Engine analysis brainrot

Sort:
Avatar of Shaikidow

How do you deal with it?

For example, in the openings, I used to use the engine only to check whether my ideas had any blunders to them or not, but nowadays I seem to be too concerned that my opponent is gonna know the exact way to obtain and keep a theoretical advantage over me (and mind you, I don't really play any really mainstream and/or theory-heavy lines), so I try to take the engine recommendations, but they're often hard to understand conceptually. Sometimes I don't even understand opening positions in which the engine says that I've got a +1 advantage!

How do I learn the ideas 'humanly' when the literature for them is lacking?

Avatar of RalphHayward

Your opponents will almost never have deep theoretical knowledge of your particular opening speciality until your grade is 2000+. I'm in the 1990s (in my two best formats; 1800s in the one I play for fun) and I still have opponents going into long thinks on move three quite a bit.

Looking at your last few games, you seem to be playing a sort of Pirc-Modern-Lion hybrid. The move orders are so varied and flexible in that one you'll not see hyper-prepped opponents until you hit a level to be playing titled players who devote time to their "opening lab".

But it's easier to say this sort of thing than believe it. I myself handle it by:

1) Look for a book on your opening. Not a database, not an article, and old-style "dead tree" paper book. Because they're usually written with explanations of the opening themes and position types. If you can see a preview, check the one you're getting actually does have that sort of writing in it.

2) Read. It. Sounds obvious, but too many opening books just "sit on shelves gathering dust".

3) If you're surprised by a strong move by your opponent, stop a moment. Think, "I know the ideas of this opening - what's my plan to get past that and minimise the damage?". If you really have read the book, you will know the ideas and be able to react well.

4) But every now and then we all fall victim to a Specially Prepared Line. I get past that by saying to myself, "Right. You got whacked there. Learn how not to let that line happen again. Now. Is my general understanding OK? [If no, do more reading] Yes. Then that's been a learning experience. Done now. nothing else to see here. let's get on with our lives. And pick myself up and dust myself off and get back on the horse.

Nothing will make you prep-proof. But how you approach your time matters. And understanding matters more than theory - when the memoriser gets to the end of his/her memorisation he or she is banjaxed if that's not accompanied by understanding of how to play the position. And your superior understanding of that will turn a large disadvantage into a small disadvantage into equality into better for you. Hopefully. After the opening, the gods have placed the middlegame.

Not sure if that'll help, hope it might.

Avatar of Shaikidow
RalphHayward wrote:

Looking at your last few games, you seem to be playing a sort of Pirc-Modern-Lion hybrid. The move orders are so varied and flexible in that one you'll not see hyper-prepped opponents until you hit a level to be playing titled players who devote time to their "opening lab".

1) Look for a book on your opening. Not a database, not an article, and old-style "dead tree" paper book. Because they're usually written with explanations of the opening themes and position types. If you can see a preview, check the one you're getting actually does have that sort of writing in it.

3) If you're surprised by a strong move by your opponent, stop a moment. Think, "I know the ideas of this opening - what's my plan to get past that and minimise the damage?". If you really have read the book, you will know the ideas and be able to react well.

What if the book isn't good enough, though?

Speaking of the aforementioned opening, I've got the Black Lion book by Van Rekom and Jansen, but I ran through some of the stuff contained therein with an engine, and it's just not properly up to snuff. In fact, there are just too many lines in which the plan to eschew castling for Black and play ...h6 and ...g5 is just plain bad according to the evaluation; and sure, maybe it's not a line that White will know how to properly play against by heart, but it's certainly not an opening I'll be able to keep relying on as my opponents get stronger. On the other hand, playing a traditional Philidor Lion with ...O-O and aiming for ...b5 as the main plan is an exercise in masochism. GM Simon Williams seems to have a course on the Lion that provides a nuanced update compared to the Van Rekom - Jansen book, but I'm in no position to purchase it now, and I don't know if it'd even be worth it, all things considered.

