I find it striking the the lowest edge of the scale is almost off limits; so far only 36 sites rated by more than 2 members have received an average lower than 2. I think this shows our rating system is somehow skewed and that it is far less informative than it could be. If we accept the fact we are rating WHS here, this still allows for some to be above and some below average, even though they can still be better than many other tourist attractions that are no WHS.
For my subjective judgment there are a few good criteria to downrate a WHS: For example if a site is incoherent, e.g. because historic destructions led to lack of integrity/authenticity, or if a site merely duplicates more significant or qualitively better WHS. I do not try to enforce my criteria here and would not like to open a pandora box whereby everyone tries to argue in favour or against a specific WHS. I am just reminding everyone we shouldn't have a rating system with 10 different levels and then just use the upper 8. It makes the scale signifcantly smaller and thereby less refined. At the end of the day it also makes such a rating as 3 less informative because so many WHS receive it.

Assif

To be honest, I wouldn't concur. My expectation of a WHS is that it's above average and more often than not it is. If the list was dominated by sites scoring 0.5-2 for me I would seriously reconsider if this is worth doing it. So some skewing is expected ;)

I do agree we could use some anchoring, i.e. some instruction what a 2.5 means -> perfectly average WHS.

I described above how I set my "average WHS" as deserving a rating of 2.5. Even an "average" WHS should of course still be "something special" in terms of sights around the World, so, assigning 2.5, isn't damning with faint praise even though it may seem to be so in these days where everyone expects to get a 1st class degree (In my day a 2.2 was considered the norm and indeed many exams were marked on a normative basis!). I also explained why I made 2.5 the centre value even though it leaves 5 points above and only 4 below - it gave more opportunity to differentiate the "better" WHS. In all honesty, 10 different levels of "goodness" is really too many to handle. There are very few of my ratings where I would "die in a ditch" to defend a value against an adjustment up or down of 0.5 (or, in the middle, even more)

I suspect that creating an unambiguous "ladder" of words expressing degrees of excellence is difficult in all languages, and it certainly is in English (which has just so many words, many of which carry slightly different connotations to different people) but, rather than deal simply with "Figures", I then tried to assign "words"/"meanings" to my values and came up with this. Where the same word applies to 2 values the difference can be explained as "strongly" and "standard" (In my old University days the difference between "Beta ++" and "Beta +?+")
"Outstanding" - 5.0 and 4.5.
"Good"/"Notable" - 4.0 and 3.5
"Average" 3.0, 2.5 and 2.0 - i.e "Average +", "Average" and "Average -" (Bearing mind that "Average" has NO negative connotations here in terms of a site not being "Good enough")
"Mediocre" or "Poor" - 1.5 (Many people use the word "Mediocre" to mean "average" - but here it means less than that even possibly, a slight query as to whether it should really BE a WHS bearing in mind that the scheme has to some extent been "highjacked" by nationalisms and "money"!)
"Not WHS material" - 1.0 and 0.5

The more WHS one has to rate then the more difficult it is to fully consider every site equally and there are still many sites which I haven't "fully" considered on a comparative basis - e.g I have felt able to try to differentiate between some of those Vineyard sites to at least pull out the "best" and "worst" but haven't (yet) been able to do the same for all those Wooden Churches. I came to the conclusion that, in all honesty, a differentiation between them of 0.5 when they were all in the "Average group" (+/-) anyway, wasn't really adding much value and was trying to offer false differentiation. It seems to me that the "purpose" of the exercise isn't to agonise over 0.5 here or there but to pull out those which are REALLY "Outstanding", "Good". "Poor" and "Not of adequate quality"! All those "average" ones aren't really of any great significance.

My current assignments stand at
5.0 - 27, 3.52% / 4.5 - 28, 3.78% / 4.0 - 55, 7.17% / 3.5 - 72, 9.39% / 3.0 - 112, 14.6% / 2.5 - 264, 34.42% / 2.0 - 134, 17.47% / 1.5 - 58, 7.6% / 1.0 - 11, 1.43%/ 0.5 - 5, 0.65%.
Average Rating = 2.74. (UK Average = 2.61). I haven't carried out other potential statistical analyses - Std Dev etc. So - a reasonable distribution curve, albeit skewed towards "Goodness" with a big mass in the middle and a shortish "tail" of "Badness".

Which raises the question as to what sort of distribution SHOULD we expect to emerge? I am pretty comfortable with only having identified 16 WHS which, in my view, probably shouldn't be inscribed whilst pulling out 56 which I regard as outstanding and another 127 Good/Notable ones.
I identify another 58 as not quite up to par (Perhaps I could expand this group a bit from the "Average -" group) and leave the mass of 510 at various degrees of average - but don't really regard those differences among this group as being that important.

