...just those links. If you are looking for a specific page or catchword you can use the search box on top.
3 KB (417 words) - 13:23, 19 April 2024
...''' data-structure fitting in one 64-bit register. For example, a bitboard can represent things like attack- and defend sets, move-target sets and so on.
* Bitboards can have a high information density.
26 KB (3,485 words) - 18:37, 12 March 2022
...free board representation that knows '''every''' rule of chess. While this can sometimes be a pain, especially implementing the more complicated rules suc
...a look at some of the [[:Category:Open Source|Open Source Engines]]. These can be very helpful when translating rather vague algorithms into specific data
2 KB (422 words) - 15:48, 17 December 2018
* [[:Category:Can|Can]] - [https://en.wikipedia.org/wiki/Tago_Mago Paperhouse], [https://en.wikip
[[Category:Can]]
4 KB (457 words) - 21:23, 21 September 2020
....org/wiki/Specification_%28technical_standard%29 specification], the rules can be expressed by [[Algorithms|algorithms]] and data structures.
9 KB (1,194 words) - 09:29, 3 July 2021
...] [[Robots|robot]] able to move its own pieces. It is still functional and can be visited at the Torres Quevedo Museum of Engineering, Institute of Civil
A detailed explanation of El Ajedrecista can be found in ''Les Automates'' by [[Henri Vigneron]] <ref>[[Henri Vigneron]]
20 KB (2,969 words) - 14:19, 14 August 2020
...can be well ordered. This gave rise to the Zermelo axiom that every class can be well ordered. In 1904 Zermelo defined the [https://en.wikipedia.org/wiki
9 KB (1,273 words) - 22:10, 1 October 2018
We can deduct square mapping on enumerating [[Files|files]] and [[Ranks|ranks]] fr
7 KB (965 words) - 18:15, 5 October 2020
We can safely replace bitwise-or by add, since unused bits are always zero:
10 KB (1,405 words) - 16:45, 20 June 2020
...ew/how-rybka-and-i-tried-to-beat-the-strongest-chess-computer-in-the-world Can a GM and Rybka beat Stockfish?] by GM [https://en.wikipedia.org/wiki/Daniel
* [https://www.talkchess.com/forum3/viewtopic.php?f=2&t=74484 Can the sardine! NNUE clobbers SF] by [[Henk Drost]], [[CCC]], July 16, 2020 »
65 KB (8,362 words) - 13:19, 1 March 2024
...ers with the score 1.5 to 0.5. Those, who remember the hot summer of 1972, can envy to the authors of Kaissa, that carried out the pair of days in the wee
...anywhere do not share and among them a lot of such where quality of result can be estimated objectively. The majority of serious experts, before engaged i
54 KB (7,939 words) - 06:55, 24 September 2023
...''[https://en.wikipedia.org/wiki/What_Computers_Can%27t_Do What Computers Can't Do]''.</ref> at MIT was challenged by Greenblatt to play a game against h
21 KB (3,150 words) - 18:33, 25 February 2021
...rchitectures with different net size give different nodes per second. CPUs can be utilized for example via BLAS and DNNL and GPUs via CUDA, cuDNN, OpenCL,
* [http://www.talkchess.com/forum3/viewtopic.php?f=2&t=69622 Can somebody explain what makes Leela as strong as Komodo?] by Chessqueen, [[CC
35 KB (4,831 words) - 15:09, 10 August 2023
...are all mailbox, independently from elements in the array for padding that can act as a [https://en.wikipedia.org/wiki/Sentinel_value sentinel value] <ref
* Suitable with the same efficiency for any size of boards, from one can be fitted on 64-bit integers to much larger. Thus it is also easier to supp
7 KB (993 words) - 12:00, 13 June 2021
...are all mailbox, independently from elements in the array for padding that can act as a [https://en.wikipedia.org/wiki/Sentinel_value sentinel value] <ref
need to figure out what pieces can go where. Let's say we have a
10 KB (1,393 words) - 21:42, 14 August 2018
Represented as such a square difference, displacement vectors can be used to index 225 (15x15) or 240 (16x15) sized lookup tables for stuff d
