Solve Any Hard Problem Faster by Thinking Mathematically

Solving a hard problem faster comes down to a small set of thinking habits. Anyone can learn them. They are not a special gift for numbers. Each habit strips a confusing situation down to its essential structure. That reveals the path through it directly. The same handful of moves crack century-old puzzles. They work just as well on a difficult negotiation, a messy household project, or a decision with too many moving parts.

How to Approach a Problem That Feels Too Big to Start

• Strip any confusing situation down to the few variables that actually matter, so the real problem becomes visible.
• Pick a problem with exactly one new piece to figure out, while the rest already makes sense.
• Turn a problem with too many moving parts into one with fewer, by treating part of it as fixed for now.
• Represent a problem a different way when progress stalls, since a fresh angle often reveals a solution pure effort could not.
• Treat every failed attempt as useful information that narrows down what the real answer must look like.
• Expect persistent difficulty as the normal cost of tackling something genuinely hard, and keep going anyway.
• Bring in someone else's fresh perspective, since a problem that resists one mind often opens up once it is shared.

Reduce Any Messy Problem to Its Essential Shape

The first move in solving anything complicated is simple. Decide what actually matters. Set the rest aside. Planning a route under a deadline comes down to speed, distance, and the stops available. The colour of the vehicle and the names of the streets do not matter, so strip them away. This is not a loss of information. It sharpens your focus onto the relationships that decide the outcome. A wall of detail becomes a clear starting point.

The same instinct is already present in everyday reasoning. Most people never think to call it analytical. Sharing a pizza among four people gives smaller slices than sharing it among three. Understanding that is genuine structural reasoning, and it needs no training. Children playing a game of naming the largest number soon spot something. Any number can be beaten by adding one. So there is no largest number at all. That is proof by contradiction, which shows a claim is true by proving its opposite impossible. It is one of the most powerful tools in formal reasoning, and children reach it through play. The habit is not rare. It just needs a deliberate context to come out.

Two simple questions turn this instinct into a repeatable practice. First, ask what the essential features of the situation are, and name them plainly. Second, ask what can be safely ignored without changing the outcome, and set it aside. The result can feel colder and more abstract than the original problem. That trade is worth making. Stripping away the specific context is what lets a new problem reveal itself as a familiar one already solved elsewhere.

Choose Problems That Stretch You Without Stopping You

Picking the right problem makes the difference between fast progress and wasted effort. The most useful problems sit just past what you can already do. They hold exactly one genuine unknown, while the rest of the structure is familiar enough to stand on. Too many unknowns and there is no foothold. Too few and nothing is learned. This applies to choosing a work project or a skill to develop next, just as it does to any formal problem. It gives you a simple test before you commit your time.

When several things change at once, ask a different question. What would this look like if one moving part were held fixed for now? A seventeenth-century astronomer tried to map the orbit of a planet that never sat still. He found his answer in a known cycle. That planet returned to the exact same position on a fixed schedule. So he could treat it as a fixed point long enough to work out everything else. The same halving logic helped a 1950s military disease-screening programme. It found one infected sample among a hundred in only seven tests. It did this by cutting the group of suspects in half again and again, rather than testing everyone. Complex problems usually hide a simpler one inside them once the right part is held still.

Making a problem deliberately easier first is a related and equally legitimate move. It is not a shortcut to feel guilty about. You can relax an exact-answer requirement. You can treat a curved relationship as roughly straight. You can accept a good-enough answer instead of a perfect one. Each counts as a valid first attempt when the situation does not truly need exact precision. Physicists joke about calculating the speed of a cow by first assuming it is a frictionless sphere. The absurd simplification gives a workable answer, and the real shape and friction can be corrected in afterwards. The same move works on ordinary planning. Allow a fractional answer first, like ten and a half people for a task. Then round sensibly. That often reveals the real structure faster than insisting on a whole number from the start.

Find the Fresh Angle That Unlocks a Stalled Problem

Changing how a problem is represented often beats pushing harder inside the current approach. Watch for one clear signal. You feel convinced the current method must be right, even as it keeps failing. That conviction is usually the sign that you need a different angle, not more effort in the same direction. An airport once tried to shorten the time between a plane landing and bags reaching the carousel. Complaints did not drop. Then staff realised the real problem. Passengers were standing still with nothing to do. Lengthening the walk from the gate meant the wait happened in motion, and that solved it. The actual delivery time never changed.

