Julia's Fairies

No.1784 (JM&AB)

James Malcom, Andrew Buchanan
(USA / Singapore)

Original Fairy problems
01.01.2023 - 31.12.2023

☝ Definitions

No. 1784 James Malcom,
Andrew Buchanan

USA / Singapore
original - 29.11.2023

black Pd4c6 Be6 Qb6 Kc7 Rh8 white Ph7g7f7e7d7 Sg8 Kf8
hdp=3.5                       7+6

Solution: (click to show/hide)

Notify of

Newest Most Voted
Inline Feedbacks
View all comments
Andrew Buchanan
Andrew Buchanan
November 30, 2023 05:44

White achieves a task that Black managed in a very different way in No.1117 (Miodrag Mladenovic 2016). The stip could be hs=4.0, but hdp is more natural and offers fresh design space, as it can handle longer forced sequences after the co-operative play. (Note by Codex convention the DP rule is switched off for hs=, but the hdp stipulation explicitly enables it. So both forms are sound.)
Viewed as hs=4.0, the problem is already C+ under Popeye, but you should really take a look at Ambrona’s work, which has been used to examine all the lichess drawn game ends. Deadpos is an extension of this work, and (independent of DP) has also been used to validate many Mrs Baird orthodox retractors from 19th century, which are outside the scope of conventional engines. Deadpos is highly robust and performant: please add it to the list of linked engines, Julia, and mark this problem as C+.
See https://www.thehoppermagazine.com/gallery for another 52 DP problems involving promotion. All are C+ Deadpos. These include 24 hdp 2.0 AUW, showing every possible order of the promotions. These are all C+ by Popeye as hs=2.5. All 24 are brevity records, since the briefest prior art appears to have been hs=3.0, and that only for 2 cases. Prior art for 17 cases is: https://pdb.dieschwalbe.de/search.jsp?expression=PROBID+IN+%27P1111686%3B+P1114575%3B+P1196085%3B+P1202258%3B+P1243486%3B+P1243490%3B+P1280600%3B+P1280601%3B+P1280602%3B+P1324564%3B+P1328415%3B+P1328416%3B+P1328417%3B+P1328536%3B+P1328579%3B+P1328583%3B+P1375917%27

Last edited 2 months ago by Andrew Buchanan
November 30, 2023 18:30

Probably all hs= , ser.s= and auto= can be designated as reaching dead positions.

Andrew Buchanan
Andrew Buchanan
December 1, 2023 06:00
Reply to  Seetharaman

Yes these categories, together with s=, and also a few h=, ser-h= & draw studies, would all reach dead positions. However, under Codex Article 17A, the DP rule does not apply except for retros, or where explicitly enabled (e.g. hdp). So all the prior art is unaffected.
In a sense, it wouldn’t matter even if DP rule applied, because the play would be the same, except the final move(s) are in the “dead zone”. However it would be distracting, so the Convention is good.
For hdp, it is thematic that interesting play does occur in the dead zone: like the players have pushed the cart to a top of a hill, and just let it roll down under gravity. So in the problem here, the second Schnoebelen capture is the zombie (auto) move after the game has ended. Zombie moves are necessary to prove that the game is really over!
The practical difference between hs= & hdp is that the latter allows longer sequences of auto play. For example: https://pdb.dieschwalbe.de/P1411205 by Per Olin & Joachim Hambros (who won retro Gold Medal at this year’s World Cup).

Would love your thoughts, please comment.x