Originally posted by Willzzz
As I tried to explain in my earlier post following the lines from a database in no way guarantees a win. A lot of the "winning moves" on the database don't actually work out against a competent opponent.
Unlike a chess engine you WILL run out of suggested moves sooner or later, and if you don't understand why you have been playing those moves up to this p ...[text shortened]... time. The skill in chess isn't about a test of memory, it's about spotting combinations.
Let me add to this with a hypothetical example of how databases differ in effect from engines, and how statistics lie.
Patzer looks at database and sees that candidate move "A" wins 99% of the time at the GM level, with players playing that move and winning 99 times and losing only once.
What patzer does not know is that the move won 99 times until the antidote was found, which lost and effectively refuted the variation, which is no longer played.
Because the variation is completely refuted, no one plays it anymore, so the original candidate move still shows a 99% success rate, even though it has been refuted.
A database is a tool, but you really have to go beyond the stats and "up and down" lines to get a true feel for the "state of the art" for any given position. It's still a judgment call, and that's why we play out the moves.
In addition, a database does not give evaluations or assessments, unless you use the database to find an annotated game which has assessments within it. Databases are raw data for the most part.