Website founded by
MatPlus.Net Forum Competitions Quick Composing TT-242 (h#3.5-N) C. 11-06-2020
You can only view this page!
|(1) Posted by Aleksey Oganesjan [Monday, May 4, 2020 10:34]|
Quick Composing TT-242 (h#3.5-N) C. 11-06-2020
Editorial board of international web project "SuperProblem" (http://superproblem.ru/index-en.html) announces a quick composing thematic tourney for helpmates in 3.5-N moves.
Awards will be published on the website http://superproblem.ru
View the announcement on the link http://superproblem.ru/htm/announcements/our_tourneys-2020.html#TT-242
|(2) Posted by Andrew Buchanan [Friday, May 8, 2020 07:19]; edited by Andrew Buchanan [20-05-08]|
Thanks very much for posting and I look forward to having a go.
I do have a question about the first example: J.-M.Loustau “Ideal-Mate Review”,1985, 1stHM
(= 3+6 )
It's a fine problem, but it's not a sufficient example of tempo. The bQ goes past wK not for tempo, but because it needs to herd it into the fold. bB avoids g3, because it would deliver check there.
This is not a matter of subjective aesthetics: I can prove it! Add bPa6. If there was Black tempo, then the problem would no longer be sound. There would be four solutions, with the bP making 0, 1, 1 & 2 hops respectively. However as Popeye confirms, there is still only the one answer.
If one adds wPa3, on the other hand, then there *are* two solutions. This proves that the "non-thematic" wK move to g7 is the only genuine tempo move in the problem but we would need a second White tempo piece in order to be thematic for the tourney. This bonus pawn test is a good yardstick to test any entries to the tourney.
|(3) Posted by Torsten Linß [Sunday, May 10, 2020 09:28]|
Also bB opens the line for bQ, so Bh2 instead of Bg3 is dual avoidance, and bQ abandons guard of h5.
AB’s pawn test is very nice.
|(4) Posted by Joost de Heer [Sunday, May 10, 2020 10:47]|
Joost de Heer, Henryk Grudzinski
(= 15+16 )
1. e3 h5 2.Qg4 hg4 3. b3 Rh3 4. Ba3 Rf3 5. h4 g6 6. h5 Bg7 7. h6 Be5 8. h7 Bh2 9. h8=Q g3 10. Qb2 f6 11. Qc1 Kf7 12. Qd1 Ke6 13. Bc1
Is 4. Ba3 a tempo move? 'Pass' would be a cook (the queen can choose a 4-move path to d1, e.g. Qe5-b5-e2-d1, because the extra bishop move isn't needed).
|(5) Posted by Andrew Buchanan [Sunday, May 10, 2020 11:54]|
That's a good question, Joost.
I think Ba3 is a tempo move, because if pass is allowed, there is at least one alternative solution. (In fact there are multiple solutions, because there are multiple routes for wQ to return home, but one alternative solution is enough.) Now the question is, what is the right tempo move? The tempo move early, while waiting for bR to vacate the h-file, means a debt must be repaid later. Out of the tempo moves, only one relieves the time constraint on Wh later, by allowing a short cut for wQ to return.
Proof games based round tempo often have multiple segments, "feast or famine". There is either too much time or not enough, for one player (or the other). The famine later doesn't deny the reality of the feast earlier.
It would be interesting to ask Francois Labelle to implement "pass" in Jacobi. Not only would this enable tempos to be trapped accurately, but also it would allow for short proof games to be picked up (e.g. PG17.5 works just as well as the stipulation PG18.0).
|(6) Posted by Joose Norri [Sunday, May 10, 2020 12:29]|
The definition of tempo in help play has never been unambiguous. Here with tempo manoeuvres included the pawn test is hardly conclusive.
|(7) Posted by Hauke Reddmann [Sunday, May 10, 2020 17:15]|
I say no - play Bc1-b2, which is no cook, ergo Ba3 has
the additional aim to vacate the route. It's played for
tempo reasons (White can neither pass nor split a move,
so any tempo move comes only in pairs), but it rather
gains a tempo.