No matter what opening I pick, it seems too taxing for me to try and play for a win if I'm not a prep monster. I keep losing to banal stuff; for example, one of my most recent games went 1. Nc3 d5 2. d4 Nf6 3. Bg5 (all according to plan), but then my opponent played 3... Ne4, and after trading on e4 and playing c3 next, the position wasn't bad for me as such, but it was extremeley boring and out of my explicit prep, so I ultimately lost. Is that why I play chess? Hell no!

I don't know how much I'd have to pay in order to have a professional help me fix my game, and I don't know if I could follow such a program without getting too obsessive or even outright addicted to chess again, so that's horrible.

Avatar of RalphHayward

@Shaikidow Seriously, until you hit 2000 or so what the Engine says does not matter one jot. At 1600-1800 I have had good results with the Three Pawns Gambit

and that particular opening is utter and complete rubbish in terms of soundness. If Black knows what he, she, or they, are doing the game should be over 0-1 in about twenty moves tops. Okay, so after a specific opponent hammered me with it twice I learned something new. But it took about a year for anyone to do that to me. I was winning not because the "book" I used was accurate (it was OK), not because the opening was sound (it wasn't), but because I knew more about how to play the position than most of my opponents (until someone did me over twice). It wasn't quite "trap chess" (even in the worst lines White has some fight) but it wasn't sound.

The point is that engine moves are often extremely counter-intuitive (the annotation, "...but no human would ever play that" crops up in computer analyses quite a bit) so a human only plays them if the human has memorised the line. Or is extremely lucky. or is a tactical genius.

Sounds like you need an opening in which people won't be able to predict your moves and "prep up", or at least an opening in which opponents are less likely to have prepped. Just for your own confidence, that is - in practice the player who knows the position better will often win even if the Engine is saying -1.5. My own preference is for 1..., b6 against almost everything, and there's a good easy Lakdawala book on it. it's not a complete panacea, but I'm one lazy lazy man.

Avatar of Shaikidow
RalphHayward wrote:

@Shaikidow Seriously, until you hit 2000 or so what the Engine says does not matter one jot.

Sounds like you need an opening in which people won't be able to predict your moves and "prep up", or at least an opening in which opponents are less likely to have prepped. Just for your own confidence, that is - in practice the player who knows the position better will often win even if the Engine is saying -1.5. My own preference is for 1..., b6 against almost everything, and there's a good easy Lakdawala book on it. it's not a complete panacea, but I'm one lazy lazy man.

1. And if I don't wanna be stuck in the 2000s or below it? What then?
I mean, if anyone plays a junk opening against me and wins, I want it to be due to their general strength and not due to my inability to get an opening advantage against them as such. I want the engine to show me that I lost due to some inaccuracies down the line and not because I got tricked in the opening! In fact, I would prefer such a scenario over one in which I'd win despite getting outprepared in the opening (by which I mean getting a dead lost position that takes exact memorisation of a long sequence of non-intuitive moves in an obscure position to prove)!

2. I almost exclusively play unpredictable openings such as the one you've mentioned (if they aren't too passive or boring), but I have to work so damn hard to get the benefits of my experience with them, so I end up losing more often than not. The Philidor Lion (my longest-staying main weapon being 1... d6 with a variety of transpositions) is one example of it, the 3... Qd8 Scandinavian is another, the Nimzowitsch Defence is yet another, the Modern Defence is yet another, and then there are such favourite niche moves of mine like 1. c3, 1. Nc3 and 1. d3 from the White side. You'd think that at least some of those moves would yield good enough results for me with clever enough transpositions (one of my main opening aims being to make my opponent regret playing Bf4 as White or ...Bf5 as Black), but it turns out that structural familiarity just isn't enough if the structure is so passive that a good position can be obtained against it just by playing natural moves. Take this opening, for example:

Sure, you may be able to swindle a person or two, but when the best moves are both forced and visually principled/harmonious, it's not a tool anyone can really make a good career out of.