Str8ts Forum

In addition to the contributions to the Weekly Extremes, some more puzzles are scattered over forum threads. As it was in the early stages, their URL strings did not contain the solutions, so here they come again with the solution substrings included. Caution: all these puzzles are chainy. Christmas...

2018 434_Regatta1 435_Regatta2 small endchain 436_Seesaw 437_Pegasus constellation 438_Lupus constellation 439_Continent Africa 440_Pleiades constellation 441_Lambda Greek letter 444_Surprise_S 444_Surprise_S2 combined count of black cells and clues only 10 2019 445_Semiquaver large white rectangle...

2016 332_OmegaMan Firstling 332_KnightsMove16 Sophomore 333_Hysteresis ultra chainy 33333_Hysteresis two more clues; still very hard 335_Blitz 339_XmasTree clues as candles 340_Roof47 clues as roof 2017 341_New_Year_Special first zero hint; set H5 low or see forum post 342_Zipfelmuetze small strree...

Hi Leren, Regarding the CanisMajor eliminations (position 0:17 node 106): The first one should be quite straightforward, where the single cell compartment A8 has only the two candidates 8 and 9. If CJ8 would have a 9, then it also would have an 8 and no candidates left in A8. So all 9´s can be remov...

Hi Leren

cross posting, didn't see your last one.

Retrieve CanisMajor and let it run with and without "hone search" setting. "Without" the solver needs chaining. But at the first chain nodes one can remove 9 from CJ8 and - here more important - the 4 from B3.

hone

@Leren Robo´s solver indeed is trial/error based once all standard tools are exhausted. Using sophisticated complexity measures, it determines the most promising undecided High/Low compartments and if this is exhausted uses candidates. I was surprised how this solver handled some of my ultra puzzles...