|(8) Posted by Viktoras Paliulionis [Sunday, May 10, 2020 22:47]|
Each tempo move in a solution must have at least one tempo-try corresponding to it. To detect tempo-tries, you can use Helpmate Analyzer. The problem by J.-M.Loustau has only one tempo-try, therefore, it has single tempo move on W1:
(Detecting of tempo-tries is a new feature of Helpmate Analyzer)
|(9) Posted by Peter Wong [Monday, May 11, 2020 05:24]|
Tempo play is shown if the ability to "pass" would result in a cook. Such a hypothetical pass is both (1) necessary and (2) sufficient for tempo play to occur. The "necessary" aspect is why Andrew's bonus pawn test works – such a pawn move is equivalent to passing. If passing doesn't help to give you an alternative solution, there's no tempo play.
The "sufficient" aspect is why Joost's PG shows tempo play. The 5-move manoeuvre Ba3-Qb2-Qc1-Qd1-Bc1 is potentially replaceable by pass-Qe5-Qb5-Qe2-Qd1. I like to distinguish between individual tempo moves and tempo manoeuvres. So to say that Ba3 is part of a tempo manoeuvre would be clearer than calling it a tempo move. The “impurity” of Ba3 clearing a path for the WQ is irrelevant in this context of whether a tempo manoeuvre is shown. Even individual tempo moves can have this kind of impurity, e.g. 1.e3! h5 2.Qxh5 Rxh5 3.e4 Rh8 shows tempo play because of the alternative 1.e4 h5 2.Qxh5 Rxh5 3.pass Rh8; it doesn't matter that 1.e3! is required to let the WQ out immediately.
|(10) Posted by Hauke Reddmann [Monday, May 11, 2020 09:10]|
So, speaking "neudeutsch", there are Vorplan-Tempos, which
are easily recognized and work on their own,
whereas 1.e3 is an Auswahl-Tempo :-)
|(11) Posted by Joose Norri [Monday, May 11, 2020 19:08]|
Just some thoughts, not necessarily relevant in the context of this tourney - or even the tempo move discussion. Misunderstandings are entirely possible.
Macleod & Schwarzkopf
(= 6+6 )
1. d6 Se8 2. Bc7 Sa7#
NAM, fs-102: "Is it possible to have three tempo moves in a H#2? If a tempo move is a completely neutral move with no effect either positive or negative then three tempo moves are obviously impossible. After Black makes the first tempo move white can mate at once. But if we take the definition of Dr Lindner that a move is a tempo move if it can be ommitted (sic) without altering the solution in any way, then 3 tempo moves are possible. For example, the first move 1.d6 can be left out and the solution continues exactly the same. So 1.d6 is clearly a tempo move. Similarly, after 1.d6 white may pass and the solution continues 2.Lxc7 Sa7# as before. Finally, after 1.d6 Se8 it is quite obvious that 2.Lc7 is a tempo move."
If Bc7 is tempo and a pass is substituted for it, then you can't go back half a move and substitute a pass for Se8 too; and d6 can be omitted only if Se8 is played. So either two Black or one white tempo move? (Solvers ATü & Kuli) If you take this position, would the problem be eligible here if short helpmates were allowed (and Black's tempos were performed by different pieces)? As I see it, here the bonus pawn test needs to be qualified, because one can't insert pass for W1 & B2.
Here we already have three ways of looking at it.
Taking NAM's reasoning, this has been extended to five halfmoves. There is also a two solution (or twins) problem by Ylijoki ca. 2004, dedicated to the writer of these lines, I think it's in StrateGems. It's not in pdb or yacpdb.
