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


(zluq zabaa) #166

So, how long will it take you to get to level 250?
aka: how much grind for the Glasses, Pacifier etc.

Getting to lvl 250 will net you all the special rewards once, including the glasses in the end.

To simplify I assume 3 types of players:
Type A: Wants to progress as quick as possible.
Type B: Wants to progress, but also wants to deliver good results.
Type C: Wants to be perfect in accuracy (haha try that, with the current control questions).

My assumed stats for the types are:

Type A: 50 EP average, needs 10s for one cycle on average.
Type B: 75 EP average, needs 20s for one cycle on average.
Type C: 100 EP average, needs 30s for one cycle on average.

Reasoning: Type A will click as little as possible. Type B will be frustrated with unsolveable control questions and only click when totally obvious. Type C will put a lot of effort into getting even the most weird control questions right. Currently I’d say Type C is a hypothetical Type, because 30% of the control questions make absolutetly no sense.

Now for the juice, how long do they need to get the Glasses:

Type A:
7 full Days and 10 Hours to get to Level 250. That is 178 hours and a bit. Or 22,25 full days of 8 hour job.

Type B:
9 full Days and 22 Hours. That is 238 hours or 29,75 full days of 8 hour job.

Type C:
11 full Days and 4 Hours. That is rougly 268 hours or 33,5 full days of 8 hour job.

Lol CCP…

Okay, what about other rewards you ask?

Reward :: Time it takes Type A / B / C

Suits (LVL 200): 115 hours / 153 hours / 173 hours
Enforcer BPC (LVL 150): 65 hours / 87 hours / 98 hours
Pacifier BPC (LVL 50): 8 hours / 10.5 hours / 12 hours

We all know #project-discovery is not to make us spacerich or anything. It makes sense. But this progression structure is just a joke.

TL;DR everyone who was afraid their Enforcer would be worthless can relax. Even if people decide not to care at all about the goals of Project Discovery (Type A) they’d still need to grind 65 hours with 5 clicks in 10 seconds or 117k mouseclicks to get it. (Which is still better than the required ~1 Million mouseclicks for a Type C player getting to LVL 250)

(Julie Morano) #167

It would be useful to be able to bookmark transits. Then try and spot the pattern.
If you mark one, then mark a cousin, and scroll to see if it looks good, and it doesn’t, you right click and lose both marks, meaning if you didnt scroll back to a transit before doing so, you can lose them.

(Imiarr Timshae) #168

You might want to fix this stuff also:

Also, I wholeheartedly agree with the guy above who ran the math on it taking 7 days minimum to get the final reward as being too long.

Whoever mapped that out is absolutely insane.

(Commander Aeris) #169

Ok true, slightly more than 0.5 days in that example, but still crazy difficult if you see the folded cloud of dots.

(Commander Aeris) #170

Looks exactly like one of the samples I had. And agreed on the need to fix these things.

(Commander Aeris) #172

It doesn’t say two planets. It says one out of two dips is a planet, the other 50% of dips is nothing, you are only imagining there being a dip there apparently…

(zluq zabaa) #173

You have two different orbits in one image. For some reason the test samples want you to only choose one - the other will count as a mistake.

(Eternal Griever) #174

I actually got one person tell me something that makes sense in regards to checking the graph, but does not make sense in another aspect.

Those examples are where we see transits of a planet AND transits of another star that is orbiting around another star (AKA eclipsing binary). In other words it makes sense to mark those other dips as not exoplanet transits, but makes no sense to punish someone for it, just because in the tutorial itself it says that a planet transit and another star transit is almost impossible to distinguish.

(zluq zabaa) #175

In this example, how are you going to differentiate between it being a planet or star? In the tutorial it vaguely says the drop in flux caused by a star to be “much greater” but in the posted image it isn’t really the case.

(Eternal Griever) #176

You just can’t, which I did point out.

(zluq zabaa) #177

Yeah that’s what I’m saying.

Anyhow I sincerely hope everyone will read my post above before attempting to grind for the higher levels.

It’s probably not worth ■■■■■■■ up your physical and mental health to grind more than a month of work days non-stop for a pair of glasses. CCP, if this was chosen on purpose, shame on whoever decided that, enticing real people to risk their real lives. :rage:

(Nazzarus) #178

keep in mind this will run for over a year. perspective.

(Kiauze) #179


Lack of sleep does that to you, was reading the red as yellow! :sweat_smile:

No more PD for me today!

(Galletem en Marland) #180

This is the second time I’ve gotten the bug where the graph displays a flat line! This seems to be a common issue.

(zluq zabaa) #181

Exactly. Okay, 365 days, on each day 1 hour, each hour just shy of 1000 mouse clicks. Do it and tell me again about perspective. Give it 10 years and just do additional 6 minutes of 100 mouse clicks a day, each day.

This should be a community effort and hence the rewards should not entice single players to go above and beyond any reasonable amount of effort. If people decide to do it because they like PD so much, they will anyway do it. Give them accolades at Fanfest, cool. But cap the rewards at some amount of necessarily invested hours per person.

Otherwise you only motivate a) heavy grinders and b) people contributing in bad ways to farm the low level rewards

(zluq zabaa) #182

A simple way out of both problems (abusing mechanics and grind to death) would of course be the following:

Bind the earned Experience Points to the Accuracy Rating in a meaningful way.

Right now there is no motivation to be more accurate then 50-60% while spam-clicking through samples and delivering “No Transit”.

(Nazzarus) #183

sitting around 95% accuracy, it’s definitely faster and worth it to be more accurate.

(Kiauze) #184

Is it? What’s your EXP/minute? Ballpark figure.

If you take around 15s per image and you’re glad with staying between 50 and 60% accuracy, EXP/minute is around 220.

(Buoytender Bob) #185

Three college degrees and a highly successful and long medical field career; yet, PD makes me feel like I just fell off the turnip wagon. Why does this current version of PD and the samples/results given remind me of participating in a scientific version of a carnivel sideshow game;it looks simple to “win”, yet the odds are stacked against you? Perhaps I am one of the lucky ones who were seeded with mainly difficult ones like have been presented in this thread. Well, I’m at around 63% at level 6; I’ll give myself until around level 12 or so to see if either the game or myself show significant improvement.

(Commander Aeris) #186

I do not believe this is a bug, simply missing data. IE telescope turned off for maintenance.