Two roads diverged in a wood...

I want your opinions regarding chess improvement.  When I started at chess.com I played a few games at a time, taking time to really analyze positions and make the best move.  Slowly over time I played more games concurrently, til I reached the point that it became worklike to go over all the games and make decisions.

Recently I read that Fischer, starting about the age of 12, played 12,000 games per year.  That's nearly 35 per day.  Obviously these games couldn't have lasted long.  This got me to thinking.

Thomas Edison was a great inventor that used the trial and error method.  Simply put, he would try something and learn from the failure, getting closer to truth as he went.  Nicola Tesla was also a great inventor.  He typically had flashes of inspiration, followed by imaginings in his head.  He would work out most problems in his mind before trying to construct the object.

What do you think is better?  Playing a lot of games, trying something over and over until you refine what works?  Or playing less games, studying the alternatives, taking the time to visualize several steps down the line, and reviewing what worked?

I agree you can do some of both.  But for example, should I play 6-8 games, of which I spend a lot of time on a couple.  Or should I play 30-40 games concurrently?  See the difference in volume?  It really becomes a time efficiency issue.

I play 12-30 turn-based games ALWAYS, and also play one or two (8,500) bullet games now and again and plenty of blitz. I've played close to 1000 serious games (OTB and correspondence/turn-based) games in the past fifteen years, but I've played more than 60,000 total games.