Then, complete neutrality - what if there are several setplays, but a tempo move reduces them to one? Perhaps it can increase their number. Then you can replace the move with a pass, but it still has an impact, it's not neutral:
Joose Norri, Gruengard MT 2002 5th HM
(= 8+11 )
1. ... ?? 2.Bg4/Bxg6 Kg4 3.Ke4 Sf6#
1. ... Bg4 2.Bxg4 Kxg4 3.Ke4 Sf6#
b) Bh3 -> d3
1. ... ?? 2. Q~ Kg5 3.Ke6 Bxc4#
1. ... Sg5 2.Qxg5+ Kxg5 3.Ke6 Bxc4#
1.- Bg4 and 1.-Sg5 are tempos, but they are very active agents.
Of course in the normal set play form zugzwang destroys the set. When a move changes everything it can be emotionally difficult to think of it as tempo (but naturally it is).
Norbert Kovacs The Chess Amateur 1922
(= 8+9 )
Adding bPh5 realizes the set Nc3#, adding wPh4 gives 1.b1R h5 2.Rb2 Sc3, as it is 1.b1R c4 2.Rb4 Sb2 3.Kb3 Sd3 4.Kxa4 Sc5#.
Abdurahmanović 1967/1978 Sahovski Glasnik / Schach-Echo Prize
(= 5+3 )
1.- Bh1 Sg4 2.Rf3#; 1.Sd5 Rf2 2.Sb4 Bd7#.
Black plays two tempos in a row, so adding 1,5 moves would make this eligible here - if they were performed by two different pieces.
And in a monocolour tempo manoeuvre one move could hypothetically make room for the next one. Then it might be that you couldn't insert a pass for the first move only.
|(12) Posted by Marko Klasinc [Monday, May 11, 2020 21:25]|
In general a tempo move in a helpmate is defined as a move made by either side in order not to disrupt the prospective mate or continuation. A reason for this can be a "zugzwang avoidance" or "waiting" tempo move (which are the purest presentations of tempo moves), but also when a piece which may reach the square immediately it moves first on other square for some reason (a sort of alternative tempo move).
Examples for 242-TT SP show different types:
Example one (quoted above) shows black (alternative) tempo moves in order to allow wK's moves. White K's tempo move is a pure "zugzwang avoidance" but not thematic for this TT as it is the only white tempo move.
Example two shows one "waiting" tempo maneuver by promoted wS and one "waiting" tempo move by wK.
Example three shows a "waiting" tempo move by wK and one pure "zugzwang avoidance" tempo move by wB.
As Viktoras mentioned his Helpmate Analyzer with the new option "Tempo-tries" is a very good tool to check if tempo move is a pure one or is it one of other types which Helpmate Analyzer usually finds as "Hesitation".
Judges don't want to narrow the theme down to only pure type of tempo moves. But a purity of purpose for tempo moves will be a strong criteria in making order, of course.
|(13) Posted by Andrew Buchanan [Tuesday, May 12, 2020 06:20]; edited by Andrew Buchanan [20-05-12]|
Dear Joose & Marko,
Thanks for your clear examples & explanation. I never would have guessed, and I think I am far from the only one here who had a very different sense of that term. If you want your meaning to prevail, you must define it properly and then communicate it. "Formalize and popularize."
The increasing existence of programs like Helpmate Analyzer, WinChloe & Meson which are trying to spot themes and tropes in programs means that we must rightly become more precise about the meaning of the terms we use. Defining the concepts (while not stomping on creative freedom) is much more important than the actual words used, but we should have the best shot at defining or redefining terms while we are nailing the concepts.
I guess a context to view your sense sympathetically is that if someone can't move from A to B, but has to go via C, then that "loses a tempo", in some sense. Time is a budget, and these "indirections" as I would have termed them are consuming time. "Pure" tempo is just one motivation for indirections.
A term like "pure tempo" is such an afterthought: it really deserves a noun in its own right. It is a huge subject. It is almost mandatory in classical retro where composers of most non-trivial problems need to build and embed their own clock out of the limited lego pieces available to them. Aside the help/retro worlds, tempo also has completely different senses in competitive stipulations (d#, s#, study...) where Conway's combinatorial game theory has given us the words "hot" and "cold" to distinguish racing positions from zugzwangs.
