wots the dif?

between:

1. using engines to play each other to create a database and then picking the best moves to play in online chess

2. using engines to pick best moves in online chess

??

seems like a wash to me 

I wouldn't do either. I'm more interested in learning how to play better.

die from what?

both have potential to teach you how to play better.

from eating shit, of course.

Doing the #1 without actually having related game in progress = legitimate opening research (a bit amateurish though)

Doing the #2 = despicable cheating and breaking the rules

......

The OP perhaps meant creating a limited tree of computer generated answers to expected opponent's top_n computer-preferred moves to great depths and then picking these from the DB.

Technically this may be OK within the rules, but it is indeed stupid idea, because if opponent is an idiot cheater or uses the same method and doesn't leave the tree, the play appears to be computer-contaminated from both sides. The ban is well-deserved then - for idiocy.

#1 is basically what GMs do. So when you consult a large pro database, you're bound to find a tree of computer-generated moves, up to a certain point.

No way you can cover entirely all your games that way though : that's only a kind of modern opening preparation.

That's why I have switched to Chess 960. I have somewhat lower rating because there is no opening database to consult but the game is completely mine from move one. I don't feel well about using an opening database. If I don't use it my opponent may, if I use I feel that someone else is playing the game for me.

So the way is Chess 960 and I would even more welcome the possibility of playing on bigger than normal board.

i was wondering why you switched.