Database moves could be deceptive?

Just had this random thought today.

The moves that show up in the chess.com database come from games that have actually been played on this site correct? But isn't that in itself a large problem for the accuracy of data?

A 2000+ rated player just today made the statement to me:

"I never analyze a move until its out of the DB. There is an art knowing how to get the best move out of the DB but the best moves are there. I win plenty of games so I will continue playing the way I know works and you can play around with the analyze board if you like. When you think your up to it I'll play you a game. I will use the DB for every move I make until out of DB and you use the analyze board for the opening moves. Any bets on who will win. Good Luck :)"

If the majority of players take that same attitude, then doesn't that mean that the most popular moves in the database will be self-perpetuated and not necessarily based on strength at all?

For example:

The database gives several options for the next move.

• 150 games, Nf6, 30%
• 120 games, e5, 30.1%
• 30 games, b6, 31%

If players do just take the quick, lazy way out and start choosing the most popular move without even bothering to analyze the board, then over time won't the gap between the first and second choice increase merely because the first choice was slightly more popular to start with, and not because it was necessarily stronger. And then once that gap starts increasing, the rate of change in the gap will also increase because more people will look at the larger popularity gap and say, "Oh, well look at how many more masters chose this move over the second option. It must be the best move then."

Sorry if I can't make it very clear, but does anybody follow where I am going with this?

Any database containing games of players below master will lead you astray in many ways. When I use the game explorer on this site, I use the database of master games. Even then, I rarely will play a move that was not among my candidates before checking the database. If the db suggests a move that I hadn't looked at, I will go through several games to look at it and get a sense of its purpose before playing it.

I prefer to use my own databases, or the New in Chess online database because it is very selective.

Using the databases has helped me take some moves that I played very frequently out of my repertoire because seeing that they did poorly made me look at them closely to see why. There is an art to getting the most out of databases. I consult the db on 365chess.com as well as ours.