...6x8, making the "off the board" index condition almost redundant, since it can be combined with the coding of the guard or sentinal squares <ref>[https://
25 KB (2,238 words) - 16:19, 25 December 2020
...6x8, making the "off the board" index condition almost redundant, since it can be combined with the coding of the guard or sentinal squares <ref>[https://
18 KB (2,661 words) - 17:16, 1 October 2020
* [http://rybkaforum.net/cgi-bin/rybkaforum/topic_show.pl?tid=30373 You can't win a WC by playing drawing lines] by Venator, [[Computer Chess Forums|Ry
20 KB (2,719 words) - 11:36, 28 January 2020
11th World Microcomputer Chess Championship, May 1991, Vancouver CAN <ref>[http://www.csvn.nl/index.php?option=com_content&view=article&id=75%3A
14 KB (1,590 words) - 11:43, 28 January 2020
...arks given by [[Morris Miller]], "A comedy of errors in which the readers can supply their own question marks, exclamation points, asterisks and stars" <
5 KB (622 words) - 12:51, 28 January 2020
...izing this concept we have found that a pseudo-dynamic position projection can be effected in a static environment on a local scale. Significant is that t
8 KB (1,171 words) - 00:17, 7 November 2019
...essence. Most programs have opening books, but I seriously doubt that any can handle [[Transposition|transpositions]]. I have never seen a program sacrif
...ss the nation will bring forth new ideas. We must learn all the lessons we can, for next year, the Russians are coming.
7 KB (1,107 words) - 00:17, 7 November 2019
...her side of the fence are the slower but "smarter" computer programs. They can't think about zillions of chess moves, so they need a lot of information ab
13 KB (1,864 words) - 14:31, 7 December 2019
...different patterns. The program is structured so that additional patterns can be easily added by a chess player unfamiliar with the details of the progra
4 KB (584 words) - 23:14, 1 January 2020
...second high failure, the first high failure remains the best move, and it can still avoid doing any more full searches.
24 KB (2,852 words) - 18:36, 25 February 2021
...[https://www.newspapers.com/newspage/107846183/ At Engineer Week at UL you can try to make a soft landing on the moon]''. [https://en.wikipedia.org/wiki/T
...[https://www.newspapers.com/newspage/107846183/ At Engineer Week at UL you can try to make a soft landing on the moon]''. [https://en.wikipedia.org/wiki/T
6 KB (885 words) - 19:52, 3 July 2021
I remember [[Ken Thompson|Ken]] once saying "If you can just hold on and not get caught by its tactics, it will eventually 'fold' a
6 KB (875 words) - 15:32, 2 April 2020
...pedia.org/wiki/Operating_system operating system]. One or multiple threads can exist within the same [[Process|process]] to share its resources such as [[
...alkchess.com/forum/viewtopic.php?t=65844 Lazy SMP >4 Thread Slowdown] by [[Can Cetin]], [[CCC]], November 29, 2017 » [[Lazy SMP]]
25 KB (3,329 words) - 21:14, 28 March 2022
...of Artificial Intelligence”. A public pronouncement of his point of view can be found in an interview with [[Jaap van den Herik|H.J. van den Herik]] hel
...this journal), but the game is limited in the types of challenges that it can offer to the AI researcher. Being a game of perfect information (both playe
76 KB (10,439 words) - 14:01, 10 August 2023
...ur queen, it's phenomenally brilliant. You are winning the game. So if you can find a way to get your queen out there and sacrifice her, well, you've won.
24 KB (3,641 words) - 21:59, 3 July 2021
...https://en.wikipedia.org/wiki/Linear_algebra linear algebra] library which can be used for [https://en.wikipedia.org/wiki/Dynamical_simulation simulating]
2 KB (285 words) - 10:11, 3 September 2018
...ings and pawns, add the square roots of the number of moves that the piece can make, counting a capture as two moves.
