Normative Expected-Value Benchmark For Choice

Issue 61 Edition 2026-03-02 5 min read

General

Sources: 1 • Confidence: Medium • Updated: 2026-03-02 19:40

Key takeaways

In a choice between a 40% chance at $1,000 and a 70% chance at $600, the higher-expected-value option is the 70%/$600 option.
A PDF slide deck for the lecture will be posted on thatneuroscienceguy.com.
Decision-making can be framed as the product of neurons firing in different brain regions, with prefrontal cortex activity linked to analytical decisions and amygdala activity linked to emotional responses.
Value is subjective and is not inherently monetary; low-priced items can carry high personal value.
People differ in decision thresholds; lower thresholds tend to produce faster decisions and higher thresholds tend to produce more indecision, and thresholds can vary across time and decision types.

In a choice between a 40% chance at $1,000 and a 70% chance at $600, the higher-expected-value option is the 70%/$600 option.
Utilitarianism frames human choice as seeking actions that increase utility (reward) and avoiding actions that decrease utility.
Expected value is computed by multiplying an outcome’s value by its probability and can be used to guide choice among options.
A simplified normative model of decision-making is to compute expected values and always choose the option with the highest expected value.
Lottery participation and casino games are negative expected value propositions, and a strict expected-value approach would recommend not playing.

A PDF slide deck for the lecture will be posted on thatneuroscienceguy.com.
The lecture audio is unedited and may include mistakes and repetitions because recording equipment failed for the live talk.
The episode is a lecture titled "Why We Do the Dumb Things We Do" and is part one of a two-part series.

Decision-making can be framed as the product of neurons firing in different brain regions, with prefrontal cortex activity linked to analytical decisions and amygdala activity linked to emotional responses.
Monkey studies attributed to the NYU Glimcher Lab reported that monkeys learn to choose higher expected value options and that neurons in lateral intraparietal cortex (area LIP) scale firing rates with expected value.
An fMRI study titled "Cultural Objects Modulate Reward Circuitry" reported that ventral striatum activity tracks participants’ attractiveness ratings of cars, with higher activity for more attractive categories (e.g., sports cars).

Value is subjective and is not inherently monetary; low-priced items can carry high personal value.
Values can change over time and even during deliberation as attention shifts between attributes and imagined experiences.

People differ in decision thresholds; lower thresholds tend to produce faster decisions and higher thresholds tend to produce more indecision, and thresholds can vary across time and decision types.

Was part two of the lecture series published, and does it materially change, qualify, or add constraints to the mechanisms introduced in part one?
Was the promised PDF slide deck actually posted on thatneuroscienceguy.com, and does it contain definitions, diagrams, or citations that tighten or correct the spoken claims?
What specific evidence (within the cited or underlying work) supports the mapping of prefrontal cortex to analytical decisions and amygdala to emotional responses for the decision types discussed?
Do the reported monkey LIP expected-value findings replicate across tasks/species and include causal manipulations that affect expected-value-based choice?
Does ventral striatum activity (as reported in the car-attractiveness fMRI study) predict behavior beyond ratings (e.g., stable choices) or generalize beyond the specific stimulus set and experimental setting?