The reference system should include an anonymous element

Several members offered suggestions for how we can improve the reference system in another thread.

Think (already mentioned and proposed on the main site) some anonymous feedback with the reference would work the best in combination with member rating. Maybe it’s an idea to have some local representatives/ambassadors so after the anonomous feedback they can try to handle and moderate the situation locally?

The trick is to design it in such a way where you can say, “I wasn’t 100% comfortable with this person” without worrying about being too harsh, facing backlash, etc. Some possibilities:

  • Checkboxes similar to the CS tag system (the one that has things like “fun”, “wanted to hang out”, “okay with not hanging out”) but more compulsory and extra stuff like “didn’t ever feel uncomfortable”
  • A ‘private’ aspect to references, where you can give a star rating for friendliness, appropriateness, etc and they are displayed only as an average after x references (you can’t see what each referee said)

The possibility of a ‘revenge reference’ is definitely something that puts people off leaving a less than positive review. Having a ‘rating’ separate from individual references would make it difficult to see exactly from whom your score went down/up.

Let’s continue the discussion here in a new thread!

1 Like

I would at once welcome some anonymous aspect. I’ve also been wondering whether there need to be more distinction than just good – neutral – bad. By now I’ve thought of a different approach. What is it about mixed feelings that make them fall in between the categories? Think of your past encounters …

For me a case that’s not really too rare is that I found someone excellent as a host but not really someone I’ve clicked with (sometimes even unlikeable for no reason but personal chemistry); or an instant friend and yet a terrible host in some aspects. I’d probably be more like to give the latter a better reference even if for the purpose of the site it’d be better the other way round. So can we maybe consciously split these issues?

If we kept the CS system, essentially, we’d have to remind people to rate the host only as a host. Above category one would get a positive reference that reads a bit terse (and this works already, there’s plenty of these references). Category two would get a neutral reference with a glowing text, which is something you almost never see. So that approach feels unrealistic.

Instead let’s really give distinct scores. I’d keep the personal texts as-is and then ask hard ratings for hosting skills, with clear examples (e.g. communication vs. misunderstandings – cleanliness vs. filth – appropriate description vs. surprising reality). If there are too many categories, skipping a question should be an option. I don’t know how to incorporate neutral scales (i.e. ok with hanging out and ok with not hanging out, each could be something to look for in different situations).

Anyway, these should contribute to anonymous average scores. Given the different categories I’d be quite happy myself to not excel in everything, I mean, I wouldn’t take offence at a low score in any one category as much as I would at an overall negative reference if I perceived it as unfair.

1 Like

I would vote against it.
In worst case, people will start suspecting each other based on the time they were at a place and the timing of the message. The content would also have to be so generic it not worth that much.

Sorry that this is so personal, but wanted to share it here to illustrate how important this is.

Imagine this: a really nice host with hundreds of references, in a first world country, who lets you use his bike all day every day, takes you to all kinds of house parties, introduces you to his male/female friends, includes you in his plans for the week, cooks dinner a few nights, and gives you space to share the bed with him since it’s the only sleeping surface in the house. Four nights of the week you come home drunk together and nothing physical happens. On your fifth and final night, he tries to touch you while you’re sleeping. You say, “Remember, I have a partner?” and he replies, “But don’t you like this?” You say “no,” and he stops. You fall asleep and leave the next morning.

I still couldn’t leave a negative reference. I was 1) just plain young and naive, 2) thought it must have been an accident and gave him the benefit of the doubt and 3) felt he was nice “overall” since he stopped after I said no. I left that guy a positive reference, and didn’t even mention the incident. I was 20. He was 31. I wasn’t that upset. I was like “eh, whatever, shit happens.”

It gets worse. Fast-forward a few months, and the same host has contacted me asking me to host his friend who is passing through my town. I agree to host her, and later on tell her the story, and she immediately relates with a similar story. She elected not to leave any reference, because she, too, felt he was “overall nice.”

I’m sharing this story as an example of how not everyone feels compelled to leave a negative reference, even when it is warranted, because of the public nature of that reference.

That guy is still walking around CS with 400+ references and no negative ones. And I’m still not going to do anything about it probably.

That’s why I believe making some component of references more anonymous could help improve safety and encourage more people to be honest about their experience.


In a system with a personal written reference and an anonymous score that bloke would get a high rating (and description) for hosting as such and a suspiciously low trust score. Given such an example it could be worth it including to be able to see a separate score just from women, for example.


Are you implying that if you found your “trust” rating slipping, you’d prefer to know if it were mainly because of references from women vs. men? This is not directed at you, but that might be another way for a host to weed out where a negative rating came from. I don’t see why splitting male/female references would benefit the guest/person searching for a host.

Over the past decade, I have learned that you can’t necessarily trust other references from women because everyone has such a wildly different idea of what’s normal. Whereas I find it trashy and upsetting if a CSer flirts with me, another girl might enjoy that attention. Viewing women’s references separately wouldn’t help me much in this regard.

