Tower of London Test

The neuropsychologist Tim Shallice designed the task in 1982 to study planning [1]. Each problem has a smallest number of moves, and the task records whether you reach the goal in that number, how many extra moves you need, and how long you think before your first move. It looks a little like the Tower of Hanoi, but the rules and the problems are different.

The number of moves is not the only thing that makes a problem hard. Problems where the goal leaves the order of the moves open, or where you first have to move a ball out of the way, take people longer to plan, even at the same number of moves [2][3][4].

• Three pegs hold 3, 2 and 1 balls. Move one ball at a time: only a ball on top can move, and never onto a full peg.
• Each problem shows the goal and the fewest moves it can be done in. Copy the goal in that many moves for a perfect solve.
• 16 problems, from 2 to 6 moves, getting harder. Each one ends after 60 seconds or 20 moves.
• Tap a ball, then tap the peg to put it on, or drag it. On a keyboard, the keys 1, 2 and 3 pick up and put down.
• A short practice problem comes first (you can skip it). Each run paints the balls in new colours, so a second try looks new but is just as hard.

We made our own set of problems. A computer search through all 36 positions of the balls checked that each problem's fewest moves is exact, and from 3 moves up every problem has only one shortest solution. The goals mix the three shapes researchers vary: all balls on one peg, two and one, and one per peg [2].

• Solved in the fewest moves: your main score, the number of perfect solves.
• Extra moves: moves beyond the fewest, on the problems you solved.
• Thinking before the first move: how long you looked before touching a ball. Researchers call this initial planning time.
• How you solved: the share of your time spent before the first move on the longer problems. A bigger share means you planned more of the solution in your head first.

In one study of 179 adults, people who spent a bigger share of their time planning before the first move tended to solve more problems in the fewest moves [5]. Telling people to plan first helped in one study [6] but not in another, where people planned longer without moving better [7]. So treat "how you solved" as a description of this run, not as a rule.

Once enough people have played, your result also shows where you stand among other players of this game, as a band: below typical, typical or above typical. A band, not an exact rank, because one run of this task says only roughly how you plan (see below). The other players are visitors of this site, not a representative sample.

The Tower of London is used in research on planning and in clinical assessment, but this page is a game, not a clinical test. It doesn't measure intelligence, and one run is a snapshot: tiredness and the device you use change how you do.

In a study of 79 young adults who did a computer Tower of London twice, two weeks apart, their move scores on the two days agreed only weakly (a correlation of 0.15), while they became faster at solving [9]. So treat one result as a rough guide, and look at several tries on different days rather than one number.