|
|
Currently my daily usage of the site via mobile either through the mobile website or either of the 2 apps is as poor as i can remember. I hate the push to the new app when it doesn't fulfil my needs.
The people on the site remain the reason i hang around. The technology is letting us all down.
I dont say this lightly as i love this site. But at the moment my user experience is significantly sub optimal. |
Originally posted by harrisoni
Currently my daily usage of the site via mobile either through the mobile website or either of the 2 apps is as poor as i can remember. I hate the push to the new app when it doesn't fulfil my needs.
The people on the site remain the reason i hang around. The technology is letting us all down.
I dont say this lightly as i love this site. But at the moment my user experience is significantly sub optimal.
agree its not great at the moment, but i suspect that's as code written by volunteers, probably undocumented is taken over by a new development team, and should settle down. but certainly not as bad as far ascim concerned as the months we were hacked and site was up and down like a yo-yo |
|
|
Originally posted by cgarvieuk
Your Probably right combining Rates(no txt) and Reviews(with txt) Could be problematic. Your right people are probably on Different scales. But so am I with say FatPhil. My concern is. What if a beer has 100 Rates(no txt) but no reviews Does that mean that beer doesnt have score. Or worse 100Rates all saying Suberb but 2 reviews(with txt) from a bad bottle Does that mean its score is low though almost everyone loved it.
First of all, I think the idea of using the terminology of Ratings (no txt) vs Reviews (with txt) rather than Ratings (with txt) vs ticks/unqualified reviews/private reviews/non qualified ratings/whatever we're calling them now (no/not enough txt) is worth exploring. I think I quite like it as a concept..
The problem with a large number of beers in the UK is that there simply isn't enough data to get a meaningful (or any) score. Huge numbers - by far the majority - do not reach the 10 rate minimum. Of those that do, many of them do not get enough to avoid their scores being heavily affected by the weighting/smoothing measures pushing them towards a middling score. So I don't have any particular objection to the "no text" scores being taken into account but they should probably have a lower weighting. That seems to me to accord with the way I use review/rating sites for other things - I'm going to pay more attention, and give greater weight, to a well written review than a simple star rating on Yelp, or whatever. |
Originally posted by chriso
Originally posted by cgarvieuk
Your Probably right combining Rates(no txt) and Reviews(with txt) Could be problematic. Your right people are probably on Different scales. But so am I with say FatPhil. My concern is. What if a beer has 100 Rates(no txt) but no reviews Does that mean that beer doesnt have score. Or worse 100Rates all saying Suberb but 2 reviews(with txt) from a bad bottle Does that mean its score is low though almost everyone loved it.
First of all, I think the idea of using the terminology of Ratings (no txt) vs Reviews (with txt) rather than Ratings (with txt) vs ticks/unqualified reviews/private reviews/non qualified ratings/whatever we're calling them now (no/not enough txt) is worth exploring. I think I quite like it as a concept..
The problem with a large number of beers in the UK is that there simply isn't enough data to get a meaningful (or any) score. Huge numbers - by far the majority - do not reach the 10 rate minimum. Of those that do, many of them do not get enough to avoid their scores being heavily affected by the weighting/smoothing measures pushing them towards a middling score. So I don't have any particular objection to the "no text" scores being taken into account but they should probably have a lower weighting. That seems to me to accord with the way I use review/rating sites for other things - I'm going to pay more attention, and give greater weight, to a well written review than a simple star rating on Yelp, or whatever.
+1
Only thing I do not understand ..
Is that we all talk / have concern about if tick vs rating will make a different RB ...
But still as it is now ticks is secret for others (privat rating) only the person giving a tick (new ticks) can see what score the beer have been given by the person self .. nobody else can set what the scoore is .. and hardly find the beer this person has scoored, so it all get a bit messy starting to make ticks count in all kind of stats.
