Hacker News Reader

Comment by w1ntermute | Hacker News Reader

Actually, with reCAPTCHA you can now get a prefab solution that can't be cracked.

reCAPTCHA, after extensive deployment in production, now becomes extremely difficult to solve from time to time. It just renders impossible to recognize scan images. User frustration is serious concern. If you care about conversion rate, DO NOT USE RECAPTCHA.

Yeah, one time I got an image of a sum. So I just put in the LaTeX equivalent, something like \Sigma_{n=1}^{100}, and it accepted it.

Since OCR was only able to parse one of the words, they have no way of checking the other (besides taking votes from a ton of submissions). Because the summation was not the word that was checked, you could enter anything.

It's an interesting problem. If they make it too clear that they don't know the right answer, quality of submissions will plummet. If they keep the current form, then usability suffers.

There's actually a game of this on 4chan.org/b/ there's a pattern to knowing which word is known and which isn't because the known word is put through a recognizable filter and the unknown one isn't (most of the time). They then put in a specific racial slur in for the unknown word, hoping that just once it'll slip through enough that it makes it into the final OCR'd text.

If you succeed at the control (the one it already thinks it knows) it takes your other answer as input for future viewers.

Only if multiple people give the same response, though.

If captcha system isn't reliable, a new system of deterring spam registration is in order.. maybe the system uses images to direct the user to move the mouse in a certain pattern and then use mouse gesture detection to ascertain if its human..

This is a BRILLIANT idea!!

What? reCAPTCHA is very predictable. There is one word that always looks the same. It's the word, and then the word blurred and slightly skewed and placed on top of itself. The other is a scanned word (or somethings symbols or mathematical equation). That's the bit where they're crowd sourcing book scanning.

You don't have to get it right, and in fact you can type in gibberish if you like, as long as you get the known one right, and the known one is very easy for a human to read.

I know, I know. I like the idea of reCAPTCHA to crowdsource a very hard problem for the common good. I know how it works in and out. It is awesome.

The only problem: things sound nice in theory do not always work well in practice. I ran into multiple occasions where reCAPTCHA presented me impossible to recognize scans, one extreme example is a printed texture pattern with not a single letter on it.

Your average, non-tech-savvy customers will never know how the system works. They shouldn't either, otherwise the whole idea of reCAPTCHA would have been defeated.

Judging by the volume of complaints about it on various support forums, it is really a bad idea to deploy it in production when you care even slightly about usability, user experience, or simply conversion rate.

Unless your application is a must-have.

Interesting, I'd never really given it a second thought. I appreciated the drop-in ease bit and hadn't considered its effect on conversion.

> "can't be cracked"

Not sure about that. Certainly tough, but crackable.

http://stackoverflow.com/questions/448963/has-recaptcha-been...

http://www.allspammedup.com/2011/01/google-recaptcha-cracked...

It looks like they've artificially weakened it:

> reCAPTCHA uses a “one-off” system. That means a letter in a word can be incorrect, and it will still be accepted by the system.

It is also only verifying 1 word out of the pair; usually the more garbled one.

Reply on news.ycombinator.com