These forums have been archived and are now read-only.

The new forums are live and can be found at https://forums.eveonline.com/

EVE Information Portal

 
  • Topic is locked indefinitely.
 

Dev blog: Project Discovery needs You!

First post First post
Author
Previous Topic Next Topic
Dave Stark
#41 - 2016-01-22 19:06:18 UTC  |  Edited by: Dave Stark
just jumped on sisi to try it out.

not quite sure what i'm meant to be doing - not sure i care either, i'm getting isk/LP for randomly clicking.

in all seriousness though - the tutorial doesn't really explain much, you're asking me to match a random picture with little icons that don't look even remotely like it... and i'm pretty sure all this isk/lp i'm getting clicking randomly is going to add up faster than any kind of multiplier i get by taking my time to try and figure out which icon is meant to match the scribble on the left.

honestly - remove the isk/lp reward. i know you want to incentivize it, but seriously i'm getting isk/lp for clicking randomly and not caring about the fact that this is meant to be helping scientists. because i'm getting isk/lp for randomly clicking and submitting complete trash, that's what people will do.

"****, need some new sisters probes - better **** up some scientists results with crap for a couple of mins while i fly to the nearest LP store station".

"fleet's waiting to form? may as well **** up some scientists results with crap and get free isk while people get ready to leave."

genuinely - if this mini game thing is meant to be serious and help research, get rid of the reward. or, only award it to people who are above say, 70%, accuracy.

edit: isk reward seems to be 1k isk per 1% of accuracy. that means 100k isk per at 100% accuracy. not a great deal of isk, not much incentive to try and do it correctly vs spamming complete crap.
LP seems to be a bit more random, though.

edit2: just tried to do it properly and get my rating up. it seems that most of them could be 1 of like 3 or 4 choices i always seem to get it wrong. i'm bad at it. this further promotes my "spam **** answers" mentality since i may as well get **** rewards quickly than take my time and still get **** answers.
Makhar
Joint Defence Force
#42 - 2016-01-22 19:06:31 UTC
Maybe I'm missing something - where/how to get started? I'm on SiSi, now what?

_ I tried real-life once.  The graphics suck._
CCP Paradox
#43 - 2016-01-22 19:11:45 UTC  |  Edited by: CCP Paradox
Makoto Priano wrote:

Is there some allowance for, say, 'common misidentifications,' as in this case? Or do I just need to work on my pattern recognition?


Remember there is a color channel selector at the bottom of the sample image. Changing the channels can help you identify the correct characteristics sometimes if you are having difficulty.

There is a seperate feedback thread directly for the feature on the Test Server Feedback forums, Project Discovery is now active on Singularity

CCP Paradox | EVE QA | Team Phenomenon

Space Magician
Datterich
State Protectorate
Caldari State
#44 - 2016-01-22 19:13:01 UTC  |  Edited by: Datterich
nicly done, though i have some critics about it:

It is not easy to match the given picture to the structure (almost as hard as if it would be in a real lab^^). The tutorial should be extended or give at least some more Information. A short video-introduction would be really helpful.

The reward-system is very inappropriate as it grants you rewards for randomly clicking. This does neither help the gameplay nor does it give any scientific benefit. It would probably be good to add some more (or a lot more) control-questions and give reward based on the accuracy-level.
I'd also add a skip button if you are unsure and don't want to ruin the statistic significance.
Makhar
Joint Defence Force
#45 - 2016-01-22 19:14:32 UTC
Makhar wrote:
Maybe I'm missing something - where/how to get started? I'm on SiSi, now what?


Nevermind I found it Neocom > Business

_ I tried real-life once.  The graphics suck._
Makoto Priano
Kirkinen-Arataka Transhuman Zenith Consulting Ltd.
Arataka Research Consortium
#46 - 2016-01-22 19:19:01 UTC
CCP Paradox wrote:
Makoto Priano wrote:

Is there some allowance for, say, 'common misidentifications,' as in this case? Or do I just need to work on my pattern recognition?


Remember there is a color channel selector at the bottom of the sample image. Changing the channels can help you identify the correct characteristics sometimes if you are having difficulty.

