---
title: "Short note: Base rate fallacy"
subject: "Data Mining"
chapter: 6
key: "semester_7/data_mining:base-rate-fallacy-note"
source: https://www.bctengineer.com/q/semester_7/data_mining/base-rate-fallacy-note
site: BCT Engineering
---

# Short note: Base rate fallacy

Data Mining · Chapter 6: Anomaly / Fraud Detection

- Asked in: year not given
- Video: [The Base Rate Fallacy](https://www.youtube.com/watch?v=J386AhqyT4s) — Gary Austin
- Answer checked by Claude Opus 5.5 (high effort) on 2026-10-14
- Page: https://www.bctengineer.com/q/semester_7/data_mining/base-rate-fallacy-note
- Whole chapter: https://www.bctengineer.com/md/semester_7/data_mining/chapter-6

## Question

Write a short note on Base rate fallacy.

## Short answer

Base rate fallacy ignores the prior probability of a rare event. With a 99% detection rate, 1% false alarm rate, and 0.001 base rate, Bayes' theorem gives P(intrusion|alarm) ≈ 0.09, so about 91% of alarms are false. Thus false alarm rate must be extremely low and accuracy is misleading.

## Answer

The **base rate fallacy** is the error of ignoring the **prior probability (base rate)** of an event when judging a conditional probability. It matters in anomaly and intrusion detection, where the real anomalies are extremely rare, so even a very accurate detector produces mostly false alarms.

#### Explanation (Bayes' theorem)
P(Anomaly | Alarm) = P(Alarm | Anomaly)·P(Anomaly) / [P(Alarm | Anomaly)·P(Anomaly) + P(Alarm | Normal)·P(Normal)]

#### Example
An intrusion detector has a detection rate (true positive rate) of 99% and a false alarm rate of 1%. Only 1 in 1000 events is a real intrusion (base rate = 0.001).

P(Intrusion | Alarm) = (0.99 × 0.001) / (0.99 × 0.001 + 0.01 × 0.999) = 0.00099 / 0.01098 ≈ **0.09**

So only about **9%** of alarms are real intrusions; 91% are false alarms, although the detector looks 99% accurate. Out of 100,000 events: 100 intrusions (99 detected) and 99,900 normal (999 false alarms), so 99 of 1098 alarms are true.

#### Consequences
- The detection rate alone is misleading; the **false alarm rate must be extremely low** (e.g. 1 in 10⁵) to get a useful detector.
- Too many false alarms waste investigation effort and users start to ignore the system.
- Evaluate with precision, ROC curves and cost-based measures, not with accuracy.

## Check yourself

The answer's key points as multiple-choice questions, with the correct option and why.

1. What is the base rate fallacy?
   A. Ignoring the prior probability of an event when judging a conditional probability
   B. Assuming all events are equally likely to occur
   C. Using too small a training sample
   D. Confusing precision with recall
   Answer: A. It is the error of neglecting the prior (base rate) of an event, e.g. the rarity of real anomalies. (Key point: Definition)

2. Why does a 99% accurate intrusion detector still give mostly false alarms?
   A. Real intrusions are extremely rare compared with normal events
   B. The detector has a low detection rate
   C. Normal events are always flagged
   D. Bayes' theorem does not apply to detectors
   Answer: A. With a base rate of 0.001, the many normal events times a 1% false alarm rate outnumber the few real intrusions detected. (Key point: Why it matters)

3. With detection rate 99%, false alarm rate 1% and base rate 0.001, what is P(Intrusion | Alarm)?
   A. About 0.99
   B. About 0.50
   C. About 0.09
   D. About 0.01
   Answer: C. 0.00099 / (0.00099 + 0.00999) = 0.00099/0.01098, about 0.09. (Key point: Worked result)

4. What must be very low to make such a detector useful?
   A. The detection rate
   B. The false alarm rate
   C. The number of events
   D. The base rate of normal events
   Answer: B. Because the base rate is tiny, the false alarm rate must be extremely small (e.g. 1 in 10^5). (Key point: Consequence)

## Related in other subjects

- [Design period, base period, demand types](https://www.bctengineer.com/question.md?key=semester_5%2Fwater_supply_engineering%3Adesign-base-period-demand-types) — Water Supply Engineering, chapter 3
- [Demand of rural municipality, base period 3 years](https://www.bctengineer.com/question.md?key=semester_5%2Fwater_supply_engineering%3Ademand-bk79) — Water Supply Engineering, chapter 3
- [Demand of village 2070 BS, base period 5 years](https://www.bctengineer.com/question.md?key=semester_5%2Fwater_supply_engineering%3Ademand-c70) — Water Supply Engineering, chapter 3
- [Rural village demand, survey 2070 BS, base year 2073](https://www.bctengineer.com/question.md?key=semester_5%2Fwater_supply_engineering%3Ademand-as70) — Water Supply Engineering, chapter 3