18 KB (2,507 words) - 18:37, 25 February 2021
...computation]" with the [[Alan Turing#TuringMachine|Turing machine]], which can be considered a model of a [https://en.wikipedia.org/wiki/Computer general
...ng_machine Turing machines] are abstract symbol-manipulating devices which can be adapted to simulate the logic of any computer algorithm, described in 19
28 KB (4,173 words) - 09:54, 14 November 2020
...ng strength]] of a chess engine or game playing program. Evaluation tuning can be applied by [[Automated Tuning#Optimization|mathematical optimization]] o
* Can modify any number of weights simultaneously - constant [https://en.wikipedi
91 KB (12,241 words) - 21:01, 28 March 2022
...ikipedia.org/wiki/Gradient gradient] isn't mathematically defined, but you can use [https://en.wikipedia.org/wiki/Difference_quotient difference quotients
...ew-7000003879/ Dell PowerEdge T620 review | ZDNet]</ref> . A local minimum can usually be computed within 6 hours, faster if only small changes to the par
28 KB (4,063 words) - 21:06, 28 March 2022
The basic theory on which one chess program can be constructed is that there exists a general characteristic of the game of
9 KB (1,350 words) - 16:17, 6 December 2020
...e it self starts, the answers for both kings are ready before the software can ask the question. </code>
17 KB (2,527 words) - 23:40, 18 February 2021
...] method for computer chess). The program is written in 486- assembler and can search 9 or ten plies in the [[Middlegame|middlegame]]. Schach participated
9 KB (1,257 words) - 10:10, 29 February 2020
...[Principal Variation|best variation]] if the [[Side to move|side to move]] can have anything captured.
...tor]] to resolve situations where something is under attack. If a decision can not be made using this score, then a secondary positional score is generate
8 KB (1,106 words) - 22:22, 16 December 2019
...een CB and Crafty last night, one CPU, no [[Pondering|pondering]] since CB can't yet ponder, 10 secs per move. Crafty won handily, typically searching 14-
18 KB (2,693 words) - 16:15, 16 November 2020
...the project PIONEER, Dr. Botvinnik used to say: “If a human chess master can make it, a computer will make it as well.” He believed in the existence o
* [[:Category:Can|Can]] - [https://en.wikipedia.org/wiki/Landed_(album) Unfinished], 1975, [https
38 KB (4,367 words) - 09:31, 28 March 2021
...ction reflects the [[Playing Strength|playing strength]] to maximize, SPSA can be used in [[Automated Tuning|automated tuning]] of [[Evaluation|evaluation
...ural networks]], is applicable to any optimization task where the gradient can be computed or approximated <ref>[[Martin Riedmiller]], [[Heinrich Braun]]
18 KB (2,427 words) - 16:42, 2 June 2021
...ust not be locked and skipped for the 62 with an extra depth reduction. 32 can not appear as index inside the 58 recursive tests in one path, and is there
6 KB (673 words) - 10:57, 22 September 2018
...are. They ensure that SOMA will move or defend any White piece which Black can capture with advantage, that he will harry his opponent's pieces (+5 for in
...[[Castling|castling]], which discourage him from leaving pieces where they can be threatened by an opponent's pawn advance, and which enable him to allow
14 KB (2,114 words) - 20:50, 8 December 2019
...pected pay-off replaces the iterated [[Minimax|minimax]]. The minimax idea can come in if we are trying to prove a theorem and we imagine that we have an
23 KB (3,383 words) - 17:08, 16 November 2020
... I can speak from experience here. In the early versions of my chess program Phoen
12 KB (1,783 words) - 17:57, 26 December 2019
'''YNCT''', (You Never Can Tell)<br/>
* [https://en.wikipedia.org/wiki/You_Never_Can_Tell You Never Can Tell (disambiguation) from Wikipedia]
4 KB (529 words) - 13:07, 26 October 2018
...has been my alter-ego when it comes to discussion of chess ideas, and what can be done to allow Hitech to understand this or that.
...rds) to do fast [[Search|searching]] and [[Evaluation|evaluation]]. HiTech can search about 175,000 [[Nodes per Second|positions/second]] and executes a [
15 KB (2,069 words) - 18:27, 27 November 2022
...de>For first time users Rebel is easy to understand. Within Rebel 10.0 you can even choose three modes to operate Rebel (novice, intermediate and expert).