With this in mind, I would appreciate if you look at the following incomplete list of PDB tempo-related keywords, many of them completely undocumented. They badly need curating. I would really appreciate if you can help systematize them. At the very least, we need a clear distinction between terms for help and for competitive formats. Some terms are duplicates, or are missing. I would be very grateful if you have the time to help, please.
An extra move used to force zugzwang and win
In a move-exchange task, White forces the initial position by n-turn steering maneuvers. At the same time, White loses a tempo in this duel between a white and a black piece, which means that the move is now rendered harmful to Black.
Tempo Creation bP
Tempo Creation wP
Parameters: following the standard PDB pattern, pieces performing the tempo action are indicated as upper case = white, lower case = black. Tempo games in different phases are separated by "space". Example: tempo play (L s) the white bishop makes a tempo move in the first phase, the black knight in the second phase.
Loss of tempo
With White to play, a sequence of moves to reach the same position with Black to play. (Help game, abbr .: TV)
Die Tempoverlustrückkehr hat eine enge Verwandtschaft mit Tempoduellen. Der einzige Unterschied besteht darin, dass bei einer Tempoverlustrückkehr nicht auf jeder Seite genau eine Figur agiert, sondern mindestens drei Themasteine aktiv ziehen.
|(14) Posted by Marko Klasinc [Tuesday, May 12, 2020 13:03]|
True, Andrew, a lot of situations connecting to tempo moves/maneuvers/play are not well defined together with inevitable differences between help-play and direct play problems.
Fadil and I found only a small number of problems with double tempo play from one side in databases, and therefore we wanted to attract as many composers as possible to participate at this TT. This was also a reason to allow all kinds of tempo play. The example one (Lousteau) was intentionally included to avoid questions what types of tempo play were allowed. We prefer "pure" tempo play but also "impure" presentations with good ideas will take place in the award. "Judge's clarificaton" from our last post was intended to clarify to the participants what was allowed.
Participating problems could hopefully make it possible to clarify some dilemmas regarding definitions.
|(15) Posted by Andrew Buchanan [Tuesday, May 12, 2020 18:34]; edited by Andrew Buchanan [20-05-12]|
A proof game is a kind of traveling salesman problem, except there are multiple salesmen, and some of them capture other salesmen. A candidate solution involve a certain number of moves for each piece in order for them to end up in the right place, and for all captures to happen. There may be floating tempi for each player which are not yet assigned. The algorithm then looks at to see whether the candidate is actually feasible as an alternating sequence of legal moves, spending the floating tempi as appropriate. Time is the fundamental resource which drives everything, and there may be tens of thousands of potential collisions which can only be avoided by spending floating tempi. And indeed all but one of these will be proved impossible. So when all the floating tempi are spent on a candidate and it has to be discarded, then we can describe it as dual elimination.
So viewed in this way *everything* is about indirections: trying to adjust a candidate to make it fit. How can one possibly quantify the number of such decisions that would be part of a solution? How is one indirection from all of these an artistic feature? I guess the answer is that in a proof game, the proof (or large parts of it) should be coherent and understandable. So there are certain indirections visible to the user. It's sort of like the difference between a very messy solution to a random chess position which just happens to be an #5, and a coherent designed solution.
But while we are in this area, I do want to end up with some decent PDB terms and definitions. Otherwise the whole exercise is just the usual writing in sand. Please can you help here.
|(16) Posted by Marjan Kovačević [Tuesday, May 12, 2020 22:35]|
Perhaps we might borrow from endgames the term "time losing" for the forms that do not present the poorest form of tempo?
In endgames it serves to describe minor dual in the process made longer way than needed.
No more posts
MatPlus.Net Forum Competitions Quick Composing TT-242 (h#3.5-N) C. 11-06-2020