A famous game is played with the numbers one through nine. The first person to hold three numbers that sum to fifteen wins. Without a strategy it feels genuinely hard. Now redraw those nine numbers into a grid where every row, column, and diagonal adds to fifteen. The game becomes tic-tac-toe, which almost everyone already knows how to win. The best opening move is taking the centre number five, and it becomes obvious once the representation changes. Nothing about the underlying numbers has shifted. Three simple habits trigger the same kind of shift. Walk away from your notes and think it through unaided. Explain the problem out loud to someone with no background in it. Or physically act out a spatial puzzle with your hands or body.

Let a Wrong Attempt Tell You What to Try Next

A failed attempt at a hard problem builds toward the answer. It does not count against you. It narrows down what a correct approach has to look like. The specific way an attempt breaks down usually points toward what was missing. An unexpected result deserves more attention than a confirmation. It means your working picture of the problem has a gap worth finding. One of the most useful discoveries in medicine began exactly this way. A scientist found a bacterial culture affected by mould, and chose to investigate the change rather than throw the sample away.

Persistent difficulty and self-doubt are the ordinary condition of working at the edge of real challenge. So is the sense that everyone around you understands more than you do. None of it is evidence that a problem is beyond you. Some problems genuinely cannot be solved with the tools available right now. Setting one aside to work on something else is a deliberate use of your time, not giving up. Progress made elsewhere can later hand you exactly the tool you were lacking. Years of quiet effort can turn out to be the reason a resistant problem suddenly gives way all at once.

Use Other People to Think Faster Than You Can Alone

Working with someone else on a hard problem multiplies your thinking. It does more than divide the labour. Pair a person who generates ideas freely with someone who rigorously tests and rejects the weak ones. Together they filter possibilities far faster than either working solo. Agree in advance never to measure whose contribution was bigger. That removes the most common source of friction in shared work. Simply explaining a difficult problem out loud helps too. It works whether or not the listener knows the subject. Putting reasoning into words exposes gaps that stay invisible while you think silently.

Open online collaboration adds another kind of reach. A hard problem is posted publicly and picked apart by many people with different specialities. That can surface an outside connection no single expert would have found. One problem sat unsolved for decades. It was eventually cracked after an anonymous online commenter pointed out a link to a completely different puzzle. The original solver dismissed the idea at first, then came back and confirmed it. These habits carry into ordinary situations too. Scaling up a meal for extra guests or working through a difficult conversation uses the same underlying moves. You simplify, you reframe, and you learn from failure wherever you are.

Go deeper with what matters to you

The source works through each of these habits in far more detail. It sets out the exact reasoning behind reducing a many-variable problem to a fixed one. It walks through the specific historical puzzles that show transformation and reframing step by step. It carries a personal account of how repeated failure builds the pattern recognition that experienced problem solvers rely on. It also tells the full story of how a decades-old open problem was finally solved, including the false starts and the years of work that followed one promising but incomplete breakthrough.

You may have a specific problem you are working on right now. It might have too many unknowns, resist a clear decision, or need breaking into a workable first step. Bring it to the chat and say what you are working through and what you have already tried. It will draw together the relevant habits and examples from the source, and shape an answer around your situation.

Where these ideas come from

These ideas come from Mathematical Thinking, an online course released on 27 January 2022. Terence Tao is a UCLA mathematics professor. He received the Fields Medal (mathematics' highest honour, awarded once every four years) in 2006. He has served on the US President's Council of Advisors on Science and Technology (a panel of experts who advise the president on scientific policy). He is widely regarded as one of the most capable working mathematicians in the world. He draws on decades of research, historical mathematics, and his own working habits throughout the course. If you would like to experience that original work in full, it is well worth seeking out directly.

What you read here is our own source, an independent work built from those ideas. Every concept has been studied, then rewritten from scratch and reshaped so it can answer your questions alongside other refined sources. Nothing from the reference work has been copied. The knowledge has been transformed, not reproduced. The reference is named clearly because the ideas deserve proper credit and because it stands on its own merits.