Dispute versus deliberate misuse

Figure 01 / 12

The evaluated population

The cohort contains 13,500 customer dispute records, of which 473 have the defined synthetic outcome: Synthetic deliberate misuse. The outcome is known by construction here. In production, label uncertainty and selection must be recorded separately.

Figure data and text version

Outcome	Count
Synthetic deliberate misuse	473
Other labeled outcomes	13,027

A refund request or dispute can arise from delivery failure, confusion, product design, or deliberate deception. The label must reflect evidence about the actual cause.

All chargebacks become fraud labels regardless of reason or final result.

All amounts, rates, capacity limits, and outcomes in this case are synthetic. The three conditions are separate assumptions for comparison. A better result in the response condition is not measured proof that the proposed control causes that improvement. The figures expose the calculation and its limits; a real deployment needs its own evidence.

Read the result

The rule flags 943 of 13,500 customer dispute records. Of those flags, 227 meet the synthetic target, giving 24.07% precision. It misses 246 target events. Under the stated cost assumptions, residual loss and operating friction total $48,598. The important result is the connection between the population, action, capacity, and outcome—not one isolated score.

Figure 02 / 12

Four outcomes of the rule

The rule flags 227 synthetic positives and 716 negatives. It misses 246 positives. A flagged item is a decision to intervene; it is not proof of fraud, a legal prohibition, or any other real-world conclusion.

Figure data and text version

Known outcome	Flagged	Not flagged
Synthetic deliberate misuse	227	246
Other outcome	716	12,311

Figure 03 / 12

Three rates with different denominators

Precision is 24.07%, recall is 47.99%, and the false-positive rate is 5.5%. Changing the denominator changes the meaning. This record keeps each numerator attached to the population from which it came.

Figure data and text version

Metric	Numerator	Denominator	Result
Precision	227	943	24.07%
Recall	227	473	47.99%
False-positive rate	716	13,027	5.5%

Figure 04 / 12

Precision changes with prevalence

This sensitivity plot holds recall at 48% and false-positive rate at 5.5%, then changes prevalence. It is an algebraic comparison, not a forecast. Even unchanged detection quality can produce a very different review queue when the base rate changes. Horizontal positions are the labeled observations or scenarios; equal spacing does not imply equal numerical increments.

Figure data and text version

Assumed prevalence	Precision %
0.1%	0.87
0.5%	4.2
1%	8.1
2%	15.12
5%	31.48
10%	49.23

Figure 05 / 12

The threshold trade-off

Six illustrative score bands use a stated pair of detection rates. Lower sensitivity can reduce false alarms but miss more target events. These points do not come from a trained model and do not establish the best operating threshold. Horizontal positions are the labeled observations or scenarios; equal spacing does not imply equal numerical increments.

Figure data and text version

Score band	True positives	False positives
Band 1	464	1,954
Band 2	445	1,042
Band 3	407	456
Band 4	341	156
Band 5	236	52
Band 6	118	13

Figure 06 / 12

A transparent loss-and-friction calculation

At $160 severity per missed synthetic positive, residual loss is $39,360. Review costs $5,658; lost contribution on false alarms is $3,580. The calculation assumes intervention prevents all flagged-positive loss and each false alarm loses the stated contribution. Relax those assumptions before applying it to a real policy.

Figure data and text version

Cost component	USD
Missed-positive loss	39,360
Review cost	5,658
False-alarm contribution	3,580

Figure 07 / 12

Observed outcomes mature over time

The final synthetic positive count is 473. Earlier observations reveal only a stated fraction. Comparing a day-1 cohort with a day-30 cohort would confuse label age with control quality. This curve models observation delay only; it does not change the final outcome. Horizontal positions are the labeled observations or scenarios; equal spacing does not imply equal numerical increments.

Figure data and text version

Days after event	Observed positives
1	85
3	166
7	284
14	388
30	473

Figure 08 / 12

Review demand and available capacity

The flag count is 943. The comparison capacity is an illustrative 540 reviews per cohort window. A mathematical rule can be coherent while its resulting workload exceeds the operating team’s capacity. Capacity is not permission to ignore an applicable mandatory control.

Figure data and text version

Queue measure	Items
Flagged for review	943
Available capacity	540
Excess demand	403

Figure 09 / 12

A feature is an observation with provenance

This evidence contract supports separate misuse from ordinary disputes. A value needs its event time, arrival time, scope, and source. Keeping unavailable evidence distinct from a measured zero prevents an outage from becoming a falsely reassuring feature.

Figure data and text version

Field	Example	Meaning
entity_ref	Dispute versus deliberate misuse	Subject of this case
event_time	2026-09-18T09:00:00Z	When the event occurred
received_time	2026-09-18T09:00:02Z	When the system learned it
signal_status	late	Evidence quality, not an outcome
label_definition	Synthetic deliberate misuse	The target used in these calculations

Figure 10 / 12

Missing evidence changes the observed population

The cells show an explicitly constructed completeness profile for three signal groups. The stress condition removes more history and device evidence. Missingness does not prove the target outcome; it changes what the decision process knows.

Figure data and text version

Signal group	Available	Missing
Identity evidence	11,880	1,620
Activity history	9,720	3,780
Context signal	8,100	5,400

Figure 11 / 12

Evidence, score, and action remain separate

The policy can use separate misuse from ordinary disputes only within its approved scope. The action record must retain which evidence was available, which model or rule ran, and which action was actually applied. The final action can differ from the score recommendation when a separate constraint applies.

Figure data and text version

Stage	Record
Observe	Dispute versus deliberate misuse: evidence as of the decision time
Evaluate	Rule flags 943 of 13,500 customer dispute records
Apply	Record action, reason, owner, and expiry
Reconcile	Join the action to later outcomes without overwriting history

Figure 12 / 12

What the result cannot establish

Observed classifications do not reveal every counterfactual. The synthetic labels make arithmetic possible, but production decline data is selected by prior policy. Keep measured outcomes, assumed prevention, and unknown alternatives separate when reporting impact.

Figure data and text version

Claim	Evidence in this case	Limit
Detected target	227 known synthetic positives flagged	Production labels may be delayed or wrong
Prevented loss	Assumed 36,320 USD	Requires an intervention-effect assumption
Customer impact	716 synthetic negatives flagged	Not every flag causes abandonment
Unobserved alternative	Outcome without the action	Needs a valid evaluation design

Connect the result to the system

Keep allegations, reviewed findings, commercial failures, and revised outcomes distinct.

Check the population, evidence, permitted action, and actual effect together. A balanced calculation can still use the wrong population; a successful response can still leave an unknown financial outcome. The case’s numerical result applies only to its stated assumptions.