A thought on strategy

Will people resign or lose because of bad position? When they have the equal materials? Most of time they won't. But they will after a concret decisive shot. This shot is tactic, not strategy.
@RamblinDave said in #16:
> #14 I think it's a mistake to think that a tactic has to mean checkmate or getting a decisive material advantage, though. For instance, suppose white would like to play some strategically desirable move like improving a piece or gaining space on the king's side, but they've moved their rook away from a1 and a black bishop is threatening the a2 pawn. They might hold off their plan until they've saved the pawn, but alternatively they might notice that it'd be tactically losing for black to take the pawn, because they can trap the bishop with b3 and win it with Ra1. In this case, the tactical idea isn't helping them to win the bishop (unless their opponent blunders), it's letting them carry out their strategic idea more quickly and hence, presumably, more effectively.

@Alientcp said in #20:
> I think i found the issue of why we cant agree.
> In reality, tactics are really low scale and are localized. While they might impact the local struggle, in reality, they have little to no impact on the grand scheme.
>
> Strategy rules always in the grand scheme of things.
>
> But the problem is that the chessboard is not big enough, and a little tactic often has more impact on other sides of the board than it should have using the proper context of the word.

I agree partially. I certainly means chess tactics only. On chessboard, one tactic is enough to bring victory. In real world,it won't wrok the same way,it is far more complicated.

@Buttercup22 said in #8:
> Strategy doesn't ensure tactics to work. In fact if it's strategy it doesn't even see the tactics. It's just following general chess principles to give you the best chances for successful tactics down the road.

Pretty reasonable thought. Strategy aim for potential tactics,or tactic building. That is my understanding.

@handsomestt said in #23:
> Pretty reasonable thought. Strategy aim for potential tactics,or tactic building. That is my understanding.

Yea I reread the whole thread and I agree with almost everything you said except for one point.
> On chessboard, one tactic is enough to bring victory. In real world,it won't wrok the same way,it is far more complicated.

I think in life that is often the case too. Think of a fight. One good blow can stun the opponent and then allow a series of blows to finish him off. Or if there is wrestling, it just takes a headlock one time to finish your opponent. Or one little mistake or lapse in concentration in sword fighting can result in a stab that does you in.
In other areas too like in politics. Often one scandal can be enough to end your career. These days it can be one tweet or allegation from your past. Certainly that's not always the case like Donald Trump or a Canadian version Rob Ford , but it can work that way.

We had this fight between strategy and tactics in the 80s when chess programmers had to choose how many (and which) positional rules to include in their software.

Some programmers tried a pure brute force approach, with their programmes being optimized on going through the search tree as fast as possible. To achieve this they had only as little positional evaluations in their software as possible. Ulf Rathsman comes to mind with the Mephisto MM2.

Other programmers went for the extreme opposite with a lot of positional evaluations in their engines. However, as those are time-consuming and the processors of the 80s were very slow compared to today they had to cut off moves from the search tree very early. This kind of tree search was called "selective search" because only a selection of moves was evaluated very deeply while most moves were discarded at low search depths. An example were Thomas Nitsche and Elmar Henne with the Mephisto 3.

In direct comparison the brute force programmes usually had the upper hand because the selective approach proved to be tactically vulnerable. Often the selective programme would reach a better position but then fall for a mating attack.

But pure brute force wasn't the way to go as well as those programmes would often violate simple strategical rules and land in a bad position. In general the best programmes were those who found a good balance between tactics and strategy. I tend to think this is valid also for the grey matter processor in human skulls.

@Buttercup22 said in #24:
> Yea I reread the whole thread and I agree with almost everything you said except for one point.
> I think in life that is often the case too. Think of a fight. One good blow can stun the opponent and then allow a series of blows to finish him off. Or if there is wrestling, it just takes a headlock one time to finish your opponent. Or one little mistake or lapse in concentration in sword fighting can result in a stab that does you in.
> In other areas too like in politics. Often one scandal can be enough to end your career. These days it can be one tweet or allegation from your past. Certainly that's not always the case like Donald Trump or a Canadian version Rob Ford , but it can work that way.

I am glad we agree on the important part. As to the disputable part, let us keep the opinion respectively.

@Katzenschinken said in #25:
> We had this fight between strategy and tactics in the 80s when chess programmers had to choose how many (and which) positional rules to include in their software.
>
> Some programmers tried a pure brute force approach, with their programmes being optimized on going through the search tree as fast as possible. To achieve this they had only as little positional evaluations in their software as possible. Ulf Rathsman comes to mind with the Mephisto MM2.
>
> Other programmers went for the extreme opposite with a lot of positional evaluations in their engines. However, as those are time-consuming and the processors of the 80s were very slow compared to today they had to cut off moves from the search tree very early. This kind of tree search was called "selective search" because only a selection of moves was evaluated very deeply while most moves were discarded at low search depths. An example were Thomas Nitsche and Elmar Henne with the Mephisto 3.
>
> In direct comparison the brute force programmes usually had the upper hand because the selective approach proved to be tactically vulnerable. Often the selective programme would reach a better position but then fall for a mating attack.
>
> But pure brute force wasn't the way to go as well as those programmes would often violate simple strategical rules and land in a bad position. In general the best programmes were those who found a good balance between tactics and strategy. I tend to think this is valid also for the grey matter processor in human skulls.

WOW, it's really professional and deep. Luckly we have alpha zero now, it is supposed to be almost perfect.

If you can calculate 30 moves deep "free of tactical accidents" - then strategy matters^^. AlphaZero

@handsomestt said in #21:
> Will people resign or lose because of bad position? When they have the equal materials? Most of time they won't. But they will after a concret decisive shot. This shot is tactic, not strategy.

I don't really see why you'd consider a discovered check (say) to be a tactic if it gets played on the board and leads to an instant win, but not if it's a threat that forces the opponent to make concessions to avoid it, or if it's played on the board by leads to a smaller and less decisive advantage. You're still using the same idea.

You might be interested in looking at some of the positions from the STS. They are suppose to be example of positions in which strategic decisions are required.

The Strategic Test Suite (STS) is a series of themed test suites by Dann Corbit and Swaminathan Natarajan initially created circa 2008. Ferdinand Mosca updated the suite and wrote an STS-rating system. That code was last updated in 2019.

The positions were designed to aid chess engine programmers in evaluating their program's understanding of strategical concepts. The positions are also interesting for evaluation by humans!

The EPD file from Mosca has 1,500 positions organized in themes with 100 positions per theme.

When I ran the STS-rating system I obtained the following results:

Engine.....Version......Score (%).....Rating
Stockfish..........14...........84.7%........3530
Komodo.....12.1.1...........81.4%........3382
Lc0............v0.27.0...........61.8%........2507

I wrote a series of blogs on some of the positions that Stockfish failed to find the "best move" per that EPD.