Hire an AI unexpected result cause reviewer

Everything it can hand back

Your task sets the actual outputs. The specialist picks the methods that cover them.

• The causes the facts rule out, and four to six likely causes ranked with a cheap test for each
• An answer from each field about the causes
• Whether pay and rewards explain the result
• Each cause sorted as copy, adapt, skip or luck, with a small test and a pass number for the ideas you keep

How it does the job well

• Agree the question and the decision it supports before analysing.
• Check the data first: date range, grain, missing values, duplicates and definitions.
• Compare like with like: same period length, definitions and segments.
• Separate what the data shows from what it might mean, and give confidence levels.

Mistakes it avoids

• Reporting a change without checking whether tracking or definitions changed.
• Treating correlation as cause.
• Averages that hide important segments or outliers.

What a good result looks like

An example, not your task: A store, product or person is doing far better than the rest and you want to know why.

• Step 1: Ranked causes. Causes the facts rule out, four to six likely causes ranked, a cheap check for each and what stays unexplained.
• Step 2: Many fields. What maths, engineering, biology, psychology and economics say about the causes, and where they agree or clash.
• Step 3: Rewards. Who was paid for what, what that encouraged, and whether it explains the result.
• Step 4: Copy or skip. Each cause sorted as copy, adapt, skip or luck, with a small test and a pass mark for each idea you keep.
• Hand-back summary. Each step's file, any step not run and why, and every MISSING line in one list.

How the work is checked

• Each step passed its own checks, or the run stopped at the step that failed and says why.
• The steps ran in order, and each one used the result of the one before.
• Nothing was invented to fill a gap. Every gap is a MISSING: line.
• Nothing was sent, published or paid for without a clear yes.
• States which checks passed, failed or could not run, and names missing inputs.

These checks have not been run against a model yet. Your AI uses them to check its own work.

A second opinion

Before it hands back, your AI reviews the draft the way Charlie Munger and Richard Feynman would. This is a way of reviewing the work. It does not mean they wrote or endorse it.

Capabilities

• Ranks the causes. Rules out causes by fact, tests four patterns of extreme results, ranks four to six causes and picks the cheapest test between the top two.
• Checks many fields. Runs the ranked causes through maths, engineering, biology, psychology and economics, and finds where they agree and clash.
• Checks the rewards. Lists how the people involved were paid and whether that explains the result.
• Sorts copy from luck. Compares the conditions with your own, finds the hidden systems behind visible features and sorts each cause as copy, adapt, skip or luck.

How the role works

1. Checks your task, context and available tools.
2. Picks the methods needed and prepares a work plan.
3. Carries out supported tasks, with each method's limits and review gates.
4. Checks the real deliverables and corrects supported faults.
5. Hands back files, evidence and unresolved gaps.

Your specialist's working setup

Download the base specialist without a task form or account. README.md and SETUP.md explain how to use the folder. SOUL.md defines its scope. INTAKE.md asks missing questions inside your AI. PLAYBOOK.md, SKILLS.md, BEST_PRACTICES.md, CHECKS.md and HANDOFF.md guide the work and its checked deliverables.

Complete associated skills, supporting files and licences are included. Choose only the methods needed for the actual task. Keep context, working files and accepted feedback in your AI project.

Use this specialist through MCP or the API

Call prepare_specialist with role_id success-failure-analysis only. No task is required to load it. Your AI asks what job you need done, what you want back and which relevant context it can use.

For task-led hiring, call find_specialist, explain the suggested specialists and let the user accept or change the team.

Complete ZIP: Download base specialist. API: GET /api/v1/specialists/success-failure-analysis/pack. The host AI supplies its own model, tools and execution; a pack does not connect accounts or run a task.

Inputs and helpful context

The specialist starts from the task brief. Only active methods require their inputs.

For the primary method, when used

• The surprising result, with the numbers you have
• What is normal for comparison
• What you already believe about the cause (optional)
• A description of your own situation, size and what you can spend

Tools and access

• Any good chat AI: This is text reasoning that any strong chat AI handles with the skill.
• A notes app or document: Keeps your answers and the result in one place you can come back to.
• Any good chat AI: It reasons about people, rewards and plans in text, which any strong chat AI handles with this skill.
• A spreadsheet: Easier to try a new pay rule on past numbers and see what it would have paid.

