News
GBR: Updates
Thanks everyone for sticking with me through the format changes. I have some ideas I am working through. The new Cuusoo tools I have been hinting at are coming along well but have been delayed by Cuusoo going down over the weekend. Maybe I will have the first new utility later this week, but it is hard to predict events with Free Comic Book day and Star Wars Day both coming up this Saturday.Cuusoo: Views In Comments Out
Well, Cuusoo has finally done it. They swapped out Comments for Views in the carousel of projects in the discovery page. Projects heavy on comments are obviously disappointed and projects heavy on views are happy. I can understand the move, views represent the projects ability to get people to stop and check it out while comments are largely ignored by people outside Cuusoo and even a relatively small group of fans can generate large comment counts.Ultimately, Cuusoo is encouraging certain behaviors by rewarding it (a spot on the carousel). In my opinion comments tend to attract the existing Cuusoo crowd (who have either voted for most of the high commented projects or have chosen not to) while strategies to increase views bring in new users who have a greater potential to support projects. Of course that is my conjecture, but if I could choose between 100 comments or 100 new views, I would go with the new views every time.
Either way, I would like to see Cuusoo reward this "activity" carousel spot based on Views per X. Where "X" is some kind of time frame: Hour, Day, Week, or Month. I think my preference would be Day or Week, but right now with it being absolute total, Poptropica is going to sit sit pretty with its 1 million plus views pretty much until it gets retired at 10,000 supporters. If this was most views in a week, then we would see a lot more versatility up there.
Crowdfunded
Remote Control AA Battery
The Tethercell is a fun concept. This device is a AA battery form factory for a AAA battery. The excess volume is used to add brains and bluetooth.The idea is that you put this "battery" into any device that can take a AA battery and then you can turn that device on and off with your smart phone or tablet. You can also set a timer for the battery and monitory its strength.
As a parent I can think of a few devices I would like to have the ability to turn off at will ;) but this has some fun applications for Power Functions. To be clear this is an on/off switch not a polarity control. So you can turn on lights, audio, motors, etc...but you can't make the motors go a different way or change their speed.
So, this would work great for the any build you have where you want EL wire or single direction motors were it might be difficult to get to the power supply. The 10178-1: Motorized Walking AT-AT is the example the immediately springs to mind for me.What I really see this being ideal for though, for serious Lego enthusiasts, is the Cons. You can put some nice EL, LED, audio, and motor effects into your build with one of the these you can turn them on and off without taking the whole thing apart or even getting behind the partition.
$35 will get you one unit.
Short Order Heroes: Light RPG and Character Generator
The content of this project is simple, it is a deck of cards with an adjective and a number.
To create characters you draw three cards to get a collections of adjectives. NPC's you encounter while threading your tale get from three to one attributes. When the conflict arises in the tale, the DM and the Player discuss how the attributes of all the characters affect the conflict. So if your character is "Cute" and "Foolish" and you are trying to get past a guard who is "Clumsy" and "Romantic" you could try straight diplomacy which no adjective affects ( difficulty 4). You could try to leap past the guard who is clumsy (difficulty 3), or you could try to charm the romantic guard with your cuteness (difficulty 2). You determine the result of the conflict by drawing a new card and using the number. If the number is equal to or greater than the difficulty, the character "wins" the conflict. The adjective of the card adds flavor to how the action is actually performed. For instance if the player attempts to charm the guard and draws "Wise" with a 6, the player gets by the guard, perhaps by having a disarming conversation about philosophy or maybe just complimenting the guard on their wisdom.
The game looks like it will be a lot of fun for low key RPG activities. It can also of course be used with more traditional RPGs to add some quick flavor or to more open ended story telling games (like Tell Tale) to give it more direction.
Threadless
Cuusoo
Pick of the Week
Pegasus Automatonby Amida
Support Level: 32
First off this Pegasus is a really great design. It looks both Iconic and organic at the same time. The droid elements are great for the ends of the legs.
The real art, is the dynamism though. I have made it clear that I am a sucker for animation of Lego builds and this one is mighty fine.
Pegasus Automaton from Amida Na on Vimeo.
Honorable Mention
Nutcracker Figurinesby Mmbace
Support Level: 19
These look amazing. Nutcrackers are a great subject for a Lego build because even though they are "human" their angular features and cylindrical elements are ideal for the nature of Lego shapes.
Combinators: Constructors
by Pittstop
Support Level: 21
I like this project concept a lot. It is a collection of construction equipment micro-builds that snap together to form a single robot/mech all Voltron style. Not Devastator style because the individual construction vehicles do not transform. This project has a solid presentation and has a very good scale to it.
