Spotlight randomness

[Joe Ross]

Since Brandon is presumably busy at IRI, I figured it would be a good time to give hime something to look into

I seem to keep seeing the same spotlights over and over. Now, either it's me, or they aren't exactly random. Did something change in one of the upgrades?

For those who are interested, here are 2 previous threads on the issue: http://www.chiefdelphi.com/forums/sh...ad.php?t=16319 and http://www.chiefdelphi.com/forums/sh...ad.php?t=20522

[Astronouth7303]

I've wondered about that. it seemse to be weighted towards the most recent quotes.

ps- anyone who decides to run a script again, just load http://www.chiefdelphi.com/forums/xml/spotlight_js.php as that is what is used on external websites. Like this:

<script type="text/javascript" 
src="http://www.chiefdelphi.com/forums/xml/spotlight_js.php"></script>
<p> Press 'Refresh' to see another quote!

[edit]that link is from http://www.chiefdelphi.com/forums/extras.php?[/edit]

[shyra1353]

The link above that Astronouth7303 posted seems to show randomer spotlights than the ones at the top of CD ... like there are some there that i have not (or do not remembering seeing at the top of the CD pages ...)

[mtrawls]

Originally Posted by Joe Ross

I seem to keep seeing the same spotlights over and over. Now, either it's me, or they aren't exactly random.

Well geesh ... of course it's not random ... it's pseudorandom.

And really, it's not like there's a lot you could say about pseudorandom numbers or anything. Now, I'm no statistical wizard, and it's been some time since I've been through the specific TAOCP tome, but how about a chi square test, or some other statistical tool to determine just how "random" the psuedorandom number generator is being.

[Brandon Martus]

Thanks, Joe. I <3 extra work.

Just kidding, of course.

I don't have time (or energy) to look at it now in much detail, but I did see that there were inconsistencies in the SQL grabbing 20 random rows. I made them consistent .. changing the ones that show on the top of the page to be more like the ones in the random javascript. Let me know if you notice any change (better, worse, none) and I'll look into it more this week.

[Joe Ross]

It does seem to be better. I haven't noticed anything repeated, and I've seen a lot recently that didn't seem to come up before.