators for that mission.

  11. 11

    A human draws the conclusion.

    The Alice Report is decision support for a human debrief, not an automated decision about anyone. Alice leads the debrief with the report on screen and asks the team to look at the Moments before it looks at the scores.

  12. 12

    Validation in progress.

    We are testing whether session scores predict what teams report 30 and 90 days later through the Pulse-Test and the completion of their 90-Day Playbook, and we are checking Alice’s indicator scoring against the archived human scores from 2024-2025 on the same indicator set. The study is in progress. We will publish the figures when it is complete and not before, and this page will say “in progress” until then.

06 · By design

What we do not measure

These are design decisions, not limitations we hope to remove.

  • No facial recognition. No emotion recognition.

    The camera exists so the team can see each other. Alice never processes the video. Nothing in the pipeline takes a face as input.

  • No inferred emotion from voice.

    Tone is used for one thing: tone under disagreement, scored as constructive or not from the words. Alice does not claim to know how anyone felt.

  • No personality typing.

    The eight models are a scoring key, not a label. The report never says what someone is.

  • No sentiment scores, no engagement scores.

    A count of interruptions is a count. It is not turned into a mood.

  • Nothing outside the Room.

    No calendars, no email, no chat history, no HRIS data in, no screen capture. The session is the only input.

  • No automated decision about a person.

    Access is role-based; each participant receives their own Gap Card; the sponsor sees what the access policy set at onboarding allows. The report informs a human conversation.

Posture, for Legal

Alice listens to what people say and do. She does not read faces or infer emotions; workplace emotion recognition is off-limits under the EU AI Act, and Alice does not do it by design. The report is decision support for a human debrief, not an automated decision about anyone. Participants consent before the Room opens.

Audio is processed during the session; raw recordings are deleted within 24 hours unless the client opts in. Encryption in transit and at rest. GDPR-ready processing. SOC 2 Type II in progress. The full list, in the same words, is on /security.

07 · Our thesis

Seventeen and eighty-three

“Technology-driven team building is only 17% of the success. 83% is data for decisions.”

The session is the 17%. It has to be good, because without a real plot and a real clock there is nothing to count. The 83% is the right column: 92 indicators, 25 competencies, 7 dimensions, a cohort percentile, a Moment behind every score and a 90-Day Playbook with an owner on every step. Teams remember the Burning Library. Organizations change because of the report.

08 · Sources

Sources

External figures used on this site, with their sources as cited; the eight models and their authors; and where the internal numbers come from.

External figures used on this site

  • 77% of employees are not engaged Gallup, State of the Global Workplace
  • 82% of companies get manager selection wrong Gallup, State of the American Manager
  • 18% of annual salary lost per disengaged employee Gallup
  • 50-200% of annual salary: the cost of a wrong manager appointment SHRM / Forbes
  • 2.8 hours per employee per week lost to workplace conflict CPP Inc., Global Human Capital Report
  • 75% vs 5% retention from simulation-based learning versus lectures Industry research; directional

The eight models and their authors

  • MBTIKatharine Cook Briggs and Isabel Briggs Myers, building on C. G. Jung’s psychological types
  • Process Communication ModelTaibi Kahler
  • Belbin Team RolesR. Meredith Belbin, Management Teams: Why They Succeed or Fail
  • Drexler-Sibbet Team Performance ModelAllan Drexler and David Sibbet
  • Situational LeadershipPaul Hersey and Ken Blanchard
  • The Five Dysfunctions of a TeamPatrick Lencioni
  • The Six Types of Working GeniusPatrick Lencioni
  • High-Performance Team modelPeter Hawkins, Leadership Team Coaching

Names are trademarks of their respective owners. Alice synthesizes through these models and is not affiliated with or certified by them. Alice's own validation study is in progress; no validation result is claimed on this page.

Internal sources

  • The indicator set, by number and name: the Method and Confidence page of any Alice Report lists the indicators that fired for that session.
  • The human-assessed edition, 2024-2025: sessions run on the same method with a human director and two human assessors. Names withheld at the client's request.

Before you decide what this is

The sample report is open

Look at the report before you decide what this is. The sample is a synthetic team of eight, labeled as such on every page, with the Moments, the Captain Log, the Card Ledger and the Method and Confidence page in full.

Open the sample report Talk to the Chief Methodologist

Method call: 30 minutes, free. A manager replies within one business day.

Sample Alice Report · synthetic team · sample names

Northwind Logistics leadership team · Timefold · 8 participants · 171 minutes · 4 of 4 eras

Team Potential
52nd-66thcohort percentile band · confidence high
Top strength
PressureDimension 6 · 71st percentile · 3.8 of 5
Top risk
Warning$102,000Decision bottleneck at the COO · per year · estimate
90-Day Playbook
12 steps3 sprints · checkpoints at days 30, 60 and 90 · an owner on every step

Report generated 14 min 32 s after the last decision

Open the sample report