The Viral Size Comparison Game, Explained

If you have watched a streamer drag a T. rex across a blueprint until it looked about the length of a school bus, you have seen the size comparison game. It went around Twitch and TikTok this month the way the 0.1-second song game did in August, and for the same reason: one round is ten seconds long, the guess is public, and the reveal is either vindication or comedy. This is what the game is, why it is harder than it looks, and where to play it without a daily cap.

Two outlines share a board. The left one is the reference, with its measured size written on it: a school bus, an adult man, the Statue of Liberty. The right one is the target, and it starts at the wrong size. You drag its corner, or slide a ruler under the board, until the target looks the right size beside the reference, and lock it in. The target then grows or shrinks to its measured size, leaving a dashed outline where you guessed.

That is the whole mechanic, and it is why the game travels so well on video. There is nothing to explain, the guess is visible on screen before the answer is, and the answer arrives as motion rather than as a number.

A guess is judged by how many times too big or too small it was, not by how many metres it missed by. Doubling the true size and halving it are the same mistake, and both score the same. In Size It Up, within 5% of the measured figure is a perfect round, twice off is 37 out of 100, and four times off in either direction scores nothing. Five rounds make a run, the last two count triple, and a flawless run is 1,000.

The ratio is what makes a blue whale and a beetle fair questions on the same board. It is also what makes the game feel unfair the first few times: a guess that looked reasonable can come back as twice off, and twice off is most of the round gone.

The reason the reveal is so often a surprise has a name. Harry Hollingworth measured it in 1910: people judging a quantity drift toward the middle of the range they have in mind, so very large things get guessed too small and very small things too big. Area makes it worse, because a shape twice as big does not look twice as big. Almost everyone's first ten rounds miss the same way, which is why the dashed outline is worth watching. It tells you your direction.

The daily version deals five comparisons a day and then makes you wait. Size It Up Unlimited is Daily's own take on the format, with its own outlines, figures and scoring: five fresh pairs a run, from animals, buildings, vehicles, planets and places, as often as you like, free, in the browser.

It is the one ohnePixel's chat kept asking for a link to: the same outline-and-yardstick round, without the cap. If you came from a clip, the runs carry on after the stream does.

Count in references, not in metres. Ask how many of the left shape would fit along the right one, then check the height as well. Learn what is being measured, because bridges are measured by their main span and some statues include the pedestal. And push past your instinct at the extremes: a guess that already feels enormous for a whale is probably still short. There is a longer guide to getting better at Size It Up, with the yardsticks worth memorising.

The size comparison game is a guess the size game with one round shape: an outline, a yardstick, a drag and a reveal, scored as a ratio. It is quick, it is public, and it is wrong in the same direction for everyone until they learn their lean. Play it without a limit, and read the outline the reveal leaves behind.

Hollingworth, H. L., The Central Tendency of Judgment (Journal of Philosophy, Psychology and Scientific Methods, 1910).

Stevens, S. S., Psychophysics: Introduction to Its Perceptual, Neural, and Social Prospects (Wiley, 1975).