How important is a good opening? Depends. Mednis said there is a German expression to the effect that a good opening means the game is half won. That's really true in high level play. Most of my games these days are "rapid" games on LSS where many opponents have a lot of games going and play quickly (sometimes 2-3 moves a day) and I have gotten away with some openings that wouldn't stand a chance in serious play...moves like 2.a3 and 2.Bc4 against the Sicilian and the Urusov Gambit. But in OTB play, how important are openings for those of us who are rating-challenged? Obviously, coming out of the opening with a good position is better than coming out of it with a bad one, plus it does give us a psychological boost to know we stand better.
Capablanca considered the main principle to be rapid and efficient development. That sounds simple enough, but he also added that the pieces have to be put in the right places. That last little statement, put in the right places, probably means that most of us are going to play the opening like we play the rest of the game.
The great Hungarian GM Lajos Portisch probably gave the best advice for non-masters when he said the only task in the opening was to reach a playable middlegame. That's something we can hope to accomplish.
Mednis was of the opinion that non-masters don't have to play the latest theoretical lines and advised playing whatever opening one likes and understands. He also gave good advice when he said it's foolhardy to voluntarily choose a line where, if your opponent plays correctly, you hand over the advantage. That would eliminate a lot of openings and gambits known to be unsound.
He also said that the ultimate key to successful play is understanding chess. That knocks a lot of us out when it comes to reaching a high level! It takes a fair amount of specific knowledge to play any opening and, as we all know, it's of greater practical value to understand how to play good opening moves than just memorizing a lot of variations.
There are three areas of significance in opening play: 1) King safety 2) piece development and 3) control of the center. The first two areas are self-evident, but the value and importance of the center is not sufficiently appreciated. The importance of center control has been known to average players since the days of Steinitz and the Hypermoderns deepened our understanding by demonstrating that what mattered was control, not necessarily occupation.
What are the best moves according to Mednis? White has five: 1.e4, 1.d4, 1.c4, 1.Nf3 and 1.g3.
What if you don't want to play any of them? Mednis recommends playing any of the three "mediocre" moves. These are moves that have positive features, but also have inherent deficiencies. These are: 1) 1.b4 with the plan of 2.Bb2 aiming at d4 and e5. 2) Nc3 which influences the center, but precludes utilizing the c-Pawn. 3) 1.f4 which does nothing for development and slightly weakens the K-side, but it does control e4. Don't play anything else!
As black, good moves against 1.e4 are: 1...e5, 1...c5, 1...c6, 1...d6, 1...Nf6, 1...g6 and 1...e6. Against 1.d4 you can play: 1...d5, 1...c5, 1...c6, 1...d6, 1...d5, 1...Nf6 and 1...g6.
If you want to take your opponent out of the book, and probably yourself along with him, you can play mediocre moves: Against 1.e4 you can play: 1...Nc6 and 1...d5. Against 1.d4 you can play 1...Nc6 and 1...f5.
Mednis believed that if you play according to opening principles you should be fine, but as play develops the position gets more complicated and move selection requires concrete thinking. But even then most opening moves that are unmindful of opening principles are probably inferior. Remember, he was talking about average players, not GMs. I got a good chuckle out of one post where some guy claimed 1.e4 e5 2.Qh5 was playable because Nakmura has successfully played in on occasion. First he played it in blitz games for fun and second the guy is rated well over 2700, so what would you expect?
Chapters 5 and 6 cover the Sicilian but space precludes an in depth look here. Chapter 9 was quite interesting: Bad Moves: How Not to Play Them. Because the Sicilian Defense is so popular, one sample game, Karpov vs. Korchnoi from their 1974 match, involving the Yugoslav Attack was instructive. Karpov and Korchnoi played two matches, 1974 and 1978.
It was well known that Karpov had excellent results against the Dragon, so why did Korchnoi play it? Possibly because the previous year Korchnoi had defeated Karpov in a Dragon and believed he might not be up on latest theory. That wasn't the case though and after the game Korchnoi didn't want to talk about it, saying it wasn't a real game, but a result of home preparation. Sour grapes! Korchnoi had as much opportunity to do his homework as Karpov.
This game is game two of their 1974 match. I am giving it with a slightly abridged version of Mednis' explanatory notes plus a few observations by engines. Mednis was mainly concerned with the opening while Stockfish and Komodo answered a few questions on alternate lines. Also, Yasser Seirawan's analysis on Youtube HERE is really good and I recommend you spend about 45 minutes watching it.