Tell HN: The problem with BangWithFriends - it's not anonymous
Facebook Graph search allows you to search everyone who uses this app. Be it your friends, people who aren't your friends, just about anyone.
Example criteria:
"Single women I am not friends with who use BangWithFriends"
"My friends who use BangWithFriends" This is an interesting find. I don't necessarily care about "BangWithFriends", but I am curious if you still show up on those searches if you set the privacy to "Only me" when you signup in any app. Does that hide you from searches on the Open Graph (in respect to app usage)? I ask this because I frequently signup with Facebook on some apps, and purposely put it to "Only me" privacy, to keep my activity hidden from my Facebook friends. I would hope that doing so, would keep me out of those types of Open Graph searches. Can anyone confirm/deny this (I don't have Open Graph Search, yet)? By changing the app privacy to "Only me" it doesn't appear in Open Graph searches Now the app developers have to see if they can have that as the default privacy setting for anyone who uses the app and this should be truly private info after that. You can set that default. See screenshot from https://developers.facebook.com/apps/[APP_ID]/permissions: http://bluetide.pro/k6sP/2HAHipbX - Just change "Default Activity Privacy" to "Only me". More info: https://developers.facebook.com/docs/guides/appcenter/#autho... Ask HN: What is BangWithFriends? Edit: a Facebook app that where you can click on friends you want to sleep with. If they have the app, they are notified. I would think this would be more cool if both of you are only notified if you both signify that you want to sleep with each other. This could also be toned down to just signify romantic interest to get more engagement. It's a new service that has a mission to connect people, increase natural population growth and make the world more open place. > increase natural population growth A booty call app seems like a terrible place to promote parenting. Why not? Starting from just $85/mo, http://kidswithout.us/ will raise them until they're 18. I was sadly expecting that to work. YC14. BangWithFriends seems like a new implementation of something I've seen a couple times in the last few decades. This is the first time I've seen it on FB. I ran into this years ago in it's email incarnation. I think it was something like you'd get an email that said someone had a crush on you and then you are supposed to put in emails of other people and if there was a match you'd be notified. The problem was you didn't know who sent the email so you'd probably put in the emails of all your female contacts (or vice versa, etc) and then all of them would get emails. Logically it was an email ponzi scheme where if you didn't see the inherent problem from the beginning you probably ended up really confusing a bunch of your friends. My first experience with this concept was in the late nineties called something like Crushmail. Here are some Hacker News threads about similar implementations: The fatal flaw in the model is "I'm going to click on all my friends to see who signals interest in me". That's not the fatal flaw. The fatal flaw is that all the men in a social circle will click on every friend to see who signals interest. All the women will click on the same one or two individuals. In other words, the fatal flaw is that it will simply replicate the dynamics and frustrations of high school. Yes, it gets tiring servicing so many fine ladies Fatal flaw or working as intended? The premise of the app is that with a contrived double-blind method, you can turbo past the awkward stage of signalling interest in someone. It's flawed because if I click all my friends just to see who clicked me, then find friends X, Y and Z clicked me, I still don't know if they actually want to have sex with me, or were just doing the same thing I did. We haven't skipped an awkward step at all. In a way, it's worse than that, because if I sincerely click on friend X, and X clicks me, then I assume that means she wants to fuck me. If she was just seeing who would signal interest in her, we now have an excruciatingly awkward situation where I've said out loud "I want to have sex with you", and it's not mutual. Surely an easy way to avoid this is to limit the number of clicks you can have over a period of time or you have a list of 5 people you would like to bang. You can add/remove people from that list. If you you are both on each other's list you are notified but you have to be on the list at the same time. That way you know that you are in the top 5 people that your friend wants to hook up with. I also think that they should give you the number of people that have you on their lists. I only have one friend(male) using the app but when I log in, it shows a small subset of my female friends. Any idea how it selects them? I initially thought that they were the only females using the app but this is obviously not the case. Yeah but if you are clicking on everyone to see who clicks on you, you don't have to pursue anything. If someone clicks on you and you don't want to sleep with them, you wouldn't reach out to them about it, no? So you don't reach out to anyone, waiting for them to reach out to you to express their interest. And now you get to the point where both are waiting forever for the other to make the first move to avoid embarrassment. Well, exactly. This is why the app fails in its basic model: it doesn't help you past that awkward first step of signalling interest. Even after getting a mutual thumbs up, you still don't know if there's a real hookup there. >we now have an excruciatingly awkward situation where I've said out loud "I want to have sex with you", and it's not mutual. I'm pretty sure people would be aware of that downside beforehand, and would factor that in as a possible risk when comparing it to the benefit of simply satisfying their curiosity. It's not foolproof, but it isn't exactly a crapshoot either. Why can't the reverse be true also? Maybe you clicked around and wanted to see which girls had interest in you. Maybe you both clicked around, and both of you think you each have interest in each other. How is that a fatal flaw (or a flaw at all). By clicking on all of your friends, you are signalling that you want to have sex with any/all of your friends. At this point, if any of your friends signal a want to have sex with you, the service will see that two people want to have sex with each other and tell them. If you clicked through all of your friends to see which ones clicked on you (even if you do not