There is a seperate feedback thread directly for the feature on the Test Server Feedback forums, Project Discovery is now active on Singularity


Right-o! Definitely using that. I guess what caught me up is remembering that, indeed, some staining of objects will be visible through the nucleus, despite not being part of it.

Also, a blue-only channel or a blue-red channel might also be helpful. ;)

Itsukame-Zainou Hyperspatial Inquiries: exploring the edge of the known, advancing the state of the art. Would you like to know more?
Aeril Malkyre
Knights of the Ouroboros
#47 - 2016-01-22 19:20:29 UTC
I am so glad to see this moving forward. Cheers to CCP for helping out with real science :)
Sarmatiko
#48 - 2016-01-22 19:24:11 UTC
It's easy enough...once you figured it out.

But I already failed twice with the Nucleus\Nucleoplasm recognition: http://i.imgur.com/2ZU2Mmr.jpg
As you can see Nucleus wasn't correct "answer" despite description ("overlaps with blue marker and evenly stained") and correct was Nucleoplasm (green should color everything, except small circles, which is not the case on this image).
CCP Paradox
#49 - 2016-01-22 19:32:29 UTC
Sarmatiko wrote:
It's easy enough...once you figured it out.

But I already failed twice with the Nucleus\Nucleoplasm recognition: http://i.imgur.com/2ZU2Mmr.jpg
As you can see Nucleus wasn't correct "answer" despite description ("overlaps with blue marker and evenly stained") and correct was Nucleoplasm (green should color everything, except small circles, which is not the case on this image).


In that example, it is only the nucleus, as soon as you see the green outside the nucleus also it is something else. Remember to use the color channels to see, especially for times there are green inside and out of the nucleus.

CCP Paradox | EVE QA | Team Phenomenon

Space Magician
Dave Stark
#50 - 2016-01-22 19:36:57 UTC
make the hexagons bigger. it's really difficult to see what each option is.

also, when you mouse over them, make the window pop up somewhere other than on top of every other hexagon.

these 2 things would make it significantly easier to try and compare your options with the image that you're presented.
Old Man Sam
Fishy Old Men in Space
#51 - 2016-01-22 19:52:40 UTC
So far I'm on tutorial sample 2, but this is interesting.
Some requests for the UI:
1. Let me make it bigger. I have a big monitor, yet I'm staring at a small square. Enhance!
2. Preload all the colour images. chances are if I'm doing it right I'm going to need all of them anyway.
3. let me map the colour buttons to number buttons or something. I don't want to move the mouse (moving the image) or lock the image (incase I want to move again, and forget where my mouse is.)

Looks like an interesting minigame. Something you can pickup and drop in an instant, unlike most EVE activity's.
Samsara Toldya
Academy of Contradictory Behaviour
#52 - 2016-01-22 20:04:39 UTC
Sometimes it is hard to decide between 2 options... like "all green with minor black dots" or "mostly black with 20-40 green dots".

In my opinion the example pictures shown on mouse-over don't match the magnification / sharpness / resolution compared to the sample on the left side.
After I took some screenshots and let Twitter crop / compress them to ugly small jpgs it was much easier for me to see why I was wrong, yet ingame it's still hard to see a difference.

Maybe I'll have to play around with monitor gamma, brightness, contrast to get better results.

I second that a wrong answer shouldn't give ISK/LP as it makes giving the right answer less rewarding.
EVE - wrong decision and you'll wake up in your home station.

Fullscreen would be nice, too.
Sheeana Harb
Deep Core Mining Inc.
Caldari State
#53 - 2016-01-22 20:08:45 UTC
I hope this feature will enter live server as soon as possible.

If it does, I will start using it. Both because it's fun and also because it helps a greater cause.
If Project Discovery gets implemented, I won't be upset because something else could've been improved / fixed.

And also I don't think it break EVE's theme nor it would pull me out of the game.

And lastly, releasing Project Discovery, I don't believe CCP is selling out their vision / game design.

