UKC

I wonder what the formula is for UKC photo ranking in terms of total number of votes and average vote. Maybe someone from UKC could tell me. In particular, I've been wondering under what conditions a vote of a given value would increase or decrease the ranking. Note that this is an entirely geeky musing out of mathematical interest rather than a desire to vote tactically or vindictively (I generally vote 5 for photos I love and don't vote at all otherwise).

In reply to Robert Durran:

I've always assumed, perhaps naively, that the top rankings are based purely on the number of 5 star votes a picture gets, irrespective of lower rankings.

In reply to Robert Durran:

(I generally vote 5 for photos I love and don't vote at all otherwise).

Not strictly true...
(65% superb - 9% good - 7% average - 6% poor - 10% rubbish)

Altho why I care I'm not sure? Bored/time on my hands I suppose? I'm with Gordon re votes BTW.

In reply to Gordon Stainforth:

No, if you look at the weekly top ten it says "The 'ranking' is calculated from the average rating of photos in each gallery, as well as the total number of votes made" and the top ten are in order of ranking.

In reply to Padraig:

That is historical. I've made a conscious decision fairly recently to resist the temptation to vote negatively however much I hate a photo and I've never voted for photos which are just downright poor by any standards.

In reply to Robert Durran:

I thought you were talking about top pictures rather than top galleries.

In reply to Gordon Stainforth:


I am talking about top pictures, but sorry, yes, you are right, the line I quoted does refer to galleries. But individual photos have rankings which are not necessarily whole numbers so are presumably not just a 5 vote count.

In reply to Robert Durran:

I've never voted for photos which are just downright poor by any standards

The stats say otherwise I'm afraid.

In reply to Padraig:

No they don't. The low votes were for photos I personally hated; they may well have been technically brilliant and loved by others, but simply weren't my cup of tea.

In reply to Robert Durran:


I'm afraid that we aren't able to answer that question at the moment. It was designed by Nick Smith and he isn't involved with the site any more. It is quite complicated and I haven't yet asked Paul to figure it out since we have so many other things that need doing.

Alan

In reply to Robert Durran:


You'd not be voting 'negatively', you'd just giving it a lower vote than a 5 or presumably, in your own case, a 3. Personally, I don't see the point in voting for anything, if you don't vote across the full range of scores. My scoring range is: 3% superb - 22% good - 53% average - 17% poor - 3% rubbish.

In reply to Alan James - Rockfax:


If you give me some data (maybe the number of votes of each value and the ranking) for, say 15 photos, I'd like to have fun trying to work it out for you!

In reply to Fraser:


I wouldn't consider voting 1 for a photo which any sane person would consider rubbish as negative voting. What I meant was voting 1 for a photo which I personally dislike because of it's "style", such as enhanced colours which other people might like; I don't do this any more because it just seems petty. I don't see why just voting 5 for great photos which I like is pointless at all.

In reply to Robert Durran:


Gottcha.


Because it skews the relative stats.

In reply to Fraser:


I don't think it skews any stats at all unless there is some voting pattern (presumably voting for all photos and in a particular ratio of scores) which everyone is supposed to stick to. I'm sure everyone has their own voting habits though, so there is not really anything to skew.

In reply to Alan James - Rockfax:

No need to figure it out , just post the formula, ( and they will criticise it to death