2017-07-11 - Issue with Project Discovery Evaluation Set

official
project-discovery

(Kiauze) #64

They said they would post a new update once the fix was implemented, so we should really wait for that.


(elise densi) #65

and also i still have no clue how this works couse the tutorial was go “good”


(Tom Forager) #66

She’s dead, jim!


(Naskir) #68

I think i found the best evidence that something is wrong with the System:


Only the Second HALF [12,24] of the intervall is correct and the first HALF [0,11] is wrong…

So either its something with the error margine calculation OR the Sample itself dosent contain a continues frequenzy of the signal and would be therfore unable to ever be correctly solved. [Since the system only allows us to solve continues frequenzys] … but who am i to guess

PS: To say something postiv, if findthe idea behinde Project discovery amazing !


(Manny Ortega) #69

Been trying the thing out for a little while, cant seem to get the damn things to line up properly - get failed analysis cause I cant seem to see the patterns happening in there.

Maybe a suggestion also (to stop mouse scroll wheelitis and/or click n drag ‘itis’) would be to put a slider bar in there so you can decrease/increase the days without getting cramp all the time, not to mention a better way of going back to do the tutorial if you need to (something better than a tiny ‘?’ would be more appropriate)…

Another suggestion might be to make the minigame an 'out of EvE game so that ppl can use the whole screen for it instead of a portion of it like current (logon etc can be made just like logging on to here to save issues that way).

Going to wait till it is fixed properly and start again later …


(Tom Forager) #70

I got one right and now the difficulty seems to have corrected itself.


(Gempei) #71

I give a try after dt … 10 fails from 10, got from 58% to 44%. Same confusing samples and for change with some wtf moments :confused: … like this.


(Ann Barassova) #72

Another very challenging example :wink:


(Helfeln Meathead) #73

No reset after DT
15.4 % and dropping.


(Rena Tivianne) #74

Can you show any of the ones you are getting?

The difficulty for me hasn’t changed, I’m still getting ones that look like someone has just said “yep, definitely a planet there, can’t you see the pattern!” and there’s just no sense to what it’s saying, can’t learn from something that isn’t playing fair.

Additionally, where you have over 50% correct and others are millimetres out, you still get punished, some examples shown above in this thread.

Also no reset.


(CCP Sledgehammer) #75

Hi folks,

Firstly, thanks for your patience. I’ve just come from a meeting with MMOS about their proposed solutions. We’ve come to an agreement on what we will be changing, we’re just waiting on their changes for review purposes. I’ll keep you guys informed as we have more information.

Just to address concerns, the changes to Project Discovery will be happening at an API level and don’t require TQ downtime, we have not yet performed the Accuracy reset which will happen at the same time as the other changes to the API. We still estimate this to happen today.


(Rena Tivianne) #76

Awesome :slight_smile:


(Tom Forager) #77

Sure. These are recent ones. First was “No Transit”, second/third are around the difficulty I am getting now. I lucked-out and got one of the really ■■■■■■-up ones correct, then it kind of reset. I don’t know why.




(Rena Tivianne) #78

That first one is like 70% of the ones I am getting but it returns as failed because it tells me there are planets. The others, I wish I had some like that! I can clearly see the pattern.


(Tom Forager) #79

Try to get one of them perfectly right :(. I was stuck like that from 53% down to 2%, then I got one right and not sure of why it worked.


(Eternal Griever) #80

While I have encountered these confusing few graphs where it is very hard to analyze them right I’m not entirely sure how people have so much trouble as to get down below even 10%… I’ve been toying around with PD and while yes, there are hard graphs every now and then it really isn’t that bad. I can’t help but feel like it’s just people who can’t analyze these graphs quite well that are just screaming loudest at the moment. I’ve seen multiple instances when there is no transit at all and people mark it as one just because after detrending you have a drop. That and people marking peaks as transits.

I’m more interested if there will be any feedback like with previous iteration of PD where we would have scientists analyze some of the images with us. Especially when it comes to those tricky examples.

Maybe because for people below 50% they want them to focus and start seeing trends when they are instead of just blindly going with no transits.

Why the hell would it be a BPO if it is a ship that is supposed to be released in LIMITED numbers? Just because if you open description of the item and it throws one with unlimited runs doesn’t mean you are actually going to get that. Plus the icon in PD is obviously lighter blue than a PBO.


(CCP Sledgehammer) #82

The drop from PD is a BPC, there is an outstanding issue where they show as BPOs in the redeem queue.


Project Discovery pacifier blueprint
Infinite BP help?
(Svetlana Moscovich) #83

Eww. Man, I was all excited to waste hours getting up there and getting a BPO and starting flying them full-time. Not so excited now.


(Nazzarus) #84

This makes me sad, that’s an awful lot of work for a cruiser BPC. Am I going to have to run this on multiple characters to get multiple BPCs?

It would be great if this were re-evaluated.


(Predator Ace) #85

FIX already was deployed or not ? I still have accurancy 15 %