When I search on download.com for a software, I usually consider items with an high user rating before others. slimKEYS is listed on download.com, but I'm still waiting for the first user to rate it. Looking at TuneSleeve, a friend's tool for importing album art, you might ask yourself if this whole user rating stuff is really working.
The global rating is 2.5:
But these 4 votes are:
I'm good enough in maths to say this averages to 4, not 2.5. How do they count at download.com? Now, if you click on the "[?]" at the bottom of the review list, you get this explanation:
What is the review score?
The overall review score is based on a number of factors, including the overall value of the review, the number of helpfulness votes it has received, and the number of reviews submitted by the reviewer.
Fine. That makes sense. Let's check those reviews again:
- The reviewer made 53 reviews.
- 1 of 2 users said the review was helpful.
- The reviewer made a single review.
- 2 of 2 users said the review was helpful.
- The reviewer made a single review.
- 1 of 1 user said the review was helpful.
- The reviewer made a single review.
- 2 of 2 users said the review was helpful.
Ok, I get it. Because the first reviewer is highly active, his review gets much more weight than the other three. I imagine this helps eliminate people trying to fake good reviews for their own software, but it also renders reviews with low number of votes totally useless. In fact, I don't see how it can stop someone from faking reviews for his software. You just have to vote for other random software. Why not vote low scores on all your competition! By trying to circumvent cheating, download.com is changing the real picture, and can hurt small software makers as much as they can help them. It's not fair.