Page 1 of 1

What percentage of real movies viewed are you clicking on?

Posted: Wed Aug 16, 2006 1:58 am
by pointyhat
Count the number of entries in you event list, divide by the the number of REAL movies viewed and multiply by 100. If the project estimates are right, they would expect 45 real tracks in the 700,000 movies they will create, which is the same as saying 0.006% of movies are expected to contain a track.

All the best

Posted: Wed Aug 16, 2006 2:32 am
by Mighty Pete
That will not work.. You would need to know how many movies you clicked first and right now that information is not available...

If 50 people voted for the same movie it's still only one movie..

Posted: Wed Aug 16, 2006 3:40 am
by pointyhat
Pete, this poll is aimed at getting a handle on what percentage of the movies that people view they click on, it makes no difference whether they clicked on them first or not. When I use the phrase REAL movie I mean that you should use the number of Total Real movies viewed from the VM page rather than the total number of movies viewed given on the events page which counts calibration movies as well.

If the Project's estimates are correct, 0.006% of movies contain a real track. I suspect that most people, myself included are clicking on a much higher percentage of the movies they view, (I am currently clicking on about 1.5%). I thought it would be interesting to see what kinds of percentages other people are clicking on.

All the best

Posted: Wed Aug 16, 2006 5:40 am
by DustBuster
Hi PH.

You might consider adding some headroom for events clicked that are non-tracks (i.e. "this looks interesting"). I've clicked on a few I know are not tracks I felt should be looked at.
Without reviewing every event on mylist I wouldn't know the exact number- and without knowing how many real non-track events there are, it would be difficult to come up with an accruate modifier.
I haven't seen any predictions on the expected number of non-track events; but if it's out there, fjgiie probably knows where.

Good luck.

Posted: Wed Aug 16, 2006 6:18 am
by pointyhat
Hi DB,
That’s a very interesting suggestion, but perhaps would be better answered by a separate poll:

"Of the real movies you have clicked on what percentage contain features that you don't think are tracks that look like the training movies, but did contain unexpected features that you wished to bring to the teams' attention"

If someone sets this poll up I'd certainly vote. If I had voted on this in my first three days of participation I would have said 10%, now, two weeks in, I'm thinking 95%.

All the best

Posted: Wed Aug 16, 2006 9:26 am
by photon
Good question. I'm running at about 4%, though only 1% are ones that I would say are a track that look like the tests. The others are similar in that they are below the surface, but are small enough that they show as a dot rather than a circle; so either they're very small tracks or pre-existing flaws in the aerogel.

Non-track Numbers

Posted: Wed Aug 16, 2006 10:56 am
by fjgiie
Hi DustBuster,
DustBuster wrote:I haven't seen any predictions on the expected number of non-track events
No, me neither.
Roughly, just throwing out a number here, 8500 clicked events are not tracks in the top 100 Stardusters.

I have clicked on 85 focus movies. One was a Davy Crockett bowie knife, one I believe is a real track, 83 are bad clicks made by me. My score is the average score. 85 times 100 equals 8,500.
This 8,500 seems low to me but we have only been searching two weeks.
16 times as much by other searchers = 136,000
9 times 'till end of year = over a million (the real number probably larger )

45 real tracks clicked by 2008.

So by the end of this year, a rough estimate is 1,224,000 non tracks clicked by us. Aren't we proud? :oops:

Edit: maybe divide that by 100 to get #of focus movies