Re:Google admits to running experiments which remove some me...
by ericbarrett on Wednesday January 13, 2021 @10:50AM
(#25763491)
My guess is that “tens of thousands” is the rough SELECT count(*) from their systems. And I’d bet most of these experiments are old Boolean feature flags that were essentially acceptance tests, the final hurdle before global launch. For example, seeing that your new Gmail POP3 settings workflow drives engagement down 50% is a pretty clear indication you’ve got a bug that wasn’t caught in development. The interesting tests would be buried between such mundanities.
This kind of information would also be highly proprietary so I doubt such a list would see the light of day.
Re:Google admits to running experiments which remove some me...
by kevincox on Wednesday January 13, 2021 @09:31AM
(#25762364)
Sure, on any particular day it could be more or less helpful. But It if they weren't experimenting for years prior it would probably be even less helpful than a bad experiment of today.
(Of course this depends a lot on how your definition of "helpful" aligns with Google's)
Re:Google admits to running experiments which remove some me...
by lazylizard on Wednesday January 13, 2021 @09:13AM
(#25762143)
they're trying to find out/tune what is accurate/helpful right?
Re:Google admits to running experiments which remove some me...
by Arnt on Wednesday January 13, 2021 @09:59AM
(#25762740)
Normally, yes.
In this case I suspect (EDIT: no, I just guess) that they might be trying to find out how much their advertising take changes if they include/exclude/boost/depress/whatever the search results that link to Murdoch sites and other major Australian media sites.
Assuming it's not just simple A/B testing, I'd be interested in a list of these tens of thousands of annual experiments to see what Google has been up to. It's messed up to think that even just searching Google on any particular day could be more or less accurate/helpful because we're in the middle of some experiment we never explicitly consented to.