"Unfair" calibration movies??!

Discuss your experiences with and ideas about Stardust@home here.

Moderators: Stardust@home Team, DustMods

Chuck Crisler
Posts: 30
Joined: Wed Aug 02, 2006 9:44 am
Location: Windham, NH

scores really don't matter

Post by Chuck Crisler »

Like many other dusters, I have been annoyed by the somewhat ridiculous CMs. The explanation that they are useful to determine detection thresholds makes sense. However, regardless of what 'they' say, the CMs are still training movies, whether intentionally or not.

I have been thinking about this for some time (NOT when dusting - that causes mistakes!). Scores don't matter. At the end of the project the *ONLY* thing that matters is if you saw and/or were the first to report a track. If you only looked at a couple of movies and had lousy stats but were lucky enough to get a track, then you are golden. (Obviously, more movies == greater chances.)

Yes I grumble when I miss a CM, click back and *STILL* can't find it. But that really doesn't matter. Yes, the CMs do keep me on my toes. I can only do about 200 movies a sitting because of concentration/boredom and lack of time. But I do try to make my meager contribution worthwhile, though my stats aren't anything to write home about :-(.
wickedKlown
Posts: 4
Joined: Tue Jun 16, 2009 5:03 am
Location: South Australia

Post by wickedKlown »

Don't know what everyone's problem is. I think the examples were ok and gave enough variation to show that you need to look at everything very carefully and not pre-visualize. Even a very faint streak in the gel could possibly be a track. It should make you think and concentrate.

Given the massive budgets required for space related projects, I think that the volunteers that work on the project need to be vetted to ensure that they have a genuine interest in aiding the project, and are able to achieve the aims of the project. Pretty normal really, even in the workplace.

It may seem unfair but it isn't. How many workplaces will let you come back for a job you failed to get. Here you can review your results, see where you missed, and then retry. The test ain't that hard, I failed the first 7/10 and passed the second 9/10.

Just my views.
jsmaje
Posts: 616
Joined: Tue Aug 15, 2006 8:39 am
Location: Manchester UK

Post by jsmaje »

wickedKlown wrote:Don't know what everyone's problem is. I think the examples were ok and gave enough variation to show that you need to look at everything very carefully and not pre-visualize. Even a very faint streak in the gel could possibly be a track. It should make you think and concentrate.... The test ain't that hard, I failed the first 7/10 and passed the second 9/10.
Hi wickedKlown,

From what you say I presume your comments are about the recruitment Training Session rather than the 1-in-4 'Calibration Movies (CMs)' randomly presented when regularly dusting :?: .

Having as a result just reviewed the now-3-yr-old training session movies I was struck by their poor resolution, limited cover of the varied artefacts to expect, and inclusion of just a few ?cometary/interplanetary/experimental lab tracks, thus bearing only a distant resemblance to the current phase 2 movies and CMs.

As it happens, the phase 2 CMs are in fact based on just a single computer-manipulatedphase 1 unproven interstellar candidate. That this may sometimes have been rather inappropriately done, e.g. regarding vertical level, etc., has been one of the issues raised in this particular and other forum topics.

Meanwhile, given changing expectations it seems that all of us, team, newbies & oldies, may be back to 'square one' anyway, and all eyes should be on the look out for rare unexpected features regardless of any such so-called CMs (which of course are only there to act as a standard measure of dusters' sensitivity/specificity rather than as examples necessarily to look for).

John
Chuck Crisler
Posts: 30
Joined: Wed Aug 02, 2006 9:44 am
Location: Windham, NH

Where is it???

Post by Chuck Crisler »

Where is the 'track' in movie ID 9250682V1? If it is the spec at 268, 328 - well that looks more like dust on the lens than a track. Everything on the image shifts with the up/down adjustment EXCEPT that spot. Nothing else is in focus very long. If we are supposed to be looking for things that *COULD* be that subtle, then at least 4 out of every 5 images should be tagged 'Bad Focus' because you simply *CAN'T* see that kind of resolution.

Chuck
DTF
Posts: 33
Joined: Fri Aug 01, 2008 9:47 pm
Location: USA

Re: Where is it???

Post by DTF »

Chuck Crisler wrote:Where is the 'track' in movie ID 9250682V1? If it is the spec at 268, 328 - well that looks more like dust on the lens than a track.
Chuck, 9250682V1 is not a calibration movie. Nothing you click on this movie should have changed your score.

DTF
Gamma^Ray
Posts: 46
Joined: Wed May 17, 2006 11:13 pm
Location: USA
Contact:

Re: "Unfair" calibration movies??!

Post by Gamma^Ray »

Gotta say, although I understand the concept of making things abit harder than Stardust 1 (Was in the top 100 at the end of it), I find some of the movies
now to be just basic tricks which make me at least, hesitate to mark a "Possible Target" as after getting marked with an incorrect for what I say are
absolute possible tracks, I start thinking many of them are the cm's when they are or might not be. I guess my biggest issue is that it makes me feel
as if I'm being punished for what I feel is a actual possiblel track...esepcially after going over the test movies again and so on. I still enjoy it and all, but each
time I get caught by these tricky cm's and my score goes down again..I get to where I get frustrated and stop for awhile (As it loses its fun for then).

Just my opinion on it but still enjoy it. :)
GR
A.Einstein wrote:"The only reason for time is so that everything doesn't happen at once."
Post Reply