Author Options:

The World's worst Instructables? (Updated) Answered

I've posted this as a review of how the current rating system works at the low end of the scale, with reference to my 2007 post
Reminder of the scale:
0.5 = Worthless
1.0 = Poor
1.5 = Some flaws
2.0 = Almost average
2.5 = Average (we don't go much higher than this here)

Lowest rated Instructables as of today:

extreme mini knex gun
"How on earth is this rated higher than a buynch of my stuff??????"
This one seemed to be very unpopular, 60 ratings brought it down to 1.10 but people were aiming for 0.5

Author "Hold it over a trash can and beat the living () out of the top"
Generally viewed as one of the most moronic Instructables ever, 154 ratings gave this 1.16 (many of which have been translated from the old system).

small-knex-gun 2
"my first instructionable"
Not popular at all, 27 ratings took this to 1.27

How to make a flashdark
"complete waste of instructables.com storage"
They didn't get the joke... 23 ratings 1.30

"why did you chose knex as your one and only keyword?" (knex now appears twice in tags)
32 ratings 1.36

smallest knex gun EVER
It's only got 2 pieces, which was not appreciated by the KNEX community
Author: "come on people plz leave me alone
1.37 from 41 ratings

The coolest virus ever!
"virus creators are the bane of all existence. Sorry, but I'm gonna have to ask you to rot somewhere uncomfortable."
17 ratings 1.40

"You can't really just stick screws to stuff and call it steampunk."
21 ratings, 1.45

"go back into the woods and practice your banjo."
20 ratings 1.56
July 06

Author: "yeah, well this is my first instructable.. I've made ones that shoot some 37 feet or so, but they're too complicated for me to understand how I built them. I am almost annoyed at myself for putting on such a crappy gun. oh, well, I don't really care"
27 ratings 1.58

And what of the The World's worst Instructables 2007? Only the Snapple and tongue Instructables haven't been pushed into a corner.

https://www.instructables.com/id/E8FS0ZXF1B3R8DV/ (out of sight?)
"i smell summin and it aint laser burns.. I lived on a farm once and waded through alot of it."
33 to 1.42

https://www.instructables.com/id/EDW2YEMF0ZSTFDK/ (out of sight?)
"This is precisely the same instructable as this one, complete with the same spelling errors and photos, except (your) instructable only has thumbnails of the blurry photos."
44 ratings 1.42

https://www.instructables.com/id/EFGG7JWF18DW9NB/ (out of sight?)
"can we all PLEASE stop talking about weed?"
126 ratings 1.94

https://www.instructables.com/id/EWOYNZ04N8EWIJLOX6/ (out of sight?)
"How come such a poorly made Instructable can be so popular?"
115 ratings 2.13, that's closer to average than "almost"

"Next you'll b doin it to your ()"
88 ratings 2.56 (above average)

The current system doesn't seem to be giving us a good distribution if nothing has yet managed to hit "poor" (even when a group of people were trying to rate beneath it)

Statistics as of 9th March:
5.00-4.5 16 0.07%
4.49-4.0 362 1.6%
3.99-3.5 3269 14.7%
3.49-3.0 10238 46.1%
2.99-2.5 4556 20.5%
2.49-2.0 853 3.8%
1.99-1.5 111 0.5%
1.49-1.0 8 0.04%
0.99-0.5 0 0%
unrated 2787 12.6%

81% of all Instructables are rated between 2.5 and 4.0 (93 % if you exclude unrated Instructables)



With the great deal of these I agree but with some I don't.. There are worst ones..

Please tell us, this is simply a rating review. L

Since large numbers of projects - good or bad - go completely un-rated, I don't think the current system is a useful measure of... well, anything.

Sometimes it is- it measures how much people care about rating it. If I see an unrated recipe, I know it's probably unexciting- a 4* recipe I will read because for people to rate a simple Instructable up so much it is probably very amusing or otherwise better than the average, a 2.5* recipe is almost certainly rubbish or terribly written so I won't.

If the topic of the Instructable is something controversial and it has 2.5 stars, I assume someone has taken offence to it and downrated, so will probably read it out of curiosity :) but otherwise a rating of 2.5* or below instantly says "this is probably not worth reading".

