Phantom upvoting; caused by search engine spiders? Solve w/ nofollow?

[domino66]

Vote counts on my Q&A site (users can upvote various answers) have always been perplexing. Super low-content answers are getting upvoted to suspiciously high numbers. I always felt like it was some sort of bot that must be crawling my site and upvoting stuff.

Dawned on me today that it could be Google and other search engine spiders who "click" the Upvote button just as though it were any other link...and even though it doesn't lead to an internal page, the net effect was that it would increase the Vote Count. Which leads to screwy results when low-content stuff is getting voted as high as great content.

Is it likely that's what's going on?
Is it standard to use a nofollow attribute on things like Voting or FB Like buttons that you don't necessarily want search spiders to "click"?

[netmeg]

You might be able to use event tracking in Google Analytics to see if the click activity shows up there.

[lucy24]

I don't see how "nofollow" could ever make a difference. It doesn't mean "don't request the file at the other end of this link", it just means "don't tell them I sent you". And that's if the link involves a URL that can be requested.

Now, if this is all done with scripting, then there's a completely different possibility. Say your search engine is acting on javascript-- as the major search engines now do. If there's a series of possible selections, and the search engine sets out to try each one in order, but your site is coded so only the first selection from a given IP will be recognized ... then everything gets upvoted and nothing ever gets downvoted.

[domino66]

What you wrote about 'nofollow' makes sense -- I guess it's not a directive to spiders to *not* "click" a link.

But actually, my site has no 'downvoting' option...all you can do is basically "Vote" an answer (which is equivalent to "Liking" it)...so that when you sort the answers in a thread by "best rating", the ones with the most "Votes" are designated as best.

So it's not a matter of things getting upvoted but never downvoted...it's just that there's some sort of bot that's clicking the Vote button beneath answers. So what would be the accepted method to prevent a spider from "requesting the file" at the other end of the Voting button?

[Planet13]

Maybe now is the time to implement some sort of "captcha" like device on the up-voting button...

One of my sites used to get tons of spam, and even though the spam was caught after it was submitted, I really just didn't want it to get through in the first place.

After I implemented a captcha like tool, it stopped getting all the spam submitted.

I am sure it probably also reduced the number of LEGITIMATE comments, too. But I figured the only people who would be dissuaded by a user-friendly captcha would be people who don't really care about the comment they had just made.

[lucy24]

What's at the other end of your "like" button? Is it an actual URL that can be roboted-out? Or does something visibly happen on the original page? Can you teach your code to ignore requests ("clicks") from selected IPs and/or UAs?

the only people who would be dissuaded by a user-friendly captcha would be people who don't really care about the comment they had just made

thedailywtf used to have a captcha that involved retyping a single Latin word. Based on user comments over the years, it seems like a lot of people actively enjoyed the captcha :) (They're now on a different platform, where users get their enjoyment from complaining about the platform itself.)