Fermi vs Other Daily Estimation Games: Which One Should You Play?

Quick Answer

Fermi is one of the strongest daily estimation games if you want real-world numerical questions that reward reasoning rather than trivia recall. Each day gives you three questions, one numerical guess per question, and a score based on how many times higher or lower your estimate was than the real answer.

Its main advantage is the balance between depth and speed. Three questions are enough to encourage actual back-of-the-envelope reasoning, but a daily round is still short. After guessing, you see the real answer, how far off you were, how other players performed, and an example of the napkin math that could have produced a reasonable estimate.

Other estimation games emphasize different skills. Fermidle cares more about explaining your reasoning. Estimania feels closer to numerical trivia. Give or Take tests confidence ranges. Magnitudle compresses the idea into one daily magnitude guess. Amountle is visual and fast. Ballpark focuses on intuitive scale through sliders.

If you specifically want to become better at answering questions such as How many iPhones have ever been produced? or How many steps does a mail carrier take in a year?, Fermi is the clearest all-round starting point.

What Is Fermi?

Fermi is a daily estimation puzzle based on Fermi estimation: using simple assumptions, rough arithmetic, and real-world intuition to approximate a number you probably do not know directly.

Every day you receive three questions.

A question might ask something like:

How many Oreos are eaten in the US each day?

You are not expected to have memorized the statistic.

Instead, you might reason:

US population × percentage who eat Oreos × average Oreos per eater ÷ days

Your assumptions will be imperfect, but if they are reasonable, the final estimate can land surprisingly close.

That is the central idea behind Fermi: getting into the correct numerical neighborhood matters more than knowing the exact fact.

How Fermi Works

The daily flow is simple.

1. Read the Question

You get a real-world quantity with a numerical answer.

You cannot search for the answer if you want to play as intended. The challenge is to work it out from what you already know.

2. Make One Guess

Enter a single number and lock it in.

Fermi includes a calculator, so arithmetic itself is not supposed to be the main obstacle. The harder task is deciding which assumptions to make.

3. See the Real Answer

After submitting, Fermi immediately reveals the reference answer.

You also see whether you were too high or too low.

4. Check Your Factor Error

Fermi scores guesses by their multiplicative distance from the answer.

If the true answer is 1,000:

• Guess 1,000 → 1×
• Guess 2,000 → 2×
• Guess 500 → 2×
• Guess 10,000 → 10×
• Guess 100 → 10×

Lower is better, with 1× representing a perfect estimate.

This is much fairer for estimation than ordinary absolute error. Missing 100 by 1,000 is disastrous, while missing 100 billion by 1,000 is essentially perfect. Factor scoring recognizes that difference.

5. Review the Napkin Math

One of Fermi's most useful features is the explanation after the answer.

Instead of simply saying you were wrong, it shows an example estimation path that could have gotten you reasonably close.

This turns each puzzle into a small lesson in numerical reasoning.

6. Finish All Three Questions

Your daily result combines the three estimates into an overall factor score.

You can then compare your performance with other players and see where your guesses fell relative to the crowd.

Why Factor Error Matters

Understanding Fermi's scoring changes how you should play.

Suppose the answer could plausibly be anywhere between 1 million and 100 million.

Guessing 50,000,001 simply because it is the arithmetic midpoint is not necessarily the best approach.

For quantities spanning orders of magnitude, think logarithmically.

The midpoint between:

1 million

and

100 million

on a multiplicative scale is roughly:

10 million

because:

1M × 10 = 10M

and

10M × 10 = 100M

Fermi rewards this type of magnitude thinking.

Fermi vs Fermidle

Fermidle is probably the closest conceptual comparison because it also explicitly uses Fermi estimation.

The difference is emphasis.

Fermi

• Three questions per day
• One numerical estimate for each
• Factor-based scoring
• Built-in calculator
• Crowd comparison
• Example napkin math after answering
• Focus on getting the number close

Fermidle

• One main Fermi question per day
• You provide a numerical estimate
• You can explain your reasoning
• An AI judge evaluates the reasoning process
• Performance contributes to an ELO-style rating
• Stronger emphasis on how you reached the answer

Choose Fermi if you primarily enjoy making several estimates and comparing accuracy.

