The following game, played in an e-mail tournament, shows the difficulty one faces even when trying to evaluate a position with an engine. The database games showed a big advantage for White in the games played from the position after my 13.Qe2, but a closer examination of the games revealed Black losing when he weakened his position by advancing his Q-side P’s or when he played …e5 which opened up the position and White ended up with a strong K-side attack.
But what if Black doesn’t do either as my opponent did in this game? True, Black can’t undertake any kind of attack if he doesn’t do something, but at the same time his position is very solid and it’s difficult for White to initiate any kind of attack either.
The critical position seems to be at White’s 16th move where he must defend against mate with either 16.f3 or 16.f4. The first move was the most popular, being played 8 times in my database, while 16.f4 was played only once. Honestly, this is what influenced my choice, but as it turned out Black refused to loosen up his position and I couldn’t find any way to get my heavy pieces into the attack because they were blocked by the P on f3. So, it would seem that if White wants to play for a win he needs to play 16.f4 which leads to very dynamic positions that look to offer good chances for both sides. In any case if I played 16.f4 it didn't look like I was going to be able to crush Black with the ease the database statistics suggested.
At move 19 I could have declined the draw with 19.Bf4 but even though the engines say White is slightly better I was concerned that in the long term my hanging P’s would turn out to be a weakness. The most common plan would be to push the d-pawn in order to open things up for the bishops but in this game I felt I needed the addition of a N to help support that advance. But both N’s had been traded so it seemed to me that the hanging P’s might turn out to be a weakness. It just didn’t seem, to my rating challenged eye, that White’s position offered any real winning chances so I decided to be satisfied with a draw against my higher rated opponent. But of course, because I am rating challenged, I could be completely wrong.