22 KB (3,106 words) - 17:47, 11 August 2023
...function]] is complex and hence a little slow. Positional [[Score|scores]] can "easily exceed several Pawns". Separate evaluation terms are used for [[Ope
9 KB (1,287 words) - 14:44, 29 November 2020
...re, the score of each move is now the score of the worst that the opponent can do.
17 KB (2,338 words) - 17:34, 15 January 2022
...f negaMax is only returning an optimal score, then just how is it that you can know which particular move this score is related to? These two questions ar
3 KB (435 words) - 08:40, 27 April 2018
...know that we can achieve at least that, so anything that is clearly worse can be ignored.
...e go to a search [[Depth|depth]] of 3 or greater, because now both players can make choices affecting the game tree. Now we have to maintain both a [[Lowe
37 KB (5,099 words) - 16:20, 1 December 2021
* [http://www.talkchess.com/forum/viewtopic.php?t=48356 LMR at CUT nodes can be arbitrarily bad!] by [[Michel Van den Bergh]], [[CCC]], June 20, 2013 »
15 KB (2,159 words) - 11:21, 31 May 2021
...Usually any given type of [[Neural Networks|neural network]] architecture can be employed in any of those tasks.
...ame but the final result (still) unknown. A form of reinforcement learning can be used for unsupervised learning, where an [https://en.wikipedia.org/wiki/
137 KB (18,484 words) - 00:18, 19 December 2021
...s later, but is conceptually simple, and we will first show how a computer can be programmed for such a strategy.
3 KB (408 words) - 13:52, 7 December 2019
...constituting an analysis of variations, until a position is reached which can be evaluated. The final choice depends on an acceptance value and if a move
7 KB (1,048 words) - 13:54, 7 December 2019
* [https://www.stmintz.com/ccc/index.php?id=93686 Can we use hash table at root?] by Tim, [[CCC]], January 31, 2000 » [[Transpos
5 KB (631 words) - 21:04, 8 November 2021
...o window]] search centered around alpha is conducted to test if a new move can be better. If so, with respect to the null window but not with respect to t
...second high failure, the first high failure remains the best move, and it can still avoid doing any more full searches.
28 KB (3,999 words) - 00:15, 1 February 2022
...cutoff, a [[Best Move|best]] (or good enough) move from a previous search can improve [[Move Ordering|move ordering]], and save search time. This is espe
...ally a 64-bit number, and with current hardware limitations, no hash table can be large enough to accommodate it. Therefor to calculate the address or ind
60 KB (8,162 words) - 21:35, 28 April 2022
...ne becomes unnecessary. This way, also the results from the partial search can be accepted - though in case of a severe drop of the score it is wise to al
13 KB (1,843 words) - 23:44, 4 October 2019
...aluation|static exchange evaluation]] < 0, [[Delta Pruning|delta pruning]] can be used for that purpose.
...t exceeds [[Alpha|alpha]], to see if any [[Tactical Moves|tactical moves]] can increase [[Alpha|alpha]].
25 KB (3,459 words) - 10:30, 15 June 2021
...efining one [[Ply|ply]] to be a number greater than one. Then an extension can be added that does not yet extend the search, but further down the tree may
14 KB (1,884 words) - 21:12, 20 November 2020
...otes too little effort to critical lines. The singular extension heuristic can be seen as a solution to this problem.
...of flattery"; as to why the author(s) of the clone don't openly admit it I can't say. But the specific idea I'm talking about here is attributed to the cl
16 KB (2,280 words) - 17:09, 16 November 2020
...has been my alter-ego when it comes to discussion of chess ideas, and what can be done to allow HiTech to understand this or that.
13 KB (1,716 words) - 17:09, 16 November 2020
...hould we be searching a line that already is equal to or longer than n, we can impossibly increase alpha even if another mate was found. This means we cou
...root, the opponent can impossibly be evaluated > SCORE_MATE - n. Thus beta can be set to this value should it be higher.
5 KB (719 words) - 11:57, 12 May 2018
...hing on his turn, that is, by playing a "No Move" and then seeing if White can enforce an immediate checkmate.