Choose Fermidle if the interesting part for you is explicitly breaking down assumptions and having the reasoning itself evaluated.

A player trying to improve at consulting interviews, forecasting, market sizing, or quantitative reasoning may actually benefit from playing both. Fermi gives more daily repetitions, while Fermidle encourages more deliberate decomposition.

Fermi vs Estimania

Estimania uses a broader numerical-trivia format.

You receive five questions per day across areas such as science, geography, history, and sport. Each answer earns a score from 0 to 100 based on logarithmic distance from the correct value.

A perfect estimate earns the maximum score, while being approximately 10× away reaches zero.

The important distinction is that Estimania can feel more like a mixture of numerical trivia and estimation, while Fermi is explicitly built around questions that invite napkin math.

Consider two question styles:

What is the population of a particular country?

versus:

How many takeaway coffee cups are used in that country every year?

The first may reward factual knowledge heavily.

The second naturally encourages decomposition:

population × coffee drinkers × purchases per week × weeks per year

Fermi tends to lean toward the second experience.

Choose Estimania if you like general knowledge mixed with numerical intuition.

Choose Fermi if you want questions where working backward from assumptions is a major part of the fun.

Fermi vs Give or Take

Give or Take changes the fundamental action from choosing a single number to choosing a range.

Each day presents five real-world questions. Instead of saying:

I think the answer is 2.4 million

you might say:

I think it is between 1.5 million and 4 million.

A narrow range can earn more points if it contains the answer, but an overly confident range can miss completely and score nothing.

That means Give or Take tests something Fermi largely does not:

confidence calibration.

You have to decide not only what you think the answer is, but how uncertain you are.

For example, suppose you estimate a city's annual tourist count at 5 million.

In Fermi, you submit 5 million.

In Give or Take, you must decide whether your defensible range is:

4M–6M

or

2M–12M

The first demonstrates confidence but risks missing. The second is safer but less impressive.

Choose Fermi for point estimates.

Choose Give or Take if you enjoy probability, calibration, forecasting, or deciding how much confidence to place behind an estimate.

Fermi vs Magnitudle

Magnitudle is much more compact.

Its daily format centers on one real-world number and one guess.

The challenge is primarily understanding magnitude. You might estimate a population, scientific quantity, distance, weight, or other large-scale fact.

This makes Magnitudle ideal for players who want an estimation puzzle that can be completed almost instantly.

Fermi provides more room for deliberate reasoning because you receive three questions and can use a calculator before committing.

The difference can be summarized as:

Magnitudle: Can you sense the scale?

Fermi: Can you build a reasonable estimate?

If your daily-game routine is already crowded, Magnitudle is easier to slot in.

If estimation itself is the activity you want to practice, Fermi has more substance.

Fermi vs Amountle

Amountle is technically an estimation game, but the skill it tests is quite different.

Instead of estimating an unseen real-world statistic, you watch objects fall into a container and estimate how many are present.

The daily game uses several fast rounds, with only a short period to submit each estimate.

This tests:

• Visual quantity perception
• Grouping
• Approximate counting
• Attention
• Speed

Fermi instead tests:

• Real-world knowledge
• Assumption making
• Arithmetic decomposition
• Order-of-magnitude intuition
• Numerical reasoning

Someone could be excellent at Amountle and still struggle badly with Fermi.

If you enjoy visual games, Amountle is likely more immediately satisfying.

If you enjoy questions that make you stop and calculate on the back of an imaginary envelope, Fermi is the better fit.

Fermi vs Ballpark

Ballpark presents several daily questions about real-world scale and has you place your estimate using a slider.

Questions can cover geography, science, culture, sport, economics, and everyday quantities.

Ballpark feels particularly approachable because the interface encourages intuitive positioning rather than making every puzzle feel like a math problem.

It is excellent for building a general sense of whether something is:

hundreds

thousands

millions

or

billions.

Fermi is more explicit about the reasoning process. Its questions invite you to create mini-models rather than simply place a value on a scale.

Choose Ballpark for a relaxed daily intuition test.

