Thanks. What I'm trying to work out is whether lines not discussed by McDonald are likely to lead, not just to variations that he does not give, but to position types quite different from those that he explains. For example, if 6...Nc6 7.Be2 Bd7 8.0-0 g6 is just a slightly inferior classical Dragon for Black, I would need to acquire some basic grasp of how to play a classical Dragon. But if it's just bad (e.g. because of 9.Nd5) then I don't need to know how to play the classical Dragon, I just need to know how to deal with this particular line. The impression I get from TN's post is that at my current level (1750) I'm not likely to find myself completely at sea if I use the book as my guide and my opponent plays a sideline -- which is reassuring.

TN's last sentence is intriguing!

Well, 6.Be3 Nc6 7.Be2 e5 8.Nb3 is a better than usual version of the 6.Be2 system as Black is committed to an early ...Nc6 (when he would rather start with ...Be6 or ...Nbd7). 7...Bd7 doesn't seem a problem either - you could play Nb3 which is the general response when Black plays this way (as there's no ...Nxd4/...Bc6 any more to activate the bishop), or even try 8.g4!? to exploit the fact that the Black knight can't go to d7 after g5 (although  in this exact position 8...Nxd4 9.Bxd4 Bc6 is probably a good reply, which is all the more reason for starting with Nb3!).

The McDonald book is quite decent for club players, and has the advantage that he doesn't give you more than you really need to know up to say 1800 level. But beyond that point you're going to need to learn the lines and understand the positions in more depth (from either a database, coach or more recent book, or all of the above). I think you'll find it very helpful to choose a few model players for White's side and study their games in the systems that they are specialists in and that you are thinking of playing. They don't have to be rated over 2750 either - 2600+ guys who play a lot of open tournaments might choose variations more suited to your level of play and opposition than the really top guys.

If I had to suggest an 'easy' system for someone to learn against the Najdorf, I'd go with 6.h3 - at least you are fairly sure of your next move! The best Black can get against the move is demonstrated in Nakamura-Vachier Lagrave, Sinquefield Cup 2014 (still being played as I type this post) and White has some good alternatives before that (which you can find for yourself in the database). The only problem is that there isn't a good source that nicely explains and summarises the key ideas for both sides - well, that will change soon, but I should cut the spoiler here  .

I've decided it's time I started playing open Sicilians as White. And I thought I might use McDonald's "Starting Out: 1.e4!" as an outline repertoire, plugging the holes (of which I know there are many) as and when I find them.

I decided to start with the Najdorf. McDonald gives 6.Be3 e6 7.Be2, followed by f4 and 0-0. He also deals with 6...e5, 6...Ng4 and (by transposition) 6...g6. He doesn't appear to deal with 6...Nc6, which in my database is the 4th commonest move. OK, no problem, it's a hole, let's work out what he would have recommended if he had had space to cover it.

Not 7.f3, presumably, because after 7...e6 we are in an English attack against the Scheveningen, which he does not recommend.

7.Be2 would transpose into his lines after 7...e6. And 7...g6 doesn't scare us because 8.Nd5 seems to be strong. But if 7...Bd7 and 8...g6, we can't play the Yugoslav attack as he recommends. And if 7...e5 we are out of his line against 6...e5.

So I have two questions. First, is it possible to stay broadly within McDonald's general approach after 6...Nc6? And, more important, is this just one example of fundamental problems in his choice of lines? I know I will keep finding holes if I stick with this book, but will I keep finding holes that can't be plugged in a way that's consistent with the lines that the book does give?