SF 11 prefer 1. g3 at depth 48 in chess24

[Nay Lin Tun]

Afaik , SF prefer
1. e4 in low depth <30
2. d4 in higher depth >30
3. Nf3 depth 45+
4. c4 in depth 50+

[carldaman]

Food for thought there - maybe Lyudmil was right after all.
Or maybe, further above depth 50, it will revert to e4.

[Ovyron]

This just shows the little progress engines have made in the opening phase and why opening books' value has actually gone up over the years (so much that a good book can be much more effective than doubling your hardware's speed )

[Cornfed]

I've opened almost exclusively 1.g3 for some 20 yrs. It is about time the engines caught up.

[Ovyron]

It's not that 1.g3 is bad, it's that if you're going to play g3, you can maximize your winning chances by playing it later, and minimize the positions you have to check to make it work (it's simple to prove that 1.d4 x 2.g3 is better than 1.g3 x 2.d4. Because 1.g3 e5 2.d4 is significantly worse than 1.d4 e5 - here you capture e5 instead of bringing g3.)

There's variations where white's light Bishop is best on Bb5 or Bc4 (or even Be2) instead of on g2. It's better to see what variations develop before committing to g3+Bg2 (all variations with g3 without Bg2 are terrible.)

Realistically I'd put 1.g3 below 1.Nc3, if I'm right I'm an example of a puny human still being better than engines in the opening