Choose Fermi if you want to understand why your estimate should be approximately correct.

Fermi vs Interval-Style Games

Interval estimation games introduce a different strategic problem.

Instead of producing your best single guess, you try to construct the smallest defensible interval that contains the real answer.

Suppose the question asks for the annual number of visitors to a famous attraction.

A Fermi approach might produce:

12 million

An interval approach might produce:

8 million–16 million

The second answer communicates both your estimate and your uncertainty.

This resembles forecasting and statistical calibration more closely than ordinary number guessing.

Fermi is easier to understand and faster to play, while interval formats reward players who want to think explicitly about uncertainty.

Comparison at a Glance

Game	Daily Format	What You Submit	Main Skill	Best For
Fermi	3 questions	One number	Napkin math and magnitude	Balanced estimation practice
Fermidle	1 question	Number + reasoning	Explicit reasoning	Deep Fermi practice
Estimania	5 questions	One number	Numerical trivia + scale	General knowledge fans
Give or Take	5 questions	A range	Confidence calibration	Forecasting-minded players
Magnitudle	1 question	One number	Magnitude intuition	Very quick daily play
Amountle	Multiple visual rounds	Object count	Visual estimation	Fast visual puzzle fans
Ballpark	5 challenges	Slider estimate	Sense of scale	Casual numerical intuition

The games overlap, but they are not interchangeable.

A Better Way to Play Fermi

The biggest beginner mistake is entering the first number that feels plausible.

Instead, build the estimate from smaller quantities you can reason about.

Suppose the question is:

How many slices of pizza are eaten in New York City every day?

You probably do not know the answer.

But you can construct one.

Step 1: Estimate the Relevant Population

New York City has roughly:

8 million people

Step 2: Estimate Participation

Maybe an average resident eats pizza equivalent to two slices per week.

That gives:

8M × 2 = 16M slices per week

Step 3: Convert to Days

16M ÷ 7 ≈ 2.3M slices per day

Step 4: Adjust for Visitors

Tourists and commuters may push the number higher.

You might settle on:

2.7 million

The individual assumptions can all be wrong without the final estimate being terrible. Errors often partially cancel.

That is exactly why Fermi estimation works.

Break Questions Into Multiplication Chains

Many Fermi questions become easier when rewritten as:

number of people × frequency × amount

or:

number of locations × activity per location

or:

rate × time

or:

volume ÷ volume per object

For example:

How many tennis balls fit inside a school bus?

can become:

usable bus volume ÷ volume per tennis ball × packing efficiency

You do not need precise inputs. You need inputs that are approximately sensible.

Start With What You Actually Know

Do not invent five unknown numbers when one familiar reference point can anchor the whole problem.

If asked about smartphones, you might know:

• World population is around several billion
• Smartphone ownership is widespread
• Phones are replaced every few years

Those three facts may already put you in the right order of magnitude.

Strong Fermi players are not walking encyclopedias. They are good at connecting a small number of known quantities to an unknown one.

Use Powers of Ten Before Precise Arithmetic

Before calculating 37.4 million × 0.63 × 18.5, first decide whether the answer should roughly be:

10 million

100 million

or

1 billion.

Magnitude errors are usually much more damaging than small arithmetic errors.

Fermi's multiplicative scoring reinforces this.

Being 15% wrong is excellent.

Being 100× wrong because you confused daily and yearly units is not.

Watch Your Units

Unit mistakes destroy otherwise good estimates.

Common traps include:

• Per day vs per year
• Thousands vs millions
• Kilograms vs tonnes
• Miles vs kilometers
• Minutes vs hours
• Individual people vs households
• Revenue vs number of transactions

Write the units into your mental calculation.

For example:

people × coffees/person/day × 365 days/year

is safer than multiplying three unlabeled numbers.

Use Upper and Lower Bounds When Stuck

Even though Fermi ultimately requires one number, thinking in ranges is extremely useful.

Suppose you believe the answer cannot reasonably be below 100,000 or above 10 million.

You now know the scale.

A sensible central estimate on a multiplicative scale is approximately:

1 million

This technique prevents wildly overconfident guesses.