So in my view let people see what scoore a person have been give by ticking or rating will be the first thing to do otherwise i do not see the point in letting a tick have weight in any stat. |
The ticks as they stand are too blunt against the ratings as you can only tick a whole number (1,2,3,4 or 5). If the option to rate with the scoring as is and a no text option (or simply where the beer was drunk and has it was dispensed) I would've done that. It simply wasn't offered. That could give a meaningful score to the beer as it is on the same lines as the current review ratings.
The system exists for places and is not policed well and open to homerism where a lot of places get high scores and no text. I'm happy to review the places as the USP of ratebeer for me. The beers I just wanted a record of what I have tried and if I liked it (1 being awful and 5 being great!) and this suits me. The stats being added on meaning I sort of count overall to the whole ratebeer stats is a massive bonus to me personally and has been a long time coming but I can see how the reviewers are nervous about devaluing the ratings they all spend long hours entering.
I'm not sure it'll be a perfect solution and, as Chriso said, there are a lot of beers that will never qualify as the ratebeer community simply isn't big enough to get 10 ratings from every beer, especially with the amount of beer being produced.
I think the site is heading on the right path but there is a fine line to tread to keep everyone happy. Just fix the city maps next and I'll be a happy rater! |
Originally posted by chriso
Originally posted by cgarvieuk
Your Probably right combining Rates(no txt) and Reviews(with txt) Could be problematic. Your right people are probably on Different scales. But so am I with say FatPhil. My concern is. What if a beer has 100 Rates(no txt) but no reviews Does that mean that beer doesnt have score. Or worse 100Rates all saying Suberb but 2 reviews(with txt) from a bad bottle Does that mean its score is low though almost everyone loved it.
First of all, I think the idea of using the terminology of Ratings (no txt) vs Reviews (with txt) rather than Ratings (with txt) vs ticks/unqualified reviews/private reviews/non qualified ratings/whatever we're calling them now (no/not enough txt) is worth exploring. I think I quite like it as a concept..
The problem with a large number of beers in the UK is that there simply isn't enough data to get a meaningful (or any) score. Huge numbers - by far the majority - do not reach the 10 rate minimum. Of those that do, many of them do not get enough to avoid their scores being heavily affected by the weighting/smoothing measures pushing them towards a middling score. So I don't have any particular objection to the "no text" scores being taken into account but they should probably have a lower weighting. That seems to me to accord with the way I use review/rating sites for other things - I'm going to pay more attention, and give greater weight, to a well written review than a simple star rating on Yelp, or whatever.
THanks Chris. I think the terminoligy i used. Is a little more excompasing.
and i quite like the idea that rates carry a little less weight than reviews. Not sure of teh technicallities of that. BUt if its Clearly started in user guidelines along with an explaination such as youve given, i think its fair, and anyone that want even one rate to carry more weight only need put that little bit more effort in turnin it into a review
I suspect we will see some Reviews after the fact, but given that RB and its users have many foibles i thik thats an acceptable solution
|
|
|
Originally posted by WingmanWillis
The ticks as they stand are too blunt against the ratings as you can only tick a whole number (1,2,3,4 or 5).
actually not true, if your using the app, you can tick at any ratable value, so i suspect web may follow at some point
|
Originally posted by cgarvieuk
Originally posted by WingmanWillis
The ticks as they stand are too blunt against the ratings as you can only tick a whole number (1,2,3,4 or 5).
actually not true, if your using the app, you can tick at any ratable value, so i suspect web may follow at some point
I hope so. They need to be aligned. It seems daft to me to "tick" on different scales on the app and website. I do much of my RateBeer work on a PC at home - not least because I often need to add beers and do some admin work while I'm catching up with my ticks. When I was trying out the app, I wanted to "tick" beers with the 0.1 point scale so I had to have the website open on my PC and a tablet with the app open alongside it. Not very convenient. |
Originally posted by WingmanWillis
I'm happy to review the places as the USP of ratebeer for me.
Good point. There aren't really any decent place and widely used place review sites - in the UK at least - these days. The link up of that with beer availability and ticking is pretty unique. We could be making more of all this. |
