Word games translate naturally into daily puzzle format. Most people have an intuitive sense of what sounds like a real word, which gives them traction on a puzzle even when they don't know the answer yet. That phonological instinct is the foundation Wordle is built on.
Number games have a different challenge to solve. There's no equivalent instinct for quantities. When you're faced with a numerical puzzle, you need prior knowledge, deductive logic, or a way to reason from first principles - and most people's relationship with large numbers is shakier than their relationship with language.
The number-based daily games that have emerged have each taken a different approach to this problem. They're worth understanding on their own terms rather than just as Wordle variants.
Nerdle: Equations as the Puzzle Unit
Nerdle replaces the five-letter word with a mathematical equation. The answer is always a valid calculation - something like `48/8+2=8` or `5*9-3=42`. You guess equations, and coloured feedback tells you which numbers and operators are correct and in the right position.
The challenge in Nerdle is dual: you need to think about arithmetic at both the symbol level (which digits and operators appear?) and the equation level (does this form a valid calculation?). You're doing constraint-narrowing similar to Wordle, but across a different kind of logical space.
It's satisfying if mental arithmetic comes naturally to you. For people who find arithmetic effortful, the game will feel more like work than play - the feedback mechanism is clean, but the domain itself creates a higher barrier to entry than language does.
Primel: The Constraint as the Game
Primel is a Wordle variant where the answer is always a five-digit prime number, and your guesses must also be prime. Coloured feedback works exactly like Wordle's: right digit right place, right digit wrong place, absent digit.
The prime constraint dramatically restricts the option space, which sounds like it would make the game easier. In practice it doesn't, because most people don't have strong intuitions about which five-digit numbers are prime. You either need to know your primes, or you need to check each guess before submitting.
There's a small but passionate audience for this game - people who genuinely enjoy prime numbers as objects of interest. For everyone else, the constraint feels like an added friction rather than an added dimension.
Magnitudle: Estimation as the Mechanic
Magnitudle is the odd one out in this category because it doesn't ask you to manipulate numbers or test arithmetic - it asks you to estimate real-world quantities.
Each day, the question is something like: "How many managed honeybees are there in the world?" or "How many words are in the Harry Potter series?" You type your best guess, and your score reflects how close you are in order of magnitude - not exact precision.
There's no calculation involved. The challenge is entirely reasoning: what do I know about this topic, what can I infer, and what's my best estimate of the right scale? That openness is both what makes it different and what makes it harder to describe. Nerdle has a defined method - do the arithmetic. Primel has a defined method - test primes. Magnitudle doesn't have a method, because the question is different every day.
When I was designing it, I was trying to build a number game that didn't feel like a math test. You can't study for Magnitudle. Getting something spectacularly wrong doesn't mean you failed to recall a fact - it means you learned something surprising about the world. That reframing of what it means to be wrong felt important. Getting a trivia question wrong means you didn't know something. Getting an estimation wrong means your mental model of some quantity was off in an interesting direction.
Why Number Games Are Harder to Love
The honest reason is that most people have a complicated history with numbers. Vocabulary belongs to everyone - it's the medium of daily life, and facility with language feels natural. Numbers, especially large ones, carry associations with school, difficulty, and being "not a maths person."
Estimation games sidestep this in a way arithmetic games don't, because the skill being tested isn't arithmetic or number theory - it's scale intuition. The questions aren't asking you to calculate anything. They're asking you to think about how big things are, which is more like a spatial or conceptual skill than a mathematical one.
But even so, the number game format starts with less goodwill than word games do. It requires players to develop new reference points - for what a billion feels like, for what "global scale" implies, for how to sanity-check a large estimate. That's worth doing, but it takes some investment.
For a deeper look at how estimation actually works as a technique, see What Is an Estimation Game? And for how Magnitudle scores guesses in a way that makes the investment worthwhile, see How Does Magnitudle Scoring Work?