absolutely absolutely. id say dont not know what you are doing though. bear it mind if anything goes wrong you will be spinning out millions of orders at the wrong price so dont make that happen (knight=bankcrupt). and then also its very very risky and how would you close something that you are responsible for because everyone is dispersed, markets are dysfunctional.
It really depends on the strategy. I never scale in since my trades don't last long enough. I have small stops and small targets. If i am wrong i cut my losses and move on to the next trade. The larger the target/timetable of the trade the better the opportunity to scale in. On the other hand, a lot of losing traders add on to a losing position in the hope it turns around. Which of course is one of the the worst things to do.
I scale in a strategy I use cause sometimes the extension is not big, but I always know where is the end of my scale (= the best entry I can have in a trade)
well i mean can you sort a tensor with the number of calculations order(nlogn). no. so dont worsen your average entry level without the algorithmic power to compensate.