...that sometimes an extension of the game tree by introducing of dummy move can lead to a reduction of the search tree. In positions with material advantag
40 KB (5,641 words) - 18:55, 25 December 2020
...In principle, forward pruning at an expected All-node is not dangerous but can trigger unnecessary re-searches.
7 KB (953 words) - 23:30, 2 November 2019
...and this pruning technique does not make much difference in most cases, it can significantly decrease the search effort when move ordering heuristics fail
7 KB (968 words) - 23:21, 14 December 2019
Some processing power may be saved by testing if '''any''' move can improve over alpha. Then in truly hopeless nodes we don't do move generatio
// The next three lines test if alpha can be improved by greatest
3 KB (418 words) - 10:37, 25 September 2020
* [http://www.talkchess.com/forum/viewtopic.php?t=30036 Can someone explain this?] by [[Michel Van den Bergh]], [[CCC]], October 07, 20
1 KB (164 words) - 18:59, 29 April 2018
...> used open source implementations based on the [[History Heuristic]]. LMR can often reduce the [[Branching Factor#EffectiveBranchingFactor|effective bran
...r in pv nodes), [[Depth|depth]] and move number. Here some sample formulas can be viewed:
17 KB (2,291 words) - 11:20, 4 April 2022
| if we are doing fairly badly, we can reduce the depth |
14 KB (1,941 words) - 20:31, 6 November 2020
...lstice winter solstice] or [https://en.wikipedia.org/wiki/Equinox equinox] can be observed, by means of further bearing points also [https://en.wikipedia.
5 KB (765 words) - 19:57, 6 August 2018
...same score for the (in that case redundant) re-search, which most programs can not guarantee due to possible [[Extensions|extensions]] <ref>[https://www.s
...he value returned by the null window search instead of alpha. But then you can get a fail-low on the research due to search anonomalies. If that happens N
12 KB (1,752 words) - 20:59, 2 March 2020
...-Max]] (which does have no move ordering at all, not even a move list, and can only search moves as it generates them) is able to compete with 'serious' e
9 KB (1,332 words) - 10:40, 25 September 2020
..."winning" move. For example, PxQ will always be a win, since the Pawn side can choose to stop the exchange after its Pawn is recaptured, and still be ahea
...[[Quiescence Search#StandPat|standing pat]] rather than to capture, which can be done by a conditional assignment, or by a [[Avoiding Branches#Max|max]]
20 KB (2,822 words) - 10:43, 1 August 2022
...[[Moves|moves]], which appear inside the [[Search|search]]. Those counters can then be used for [[Move Ordering|move ordering]] as mentioned in the [[Hist
13 KB (1,736 words) - 16:05, 12 January 2019
We believe that we can considerably improve the performance of the history heuristic in some games
6 KB (851 words) - 17:48, 19 December 2018
...r example, ETC transposes the right portion of the tree into the left. ETC can be enhanced to also transpose from left to right. At an interior node, look
7 KB (1,019 words) - 23:44, 10 January 2020
...es. In [[Alpha-Beta#MaxversusMin|max versus min alpha-beta]] a beta-cutoff can only occur for the max-player, while the min-player cuts if below or equal
4 KB (515 words) - 10:27, 7 April 2022
...voiding this position, so you have to assume that they'll do this. If they can avoid this position, there is no longer any need to search successors, sinc
You can also talk about failing high and failing low from the root position, if you
5 KB (759 words) - 10:22, 11 April 2020
...score or an upper bound on its value. This path from the d - 1 ply search can be used as the basis for the search to d ply. Often, searching the previous
6 KB (795 words) - 15:07, 30 April 2018
...]]. Both reduce a randomly ordered tree by about 80%, in favor to HH which can perform the search in much less time. However, along with [[Iterative Deepe
...e information about its environment, meaning that the evaluation functions can be reliable, and is not so complex that trying to encapsulate the process i
11 KB (1,358 words) - 23:25, 5 February 2020
