Suppose you make a antiageing cream with 'nothing but water' and another one makes a wonderfull antiageing cream with 5% retinol 2% ginkgo biloba 3% glycol acid etc
Then your system will tell that the company making water, is the best. That social ethics of that company is the best. That the water product is the least toxic...
But you seem to miss the complete point , that one actually compares the performance of a product.
Its like buying a green car that can drive chip tuned 30mph versus the same care driving 180mph and consuming three times more...... These examples actually exists, because with chiptuning, engines end to be the same in the whole series of cars, but just differ in performance and consumption through their tuning.
So how do you solve that problem of 'performance evaluation'
Help get this topic noticed by sharing it on Twitter, Facebook, or email.
EMPLOYEE0You've identified a real gap in the kinds of information that a service like GoodGuide can provide to consumers. How well a product works (its functional efficacy) along with price are the two most important factors influencing most consumers purchasing decisions. However, credible, unbiased data on functional efficacy is virtually impossible to obtain for most products. For most products, we have access to only what manufacturers themselves claim about efficacy. For a few products, third-party certifiers test products and make the results public for free (e.g., EPA's Design for the Environment certifications includes efficacy testing). Consumers Union tests products but does not make the results public without a subscription.
So GoodGuide addresses these information gaps with user recommend/avoid votes and reviews (either acquired on our site or via syndication from sites like Amazon). These reviews often contain information about efficacy, but they do not reflect standardized testing of efficacy so they are at best positive or negative signals about a product's potential performance.
When people vote fo a product you end up always after a time with an average score.
How do you average out the fact that everybody uses his own scale and you cannot compare1000 individual s ales
Mathematically there exists one system that makes this happen. We programmed it in our farmacompendium.be its based upon ratings as with games. Compare with the tennis rating. Each time you win against another player you climb inthe ranking.
So we programmed that system and it works fine. People have to compare two products. The product that wins in a comparison comes higher inthe ranking. To speed up things one compares fe with five products. That makes theranking change as you organise an us open
If you are interested i can share the software code
The code is comparable as an Elo rating system used also in chess
the only thing youcan have problems with is how fast should the system be influenced by each vote
We choose a slow adoption. But since people dont vote too much the ranking comes slowely