We are never going to escape from people voting on their personal views and these are informed by a vast range of factors - in particular their interests and the number of sites they have visited /know about as well as (for better and worse) their nationality/culture and individual characters (glass half full , glass half empty etc etc). These will impact both individual site scores and the range of those scores. Perhaps we just have to extract "value" from the exercise by identifying and trying to understand those differences (even if we conclude that they are unsupportable!) rather than by corralling everyone into a standard way of looking at the World! So - we could provide information on the range of views within the Community about sites by listing those where we are pretty well in agreement and those where we differ wildly together with the extent of that variation. Each of us individually can conclude whose views accord with our own and whose views are (in our opinion (!!) illogical. At least we have learned something about the views of others which may or may not inform our own! I might even be persuaded to find more "value" in the Flemish Belfries!

From a statistical point of view: Standard deviation is not a good indicator for "skewness" or extreme outliers. To attenuate the effect of "nationalistic votes" we could show the median instead of the mean value. The medain is more robust towards extreme outliers, and for sites with a high number of votes both should be close together.

A high standard deviation also indicates that the entire range of possible ratings has been used and reflects the wide range of our opinions and interests. Boring, if we were always in agreement. Bauhaus might be such an example, where I would expect a quite homogeneous distribution of the 14 votes.

Maybe a bit nit-picking, but: 2.5 is not the average in our rating, as we do not allow the "0". The exact average is 2.75, so 2.5 is slightly below the average, 3.0 is above.

Solivagant
I really like your scale, some minor tweaks:
* True World Wonder - 5.0.
* Outstanding - 4.5.
* Very Good - 4.0
* Notable - 3.5
* Above Average - 3.0
* Average - 2.5
* Below Average - 2.0
* Mediocre - 1.5
* Poor - 1.0
* Not WHS Material - 0.5

I think the 5.0 should be really reserved for one of the greats.


Calculating the median is not possible in mysql. Would require custom coding and for the time being I think we are still okay.


You probably meant heterogenous distribution. ;)


I actually noticed sites being scored with zero. Not sure how that came about, must be a bug.

Actually I meant a (statistically) normal distribution in contrast to a skewed distribution.


Basically, these outliers do not bother me, we have no strict rules, anyone can give his rating as she/he wants. Anyways, it only has an impact on sites with a low number of votes.

On average, the rates of European countries are undervalued compared with other countries, Italy excepted. I don't know if you have the medium rating per country, but it looks that Iran, Mexico, China, India, or USA are better marked than Germany, UK, Spain or France. The latter are they underrated or the the first ones are they overrated? Is it fair, are we fair? Are we so much used to the western culture that we rate too harsh Western sites? Or is it the consequence of a supposed WH race? In any case, it's clearly subjective

I think the European countries exhausted their list to a higher degree resulting in lower quality inscriptions. At least for Germany I would say this is the case. Meanwhile the USA still has plenty of memorable sites left.

Consequence of WH race then...
I share you opinion concerning Germany and USA. Less for low rates in Spain and France comparing with Mexico and Iran, even if there are beautiful sites in these countries. When I rate cathedrals for example, i was surprised by the low rates, but i can understand them if you consider there are too many catholic sites. I asked a question about St Denis cathedral (T), and that's what the community answered me. I guess that's why Spain and France have low rates...? But in my opinion, this kind of construction in another country, a poorer country (i mean with less WHS), would receive a better rate :) Is it fair in this case? :)

A "nit pick" of the "nit pick"! I don't think it was suggested that 2.5 was an "arithmetic average" of the available ratings - merely that it was an appropriate level to place the "Average WHS"! A case could be made out for using 3.0 (As Els did). I don't see any problem with then calculating an "average rating" since the same ratings were available to everyone.


