Additional elaboration on the marking scheme
I follow a standard when I apply the marking scheme. Marks always start at 8.0 on first impression. Getting to 9 requires a simple +1, any +1 after 9 is exponentially decreased with the rule of diminushing returns. 9 is called the barrier or upper value - the 9 will try to harden the path while forbidding the peak value of 10. Inversely, it is easy to obtain 7.0 with a single -1, but harder to reach a 6.0 (well not that hard) because of diminishing returns passed the lower mark of 7.0. The lower barrier is in our case a lot weaker than the higher barrier, so marks can drop below 5.0, but in general, they won't (cuz I try to be generous and if there's something below 5.0, then chances are that it is not even supposed to be in the best bearhugs in the first place unless they were extremes in different criteria ranges (like a very fabulous pic, but extremely small..)).
Smaller factors and aspects in the criteria level that are not discussed in the individual comments are those that generate the decimal change in the mark (±0.1 to ±0.9). Not ALL significant individual comments are displayed, so a mark can even be 6 or 9.5 without reasoning/comment. I only display individual criteria comments when I feel like it. Even if these comments were displayed, it may also be only a portion of them. They'll always be some pluses/minuses you'll never see :P cuz that's just me :PP and it'll also protect/prevent me from justifying my marks. One of the small factor NEVER displayed is called the uncertainty factor, which will be discussed in just a moment.
When both pluses and minuses are involved in a rating, then a difference is performed first, then the remaining units are bounded by the law of diminishing returns.
There's an uncertainty factor of ±0.5 per criteria except for the last criteria. Uncertainty takes into consideration the pending/floating decision I have for the criteria. It may consider comparison issues.. but I normally try to avoid that - because I said that I'm evaluating the pic as one - not compared to another.. but sometimes, comparison is inevitable. If an uncertainty factor requires more than ±0.5 for a fair grading adjustment, it will be added in the balance factor (taken into account in the criteria for factors of opportunity). The uncertainty factor is the only value that is not subjected to the law of diminishing return because it is used as a correcting value and will never cause the marks to go beyond their peaks.
Speaking about comparison, the size criteria is the only grade that can be compared with other grades.
The last criteria is called the factors of opportunity and contains the balance factor which try to correct uncertainty and/or contradictions. If you want, the balance factor contains the counter-uncertainty :P The grading for factors of opportunity is split into two parts: There's a balance factor, which is limited to a change of ±5 (which is huge). The other half is reserved for the overall bonus (-5 to 5, with most likely +0 and a rare +1/-1). In the case where a criteria should have been given a better emphasis because it was penalized in another, the criteria for factors of opportunity comes into play and adjust the mark to balance the criterias. This prevents having marks above 10 or below 0.
Unidentified bonuses also appear in factors of opportunity and they represent the biased/taboo elements that cannot be discussed publicly because of site guidelines ;) You get the idea... Let's say the wrestler has something pointing out in his trunks, you won't see me put that as a +1 'XX' comment, but instead as a +1 (unidentified) comment :P. Biased elements can also be negative.
You'll never see a 10 because there's no such thing as a perfect art/photograph regardless of the criteria.
I'm aware that there are similar aspects in the rating scheme.. for instance, it may be hard to differentiate "background" with "scenario". In the case, where a pic has received twice the amount of points for a single element, it should be normalized/negated in the balance factor.
Here's the order of precedence or more or less the algorithm:
• Criteria starts at 8.0
• Rule start: Diminishing returns starts (applies after differences/summations)
• Heavy impact (-3/+3 or worst/better - this is rare)
• Plus or minus integers (-2/-1/+1/+2)
• Plus or minus small factors (±0.1 to ±0.9)
• Rule ends: Diminishing returns ends
• Uncertainty factor ±0.5
• Trailing uncertainty factors summed for the last criteria as balance factor (±5)
Ok, enough chit-chat, hit "ENTER" to get in or hit 'BACK' if you didn't finish reading the crap that was on the previous intro page.
Usage of this site requires agreement and understanding of the terms of service - privacy policy - disclaimer.