Reply to thread

Message

<blockquote data-quote="UngeheuerLich" data-source="post: 9407933" data-attributes="member: 59057">Yes. Mean makes more sense.Bit we still don't know if they use the unweighted mean value. Or if they had some weighting or a nonlinear scale.star rating -&gt; satisfactory scoreLinear:1 -&gt; 02 -&gt; 0.253 -&gt; 0.54 -&gt; 0.755 -&gt; 1So votes of 1,2,3,4,5 would average to (0+0.25+0.5+0.75+1)/5 = 0.5They could well have used:1 -&gt; 02 -&gt; 0.33 -&gt; 0.64 -&gt; 0.85 -&gt; 1The same votes of 1,2,3,4,5 would average to (0+0.3+0.6+0.8+1)/5 = 0.54So without these values, we actually don't know what their treshold of 70% means.They could also have some flexible valies attached: depending on how many 1 stars you gave, the less they count. This would be done with weights, but this is too comolex for me to write down here.</blockquote>

[QUOTE="UngeheuerLich, post: 9407933, member: 59057"] Yes. Mean makes more sense. Bit we still don't know if they use the unweighted mean value. Or if they had some weighting or a nonlinear scale. star rating -> satisfactory score Linear: 1 -> 0 2 -> 0.25 3 -> 0.5 4 -> 0.75 5 -> 1 So votes of 1,2,3,4,5 would average to (0+0.25+0.5+0.75+1)/5 = 0.5 They could well have used: 1 -> 0 2 -> 0.3 3 -> 0.6 4 -> 0.8 5 -> 1 The same votes of 1,2,3,4,5 would average to (0+0.3+0.6+0.8+1)/5 = 0.54 So without these values, we actually don't know what their treshold of 70% means. They could also have some flexible valies attached: depending on how many 1 stars you gave, the less they count. This would be done with weights, but this is too comolex for me to write down here. [/QUOTE]

Verification