There is real creativity here, from minds both unfamiliar and unfettered with traditional Western ideas. The games also resemble those of the top computers; no idea is too unorthodox to be examined and used as long as it works.

13.Be3 may have been a mistake. 13.Bxf6 and 14.Nd2, controlling the e4 square, was played in I Zugic vs V Babula, 2000, which ended in a short draw. Once Black got in 13...e4, I'd agree with <Oceanlake> that he was better.

I'm not sure about the exact reason for 17...Re5; the move accomplishes a lot of little things, but may have been no more than a probe. Retracting the rook to a safer square was not a big problem, though; White had hardly improved his position by advancing the queenside pawns.

I'm thinking of one possibility. Here's the position after <19...Ree8>;

click for larger view

One thing White would like to do is move the ♘d2 to a better square. However, if 20.Nf1 or 20.Nb3 immediately, there's a shot for Black: 20...Rxd1 21.Rxd1 Nxg4, with a discovered attack on the unprotected ♘c3. White could avoid this with 20.Ndb1, but that's not really solving his problem.

So the idea is to bring the king over to e1 protecting the ♖d1 so the ♖c1 can continue to protect the knight. When you need to resort to ideas like that, your position probably isn't very good.

And, come to think of it, maybe that was the idea behind the Re8-e5-e8 idea--luring those pawns forward. If Black had just played ...Rd8 immediately, we'd have a position something like this:

click for larger view