Well, there are several issues (some of them which are of course beyond your control), and basic statistics do not really cover these issues. However, I re-read my post and it came across as a very grumpy one. I really do not like grumpy posts, so allow me to apologise for that right now. The printer guide is immensely useful as a central repository of reviews, and I should have been more congratulatory over the publishing this significant piece of research. Kudos and thanks for that. So, please do not take the rest of my post as a criticism, it merely explains my thinking to add a couple more refinements to this significant and important survey.
Now, of course you created a minimal sample size, and that’s important.
But a sample of 20 is not as significant as a sample of 140. You can bandy around the maths every which way you want, but basically you are looking at footfall here: if more people buy printer X, does it mean it’s good? Well, that depends on the situation surrounding it. If, for example, Makerbot came to the show before Ultimaker, and the latter overtook Makerbot on 3dhubs, then that is a very, very significant fact that outweighs the numbers themselves, even if they only have an edge of 10 printers (I’m not trying to sell Ultimakers, I’m trying to get the best printers to the forefront, and hopefully have the truly awful ones slated.)
Then you have something like the Lulzbot Taz 4 which has not been around for very long. It has come below the Witbox, which really makes you wonder, since the Wibox does not have a heated platform whereas the Taz 4 does, it’s noise level is higher than the Taz 4, it is not open-source and in fact it has a host of issues with it.
Now, of course judging between these two printers is subjective, since what I need from a printer is going to be different than someone else, which will explain this apparent anomaly. Also, somone with a Form 1 will be far more appreciative of fine quality printing than someone owning a Makerbot.
The point is, though, that most of the results made me scratch my head.
So, what I’m advocating is for two things:
1) Show the response numbers on the front page of the listing. This way people can look at them and say: “well, printer X has only 21 responses, but it ticks all the boxes for me and is very new, therefore it makes sense few people own one. I will take the risk.”
Or they could say: “Why would I buy a printer that has only been tried by 21 people when I can get one that has over 100 people using it? Surely that technology will be more mature.” (Of course, this is simply not true either, since there is a very famous brand shipping printers that are essentially very expensive land-fill.)
2) Publish the raw figures. What’s important about open-source is that the information is made available. Never mind how poorly someone might interpret it. That’s their problem.
I hope this has not come across as too grumpy. If it has, re-read the start of this post and imagine me in a hot-tub with a cool beer, smiling and waving. You will only be able to imagine that, since I do not own a hot tub, don’t like beer and and if I’m waving it’s because I need help to find my glasses.