How tall is a Falcon 9 rocket? Players nearly triple it

When 295 players of a daily size game had to draw it, the typical answer was 194 metres. Almost one in four drew it five times too tall or more.

I did not expect a rocket to grow like that. In the same game, the Hindenburg shrank to a fifth of its length, so I assumed big machines get squashed. This one went the other way.

A Falcon 9 is shorter than the building it was drawn next to

SpaceX's workhorse is 70 m from the engines to the tip of the fairing (Wikipedia). It is also very thin: 3.7 m wide, a pencil about nineteen times as tall as it is wide.

In How big? you drag an object to the size you think is right, next to a reference you know. On 8 October the reference was the Atomium in Brussels, 102 m tall. The rocket is shorter than the Atomium. Players drew it almost twice as tall. So the reference did not hold the guesses down.

My hypothesis was that rockets have no scale

Launch footage shows a rocket alone against the sky, often from miles away, with nothing next to it. My hypothesis was that without anything to compare, every big rocket ends up the same size in our heads, whatever its real height. If that were true, the other rockets in the game should land at about the same guess as the Falcon 9.

So I checked the other rockets. The SLS, NASA's moon rocket, is 98 m tall (Smithsonian). Players drew it at 210 m. The Saturn V that flew to the Moon is 111 m (NASA). Players drew it at 150 m.

Every big rocket lands near 200 metres

Put the three side by side at the same scale. The smallest real rocket got almost the tallest guess, and the tallest one got the smallest.

The hypothesis held. The real rockets range from 70 to 111 m, the guesses all sit at about 150 to 210 m. That is why the Falcon 9 nearly tripled while the Saturn V grew by only a third. A small rocket stays small, though: the V-2 from the 1940s, 14 m tall, came out at about 12 m.

The pattern is clearer when every guess is drawn. Where the outlines pile up, the picture turns bright, and for all three big rockets that bright zone sits at roughly the same height.

It is not the usual height illusion

The obvious objection is that people always overestimate heights. They do, but by much less. Looking at real tall objects, people see them 10 to 25 percent taller than they are (Yang, Dixon and Proffitt, 1999). A factor of nearly three is something else.

A better fit is an old finding about memory. When a memory is vague, an estimate slides towards the typical member of its category, a model that goes back to Huttenlocher and colleagues in 1991. My hunch is that "a big rocket" is one such category in our heads, and its typical size sits near 200 m. Real rockets are all shorter.

What this can and can't tell us

These are players who chose to play a daily browser game, one guess per page view, mostly in English. That is not a representative sample, and some will have known the answer. But when 271 of 295 miss in the same direction, the direction is hard to argue with. The three rockets come from different days and different players, so the comparison shows a pattern, not a law.

Next time you watch a launch

Look for something next to the rocket that you know the size of. Usually there is nothing, and now you know which way your eye will go. For another thing players get wrong in one direction, read why old prices get guessed too high.