Let me preface the following by saying this: I have nothing but the utmost respect for [MENTION=21]The Joker[/MENTION] and his advice on anything PC related. You have shown me nothing by incredible kindness, sincerity and assistance since the very first moment we spoke. The following is NOT BY ANY MEANS an attack against you or anyone else. The following is merely my observations and understandings based on how the article was written and how the list of cards was determined. The list, according to my understanding, is a fair and correct list, but let me get into the details.
It seems all the cards are ranked and listed based on their high Passmark Value Score, which Passmarks calculates and updates on a daily basis based on thousands on Passmark performance tests versus the Dollar price of the card. Thus, a card with a higher Passmark Value score will have more graphical power per dollar spent. All based on actual tests from within Passmark. How accurate Passmark is can probably be debated, but at least it gives an indication as to the value of each card, and as a catious search on the web showed me, one of the best ways of determining best bang for your buck.
Passmark also seems to ignore differences in manufacturer types, thus averaging out the performances of the different manufacturer cards and giving a total average performance score per card.
So we are left 10 cards, all with some of the highest performance per dollar calculations according to Passmark, making them some of the best value for money cards you can buy. If we compare the list of cards in the article, with what Joker shared in another thread on the forum, we see that all his listed cards has chipsets that feature in this list, namely the GTX 960, GTX 970 and R9 380. Furthermore, there is a difference of 5% in price between the cards listed by the Joker and within the MyGaming list. For the 960, his lists is showing the price as cheaper, and the other cards are 5% more expensive. Now, my argument is, is the cards listed by the Joker 5% better performing cards to warrant their prices, and will a consumer, looking at a PC's each with one of the listed cards in really notice the difference?
Then, the highest rated card in the list on MyGaming is the Geforce GTX 950, a card that retails at one location for R3036, and has a Passmark score of 5252. The next logical step after this card is the GTX 960, retails from Joker's list as R3870 with a Passmark score of 5938. That is a jump of 27.4% in price, with a 13.1% jump in Passmark performance. Does the jump in price justify the jump in performance, bearing in mind that this is an objective view of comparing the cards and not a subjective feeling? Personally, I would, looking at these figures, knowing that my budget is tight and that I need to get the best value I possibly can, I would settle for the GTX 950. It makes logical sense based on the scores and facts put in front of me. Is the 960 better, of course, but hey, I am a budget conscious consumer, and if the math and Passmark scores shows that I would get better bang for my bucks from the 950, it is the better value card to buy.
Again I will say, this is my understanding of how the article was put together, which makes sense, and conveys the correct message. My findings are based on how I would imagine the writer went about getting the info and deciding which cards to be ranked here. Should this whole methodology be flawed and incorrect, please let me know, and let us all know what is a better way of objectively determining the value of GPU's. Maybe what comes out of that will prove even more valuable to readers, lurkers and members for the future.
SOURCE
