whatnext 

Can't prove anything regarding who had a clear advantage in potentially the most profitable day in trading history. Even if I did I wouldn't have the balls to post it - nor the big (but circumstantial) chunk of the puzzle I saw 10 days later. FWIW it's well known that not all venues of trade were subject to the same (unheard of ?) 30% max profit limitation. It took a while to come out but Zero Hedge prove the flash crash was a result of quote stuffing (in my book at least). What I incorrectly said caused it at the time was the high frequency trading algos going from buy to sell. 100's of thousands of HFT's are sent out every second. Almost all of them cancel out but the ones that don't break up big orders which pushes up the buy limit of the order. Not great if it's your order but great if you already owned the equity. If they go from buy to sell that would be reversed and the number of liquidity providing old fashion market makers has gone down a good amount.

Right, so how does that affect underlying stocks?

Don't know if or how much each long/short of an ES contract pushes up or down the underling 500 companies of the S&P.

Quote stuffing disables the filling of new orders so only sells are filled (sells make a stock go down).

Not a 100% but think a reverse of HFT algos, from buy to sell, would push down the limits of buy price range and / or up the limits of shorting.