Explained: Why Some Marjory Stoneman Douglas Shooting Google Search Results Predate the Event

Mick West

Administrator
Staff member
The shooting at the Marjory Stoneman Douglas High School happened today, February 14th 2018. But if we search for stories about it from before today we actually find some:

https://www.google.com/search?q=florida+high+school+shooter+former+student&num=50&source=lnt&tbs=cdr:1,cd_min:,cd_max:2/13/2018&tbm=
Metabunk 2018-02-14 21-50-53.jpg

This led some excessively suspicious types to claim that the event was pre-planned with media stories pre-written. Of course this is nonsensical, but why are these stories pre-dating the incident?

It's not a new thing. It happened with the church shooting in Charleston. In fact it happens with every single news story that gets broad coverage.

The reason is simply to do with the way Google tries to extract the date from the web page. Unless the date is explicity coded in then Google will search the page for what looks most like the date. In the above case, the page at ohiol.com has the following at the top, with the correct date:
http://archive.is/Mf2uS

Metabunk 2018-02-14 21-56-23.jpg

But lower down it has a rotating carousel of earlier featured stories, and these are dated earlier.
Metabunk 2018-02-14 21-57-21.jpg

So clearly Google is getting the date from the "Marketplace" section of the page, and not from the actual article text.

The reasons can be found by looking at the underlying HTML of the page. The article date is simply there as plain text:

Code:
<span class="pub-post-date">Published: February 14, 2018 - 3:10 PM</span> | <span class="update-post-date">Updated: February 14, 2018 - 10:30 PM</span> </div>

If there were no other dates on the page, then Google might have picked that as the article date, but it's not really formatted how Google likes it. But further down the page we have:
Code:
<ul>
<li><a href="https://www.ohio.com/akron/homes/featured/complete-these-home-improvements-over-the-weekend">
<p class="empty-para">
<img src="https://www.ohio.com/wp-content/uploads/2018/02/5a81f3dd771fd.jpg" alt="" />
</p>
<span class="rpwwt-post-title">Complete These Home Improvements Over The Weekend</span>
</a>
<div class="rpwwt-post-date">February 12th, 2018 3:06 PM</div>

Here's another date, in a cleaner format. So Google picks that one.

Essentially this is just down to poor programming of the web site. A proper web site will have specific header "meta" tags that tell Google that the page is an article, and when it was published. For example this LA Time story:
http://www.latimes.com/nation/la-na-florida-school-shooting-20180214-story.html
Has a tag:
Code:
<meta name="date" content="2018-02-15T04:45:00Z"/>

The Ohio.com pages lacks a meta date tag (or a more modern meta article:published_time tag), so Google has to try to find the date itself, and it picked on the the page that was the date from something else.

You can see this effect on any story, here's another even from today, the resignation of South African President Zuma. Again Ohio.com has a bad date for the same reasons as before.
Metabunk 2018-02-14 22-17-42.jpg
 
Last edited:
Back
Top