want to have sex with them), then you simply lied. People who lie this way will inconvience those who the service paired them with once, after which point none of his friends would click on him again because they do not trust his genuine use of the system. This leaves you with the people using the service as intended. CrushMail (or something like it) circulated among my friends at some point during the late 90s. Perhaps she was fibbing, but the one girl who I think I matched with (??) said she had just clicked on everyone. So, plausible deniability wins. On a completely different note: this is basically a non-issue after one's mid-20s, by which point most people seem to have the ability to appropriately inquire after those whom they have an interest in, and have a reasonably good sense for who is eligible and who isn't. That would damage the anonymity but perhaps there are ways to reduce this effect without greatly damaging the benefit to users to other services. For example, you could have a quota so you can only select say 10% of your friends at a time. The QuickMatch system on OkCupid has the exact same fatal flaw. It's a similar sort of thing. If you rate someone highly, they are notified that "somebody" rated them highly, but they don't know who did it. If they rate you highly in turn, then you're both notified. What ends up happening is a lot of people rate everybody highly so that they find out who rated them highly, ultimately defeating the purpose of the system. Could limit the # of people you could click on to 5 or 10 or whatever. Could be a different number for guys vs. girls. Or could equalize by charging guys per click and girls click free or something. Exactly. It should also have a search functionality in case you have a thousand friends. It does work only work based on mutual notifications. I know it's bad form to link to your own website, but I wrote at length about this here: http://www.zacharyalberico.net/blog/dating-websites/ You are looking at it from the point of view of the user if the system is working as desired. From the point of view of the website, you're not driving clicks. You're not making money. You're not driving growth. It is also bad for users. You do not drive clicks. Cannot pay for marketing. So your site gains no traction. And with no traction, actual users show up, can't find the people that they know, and then leave the virtual ghost town. Therefore it sounds logical, but fails in practice. Now add in a single feature. Add notifications to people saying that X people like them. Now people have a reason to come back. A reason to click on others. You get growth. Your website can make money. And you've become a spam site which completely lost the original idea. But you're a profitable spam site. And will probably drive more actual hookups than the pure form. And this is why dating sites either fail trivially, or go on to suck. Interesting points, the chicken/egg issue is a real problem and something I think starting with a targeted community (college is the example in the post) would help alleviate. You can't tell people that X people like them because it creates a perverse incentive to click on everyone to find out who. Driving clicks shouldn't be the foundation of the idea, the partnering with local places to meet could help with the value side of things. This is how you also keep people using it (giving them a reason to check in to places) that I talked about toward the end of the post. The not making money/driving growth part reminded me a little of a meeting Page and Brin had when they were originally trying to sell their algorithm before deciding to start google. They were showing off to a current search engine how page rank returned results that were magnitudes better than the contextual search (which the other site was using). The guy they were showing it to said something along the lines of "This is too good, too fast and the results returned are too relevant. We want people to stay on the engine page as long as possible to click on ads". Sometimes the value of something isn't immediately obvious. It's more about just solving the problem first and the money would be a means to do that. I think if you can solve a common problem in a better way there is probably a way to create value out of that. For example it'd also probably work well for people meeting each other within a company, but I'd doubt any companies would go for that. I know it's bad form to link to your own website You must be new here... Thank goodness for new people. Without them, old gets old. This may also work, and won't require graph search: https://www.facebook.com/browse/friends_using_app/?app_id=17... Replace the app_id with the id of any facebook app you want to use. Nice find, it actually works for me too. > "Single women I am not friends with who use BangWithFriends" If I understand the current user base of BangWithFriends, this will return no results. You are wise beyond your years. Well, I could not resist testing it and I received many results. Interesting. Here's another problem I just discovered: "Men who live in Toronto, Ontario and who use BangWithFriends" - 65 "Females who live in Toronto, Ontario and who use BangWithFriends" - 6 Well, you have to do some refractory period math, but it's doable. The bigger problem is trying to partition the graph so everyone is willing to sleep with everyone in their partition, and the partitions are evenly sized. I'm willing to bet this is an NP-hard problem. Reminds me of Stable Matching[1] but the groups aren't the same size. I guess if we can relax the "opposite sex" requirement then we reduce the problem to a Stable Roommates Problem[2] ;) [1] http://en.wikipedia.org/wiki/Stable_matching
