4
$\begingroup$

Go to the Physics.SE main page and look at the community promotion ad. Do it now, I'll wait.......

Back yet? Okay, was it advertising Physics Overflow? Because every time I visit the main page, it's advertising PO. Last I checked, it wasn't rated high enough to be featured on our community ads. And according to meta they should be using the 2015 ads now. Furthermore, every time I refresh I believe (and others are in this belief too) that it should randomly select a new ad to show, but I always get PO being advertised.

This seems wrong. Conveniently and suspiciously wrong to boot. Thoughts?

$\endgroup$
14
  • $\begingroup$ Some browsers aggressively cache dynamic content. Have you tried forcibly reloading? $\endgroup$
    – dmckee --- ex-moderator kitten Mod
    Commented Jan 15, 2015 at 15:34
  • 2
    $\begingroup$ Note, it seems that they are using the 2014 ads currently for an unknown reason, but I'm leaving this up because PO still appears far more than random statistics should allow $\endgroup$
    – Jim
    Commented Jan 15, 2015 at 15:37
  • $\begingroup$ @dmckee tried that. All I get is PO, but Kyle Kanos claims he got something else once $\endgroup$
    – Jim
    Commented Jan 15, 2015 at 15:44
  • $\begingroup$ @dmckee Just for fun, I installed a whole new browser in a brand new virtual machine and get PO every single time. $\endgroup$
    – tpg2114
    Commented Jan 15, 2015 at 15:46
  • $\begingroup$ Waffles Crazy Peanut gets PO all the time! $\endgroup$
    – Waffle's Crazy Peanut
    Commented Jan 15, 2015 at 15:46
  • 1
    $\begingroup$ i got the unanswered question thing now $\endgroup$
    – Gowtham
    Commented Jan 15, 2015 at 15:50
  • 3
    $\begingroup$ me too. It's possible everyone get's the same thing $\endgroup$
    – Jim
    Commented Jan 15, 2015 at 15:51
  • $\begingroup$ I get a different one, but it doesn't seem to change no matter how often I reload the page. So there seems to be something wrong in any case $\endgroup$
    – Mad Scientist
    Commented Jan 15, 2015 at 15:55
  • 1
    $\begingroup$ So I just logged out and open several pages. Got unanswered questions, and SciRate. But only those over ~12 different pages. $\endgroup$
    – dmckee --- ex-moderator kitten Mod
    Commented Jan 15, 2015 at 15:56
  • $\begingroup$ @dmckee so much for random chance $\endgroup$
    – Jim
    Commented Jan 15, 2015 at 15:58
  • $\begingroup$ different browsers all give me the same ad. Time to consult the meta $\endgroup$
    – Jim
    Commented Jan 15, 2015 at 16:00
  • $\begingroup$ Meta doesn't say much about how ads are shown or how often they refresh. This could be a bug or it could be status-by-design $\endgroup$
    – Jim
    Commented Jan 15, 2015 at 16:28
  • 1
    $\begingroup$ @Jim: It would be interesting to know the actual method/procedure/formula for allocating ad time. E.g., does it depend on just the # of upvotes, or the # of upvotes minus the # of downvotes? $\endgroup$
    – Qmechanic Mod
    Commented Jan 15, 2015 at 17:19
  • 2
    $\begingroup$ We're looking into this now. $\endgroup$
    – Grace Note StaffMod
    Commented Jan 15, 2015 at 17:54

2 Answers 2

Reset to default
7
$\begingroup$

Alas, I was a little optimistic in saying that the new ad threads would be live by now.

I switched Physics over to the 2015 post, so barring any caching issues, you should be seeing new ads shortly.

$\endgroup$
5
$\begingroup$

You're not the first to notice this trend; actually, we had some discussion about it in the moderator chat room a few months ago. So I collected some data. Sampling which community ad is displayed on the home page at 10-minute intervals for a couple weeks shows that the distribution among different ads is roughly Gaussian, with no significant outliers.

histogram of ad frequency

(this omits the ad for the site's newsletter) The Physics Overflow ad was displayed more than average, but was certainly far from being the most common.

However, looking at the actual samples shows that the same community ad is displayed for a run of several hours. The exact length of the run changes according to some pattern I couldn't figure out (perhaps randomly), but the ad is not randomly chosen on each page view. I presume this is done to allow better caching. Anyway, I think this contributes to the impression that the PO ad was being displayed more than it should be. If you tend to notice one ad more than the others, then when you see it the first time, you're more likely to register it when you visit the page in the future. While that ad is the current one, then it is being displayed more often than it would be if the ads were randomly chosen at each page load, and your brain registers something fishy about that. On the other hand, during a run of any other ad, you don't notice it at all and your brain doesn't register the fact that you're going a long time without seeing the ad you focus on.

$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .