Ratings Average

<p>At the risk of being repetitious, I've previously asked about this problem. Now I have a concrete example.</p>

<p>This concerns the following image: http://www.photo.net/photo/17731676. If you look at the displays under the thumbnails on http://www.photo.net/photodb/folder?folder_id=1060624, you will see that 5 ratings were given for this image, averaging 5.40. </p>

<p>Upon looking at the details, the average is shown as 4.67, resulting from 6 ratings, but only 5 names are listed. Clearly if the first 5 ratings averaged 5.40, and the 6th rating was a 1, the new average would be 4.67. </p>

<p>So the question comes down to this. <strong>Exactly</strong> how many ratings were issued in this instance? Or am I losing my mind . . . again?</p>

<p>You are not losing your mind; your assessment of the ratings is correct. The photo.net history of ratings and what is shown of the ratings is long, convoluted, and arcane. I am in the process of examining the ratings of this photo and others and I expect you will see the overall rating of this photo change before very long. In any case, it is unwise and unhelpful to assume that the ratings of any photograph on p.n have any useful meaning; they are usually arbitrary and often capricious or spiteful.</p>

<p>Jeremy,<br>

Here is another one for you to look at.<br>

http://www.photo.net/photo/17726504<br>

It show 5 rates with 4.80 average in thumbnail view and 6 rates with a 4.17 average in image view. <br>

I discovered a new (at least for me) twist on mate rating the other day. When having a look at the 7s and 6s of the rater, I noticed that 95% of the rater's high rates went to individuals who had surnames that one associates with the rater's country or countries in the rater's region....Hmmm maybe its time to introduce a new phrase, 'ethnic cleansing mate rating!'</p>

<p>Jeremy: You certainly have pegged the overall value of ratings. The only reason I still ask for them is to increase the number of views my images get.</p>

<p>Michael,The ratings sum is not just as simple as you think,there is an algorithm in place since they introduced 1's and 2's about 3 years ago.If someone rates your image 1 or 2 that does't show in your ratings it ,does't affect the rank of the image but it affects the average.Probably only Jin knows how it works.How do i know this? Some idiot who appears on PN twice a year rated about 80 of my older images with 1's and 2's.That made me investigate the rating system.PN has the worst rating system of all photo sites but on the other hand is the most interactive.If you want views you have to make friends,followers, that will look at your photos and comment or rate them or post only masterpieces... :-). Gj Since most of the European photographers left the site,i see mainly American names in the ratings,i would't even call the site international,it's probably 80% American.</p>

<p>I really don't think the PN ratings deserve such mathematical effort being lavished upon them :-)</p>

<p>Harry: Having been involved with regulatory consumer enforcement for most of my so-called adult life, I guess I just expect information to be straightforward. When the current system was introduced, nothing was stated that would have led me or anyone else to expect averaging to be other than arithmetical averaging. Had Admin stated that the averaging would be geometrical, harmonic, or some other more esoteric form, I wouldn't have posted this thread. </p>

<p>Colin: I just prefer things to make sense whenever possible.</p>

<p>Here is traffic by country as a % of overall traffic on the site according to Alexa.com.<br>

Second column is the traffic rank of PN in the various countries. That said, this table accounts for only 70% of the traffic. </p>

<table id="demographics_div_country_table" cellspacing="0" cellpadding="0">

<tbody>

<tr data-count="1">

<td><a href="http://www.alexa.com/topsites/countries/US">United States</a></td>

<td >32.4%</td>

<td >4,340</td>

</tr>

<tr data-count="2">

<td><a href="http://www.alexa.com/topsites/countries/IN"><img src="http://pcache.alexa.com/images/flags/in.50d62cba8134c8c097d073646cda1b9b.png" alt="India Flag" /> India</a></td>

<td >12.2%</td>

<td >4,770</td>

</tr>

<tr data-count="3">

<td><a href="http://www.alexa.com/topsites/countries/GB"><img src="http://pcache.alexa.com/images/flags/gb.0894999b108830afc0733ee7b6e08310.png" alt="United Kingdom Flag" /> United Kingdom</a></td>

<td >6.0%</td>

<td >4,321</td>

</tr>

<tr data-count="4">

<td><a href="http://www.alexa.com/topsites/countries/CA"><img src="http://pcache.alexa.com/images/flags/ca.8618709a45d8d1c4d9d254c61bdf29b8.png" alt="Canada Flag" /> Canada</a></td>