You often see the classic guy with tons of positive references from women and upon closer inspection, few or no male ones. You might think, “Wow, other women have had really good experiences with this guy! I’ll be safe if I stay with him!” While the experience might be “safe” in that you don’t get raped or killed, I have usually felt uncomfortable with these hosts, who reveal pretty quickly that they are lonely, lack the company of women, and are maybe using hospex to meet women and hope for some action. Despite that, the vast majority of women still rate them positively for some reason.

So, if people got in the habit of looking at ratings by women specifically, it could create a false sense of security, the same way that the references system already does.

Not sure if this makes any sense. :slight_smile: It’s hard to explain what I mean!


Ah, no, what I thought was that it’d be an advantage to see a separate score as a guest. You might notice that someone’s rating by women was especially low and thus avoid them (or, to begin with, inspect the written references more closely).

Perhaps one’s own numeric scores don’t need to be on display. Sure it’s possible to find them out regardless, by a second profile or by asking someone else, but that way you’re at least not permanently hitting F5 to see them update. I thought it goes without saying that the ratings should get some small random fluctuation to prevent weeding out specific ratings, and likewise no display till there’s a threshold amount of received ratings.


This is exactly why some guys behave the way they do on CS. Nine times out of ten, they’re considered creeps… but they’re only really looking for lucky number 10. And truth is, yeah, she’s out there. If he wanders around the world a bit or opens his home to every woman passing through his city seeking a host, he’ll find her. Does that make him intrisically evil? From my observations, it makes him about 25 to 35 years of age, that’s about it.

Has that not been enough to ring some alarm bells? I’ve travelled with women that have found this guy in the “Accepted couch request” column and have immediately ruled them out as potential hosts. I thought it was a fair call and an easily decipherable code. Conversely, one time I was hitchhiking with a polish girl and we accepted a couch with a spanish guy that had a profile that screamed sex pest to me. All women references. Some detailed, others very short and vague. All positive. He was, however, pretty handsome and yeah, unsurprisingly, they hooked up later that evening.

1 Like

Well, it is for an experienced person. For a younger person or someone new to the platform, it could be counterintuitive. You’re saying we should expect everyone who uses the platform to be able to “decode” references?

I’ve had to educate friends on how to detect creeps on CS for a while now. It would be better if the community could detect them first :slight_smile:

1 Like

How about some “shaming” badges? I wouldn’t have a problem with a bit of shaming if it doesn’t get medieval :smile: Like if x out of your last x anonymous references clicked tried to hookup, you get a looking for love badge. And it also goes away if you get out of that ratio again.


I love getting random pointless internet badges. If we introduce such a system I’ll be tempted to game it and hit on guileless people just to decorate my profile at least once :confused:


No, I don’t expect everyone to get it right away, but @itsi’s proposal to encourage group meetings with experienced members seems like a reasonable suggestion. Community detection would be great but online, a hospex community is made up of geographically seperated individuals that don’t necessarily always see eye to eye.

I think trustroots is doing both, I think that’s a good idea, one anonymous review and one public review.

1 Like

hahahaha, love this idea, and it could actually work for the benefit too! If a surfer is also looking for love, they know where to come a knocking =)

Like right there now I am feeling uncomfortable - encouraged group meetings to be trained and how to decode if a profile is legit or a perv. (ladies can be pervy too btw)
If I see that as an “amazing feature of a new platform” I am walking away.

We need to pull ourselves back a bit and consider first time users - we were all first time users at one point. And it is through our experiences that we grew. We can’t be like overprotective parents trying to dot every i and cross every t before the words are even writter.

If it is very clear from the day one that on your profile you are going to recive a reference and some standard pointing systems then there should be no fear of revenge references or whatever. Consider Airbnb example:
etc etc

For couchers

and something like:
On a scale of 1 to 10 how well did you connect with this host/surfer?

If someone marks a one, what I would do anyway, is check out the persons profile who left the one. If I resonate with that profile then I will likely not connect with this experience. If I am the complete opposite of this profile then I’ll have a fine bloody time.

but then again, does every one fine-comb and go way too deep checking out references?


Side note regarding negative references:

Naturally I’ve always been curious about someone’s negative references. All positive references are alike, but each negative reference is negative in their own way (says Tolstoi). But what’s interesting is not necessarily what’s in them, which can be really petty stuff, classical revenge or whatnot; but how they’re being handled by the person who got them (especially in the annotations). I’ve surfed with some cool people who got some negative refs but responded to them very maturely (both to legitimate concerns (“you’re right, I fucked up”) and to entitled kids who played princess-on-the-pea).


Yeah, of course! What I meant though was that even petty, ridiculous and senseless negative references often provoke a thoughtful and informative response. And that’s sure more than 1%, far more.

Making them amendable rather than editable was one CS decision I’m really embracing.


Sorry, my bad! To edit something to me implies to change it, e.g. replace part of a ref altogether. To amend something means you can’t change it, but you can add something, like a P.S.


And in such a case, too, I’d rather have an addition to the reference rather than something replaced. Having the possibility to edit references to me still sounds like it’s good for little but mischief.