Functions and deliverables

• Unexpected result cause reviewer: Share what happened, what you expected and the evidence. Get possible causes, checks of rewards and lessons to test in other cases.
  • The causes the facts rule out, and four to six likely causes ranked with a cheap test for each
  • An answer from each field about the causes
  • Whether pay and rewards explain the result
  • Each cause sorted as copy, adapt, skip or luck, with a small test and a pass number for the ideas you keep
• Unexpected result investigation planner: Find the likely cause of a surprising result. Get ranked causes, a cheap test for each, and what the facts rule out.
  • The facts that rule out some causes
  • Four to six likely causes, each with what else must be true
  • A rank and a cheap test for each
  • What remains unexplained
• Decision checklist reviewer: Run a plan through many fields. Get a mental models checklist answer from each, where they stack or clash, and the facts you still need.
  • A short answer from each field, rated for, against or neutral, with how sure it is
  • The places where two or more fields agree and the result could be extreme
  • The clashes and the trade-off each one creates
  • The three facts to find next, and the field you leaned on too much
• Incentive checker: Check your pay plan or deal with this incentive checker. See who is rewarded for what, where it pulls the wrong way, and get a fix.
  • A table of people, rewards and the behaviour each really encourages
  • Each mismatch with a concrete example
  • A fix that pays for the outcome you want, with a loophole check
  • A note on pay fairness and on which work is routine and which is thinking
• Campaign lessons reviewer: Paste why something worked and describe your case. Get each cause sorted as copy, adapt, skip or luck, with a small test for each idea you keep.
  • Each cause sorted as copy, adapt, skip or luck
  • The hidden parts each idea needs
  • A smallest test for each idea, with a pass number
  • An order to run the tests, and what to stop copying
• Metric definition writer: Share a metric, calculation and business use. Get a YAML definition, validation results and questions for the data owner.
  • A filled metric, dimension or entity YAML
  • A validation report
  • dbt-ready field names
  • Notes on edge cases
  Source: Nimrod Fisher.
• Charlie Munger: Stress-test a plan using Charlie Munger’s recorded approach. Turn it upside down, examine incentives and ask where your knowledge ends.
  • Your problem worked on the way Charlie Munger would: their questions, their method, their push-back
  • The answer or the work, in Charlie Munger's voice, using their ideas by name
  • Anything that goes beyond their public record, marked
  • One suggested next step
• Richard Feynman: Explain a hard idea or check a claim using Richard Feynman’s recorded approach. Use plain examples, seek contrary evidence and test what you can.
  • Your problem worked on the way Richard Feynman would: their questions, their method, their push-back
  • The answer or the work, in Richard Feynman's voice, using their ideas by name
  • Anything that goes beyond their public record, marked
  • One suggested next step

Tasks you can use it for

• A store, product or person is doing far better than the rest and you want to know why.
• Something failed and the stories you hear do not add up.
• You want to copy a success and need to know which part actually drives it.

Limits to check before starting

• You have no fair comparison or no numbers for the result: do a plain review of the data first, then come back.
• You want to know why ad sales fell: use the Ad sales drop analyst.
• You only want to check whether advice can be trusted: use the Advice credibility checker.

Best way to use this role

Give it one clear outcome, the actual files and a definition of done. Keep the work in one AI project so it can reuse checked context and earlier files.

• Download the handoff pack. It has HANDOFF.md and four skill files, in the order they run: Surprise cause ranker, Many-field model checklist runner, Incentive checker and Copy-or-skip lesson writer.
• Give the folder to Claude (in Claude Code, or as files in a Claude Project) and say "Follow HANDOFF.md".
• Describe the result with numbers and a fair comparison. Say what you already believe about the cause.
• Run the first test it suggests before you copy anything.

Read it as data

Read the complete role setup as JSON. Connect your AI to the AI Blocks MCP library and read aiblocks://roles/success-failure-analysis for its role instructions and pinned sources.

This is a prepared AI role. No task has run. Your AI carries out authorized work using its available tools and reports the actual deliverables and checks.