<td >3.4%</td>

<td >3,731</td>

</tr>

<tr data-count="5">

<td><a href="http://www.alexa.com/topsites/countries/FR"><img src="http://pcache.alexa.com/images/flags/fr.c1cf1874c3305e5663547a48f6ad2d8c.png" alt="France Flag" /> France</a></td>

<td >3.4%</td>

<td >5,661</td>

</tr>

<tr data-count="6">

<td><a href="http://www.alexa.com/topsites/countries/AU"><img src="http://pcache.alexa.com/images/flags/au.2fba49c88880e9ffcff947015cb7ab9c.png" alt="Australia Flag" /> Australia</a></td>

<td >2.5%</td>

<td >4,234</td>

</tr>

<tr data-count="7">

<td><a href="http://www.alexa.com/topsites/countries/DE"><img src="http://pcache.alexa.com/images/flags/de.ddabae687ecae5edaaeb808d440543e6.png" alt="Germany Flag" /> Germany</a></td>

<td >2.4%</td>

<td >13,325</td>

</tr>

<tr data-count="8">

<td><a href="http://www.alexa.com/topsites/countries/IT"><img src="http://pcache.alexa.com/images/flags/it.784f7eb333f0591558bcce9616a3c105.png" alt="Italy Flag" /> Italy</a></td>

<td >2.4%</td>

<td >6,087</td>

</tr>

<tr data-count="9">

<td><a href="http://www.alexa.com/topsites/countries/ZA"><img src="http://pcache.alexa.com/images/flags/za.98e1044d0ffd11afc67a79f3676ba97a.png" alt="South Africa Flag" /> South Africa</a></td>

<td >1.9%</td>

<td >2,891</td>

</tr>

<tr data-count="10">

<td><a href="http://www.alexa.com/topsites/countries/CH"><img src="http://pcache.alexa.com/images/flags/ch.e67b19a7767114078cda2b3c874a5d5b.png" alt="Switzerland Flag" /> Switzerland</a></td>

<td >1.6%</td>

<td >2,443</td>

</tr>

</tbody>

</table>

<p> Michael...If you were to ask the previous administration how the rating system worked they were going to tell you off and mark you as troublemaker,till now most members don't know that 1's and 2's count.When they changed the system i complained that they left the older images open to abuse,the answer i got was " Whatever rating you get on the older images is someones opinion" take it and shut up. <br>

The new administration promised a new rating system,that was a year ago,wait and see.</p>

<p>Michael,</p>

<p>There's a legal theory that says a party to a civil lawsuit must come to court with clean hands, otherwise their case (even if it would otherwise have merit) becomes suspect. So, if you're suing for negligence but were committing a crime when you were injured, you're going to have problems. That's kind of how I think of trying to make sense of the ratings system. You're starting with both statistical and technological nonsense and then trying to make sense of it all. Don't. It won't and it can't. It's off from the start so the whole process at every stage is flawed. You say you ask for ratings to increase your views. If that's the case, just look at the number of views and don't get involved with the details and averages and updates and contradictions of the rating system itself. May be best just to get what you say you want out of the rating system and be satisfied with that. Just the views. You'll sleep better and have more time for photos! :-)</p>

<p>Fred</p>

<p>Jim, those Alexa rankings aren't useful without context. (Some would say Alexa isn't useful at all.) Most of the traffic from India is due to spammers, not participating members. The vast majority of spammers have IPs tracing to India and China. I'm only surprised that China doesn't show up as the main source of traffic, although very little valid traffic comes from there. Presumably Alexa has metrics to account for some spam traffic, but not all.</p>

<p>And the U.S. counts for far, far more traffic than the 32.4% indicated by Alexa. It's been a year or so since I last checked some more reliable sources but the U.S. probably accounts for more than two-thirds of all the valid traffic.</p>

<blockquote>

<p>"The ratings sum is not just as simple as you think,there is an algorithm in place since they introduced 1's and 2's about 3 years ago.If someone rates your image 1 or 2 that does't show in your ratings it ,does't affect the rank of the image but it affects the average."</p>

</blockquote>

<p>Harry, has anyone in admin confirmed this theory to you, or are you just guessing that some hypothetical anonymous troll who's busily doling out 1's and 2's with one hand and <a href="

underpants</a> with the other is responsible for downvoting your photos from an average of 5.8 to 5.76?</p>

<p>Fred: I'm off to bed now. Hopefully, my dreams will not be of "details and averages and updates and contradictions . . . ." </p>

<p>Lex...... I can see the name of the TROLL in my images and i can see all the 1's and 2's in his ratings.<br>