Prefer Multiplicative Midpoints

For quantities spanning large ranges, do not average endpoints arithmetically.

If your lower bound is:

10,000

and your upper bound is:

1,000,000

then the multiplicative midpoint is:

100,000

because it is 10× above the lower bound and 10× below the upper bound.

That matches Fermi's factor-based scoring much better than 505,000.

Sanity-Check the Final Number

After calculating, translate the result back into something intuitive.

Suppose you estimate that Americans eat 50 billion hamburgers per day.

Divide by roughly 350 million people:

50B ÷ 350M ≈ 143 hamburgers per person per day

Clearly impossible.

A ten-second reverse check can catch huge mistakes.

Learn From the Crowd Without Copying It

Fermi shows how other players guessed after you have committed.

This is useful feedback.

If nearly everyone clustered around 10 million and you guessed 10 billion, ask what assumption caused the discrepancy.

But the goal is not to train yourself to guess what the crowd would say.

The valuable skill is identifying why your own model was too high or too low.

Use the Napkin Math as Training Data

Do not skip the post-answer explanation.

Compare it with your own approach.

Maybe you estimated:

population × usage frequency

while the suggested solution used:

number of stores × sales per store.

Both approaches can be valid.

Seeing alternative decompositions gradually gives you a library of estimation patterns that transfer to future questions.

Common Fermi Mistakes

Guessing From Vibe Alone

Gut instinct is useful as a sanity check, but an unexplained number is fragile.

Even a two-step model is usually better.

Using Too Many Assumptions

A 12-variable model feels sophisticated but introduces 12 opportunities for error.

Use the shortest reasonable chain.

Chasing Fake Precision

An answer such as:

47,382,119

is rarely more defensible than:

50 million.

Fermi questions usually contain much more uncertainty in the assumptions than in the arithmetic.

Forgetting Scale

Million-versus-billion errors are much more important than whether the estimate should be 37 million or 42 million.

Treating Every Question as Trivia

If you happen to know the exact answer, great.

But waiting for memory to produce an obscure statistic defeats the point of the game.

When you do not know, start estimating.

Which Daily Estimation Game Should You Choose?

Choose Fermi if you want the best balance of short daily play, real-world questions, mathematical reasoning, and useful post-answer feedback.

Choose Fermidle if you want your reasoning itself to matter and enjoy building a longer-term rating.

Choose Estimania if numerical trivia and broad general knowledge are part of the appeal.

Choose Give or Take if estimating uncertainty is more interesting to you than committing to one number.

Choose Magnitudle if you want a one-question game that takes almost no time.

Choose Amountle if you prefer visual estimation and fast reactions over real-world statistics.

Choose Ballpark if you want a relaxed way to improve your intuition for scale.

There is also no reason to play only one. Because each format trains a slightly different skill, a combination can be more interesting than several near-identical Wordle-style games.

Who Is Fermi Best For?

Fermi is particularly well suited to players who enjoy:

• Estimation puzzles
• Numbers and mental math
• Real-world trivia
• Logical decomposition
• Market-sizing questions
• Consulting-style case questions
• Quantitative interviews
• Forecasting
• Daily brain games
• Comparing results with friends

You do not need advanced mathematics.

Basic multiplication, division, percentages, and a sense of scale are enough for most questions.

In fact, players who dislike formal math may still enjoy Fermi because the difficult part is often deciding what to calculate, not performing the calculation itself.

Final Take

Fermi stands out from many daily estimation games because it sits in a useful middle ground.

It is more analytical than a visual counting game, less demanding than a full calibration exercise, deeper than a single magnitude guess, and faster than a long-form estimation competition.

The three-question format is especially effective. It gives you enough attempts to notice patterns in your thinking without turning the daily puzzle into homework.

The best way to play is not to ask yourself whether you know the answer. Ask instead:

What smaller quantities do I know that could help me build the answer?

Once that habit becomes automatic, Fermi stops feeling like impossible numerical trivia and starts feeling like a compact reasoning game.

For players who enjoy discovering that an apparently unknowable number can be approximated with two or three sensible assumptions, Fermi is one of the most satisfying daily estimation formats available.