I'm still digging through the article, but do they explain how they account for a simple "more symptoms = problem" model. How much noise is in the diagnosis information that could influence the results. What am I trying to say? "If we're sure there is something wrong with you we can find out with pretty good accuracy. If we're not sure then performance goes way down" sounds right.

I have suspicions that their "Filtering process" is "Burn it and let someone else breath it"

If a market signal is accurate, there is money in making sure it isn't. The same applies to stock prices.

As far as I can tell you can no longer run Fortnite on MacOS, so I don't know what point they are making.

Be sure to change your password before playing the Patriots. I'm waiting for the DDOS attack on their Watson powered play maker.