More features like this please!
Sir Constantin
#54 - 2016-01-22 20:34:41 UTC  |  Edited by: Sir Constantin
Let us rezise the window or make it fullscreen because everything is really small and it hurts my eyes.
Give the reward (isk/lp)if people got a accuracy of over 70-80%

Tutorial samples should be on par with what we see or explained better, a youtube video of someone going through the samples and explaining them would be awesome.
Uriel Paradisi Anteovnuecci
Itsukame-Zainou Hyperspatial Inquiries Ltd.
Arataka Research Consortium
#55 - 2016-01-22 20:51:17 UTC
It's bothering me a lot that people are just clicking one option ever and over again - if there's no downside to doing so, people will just start doing that and mess it up for everyone.

Apart from that: I assume accuracy scores will update on TQ, where they don't past the already-identified samples on SISI?

"A City made of Dreams is built in heaven"
Samsara Toldya
Academy of Contradictory Behaviour
#56 - 2016-01-22 20:52:18 UTC  |  Edited by: Samsara Toldya
I'm getting to a point where after I submitted my result I'll see lots of hexagons with 6%, 50% etc.
Best guess: that's a sample without a right/wrong answer where "what does the majority think it is?" data is collected?

Great fun, but eyes start bleeding. None of the drifters or goons seem to be my father, but maybe I'm misunderstanding what I am doing ^^

Magnification - mouse movement doesn't match the movement of the picture, hard to zoom in the edges as you lose zoom as soon as you are 1px outside the square

Magnification is not strong enough to zoom in CCP Guard Twisted

If I want to move my mouse from (example) "Mitochondria" to "Cytoskeleton" too fast, the mouseover-pic of "Mitochondria" won't dissappear as mouse cursor is above it and I have to move it back. Too much delay after leaving the mouseover zone.

EDIT: Yes, please make some hardcap for how many answers you can give. Like 3 within 60s or something like that to prevent "nucleus spinning" being the new ship spinning.
Tomiko Kawase
Perkone
#57 - 2016-01-22 21:11:09 UTC
This is pretty cool, but I wish the images were larger. It's hard to match the detail in some of these.

A couple of images in the presolved section I would never guess matched up (specifically the endoplasmic reticulum one) and I don't see how it matches at all.

Bombastic vintage soul with a driving groove, brass & male vocal
Starbuilder Stasarik
Doomheim
#58 - 2016-01-22 22:10:18 UTC  |  Edited by: Starbuilder Stasarik
A number of the Tutorial samples I've been provided with don't seem to make any sense whatsoever.

Case in point: http://puu.sh/mFMRG/e8f3a722ec.jpg

The green blobs are clearly in the nucleus, but not in the nucleoli, but it's marked as being in the vesicles, despite the description of vesicles stating that they never appear in the nucleus.

There have been a number of others, but this one has been the most glaring example I've seen of "wait, what?" that I've encountered so far.

Just a quick edit to say that I'm doing the tutorials a third time. I'll screenshot all of the ones I encounter that don't seem to line up with the descriptions and examples.

Another edit... I'm not sure what the other examples are being highlighted for, now. I just did a tutorial question wrong on purpose, and it clearly shows the red X marker, but on correct ones, other examples are being highlighted in Green after submitting.

I think that's throwing me off a bit.

Is there a reason that other options show up highlighted in green after, or is that a bug?
Zappity
New Eden Tank Testing Services
#59 - 2016-01-22 23:38:36 UTC  |  Edited by: Zappity
Gotta say as an RL biologist I'm pretty excited to see this integrated into the game. Very cool.

1. Make window larger. Maybe the new map window options for dock, full screen etc.
2. Needs intro video option to explain what it is.
3. Add another seven (at least) optional tutorial slides to let people get a better feel for it if they wish.

How will people find this particular agent if they don't know to go to the Business screen?

Edit: to all of the people commenting about how the images don't match the description it doesn't really matter. This is harnessing the power of a large sample size and the correct answer will emerge after enough people have had a go.

Zappity's Adventures for a taste of lowsec and nullsec.
Cixi
#60 - 2016-01-22 23:50:54 UTC
This is awesome I can't wait to try this Big smile