Learn it → Apply it → Save as Skill
The more options you put in front of someone, the longer they take to decide. That relationship is not a straight line. Each added choice adds a shrinking amount of extra time, so the first few options cost far more than the next few.
The Dr. Wallace’s Favorite Mental Models series is designed to turn useful ways of thinking into reusable tools. Each model gives you a fast visual overview, deeper interactive learning, practical application exercises and an AI-ready Markdown skill you can save for future use.
Instead of merely reading about mental models, you can gradually build your own AI skills library: a collection of portable reasoning skills that can help ChatGPT, Claude, Grok and other AI assistants apply the mental models you value.
Every option you add places a small, predictable tax on decision time. That tax grows with the logarithm of the number of choices, not with the raw count, so going from 2 options to 4 costs more time than going from 16 to 18.
Important: doubling the options does not double the time. The curve bends, so the biggest gains usually come from cutting the first few unnecessary choices, not the last few.
The model is most useful when someone has to look at a list, a menu or a set of alternatives before they can act, and speed or completion rate matters.
Hick’s Law does not tell you which choice is correct. It tells you how the number and structure of choices affects how long someone takes to choose at all.
List every option a person must weigh before they can act, including the rarely used ones hiding in a submenu.
Group related options into categories. A large, well-organized set can feel faster than a short, disorganized one.
Remove, merge or default the options that rarely change the outcome. Keep the ones that carry real weight.
Time how long people actually take, and watch where they pause, backtrack or give up.
Open a use case to see what the model changes.
Group items by task rather than by internal category, and check whether people can jump straight to a known item instead of scanning the whole list. Alphabetical or task-based order lets people search in a pattern closer to halving the list each step, instead of reading it line by line.
Chunk the list into a few categories, pre-select the most common value, and move rare options behind a clearly labeled “more options” control rather than deleting them.
Pre-commit to routines for recurring, low-stakes choices such as meals or outfits. Removing a repeated decision point frees attention for the choices that actually deserve it.
Present four to six top-level reasons first, written the way customers describe their problem rather than how the internal team organizes it, then branch further only once narrowed.
Pre-filter proposals to a short list before the room decides, so discussion time goes to real tradeoffs. Keep a visible record of what was cut and why, so filtering does not look like suppression.
Show a small number of tiers with sensible defaults rather than every individual add-on at once. Let people expand to see detail instead of forcing them to evaluate it up front.
This is an illustrative example designed to teach the method.
Expand for the research behind the logarithmic relationship and where it holds less well.
William Edmund Hick published “On the rate of gain of information” in 1952 after experiments using ten lamps arranged in a circle, each paired with a response key. He was building on earlier observations: Franciscus Donders had reported in 1868 that reaction time rises with the number of possible responses, and Jakob Merkel found in 1885 that response time increased as the stimulus set grew larger.
Ray Hyman’s 1953 follow-up experiments, using arrangements of lights, established a close-to-linear relationship between reaction time and the information carried by the number of choices. Because of this, the formal version is often called the Hick-Hyman law.
For equally likely choices, T = a + b × log2(n + 1), where T is reaction time, n is the number of choices, and a and b are constants fitted from the data. The “plus 1” accounts for the uncertainty of whether to respond at all. The key point for everyday use: doubling the number of choices does not double the time it takes to choose.
A 2018 review by Proctor and Schneider documents real exceptions: well-learned verbal responses to familiar stimuli barely slow down as choices increase, eye movements between targets often show no increase or even a decrease, and searching an unordered list tends to follow a straight line rather than a logarithmic curve. Treat the law as a strong default for unfamiliar, deliberate choices, not a universal constant.
Hick’s Law predicts how long a decision takes, not whether it was the right one. A faster choice is not automatically a better choice.
Doubling the options does not double the time. Because the curve bends, the first few unnecessary choices usually cost more time than the last few you could trim.
Twelve flat, unsorted options can feel slower than twelve well-grouped ones. Good categories effectively shrink the number of things a person weighs at once.
Hiding an option people actually need just pushes the search somewhere else, often adding frustration, a support ticket, or an abandoned task.
A well-practiced, automatic response, such as an experienced pilot’s trained reaction or a familiar keyboard shortcut, does not slow down the same way a novel decision does.
Trimming a team’s options to speed up a meeting can quietly suppress real disagreement. Check this against Groupthink and the Abilene Paradox before calling a quick decision a good one.
legal, medical or financial rules require presenting every material option, even if that makes the choice slower.
the person is a trained specialist whose speed already comes from practice, not from a shorter menu.
the decision is high-stakes or hard to reverse, and you want people to slow down and consider more, not less.
This browser tool structures your thinking. It does not send your information anywhere or perform AI analysis.
Save as Skill converts the model into portable Markdown instructions that teach an AI when and how to apply Hick’s Law, what questions to ask, what evidence to seek, what mistakes to avoid and how to structure its analysis.
Save the skills you find useful and organize them into your own reusable library. The Markdown is intentionally model-neutral, so it can be used wherever an AI system accepts uploaded files, project knowledge, custom instructions or reusable text context. Platform capabilities differ, so this is a portable skill file rather than a claim of one-click installation.
No. It predicts time, not whether a shorter list is better. Good chunking and layout can let a large set feel fast, so the goal is matching structure to the real decision, not minimizing the number shown on screen.
People appear to subdivide a large set into smaller groups and eliminate roughly half of the remaining options at each step, a strategy that resembles a binary search. This is one reason a long, well-organized list can outperform a short, disorganized one.
When the layout of options matches how people naturally expect to respond to them, reaction time drops regardless of how many options exist. A menu arranged the way people already think about the choice can beat a shorter menu arranged in a confusing order.