During World War I, the Atlantic was in chaos. German U-boats prowled the seas, looking for Allied ships to destroy. To counter this threat, Britain turned to an unlikely tool: art. Bold, irregular geometric patterns — known as dazzle camouflage — were painted on ships to confuse enemy submariners.

The idea was simple: distort the perception of a ship’s speed, direction, and size, making it harder to target. But did it ever work? A new study revisits a century-old experiment to find out.

The original research, conducted in 1919 by Leo Blodgett, an MIT student, claimed that dazzle camouflage could mislead observers by up to 60 degrees. But when Timothy Meese and Samantha Strong of Aston University reanalyzed Blodgett’s data, they found a much more diminished effect than claimed. The dazzle effect was just 7 degrees of misdirection.

Instead, a much larger perceptual bias, unrelated to the dazzle camouflage, was at play: the “horizon effect”.

Dazzled and Confused

Dazzle camouflage was born of desperation. Submarines were a relatively new weapon at the time and were wreaking havoc on Allied shipping. No good counter had yet been invented. Traditional camouflage, designed to blend ships into the ocean, was ineffective. Instead, artists like Norman Wilkinson proposed using high-contrast patterns to disrupt the enemy’s ability to judge a ship’s course. “The only course open is to paint [each ship] in such a way as to deceive the attacker as to her size and course,” Wilkinson wrote in 1917.

U-boat commanders had to fire in anticipation of where a ship was heading and dazzle aimed to break up the constructional lines, making it difficult to identify a ship and estimate its speed, range and course. Thousands of merchant ships and hundreds of naval vessels were “dazzled” by the end of WWI. Dazzle camouflage was resurrected by the U.S. during World War II, and was used on the decks of ships as well, in an effort to confuse enemy aircraft.

It was only after the war that someone like Blodgett actually thought of doing an experiment to scientifically test this idea. Using scale models and mechanical simulations, he reported dramatic errors in observers’ perceptions of ship direction. But his methods were flawed. The study lacked modern statistical rigor, and its organization made it difficult to evaluate. Meese and Strong corrected these flaws and redid the experiment with modern tools.

The horizon beats the dazzle

What they discovered was surprising. While Blodgett had attributed the large perception errors to the dazzle patterns, Meese and Strong found that most of the misdirection came from a phenomenon they call “hysteresis.” This is a tendency for observers to perceive a ship’s direction as closer to the horizon, regardless of its actual course. This optical illusion is so powerful that even experienced naval officers fall for it.

The effect was substantial, causing errors of 19 to 23 degrees. The dazzle patterns themselves contributed only a small twist of about 7 degrees.

“The bow of the ship would sometimes twist towards the observer, not away,” said Meese, a vision scientist at Aston University. “This was unexpected and shows how complex visual perception can be.”

So the twist effect is real, but it’s much smaller than Blodgett thought. The bigger effect is this pull toward the horizon, which has nothing to do with the camouflage.

To confirm their findings, Meese and Strong conducted a new experiment using edited images of Blodgett’s original ship models. They replaced the dazzle patterns with neutral gray and asked participants to judge the ships’ directions. Both dazzle-painted ships and plain ones produced the horizon effect.

“This is a clear case where perception overrides knowledge,” said Professor Meese. “Back then, no one realized the horizon effect existed. Blodgett attributed all the confusion to dazzle camouflage. But now we know better.”

Why Did Dazzle Camouflage Fall Short?

Why didn’t dazzle camouflage live up to its promise? One possibility is that the patterns were too subtle to overcome the natural biases in human perception. Another is that the conditions at sea — changing light, weather, and waves — made it difficult for the patterns to have a consistent effect.

Its military effectiveness remains debated. While some reports claimed that dazzle-painted ships were less likely to be sunk, others found no clear evidence of its success. For instance, in the first quarter after dazzled ships entered service with the Royal Navy, around 72% of dazzled ships that were attacked were sunk, compared to around 62% of non-dazzled ships. However, during the second quarter the fortunes were reversed: 60% of attacked dazzled ships were sunk compared to 68% of no-dazzled ships. 

As for Blodgett, his work may have overestimated the power of dazzle, but it remains a fascinating chapter in the history of science and war. He was ahead of his time in trying to quantify these effects. Today’s scientists are just picking up where they left off.

The findings appeared in the journal i-Perception.