A new promising sudoku technique: The Slot Machine

Advanced methods and approaches for solving Sudoku puzzles

Re: A new promising sudoku technique: The Slot Machine

Postby StrmCkr » Tue Jul 02, 2019 6:26 pm

I don't think it's exactly SC

Single digit patterns:
Basic slot Machine is simple colouring (true false chains )
Slot machine with and or examining the over laps truths is muti colouring

Muti digit slot machine: over examines the full digit template of each digit selected instead of direct relationships (strong weak links) of the cells.
I class this as advanced Muti colouring(which was talked about years ago don't think I've seen it programed as it's combersum as it uses too much data which masks what is doing the elimination.)
and a limited 3d medusa/gem (as noted. By Space)
Some do, some teach, the rest look it up.
stormdoku
User avatar
StrmCkr
 
Posts: 1432
Joined: 05 September 2006

Re: A new promising sudoku technique: The Slot Machine

Postby SpAce » Tue Jul 02, 2019 10:50 pm

StrmCkr wrote:
I don't think it's exactly SC

Single digit patterns:
Basic slot Machine is simple colouring (true false chains )
Slot machine with and or examining the over laps truths is muti colouring

That doesn't sound fully accurate. In normal terminology, multi-coloring is still a single-digit technique, just more powerful than Simple Coloring. Both reveal X-Chains only, but only multi-coloring can reveal them all (if just those two coloring techniques are compared) -- but not in Hodoku as it only support two color-pairs (i.e. its multi-coloring implementation doesn't necessarily find all X-Chains). As far as I understand, the basic Slot Machine has the power of single-digit multi-coloring, i.e. it can find all X-Chains and not just the small subset found by Simple Coloring. It just doesn't find multi-digit chains (but the Linked Slot Machine does).

That being said, multi-coloring is obsolete as far as I'm concerned. There are much easier coloring techniques to get the same results, both for single-digits and multi-digits. I use coloring a lot, with lots of different colors (10 to be exact, because it's the max allowed by Hodoku), but I've never used multi-coloring of any kind in the true sense of the word (as in having multiple coloring clusters that are then compared with each other).

and a limited 3d medusa/gem

Just to be sure (even though you probably didn't mean it): 3D Medusa and GEM are not the same thing. GEM includes 3D Medusa but goes way beyond that and is much more powerful. The normal definition of 3D Medusa is just a multi-digit version of Simple Coloring (i.e. it only uses conjugate links), so its power is similarly limited. If someone used a multi-digit version of multi-coloring (basically multi-color 3D Medusa), it might approach the power of GEM, but would be really difficult to use (like multi-coloring in general). GEM is easy and probably more powerful than any other humanly applicable coloring method.
User avatar
SpAce
 
Posts: 2671
Joined: 22 May 2017

Re: A new promising sudoku technique: The Slot Machine

Postby StrmCkr » Wed Jul 03, 2019 4:36 am

I did mean it actually, from the links I sent you in the past regarding the direction muti colour was advancing befor it was renendered obsolete but its premises was their to be equivalent to gem and 3d MEd.
which is the definitions for. Muti colouring I use as it could use muti digits. But I don't have implemented in my Solver as that remains short chains and over sets.(nor was the advanced styles implemented into any solver that I know)
Some do, some teach, the rest look it up.
stormdoku
User avatar
StrmCkr
 
Posts: 1432
Joined: 05 September 2006

Re: A new promising sudoku technique: The Slot Machine

Postby SpAce » Wed Jul 03, 2019 5:09 pm

StrmCkr wrote:I did mean it actually, from the links I sent you in the past regarding the direction muti colour was advancing befor it was renendered obsolete but its premises was their to be equivalent to gem and 3d MEd.
which is the definitions for. Muti colouring I use as it could use muti digits. But I don't have implemented in my Solver as that remains short chains and over sets.(nor was the advanced styles implemented into any solver that I know)

Yes, I remember those discussions, but my point is still the same. Whatever the prehistoric definitions of those techniques might have been, it's mostly irrelevant today. I value your input in explaining the history of various techniques, and I find it interesting, but when discussing current techniques it just confuses people if obsolete or subjective definitions are used as if they were current and popular.

Most prominent sources are in full agreement about what's Simple Coloring, Multi-Coloring, 3D-Medusa, and GEM -- and they're all distinct and easily recognizable coloring techniques by those definitions. Using those crystal-clear definitions makes communication much easier because then there's no ambiguity about what is actually meant. It's just about keeping things simple for current users, even if the history is more complex.

Of course the same idea of multi-coloring can be used for multiple digits as well, but then it shouldn't be called just Multi-Coloring (which is a named and well-known single-digit technique). Similarly, even if non-conjugate links were at some point discussed in the context of 3D Medusa, they shouldn't be included under that name now (as they're covered by GEM, or similar extensions). Names stop having any meaning if they're mixed at will to mean anything.
User avatar
SpAce
 
Posts: 2671
Joined: 22 May 2017

Re: A new promising sudoku technique: The Slot Machine

Postby StrmCkr » Wed Jul 03, 2019 8:52 pm

Fair enough, seeing how it was written about never wide spread I see your point
Some do, some teach, the rest look it up.
stormdoku
User avatar
StrmCkr
 
Posts: 1432
Joined: 05 September 2006

Re: A new promising sudoku technique: The Slot Machine

Postby SpAce » Mon Jul 15, 2019 7:27 pm

SpAce wrote:That being said, multi-coloring is obsolete as far as I'm concerned.

I've reconsidered that position, and am not quite so sure anymore. I've now tried it a couple of times (single-digit multi-coloring, multi-digit multi-coloring example 1, and example 2), and it seems kind of fun for a change. For most practical purposes I'd still use GEM in manual solving, because it's yet more powerful and much simpler to use, but I think multi-coloring is a bit more elegant. GEM produces more straight-forward results, which makes it easy but also quite mechanical.

StrmCkr wrote:(nor was the advanced styles implemented into any solver that I know)

What about champagne's Full Tagging method? From the little I understand of it, it seems somewhat similar to 3D multi-coloring (though extending to ALSs and other almost patterns too, just like GEM). As far as I know, it's successfully implemented in at least champagne's own solver as well as Cenoman's, and produces very impressive results. It just doesn't seem nearly as applicable to manual solving as GEM.
User avatar
SpAce
 
Posts: 2671
Joined: 22 May 2017

Previous

Return to Advanced solving techniques