...quares]] of the moves. Although the network is very large, the move scores can be computed quickly, since one only has to propagate the activation for the
5 KB (725 words) - 09:47, 22 May 2021
...s-players-that-cant-be-beaten BBC - Future - The cyborg chess players that can’t be beaten] by [http://www.chrisbaraniuk.com/ Chris Baraniuk], December
...ter the games we were in the bar talking and he was saying that "computers can't play endgames, even simple endgames and they never will." And he said "I
35 KB (4,953 words) - 10:55, 21 March 2021
...with depth A against itself with depth B. The exact tournament conditions can be studied on his webpage <ref>[http://www.top-5000.nl/ply.htm Experiments
28 KB (3,741 words) - 10:51, 16 February 2022
...roduced strong dependencies between sibling nodes and discovered that this can cause search-depth pathology to disappear. While Nau was primarily concerne
12 KB (1,680 words) - 11:13, 4 April 2021
* The position can be treated differently if there is a chance of [[Repetitions|repetition]] o
* Different [[Move Ordering|move ordering]] can cause different result (different moves causing a cutoff), when it comes fr
3 KB (359 words) - 11:50, 2 May 2018
...itions are not the same if a pawn that could have been captured en passant can no longer in this manner be captured or if the right to castle has been cha
...ght not use this kind of replacement scheme. Another problem is that there can be hash table entry collisions, and they must be dealt with. I am not talki
34 KB (4,898 words) - 12:51, 21 August 2022
* [https://groups.google.com/d/msg/rec.games.chess/wqAWR7ylU4w/PTvP5DbWadcJ Can Chess Help Adapt to Life?] by [[Michael Valvo|Mike Valvo]], [[Computer Ches
...s-players-that-cant-be-beaten BBC - Future - The cyborg chess players that can’t be beaten] by [http://www.chrisbaraniuk.com/ Chris Baraniuk], December
40 KB (5,362 words) - 20:47, 22 November 2023
...arch bound without an extra bonus, to make the final convergence. Also, it can be quite instructive to experiment with different evaluation function grain
22 KB (3,042 words) - 10:31, 11 April 2021
...s]]. Depth-first examines child nodes before [[Sibling Node|siblings]] and can easily implemented with [[Recursion|recursion]] using a [[Stack|stack]] of
7 KB (955 words) - 23:55, 9 March 2019
...aeffer]], [[Wim Pijls]] and [[Arie de Bruin]] pointed out, that '''SSS'''* can be reformulated as a sequence of [[Alpha-Beta|alpha-beta]] [[Null Window|nu
...order. Insert and (in particular) delete/purge operations on the OPEN list can dominate the execution time of any program using SSS*. Despite the promise
18 KB (2,480 words) - 23:49, 9 March 2019
...aka [[NegaScout]], and [[MTD(f)]], with integers (α, α+1) or (β-1, β), can therefor only provide a [[Bound|bound]], either [[Fail-High|failing-high]]
...fficiency. Here's the way to follow why. If every root move fails low, it can do so after searching only one move at ply 2, the one move necessary to 're
4 KB (612 words) - 19:38, 22 September 2018
A null window search can never return an [[Exact Score|exact score]] with <span style="background-co
3 KB (443 words) - 16:23, 2 May 2018
You can also talk about failing high and failing low from the root position, if you
6 KB (812 words) - 20:53, 8 April 2020
...causes of instability, and I will discuss how various search enhancements can lead to instability, when I discuss those enhancements. A few other search
...ail-high indicated very clearly that the score should be 25 or better. How can you fail low?
7 KB (952 words) - 13:56, 7 December 2019
...utdegree]. If this value is not uniform, an '''average branching factor''' can be calculated <ref>[https://en.wikipedia.org/wiki/Branching_factor Branchin
...tps://www.stmintz.com/ccc/index.php?id=69483 Please, say in few words what can reduce the "branching factor"] by [[Leonid Liberman|Leonid]], [[CCC]], Sept
9 KB (1,311 words) - 18:22, 12 March 2021
which can be reformulated for both cases using [https://en.wikipedia.org/wiki/Floor_a
3 KB (445 words) - 13:56, 7 December 2019
