Engine can be used many ways

NFork

What is your attitude on engines?

People have different attitudes about engines on chess, whether they should used in preparations or not? Is engine even helpful or necessary because after all humans are playing? Some people like to underline that chess is mostly art where engine is not its own territory.

 

The reality to get assistance


There are many ways to make preparations and also play the game. I am writing what way engine can be used as help on preparations. Mostly as far as I can see it engine is useful help because it's very fast and efficient. Certainly Carlsen would be much better help but probably he is too busy so for us mortals engine is 2nd best help. Nowadays engine like Houdini 1,5, 2 or 3 are already very strong that they even understand openings very well. I remember examining a position of an opening of Najdorf where superficially black had open and weak kingside. Engine Rybka 4 evaluated it as about strong advantage for white. The problem was that it was opening theory considered as about = !

I am not totally blind with engine flaws but mostly I trust them and I very rarely see Houdini make a mistake. Still it makes very silly mistakes in endgames and people have to be careful to not trust too much on them. But for me the examples of positions when let's say white is 1 rook up but all the 16 pawns on board give no escape route for rook are too trivial positions that never occur in real games.


Mostly strong engines are good to use because:

1) Analysing your opponent playing style and seeking flaws.


I have started to build my own analytical skills in this part of preparation which is best to do first before planning the opening against specific opponent. The idea is to first let engine analyse quickly your opponent games. You have to active and search the patterns that are repeated over and over again on his games. Following questions may help:
-Does he attack overoptimistic or is he used to passively defending
-Does he blunder often when it is about his own king safety
-Is he good at using pawns or is pawn dynamics difficult for him?

Mostly my interest is on the losses opponent have got, because they give valuable information how I could win him. Also won games are important. Let's say my opponent have won many games in poisonous variation of Najdorf. What does it tell about him? Yes he may know lot of theory
or even have opening book of his pet variation, but probably I would also say that his strenght is tactical abilities.

2) Choosing opening(s) and specific lines

This phase can be done with many ways without using engine at all. You can read your opening book and choose some interesting line that is well analysed, use opening database ask your friends' assistance or if you are brave play certain line that your opponent have struggled with before and lost many games.

Opening theory is good to use in any case if you are not totally familiar with it. Even if you are you can test a game or two of the line against engine. And also check its tactical lines after the game. It's true human doesn't use brute force like Houdini. Human player doesn't calcuate everything. But this doesn't mean that the tactics engine has found couldn't be used by human player during game. It's not cheating if you do that before the game!

Now when you have chosen at least some opening you will try to play. Then you can also ponder will you trust on common opening theory or will you try some playable sideline that you have found on your own or with engine. I have used engine lines quite often and sometimes won IMs
and GMs as well. I think engine can be used and also are used creatively on opening preparations. 2 years ago I found some totally playable gambit line in Sicilian Alapin (2.c3) and decided to play it against GM on first round and I ended up winning. The line was hardly played. I checked some tactics tried and formed some strategy how it all should be continued.

I didn't have much time to preapare but it seems I was much better prepared for this game than my opponent. On last game GM Kanep around 2550 wanted to go for the same variation again as black. I changed to other line that I was prepared with Houdini got strong advantage but was tired after playing 2 tourneys (2 weeks) in a row and ended up in draw. Many things went well in my tourney for example I was full focused on the games without discracting emotions but preparations for these 2 games helped me a lot to get good result.

Here is the draw game: http://chesstempo.com/gamedb/game/3138087

I am myself basic player who can at least a bit play differently depending on my opponent. On this phase I often choose the opening and the line that I think is difficult for my opponent.

Some simplifications can be said:
-If he is aggressive player I choose boring opening for example 1.b3 or Sicilian Alapin as white.
-If he likes the fixation of his pawns and play with pieces I try to go position with fluid pawns.
-If he is sold player who doesn't like to take risks I search some proper gambit lines or aggressive play.

When analysing some sidelines of opening theory with engine it's good to quide engine a bit and also think how would at least human or your opponent play. If some line is even according engine but it seems very nasty position for your opponent side then it could be good to try!

Humans tend to avoid complications even more when uprepared. initiative plays bigger role in human play than between engines that are mostly master at tactical defending.

3) Post-game analysis: your own and then engine assistance

After the game when it's still fresh and you remember the emotions and attitude you had during  the battle it's good to analyse the game on your own. Search improvements on both sides mainly on yours and try to search mistakes even if you managed to win the game. Mistakes are often done because of emotions, attitude and sort of thing that fit in the category of psychology. Did you just miscalculate the sacrifice or were you also overeager to give a lesson to your opponent who had luck on previous game? Engine only gives tactical lines and doesn't understand your eagerness.

After own analysis it's time to check engine analysis. Some players don't like engines, but nevertheless offer almost always objective analysis of your game. It's your job to interpret analysis right way. Besides learning from your mistakes is most efficient way to improve! If engine says it's blunder then IT IS blunder - Be realistic! There is a time to celebrate magnificent win but also after that to get down to earth and think whether your opponent actually had his chances to get advantage.

FM Järvenpää

macer75

True, engines are very helpful when you're preparing before a game, or when you want to analyze a game you just played. However, they are most useful during a game, when you need help deciding what move to make.

NimzoRoy

I think they're great for spotting tactical errors. And they ain't bad for "fact-checking" analysis even by IMs and GMs, although to be reliable they need to be given way more time than is required to analyze my individual moves or games :=(

AND surprisingly enuff Fritz often indicates that GMs were quite capable of making the best moves (or alternative "best" moves by human standards) long before PCs were around. The main advantage the PCs have (at IM & GM level) is they never get tired, emotional or "careless" so their consistency tends to be better overall ie they are much (?) less prone to make a that proverbial one bad move that turns a win into a draw or a draw into a loss.

NFork
Moses2792796 wrote:
NimzoRoy wrote:

I think they're great for spotting tactical errors. And they ain't bad for "fact-checking" analysis even by IMs and GMs, although to be reliable they need to be given way more time than is required to analyze my individual moves or games :=(

AND surprisingly enuff Fritz often indicates that GMs were quite capable of making the best moves (or alternative "best" moves by human standards) long before PCs were around. The main advantage the PCs have (at IM & GM level) is they never get tired, emotional or "careless" so their consistency tends to be better overall ie they are much (?) less prone to make a that proverbial one bad move that turns a win into a draw or a draw into a loss.

I agree with this I think it is still mainly consistency that gives engines the upper hand over top players.  2700+ players still seem to have a better understanding of most positions than engines.

Yes you are right that top human players have better understanding of positions than engine. I agree. Humans ponder positions much more various ways than engines that are more "narrow-minded" although without actual mind.

 

On the other hand if I pretend to be fan of engine I could say that from perpective of engine humans don't understand chess at all and they are very poor on tactics plus chess can be played tactically.

 

I am human so naturally I try to play chess human way because that is my only chance.

 

When it's about understanding we could think the issue from analysis vs playing chess. Human are much better analysing positions than engines. Even for human players tactics are very important when it's about playing chess. I remember some titled player have said that he have seen many GMs who are poor on positional chess but no GM who is poor on tactics.

PedoneMedio

Another interesting use of engines, at least for a low level amateur like me, can be position exploration, meaning you leave the permanent analysis on and play around with moves that make sense to you in an unfamiliar position (from a "new" opening variation, or from a game of yours gone bad, etc.): by doing this you can use the engine as a teacher which points out one after the other all the misconceptions you have of a position in which your intuition is misguiding you, since your lack of experience makes you unaware of what are the tactically soft spots in your and your opponent's formations.

In this sense, engines can also help humans to understand Chass, somewhat.