Tmnt Turtle Blimp
by C3brix
Support Level: 16
Whether you are a TMNT fan or not this is one of the best uses I have ever seen for the Lego Blimp element. C3brix also did an amazing job of the custom printing!
LEGO Snow and Sand Shapes
by Nickyt13
Support Level: 86
This projects gets some serious points for originality but it is very hard to figure out what is going on from the cover image and title. The concept is to create large scale molds for packing snow into, to create Snow Blocks.
The concept is intriguing and quite popular it appears. I don't get the elements Nickyt13 picked though, A 4 x 6 x 2 brick, the dome and the new large cone? I would have showcased with the ever popular 2 x 4 brick and some more basic elements. I also think that if this idea is really viable, and Nickyt13 lives in the right climate, they should back up the idea by proving it with....milk cartons or some such then attach the video to the project...or find on the internet.
Mecanum Drive
by Fex00
Support Level: 23
Mechanum Drives are omnidirectional wheels. It is effectively a set of wheels organized into a wheel (like the picture above, imagine that!). This is a "good" idea in that I would love to play with it, but this arrangement above is 137 parts! And that is a single wheel, granted most of them are tiny but for four wheels you are looking at 548 elements and the creation of three new molds. That is a lot of investment.
I think fans of Mechanum Drives will be stuck with what we got, or second party wheels adapted for Lego.
L.M.I.R. - Roman Amphitheatre
by Bigboy99899
Support Level: 52
This L.M.I.R. series is really impressive but I really don't know how well they will transfer to actual sets. I think this is my favorite to date but when you really get down to it, this is a stadium. This means you need to stand above it, looking down for best effect and you need to populate it with a ton of figures with Roman-esque elements. Failing that and the build either looks empty (no figs) or anachronistic (any figs you can find). The stadium's basic seating holds room for 64 figures. Even at 80% seating that is still 52 attendees.
Notables
Crusaderby Laptop1655
Support Level: 5
A very clever build. High posability with a lot or rare parts. The black and green also go very well together. I think the main picture could be swapped out with something "easier" to make out though.
L.M.I.R. - Roman Bireme
by Bigboy99899
Support Level: 61
Not the first Bireme but certainly well designed with many nice features. The figure count, as always for a Bireme, is a little through the roof though. When is Lego finally going to build an honest Bireme?
Macross Valkyrie VF-1S: Skull Leader
by Daikoncat
Support Level: 51
I love Robotech/Macross and especially love any transforming Veritech builds. There is already a definitive build that is well publicized and pushed hard and still only crawling. I think the big roadblock for these projects is that the serious fans that would go for transforming Veritechs can already own them. These projects only strongly appeal to the serious Robotech/Macross and Lego fans.
Medical Imaging Department
by Goobs_6
Support Level: 22
Wow, this is some excellent work. A strong variety of equipment for minifigs to lay on while other minifigs stand next to them ;). Seriously though great work but a little too specialized for the 10k. That being said, if Goobs_6 is canny hey might be able to get a movement started in medical circles.
Ian Fleming's Chitty Chitty Bang Bang
by Ipaloosa
Support Level: 11
This is a charming projects and I hope it does well. In reviewing this project I realized that some new Chitty Chitty Bang Bang books have come out recently. An image search of Google for Chitty Chitty Bang Bang clearly shows that the the movie version is well locked into people's mindset for what CCBB "looks" like so Ipaloosa has an uphill battle.
I.M.C.T.B. Module 3: Cleopatra's Lighthouse of Alexandria
by Therealindy
Support Level: 7
See below
I.M.C.T.B. Module 2: Caesar at the Top of his Rostra
by Therealindy
Support Level: 6
Personally I think these builds are very attractive. There accuracy is also quite impressive, especially in the case of the Rostra. The anchors are a great idea for the design.
The issue with these builds, like so many in the "notables" category is that they are so specialized. These builds are not just for fans of ancient history but for fans who know it well enough to know what the Rostra is.
Week of Cuusoo
After the break is a showcase of all the new Cuusoo projects that came out in the last week.DC Comics: Minifigures Series
by Marvel-boy
Support Level: 100
Cuusoo has not to date shown any model for production other than sets. The documentation does allow for code (except MMOs) and parts as well but until I hear otherwise from Lego Cuusoo I have no faith that randomly, individually packaged minifig bags have any chance at production.
Update: Never mind, this project, has been deleted.
LEGO Snow and Sand Shapes
by Nickyt13
Support Level: 86
See above.