But now it's on 3.38. I don't see much value in rating Forum topics and even less for Answers. The title should be enough. (I don't judge things by rating) L

. I don't pay much attention to the ratings. I try to remember to rate stuff, but I don't use the numbers for anything. I must admit that I will often not bother to rate the really bad iBles.

I stick to pageviews. It doesn't determine the quality of the instructable, but it does have something to say about the interest in an instructable.

Or how many other sites are linking to it, or is featured, or a competition winner, or like IanHampton's USB stick keep popping up on the front page. L

It's fun to click stars though, so it achieves something. Amusement?

A large number is only 12.6%. But whilst I was trying to find the 3.00 / 2.99 threshold I did find that virtually all the 3.00 Instructables only had received one rating. You seem to concur with my analysis. L

Yep, I just used less maths* to come to my conclusion.

*That is, none.

And did it the hard way... when will I learn? L

hehe, One ible will eventually get into the 0.5-0.99 range. You can guess which one that is.

Ahhh the list... When they come and ask for our best and brightest, they'll get sent here. Let's not them see this ehh?

There's hardy anything down the bottom end of the score table really. L

now which of them are really rated that low ? (if you filter out those deliiberately 0.5-bombed down)

The KNEX ones are rather poor, but they received a lot of ratings designed to score them that way. Same with the rest really. The only one that shouldn't be there is the tongue-mutilation, that's been downrated because people just didn't like it. L

Our rating system is based on the work Christopher Allen posted here Systems for Collective Choice.

Yes, it is difficult to get something to a rating of 0.5, but overall the distribution of ratings is better than our old +/- systems and better than a straight average of the 5-star ratings.

I agree that the range of the old system could get a bit silly, but the current system is not giving a good distribution. With 83% of all Instructables above average (as it is defined at 2.5) the true value of "average" is somewhat unclear.
In terms of actual ratings "average" is 3.0-3.5. The whole system could do to be shifed down a bit?
For example, instead of using the siteaverage in the formula, use the value of 2.5, being the site definition of average?


Good distribution is relative. With this system, we were aiming for better distribution over our previous system, and we achieved that. Of course, there's always more to do.

The distribution is narrower, but as I've alluded to previously having more than 80% of Instructables over the site definition of average leads a person to ask " what is the real Instructables definition of "average"?" If it's not based upon regular statistics it is something arbitary and unknown (and somewhat self-congratulatory perhaps?).
We all know there's some poor quality stuff on the site, but officially speaking unrated to "poor" is 13 %, and people can't force a "poor" even if they try. For rated posts not even 5% are below average.


Compared to a straight average, the distribution is wider. The current distribution isn't directly comparable to the +/- system. When we translated we took +s to equal 4s and -s to equal 1. The numeric value of the site average is computed regularly, and I don't know what it is offhand. It does deviate from the suggested "average" value; the only thing that will push the site's actual average rating towards the suggested average of 2.5 in the current system, is the ratings themselves.

Thanks for the feedback - I hope I've not been ambiguous between average (normal distribution) as compared to average (defined as 2.5 by the rating system) L

There's a problem with your list. I don't see, "How to headbang" Read it, rate it, add it to your list.

Hmm, that didn't show up, or I missed it. Thanks for bringing that on to my attention! L

Heres a tip. use the random button. Your bound to find a terrible instructable.

Some of the comments are amusing, but I was mainly interested in the rating system L

Additional: I rated something 3.5 which was a bit above 3 - this actually decreased the overall rating. It would seem that we're obliged to go to the ends of the scale to show our support / disdain. This is effectively the same as the old +/- system but heavily-damped


I think what's happening here is that the rating of the Instructable is out of date (for whatever reason). When you rate something, we update the rating and it's possible that we're taking your rating into account as well as some other inputs. However, I agree that when the rating moves in the opposite direction as your input, that is counter intuitive and a bad interface.

There's another case where this can happen: If another user rates a project in between the time when you load the page and you make your rating.

However, there still might be something up, so if anyone sees the rating moving in the opposite direction as your rating, please post:
  • URL of the instructable
  • previous rating
  • your rating
  • updated rating
  • the time this happened, as close as you can tell