What puzzles me more than anything is WHY someone is rating old images that were shot 7-8 years ago with clunky 3 or 4 Mpixel archaic cameras now,when we have 24 and 36 Mpixel cameras?<br>

And more,the ratings of the individual image are affected but not the photographers average.<br>

And as i am arguing with the biggest human encyclopedia (who 7 months ago didn't know that 1's and 2's count) i am probably wrong.</p>

<blockquote>

<p>WHY someone is rating old images that were shot 7-8 years ago with clunky 3 or 4 Mpixel archaic cameras now,when we have 24 and 36 Mpixel cameras?</p>

</blockquote>

<p>Maybe they're looking at content, moment, photographic qualities such as light, texture, use of shadow, and not counting or caring much about Mpixel count and its effects.</p>

<p>That they chose to rate (unless it's a bot) a photo shows at least that much interest in the photo, which is more than we can say for a lot of photos here that get NO attention. </p>

<p>If you think someone maliciously rated a bunch of your photos with 1s and 2s, you could notify administration and have them looked into. Or, you can chalk it up to someone else's stupidity and ignore it. If your ratings went down, so what?</p>

<p>Fred<br>

Do you know how the rating system works?</p>

<p>Whether I know how the rating system works, and to what I extent I know it, is irrelevant in terms of the comments I've made. I'll stick to what I said and not hypothesize about or make inquiries into who knows what about the ratings system. My point in the post above was that people are rating based on many things besides Mpixel count. Why that would encourage a question about my personal knowledge of the rating system is beyond me.</p>

<p>Because Michael wants information about the rating system,If you don't know how it works WHY are participating in this thread?</p>

<blockquote>

<p>If you don't know how it works</p>

</blockquote>

<p>Harry, I didn't say I don't know how it works.</p>

<blockquote>

<p>WHY are participating in this thread</p>

</blockquote>

<p>Because I want to and have a perspective to add. Anyone reading has the freedom to consider what I say or not.</p>

<blockquote>

<p>"And as i am arguing with the biggest human encyclopedia (who 7 months ago didn't know that 1's and 2's count) i am probably wrong."</p>

</blockquote>

<p>And it still hasn't been confirmed by anyone in admin that ratings of 1 and 2 are factored into the average of <strong><em>ratings received</em></strong>. The last time anyone in admin who actually knows how the system works explained things publicly:</p>

<ul>

<li>A tally of ratings of 1 and 2 was visible on the <em>rater's</em> page of ratings <em>given</em>.</li>

<li>Ratings given of 1 and 2 were figured into the <em>rater's average</em>.</li>

<li>Ratings received of 1 and 2 were <em><strong>not</strong></em> figured into the recipient's average of <strong><em>received ratings</em></strong>.</li>

</ul>

<p>That was the last official public explanation given by anyone in admin about how the ratings were figured.</p>

 

<p>The fact that you can see that some presumed troll or underpants gnome has given ratings of 1 and 2 does not necessarily mean it is calculated into your average. Only folks in admin with access to that data would know for certain.</p>

<p>If you have any evidence to the contrary it would be interesting to see it. However, since recipients of ratings cannot accurately determine who gave which ratings, of the named persons giving ratings, anything else is guesswork.</p>

<p>Incidentally, while I've been called a "human encyclopedia" before, it's usually without the pejorative implication. But if you feel that epithets strengthen your argument, help yourself. However to be accurate I am a repository of uncategorized randomly accessed useless trivia, not a properly indexed human encyclopedia.</p>

 

<blockquote>

<p>"What puzzles me more than anything is WHY someone is rating old images that were shot 7-8 years ago with clunky 3 or 4 Mpixel archaic cameras now,when we have 24 and 36 Mpixel cameras?"</p>

</blockquote>

<p>What puzzles me is what anyone would believe photography should be rated on the materials used, rather than on overall aesthetics, composition, mastery of technique, accomplishment of intention, and presentation? If so, perhaps we should delete all of our older photos taken with inferior materials, including cameras with too few pickles and any film smaller than 8x10.</p>

Even in a half-megapixel image on the web, the superiority of a 20+ megapixel camera should be obvious. ;-)

<p>The web is sour enough without more megapickles adding to the mess.</p>

<p>Lex - What . . . no sweet gherkin variety of megapickles?</p>

"...we should delete all of our older photos taken with inferior materials, including cameras with too few pickles and any film smaller than 8x10."

 

At last someone has finally come up with a good suggestion.