L.M.I.R. - Roman Bireme
by Bigboy99899
Support Level: 61
See above.
As I commented on the Birds project earlier this week (I can't manage to log in as myself here), I am pretty irritated by the views-in-comments-out move.
ReplyDeleteThe obvious reason is, of course, that this is unfair because the rich are getting richer. And we have enough of that on CUUSOO already. But that is actually not my beef.
My beef is the less obvious, but more important, fact that this actually puts the focus on projects that, for one reason or another, are subpar. By definition, objectively.
Remember for a second that these are not just the projects that have the most views; these are the projects that have the most views *and yet still haven't managed to get to 10k*. In other words, their ratio of supporters to viewers is relatively small. Which is a bad sign, not a good one.
If you check out any number of projects at random, you'll find that the solid ones, no matter where they stand, tend to have a ratio of some 15-20 views per supporter.
For the now-advertised top three projects by views, however, the numbers are 200, 40, and 34, respectively.
Views per week/day would not fix that, either. If some project suddenly gets ten thousand new views in a single day, but only three and a half new supporters, there is no point of trying to get it more views still by featuring it on the home page. It already has way more attention than anything else, no need to pile up. If 250000 fans of a particular brand can't get their project off the ground — well, who will?
So I think the X in your example should be not a time period, but the number of supporters. Then projects that really strike the nerve with people would get more exposure, no matter where they stand. A project with 5000 supporters after 50000 views would get the same treatment as a project with 100 supporters after 1000 views.
This would highlight the projects that truly have the highest potential. Across the board.
I respect anyone who is passionate about Cuusoo and more so anyone who is passionate enough to write about it.
DeleteI find which project shows up on the Discovery page to ultimately be of little value. From my observations, the vast majority of project supporters for projects that get to 10k come directly to a specific project, support it, and leave (Never seeing the Discovery Page or any other part of Cuusoo). A small fraction of these people coming directly for a specific project look around, maybe do a search or click on suggested projects (Never seeing the Discovery page). In the end the The vast majority of people who see the Discovery page are people with vested interest in hanging out and seeing what Cuusoo has to offer and do so more than once.
Cuusoo has stated that their intent in the change is because "Promoting the number of page views is a much better metric of the overall popularity of a project/concept than the number of comments." I see nothing wrong with their logic in this. It stands to reason that if one million people have clicked on a link for something about Poptropica, and Galaxy command has a group of very passionate individuals who collectively have made nearly 6000 comments, that despite the passion of the GC fans, Poptropica is more popular.
Similarly my statement about Views per week in not to showcase the passion of individuals but to increase the variety of projects that will be shows to the Cuusoo regulars. To show what is trending. As is, we will see very little variation in the Carousel. Sure, Poptropica enjoyed insane surges of views as did LoL but that is old news, but their footprint is, currently permanent and stagnating. But YES, does accurately showcase the popularity of the concept (and connectivity to the internet). My offered suggestion is to show us snippets of popularity instead.
Now, all this stands to reason as long as people are not being rick rolled into viewing the projects. As long as they see a link appropriately titled and then click on said link, the argument can be made that the viewership is indicating how powerful the concept is at getting people to give it a minute of their time.
Your model of views vs support reflects this for the most part but in this case your are making many more assumptions. You really don't know what is happening here. By this Logic, Purdue Pete was at one point the best thing to ever happen to Cuusoo. Not because of your argument explicitly but because of the surgical accuracy of their fan base. Every person going to PP was going there to vote for it. I similarly accomplished this with Ogre as the Kickstarter campaign directly targeted the project.
Interesting, there is a limit to the number of characters in a comment, did not realize this....so continuing...
DeleteAnother flaw in your argument is that we don't know the explicit details of why people are not supporting. Many fans of Poptropica may not realize the point of supporting for instance. From the FTL project I KNOW many people NEVER read the project details. So projects that fail to point out what Cuusoo is, or an uneducated audience really hurts its chances.
There is another great reason why Poptropica has such large numbers of views and such low support. It is targeted at 7-8 year olds. They can click on the link from Poptropica but they are not allowed to vote.
I think what you should really be asking for is a logical regression model that predicts what people will like based on others who supported the project that brought them to Cuusoo.
Most people supporting the popular projects show up, support, see the suggested projects, ignore them, and leave.
This makes perfect sense because the suggested projects are little better than random. A logical regression model could see say "Hey, you are here to see X, well, people who supported X also supported Y, check it out." Each project the user supports only makes the predictions better.
