He's probably responding to <The rejoinder 16.Bd2 is winning for White.>

<JohnBoy: <CHC> - I believe <mike1> is merely objecting to <perf>’s “16.Bd2 is winning”. Better... certainly. Winning... not quite.>

Having read the comments, and seen the computer annotations that keypusher mentioned, it looks like I was premature in writing off Portisch's chances.

SF, in the annotations, gives 11.Bxh7+ as ⩲ +1.12.

Its next comment is 15...Ne4?, it prefers 15...Ba5 16.Bd2 etc which it says is = +0.45.

click for larger view

threatening 18.Bf6 ..

STOCKFISH PLAYERS LAZEY

The tournament bulletin (published in daily installments after each round) mentions that Matanovic (who commented on the games for the audience) expected Portisch not to allow Bxh7+ .. by playing 10... Bxd5 when after 11.exf6 Qxf6

click for larger view

You just never know during an OTB game

I'm definitely with <perf> here

The score between these two heavily favoured Portisch.

This one...

J H Donner vs Portisch, 1968

....was used by Shamkovich in his Batsford edition of 'The Modern Chess Sacrifice' (page 111) highlighting a positional pawn sac to keep a King uncastled (move 17.Bd3).

I think that "No" is the accurate assessment, at least from a human perspective. An evaluation of [+0.40] is considered an advantage by an engine, but humans don't have centipawn resolution; probably more in the neighborhood of [0.50]. Chess GUIs (engines don't make a qualitative assessments) typically assess evaluations in the [0.00] to [+0.49] as equal, evaluations from [+0.50] to [+0.99] as White standing slightly better, evaluations from [+1.00] to [+1.49] or [1.99] as White having a definite advantage, and evaluations greater than [+1.50] or [+2.00] as White having a winning advantage.

And you would probably be wrong, same as the old days. :-)

Seriously, it depends on whether the engine's evaluation and line are presented like coming from the gods, without any additional verification, as unfortunately it's often done. I am occasionally guilty of the same thing although it's usually as a result of time constraints or some such (yeah, I know, if I haven't checked the engine's line I shouldn't be posting at all). In that case I would agree with you.

But if the engine's line is verified by reviewing each move and investigating alternative moves, and if the impact of the horizon effect is mitigated by forward sliding or by at least what I like to call "forward leaping", then I would put my money on the engine's evaluation of the position, even if the human that disagreed with the engine was Magnus Carlsen. It doesn't necessarily mean that the engine will be right all of the time, but on the average it will be.

No doubt in the Sierra Nevada:

More'n likely. (laughs)

<....Seriously, it depends on whether the engine's evaluation and line are presented like coming from the gods, without any additional verification, as unfortunately it's often done....>

That very often happens, which was why I was nettled by the post above.

<....I am occasionally guilty of the same thing although it's usually as a result of time constraints or some such (yeah, I know, if I haven't checked the engine's line I shouldn't be posting at all). In that case I would agree with you....>

Actually, I find your posts quite reasonable on the whole and leavened with humour--a most welcome trait in these days which so often feature just those strings of analysis.

<....And if you don't think that using engines properly and spending time validating their results and exploring promising looking alternative moves and lines requires thinking for yourself, then you probably haven't done it, or at least haven't done it very often. Although it does sometime require a somewhat different kind of thinking.>

At best, engines are a tool to improve one's OTB play. Chess is after all, meant to be played OTB

Using a computer to actually play chess is borderline cheating IMO

Pros: It really is a fantastic tool if you use it properly, i.e. by calculating a line and the using it to check if you were right