Most sites will only have a relatively "low" number of votes - low enough anyway to distort a simple "average" (which is the only "analysis" figure shown at the moment) if there are significant outliers. Bearing in mind also that the majority of sites are going to be regarded as "average" so there will be a centralising tendency if that it the only figure "on show". There are only 820 registered "Community members" and only 316 of these have visited 100 or more sites. Only 400 sites have been visited by more than 100 people. Many of these people are not active anymore or won't join this "rating" activity. We might get c50 people doing reasonable numbers of ratings (and some of these might only "rate" their favourite sites providing a further "Skew") but even that would leave many sites rated by fewer than 10 people ?? I think it IS worth trying to highlight the spread of voting beyond mere "Average" figure - how and where is another matter. Not particularly to "attenuate the effect of "nationalistic votes"" - there may be a whole range of reasons for a wide spread and I don't think we should attempt to force people into voting in a particular way. But we should try to provide info on the "make up" of votes


As are ALL the votes - but I agree that it is an "issue" to consider both when rating AND when looking at the ratings!! I have noted that a number of people have consistently voted some UK sites higher than I have. This may be because I have leaned over backwards to avoid "nationalistic voting". It may be because others value certain UK sites for their "Foreignness" or international significance in a way I can't - and the same could apply vv for me looking at other countries. Since the majority of us "voters" are "European" that could certainly put European sites at a comparative "disadvantage" but it isn't something I am much concerned about - I don't think it will result in the "creme de la creme" being undervalued. And there isn't much else in the WHS system which is "unfair" to European countries!! We can't avoid the fact that there IS a lot of "duplication" within European WHS (Wooden Churches, Cathedrals, Vineyards, Palaces, Castles, Mediaeval Towns, Mines etc etc) - I don't think I have treated these any more harshly than the duplicated Arab Medinas, Tropical Forests, African Cultural Landscapes - there are just more of them!

However we finally decide to provide an analysis of ratings by site then doing the same for "Countries" would certainly be of interest. I have already suggested doing it for "Categories" and Natural/Cultural/Mixed as well in order to provide info on how we "value" different types of site. I would be interested to see what the average and "spread" for Vineyards and Medinas was!


It is up to the "Community" and Els to decide whether, and if so, how, to provide "guidance" on the meaning of ratings. It might be regarded as a bit late now or not likely to make any difference!!!

In a way it's logical, Europe, the "Old continent", has a distant culture which has been much more preserved than other parts in the world. Europe dominated the whole world and its influence is historically considerable. I don't want to shock anybody, but you can't find the same potential cultural WHS in Argentina for example. You can find this kind of influence in Asia, above all in China and India, but even in this case, the density of potential WHS is, in my opinion, less important than Europe.

Perhaps because "Europe" insists on treating each of its little "states" as separate entities which justify equal billing when it comes to WHS??
I think of Wooden Churches separately inscribed for Slovakia, Poland, Ukraine and all those separate vineyards in France/Switz/Italy/Hungary within a few hundred kms of each other. What would we say if China insisted on inscribing every different Tea producing region within a similar area! I am sure they regard the differences in Tea flavour and types as being as significant as France does for its vineyards. Very often European lack of appreciation for the sites of "far away" countries is to some extent a reflection of our lack of knowledge and appreciation of those countries! I have noted a similar issue regarding date palm groves in N Africa etc etc - perhaps Arab countries should inscribe the producing area for every type of Date??

We have just returned from Thailand - I still only vaguely appreciate the wide variety of styles of "Wat" in that country (and its near neighbours) but i have no doubt that they could come up with as many different "styles" and periods as multiple European countries have for "Cathedrals" and populate the WHS list just as fully. We can't just sit in Europe and dismiss such countries as not having cultural sites as "densely" situated as in Europe!

Yes, you're right, i agree, they could associate some cathedrals and churches (vineyards, not sure). But is it the reason of our low rates? Does it mean these sites are not interesting comparing with other sites in other countries?
In addition, some countries have clearly more potential sites than they currently have (Myanmar first), but other countries like China, South Korea, Japan, India have maybe the same situations? They could associate different places in one site (like Hills and forts of Rajasthan)

The biggest countries have also huge natural WHS, it's completely different with Europe. Hence their density of WHS is weak.

Guys in my opinion you overthink the whole rating thing. Fact is that there is the possibility to rate WHS with 0.5 till 5 stars. Basically the same thing you can do on many other platforms, like tripadvisor.

Why people give as many stars as they give is pretty individual and hard to steer. E.g. I rather value how much I enjoyed visiting a site and not so much if it really has OUV. And it's very subjective and there might be external factors involved on how I came to my rating. So what. Same goes for the millions of people voting each day on countless platforms.

Coming from the one true democracy in Europe I'd say: just let the people vote. The majority decides and we accept the results (even if that means that a beloved Bauhaus site gets a low rating).

I haven't read the entire thread, so forgive me if this has been addressed, but is there a way to see the entire list and rate sites in one fell swoop, instead of opening each site's link? I'd be inclined to rate sites (especially to call out some clunkers -- sacred forests, I'm thinking of you) if I could do so in bulk, but not if I have to open each link. Thanks