You see the beauty of this model is that it uses the crowd wisdom of the Cuusoo regulars to suggest the best projects to the people flitting through.
Btw, I used to track and report on Support to View ratios because at the time I thought it would have merit like you do now. I stopped reporting it because I found that the evidence did not support the theory (for my opinion on what the best stuff was).
Cheers!
Thank you for your thoughts, Glen. Of course we never know where the cold numbers actually come from; what we do with them is always an approximation and an extrapolation. However, it is a given that extrapolating from two numbers, combining two different things, is always a safer bet than taking just one naked number and running with it.
ReplyDeleteYou have identified two flaws with the sup/view metric. I will address them in reverse order. You say that "we don't know the explicit details of why people are not supporting". Very true. But that is entirely irrelevant, because of the way CUUSOO is set up. Every project *must* get to 10k. If seven billion people are huge fans of LoL, and visit the project, but only leave 9000 votes, it won't get a pass. And as you said yourself, those people are unlikely to return to the site ever again just to see the error of their ways.
Yes, we don't know people's reasons for merely viewing. But we also do not care, by design. Never did. I'm not saying that's a good thing, but as long as we don't care, we don't care. So we should be upfront about it rather than pretending that somehow the views do make a difference after all.
The other flaw you identified was PP. Yes, there is no denying, by the sup/view metric PP was an outlier. But it was an outlier anyway, every which way. That was its whole point. People set out to create an outlier, and so they did. It lasted for a whole of one day and then was through, so no damage done if we had featured it for that one day in the proposed sup/view tab. In fact if anything, it would have immediately brought its ratio down, better reflecting its *actual* potential.
You see, that's the beauty of the sup/view system: it is entirely self-regulating. If a project with a high ratio gets featured, gets a ton of new views, but not many supporters, then its ratio gets worse all by itself, and it makes room for another project. And if its ratio gets better thanks to the additional exposure -- well, that means that its appeal to the public at large is even higher than expected, so it totally deserves to stay in the top spot and sail on smoothly.
This system would also be much harder to game than just views. Anyone can get his project a thousand "new" views simply by reloading the page a thousand times. And that number is then there to stay. It never drops down again. With some dedication and a bit of spare time, you could get all your projects ahead of LoL. But in order to game the s/v metric, you'd have to either artificially increase your number of supporters by creating sockpuppet accounts (which are already detected and removed by the CUUSOO team), or artificially increase the view count of *every single* competing project (which cannot be pulled off manually, and stands no chance to go unnoticed if attempted programmatically).
You say you used to report on sup/view and found no merit in it. But that's precisely because only *you* were looking at it (and some of your readers). That's not enough to fix skewed stats. But if instead those projects got featured, and exposed to a wider audience, then the skewed stats would have fixed themselves in the blink of an eye.
Again, the system is self-regulating. But it is only self-regulating if it's actually in place. Then it can guarantee a high influx of reviewers on a large number of projects, no matter where they stand, and fix the stats across the board, whichever way they were skewed and for whatever reason. It's a win-win.
Oh, and I am actually keeping tabs on a number of random projects, and even maintaining a spreadsheet of nine projects that have nothing in common other than having started on the same day, monitoring their sup/view figures, with a nice graph and all. The projects that come out on top are precisely the ones you'd expect to, and the projects lagging behind are precisely the ones that could use improvement in one regard or another. And if they had more exposure, the numbers would be more accurate still, separating the wheat from the chaff even more effectively.
ReplyDeleteAs I mentioned above, the way the system is set up right now, it's the plain boring "the rich are getting richer" -- a concern many people are very quick to voice. Now here's the twist: I don't actually think that's necessarily a bad thing. And in fact it may have been TLG's full intention -- you do want to be pushing *some* project, and if you spread too thin, you won't have a single candidate for the next review.
So make no mistake, we absolutely have to be piling up attention on some projects. That is not a problem. The real problem is, how we pick said projects. And what we are doing right now is pile up attention on projects that have proven, as a matter of fact, that the insanest amounts of attention still can't get them to 10k.
(Again, we can come up with excuses or even solid reasons, but at the end of the day all that matters is that they just don't reach the one and only threshold that every project absolutely must.)
At the same time, there are dozens of projects that are reliably getting seven new supporters for every hundred views, and have always been. It just takes them forever to so much as get ten views.
So what I'm saying here is not the usual boring, idealistic, meaningless "High time we rethink our priorities". What I'm saying is, our priorities are just fine, but it's high time we adapt the system such that it actually reflects and caters to these priorities we always had in mind, from day one.