[2] http://en.wikipedia.org/wiki/Stable_roommates_problem that was my first though, when i saw this website. "another website, which will be only used by men kind" ... and nobody got banged ever... ========= but damn now i want access to the open graph search :D just to play around with it. For people who don't have access to Graph searches yet. Just do a BangWithFriends search in app and you will see something like "xxx, xxx, and xx Other friends are using this app." under the app description. It will return the same list at the Graph search. That's not the real problem with BangWithFriends. That's just a symptom or side-effect of the real problem. As the idiom goes, if you lie down with dogs, you will get up with fleas. True, but doesn't this apply to Facebook in general? The graph search functionality wasn't created by BwF. It's been clear since the beginning of Facebook that they are eager to reshare any info you share with them. I don't see the problem here. You can't find out who wants to bang who, right? You only can find out that people want to bang. Which is true since the big bang 13.8 billion years ago. (well, humans evolved later, i know) I don't see a problem if you are single and don't care what potential employers think. And sure, if you aren't single then you shouldn't have the app installed in the first place (preempting this argument). But it still makes Facebook more of a landmine for less scrupulous users and they do make up a significant part of the population. > I don't see a problem if you [...] don't care what potential employers think Wait, what? My employer cares that I have sex? My only point is that some may care, I have no idea if yours does or not. Personally, I wouldn't work for an employer that looked for such things, but I am also a very private person and would never use an app like that. Does the Graph API allow you to find all app's being used by a specific person? This topic keeps coming up. If you can see all app's installed by someone then you might have the point. If you can't then an employer would need to search for this specific app and find your account. I guess it is possible that some employers may have a blacklist of hundreds of app's but I would have thought it would be more trouble to maintain compared to the value of having it. Especially considering how tomorrow I could launch "BangWithBuddies" and this app would need to be found, flagged as not appropriate for employees and added to a search list. If an employer is going to this much trouble to scan for what they believe is moral correctness then they probably aren't worth working for. Looks like indeed, you can! Use the graph search for "Apps that _user_ uses", or it looks like you might be able to use this url (might have to have graph search access) https://www.facebook.com/search/**userid**/apps-used You are probably correct and I agree with you about working for employers like that. I admit to not being a facebook expert and probably shouldn't have commented. But it seemed obvious to me that there are problems, contrary to what the OP says, for some people. "My friends who have searched 'My friends who use BangWithFriends'" I have graph search. Tried this, got four results, all male. So yes, this seems to be true. Those crazy kids with their crazy ideas about how to love each other. Good for them. I can imagine Version 2.0 being even more general (a craigslist of discreet hypothetical matching) for more than just sexual arrangements. But continuing the sexual theme, obviously the next logical step is N-way group hookups. It occurs to me that something similar could also be implemented in a distributed manner (for use offline or with decentralized P2P networks, etc...) using asymmetric encryption and split keys. Only on HN would sex involve asymmetric encryption and split keys. I kid, but I'd wait to see if this 'crazy idea' is successful (top tip: it won't be) before planning the second iteration. > (top tip: it won't be) You seriously think so? Casual sex has been a major use of Facebook for quite a long time. This particular approach may not catch on, but I guarantee you that some variation of this idea eventually will work. I don't think casual sex has been a major use of Facebook for a while- certainly during it's college-only days that could have been the case, but less so in a more social network aware world. I think something like BangWithFriends is doomed to fail simply because of the visibility. People can see if you use the app, which would embarrass many. And like almost every single online dating/sexual site out there, the number of men dramatically outnumbers the number of women. That doesn't mean that a majority of participants want to publicly identify themselves as such, with the default privacy settings and erosion Facebook encourages. Perhaps if it was more subtle and used less targeted/classless language. After seeing this thread, I tried looking for ways to do bangWithFriends in a cryptographicly private way. Eventually I googled "bangwithfriends cryptography" and the first result was this thread. "split keys" That just sounds painful. This app is interesting however, users are signing up for this with the idea that no one among their friends get to know about it. Unfortunately, facebook enthusiastically shows everyone using various apps. Not sure if facebook has a privacy setting to prevent others from knowing the apps I use but if they do then the developers of this app should inform this to their users. If facebook does not have a way for me to hide from my friends the apps I use then IMO, facebook should have this option. And then there's this search, which returns lots of results: http://f.cl.ly/items/1D3u380L3z3R2E0v101H/Married%20people%2... No one I'm friends with is using it! That's so disappointing. I have just tried to join http://www.BangFriends.com but it says i will receive an email shortly because they are still in BETA mode. Maybe they are adding the default privacy option so people cant monitor you on facebook via Open Graph? Great find. I had a similar issue with a dumb app I built a long time ago that was more of a learning experiment / sick joke than anything real. It was called Rubbed Out and the point was to list all of your friends you had thought about while masterbating. With bad permissions, or a small sample size, it gets really scary. The problem with BangWithFriends is that there are no females on there that are down to bang. Every person I've ever seen login has a multitude of male friends "using this app" and little to no female friends using it. This kind of app is completely fucked from the start bc of the gender disparity. This was exactly the same problem when AirTime tried to use the Chat Roulette concept to create an app "to meet new people on your social graph". Using FB connect killed all the fun in that. You can tell by the disclaimer they're now using on their landing page. I agree, it is not anonymous. We created a version that is more suitable for Facebook - for people looking for serious relationships instead of casual encounters. We have got some good support in our beta launch. I wonder what it would take to implement BangWithFriends with cryptographicly assured privacy. When I have the chance, I'll break open my crypto book and see if this problem is less impossible to solve than it seems. lol yea thats the only problem with this app.. Disgusting. Obviously, the answer is to have everyone sign up for BangWithFriends. Then anonymity is assured.
