Pic A Fight: Instagram meets FaceMash | Hacker News Reader

Pic A Fight: Instagram meets FaceMash

picafight.com

98 pointsjazzychad37 comments

What's the purpose? Like, what are my votes to be based on? FaceMash was based on hotness, or something. What is this? I see a picture of a C3-P0 helmet versus a picture of a VoIP phone. Am I supposed to just vote for the picture that somehow touches me the most?

Yep. Nothing too serious, just some fun; which ever one is more visually appealing to you, or (as you say) touches you the most. It will be interesting to see which pics rise to the top this way.

It might be good to state that somewhere. When I looked at the site I felt like I was being called on to answer a question, but I couldn't figure out what it was. Led me to skip a lot of comparisons. Once I decided it didn't really matter it was a lot more relaxing.

Also, gorgeous layout!

I think a little more instruction to the user would be helpful. I kept picking which picture I would rather be in a fight against until I came here and read the comments.

Hi, I've a identical project but for flickr photos: elograde.com with over 200k comparisons so far and some very quality photos. tho the flickr community has not really taken to it too much. (been up about a year now)

kalendae: Nice site. I think the major thing that separates Paul & Chad's site and your's is that you can start voting without any sign up whatsoever.

Perhaps you three should collaborate or you could simply remove the registration portion and let people start voting from the homepage.

you are absolutely right, most hotornot sites have that kind of low barrier to entry. my problem is that people who vote but don't submit photos don't really add to the project (not an ad site like hotornot and especially if they vote poorly and are taken out, the votes have to double up for them to have meaning to prevent random voting etc), right now there are actually way more votes than needed in proportion to people who submitted their photos.

but yeah it might need some re-examining.

Wow, i was in the middle of making one of these. Ha. should i stop? i was almost done.

Cute, I like the overall design a bit more than the voting to be honest. I did something similar over the weekend, plotting instagram pictures on a map. You can see it at http://insta.notphil.com/map.html - click on a place on the map and then on the pins that appear to see pictures taken there.

notphilatall: I get "oops" on every click. :( But, I'd LOVE to see my Instagram photos on a Map!

Chad, you are a machine.

This is fun. I'm curious; what technologies are being used in the backend?

Nothing special.... rails 3, ruby 1.9.2 (this is my first rails app ever!), mysql, hosted on aws. stammy used Sass for the css styling. I wrote a few scripts to spin up more servers as load increases.

The idea isnt inspired imho. That being said, allowing the user to vote using the arrow keys is. The fact that I could vote with barely any hand movement got me through ~10 votes which is 8 more that I wouldve had I been required to click on each image.

Also, the design was simple, clean and effective. Good job.

most ideas aren't, it's the implementations that are. we added the arrow key voting to reduce friction for exactly the reason you stated. Stammy did the design, all props to him for that. we hope you had some fun with it, that's what it was made for.

Do you limit the number of photos? Seems like if the number of photos in the pool far exceeds the number of votes coming in, each photo would only ever have a handful of battles. I've build a similar voting system before but the purpose was to find a winner amongst all entries.

hey daniel, long time no see! we're probably going to cull the pics and delete the ones that fall below a certain elo/win/loss threshhold, so it should keep the quality of the pics high. and we can adjust the threshhold so that it keeps the total number of pics from getting too huge.

I like the design, very clean. I did something similar at http://mugshotwars.com/

Elo is a fun algorithm. You can read about my implementation at http://campedersen.com/

I had an idea of building a similar dueling site using memes. Actually, I even designed and built the site, but decided to ditch the project it just before paying for a domain.

The reason? I didn't want to turn into that guy who runs lolcats franchise.

bad call.

You may be correct, but aside from wasting 2 weeks to plan the site, 2 weeks to the design the site, and another month to build that site, I don't particularly feel any other regrets as looking from mine point of view, there isn't anything more important than to achieve inner peace with oneself.

Nice! Give it a day then see how it compares to the Flickr interesting photos. I bet there's more off beat and a better variety of photos. Human vs. Machine.

Are you sharing what rises to the top via some feed? I'd love to plot them on a map.

Looks great. I've been working on something similar for iOS, but with themed challenges. There's a lot of potential here.

http://photochallengeapp.com

> The site got twenty-two hundred hits within two hours? Thousand. Twenty-two Thousand.

Is this generated dynamically? If so, awesome. If not, still cool.

This is a quotation from The Social Network. It's static. However, at the current rate, we could easily be passing that number... yikes! According to mixpanel, we've processed nearly 18k battles in less than one hour.

Nice, though to be fair your potential user base includes all of the internet vs. just Harvard students.

Quite true, but of all the Twitter/Social-related websites I've built (a lot), this one has seen more hits and buzz than all the rest already. Please allow me to be excited :)

Nice concept - not only as a bit of fun, but as visual feedback as to what types of photos your followers prefer to see from you.

One constructive criticism: it makes it too easy to just mash the left and right arrows. Take that feature away :)

Cool stuff. Are you doing any face detection or do you just pick two random Instagram pics?

Completely random at this point.

Cool idea! Just out of interest, how long did it take you to build it?

Do you have a particular algorithm when selecting pictures?

selecting is random, rating is http://en.wikipedia.org/wiki/Elo_rating_system

I am surprised how people spend too much time on useless things instead of solving really interesting problems. You can at least use some machine learning to present a pictures of increased interest to each user. Or present pr0n pictures based on personal taste.

I am surprised how many people spend too much time trashing other people's work. Also, I think the entire gaming industry would like to have a word with you.

I am gamer. Do you created any interesting game, or just this useless site?

Reply on news.ycombinator.com