Reverse Turing tests and accessibility

Reverse Turing tests, or CAPTCHAs, have become an unpleasant practice that I hope disappears soon.

The truth is that several CAPTCHAs have made me feel like a complete idiot, to the point where I wonder whether my problem is my astigmatism, or whether I’m simply incapable of recognizing images.

At Factor Humano there is a very thorough article on CAPTCHAs from the accessibility point of view: http://www.webstudio.cl/blog/captchas-y-accesibilidad/.

There are other ways to fight spam. If what worries you is security, CAPTCHAs are not for that, as I explained earlier.

In any case, if what you want is to verify that there’s a human on the other end, there are 2 ways.

• The most expensive is to use biometrics, especially cancelable biometrics.
• The cheapest is to use mouse movement events. For example, you can draw a grid of numbers on colored cells and ask the user to click on one of the numbers, or on a specific sequence.

One of the best CAPTCHAs I’ve seen is, ironically, the one used by that SPAM generator called BlogClicker. The image shows the concept:

It would be even better without distorting the requested number.

It is very hard for a script to simulate mouse clicks (truth is, I can’t think of how to do it, but there are always clever hackers).

Either way, I think this practice should disappear. What we need to do is find better ways to secure our users’ trust, by other means. The new ideas around the laws of identity will probably lead us to actually stop spam.