The Limits of Visual Magic: Part One

In March of this year, as part of a larger scale focus group test, the following trick was performed for 42 individuals.

Approximately half of the people (22) saw the trick pretty much as demonstrated in the GIF above.

The other 20 people didn’t see the card change visually. In other words, they were shown the face of the card, then the card was turned back towards them, then turned around and it was a new card, turned away, and finally turned back around to show a third card.

Each spectator rated the trick on a scale of 1-10 as far as how impossible the trick seemed, and gave another rating for how enjoyable the trick was. In addition, they were asked a free response question regarding any ideas they might have as to how the trick was accomplished.

(By the way, after years of testing and trying to come up with a system to standardize reactions and keep the ratings from being bundled in the 7-9 zone, we hit on a simple and obvious solution. I’ll go into that in a future post sometime.)

For the Visual Change -

Impossibility Score: 6.8

Enjoyment Score: 6.2

Where 5 is meant to be “average” for both of these categories, this trick ranked somewhat above average for how impossible it seemed and how much they enjoyed it. That’s not too bad given that there was nothing to the presentation at all. It was literally just, “Take a look at this,” essentially.

For the Implied Change

Impossibility Score: 7.3

Enjoyment Score: 6.6

Both scores jumped a little. Not a crazy amount. But these scores are actually pretty decent for a quick trick performed with no presentation. You’re not really going to get in the 8s or 9s with such a trick, in my experience of testing.

The sample size wasn’t large enough that I would bet my life on the idea that the implied version is definitely better than the visual change. But I am confident the implied version isn’t significantly weaker than the visual version.

As magicians, we’re always drawn to visual magic. “Visual” is a term that’s used to sell tricks. And I think it’s difficult—or, at least, it’s difficult for me—to think the visual version of a magic trick wouldn’t be stronger than the implied version in almost all cases. Surely seeing the magic would make for a stronger, more entertaining, trick. Right?

I looked further into the feedback and noticed something which I thought provided an answer.

When looking at the suggested methods for the visual version, 20 out of 22 said that it was likely a “trick card” or words to that effect. They didn’t mention anything about the mechanics, or suggest that they saw that it was a trick card. So I don’t think they registered the flap card… at least not consciously. It’s just, when you see a card changing to other cards so openly, there’s really no other possibility.

With the implied version, a lot of people guessed that it was a “trick card” as well. 17 of the 20 respondents put that as a possibly. But 18 of the 20 also thought it could have been “sleight of hand”—secretly switching cards in and out.

You would think having more options for how a trick is done would make it seem less impossible and therefore less entertaining. But that wasn’t the case.

In a way, I thought we had just proven the Too Perfect Theory. When there was only one possibility for people to consider for the method, the trick was weaker than a trick with multiple potential methodologies.

It probably comes down to a question of certainty. I think if we had asked the participant to rate their certainty of how the trick was done, the visual group probably would have stated that they were fairly certain that “trick card” was the answer.

But the implied group likely wouldn’t have been so certain. For them the method might have been a toss-up between the card being gimmicked or the magician secretly switching cards.

That’s the first weakness of very visual tricks—they often lead you to the one and only correct method because it so clearly can’t be anything else. Look at the gif above. No one would suggest that’s sleight of hand. So the card itself becomes the only other option.

But I don’t think that’s the only issue with visual magic. So I wanted to try the test again, but this time I wanted to remove the “too perfect” issue. This time, the card would change in either a visual or implied manner and then that card would be handed directly to the participants to examine. How then would they process the visual or implied effect? That’s coming tomorrow.