The New Sudoku Players' Forum

[tso]

I was at the bookstore yesterday and was appalled by the spate of Sudoku books, each of which offering nothing but a bunch of puzzles (typically about 100, but as many as 240), the answers, and a few pages that giving the least possible information to assuage they're guilt for cashing in without adding *anything* to the subject. If the do contain any information, it's usually incomplete, inaccurate or plain wrong.

Of course, regardless of how they're rated, they're nearly all fairly easy, no more than a Pappocom "HARD". The majority are not hand made, just spit out by a computer program. Some of the books are "Official", in the same sense that I am a "Martian". Will Shortz, the up-until-know respected puzzle expert has shamelessly lent his name to several books on Sudoku -- as a man who specializes in WORD puzzles, why shouldn't he? He calls them "wordless crossword puzzles". Pathetic.

Ok, I know people are gonna cash in and no one cares about truth and accuracy in actually important subjects, so it would be silly to expect it in our silly little puzzle world. I'm not who they're marketing to anyway.

Still, it would be nice if some of the books offered *anything* other than printouts of 100 random easy puzzles.

How about:

1) Include at least some actually hard puzzles, up to and including those that most humans cannot cannot crack.

2) Instead of just showing the solved grids in the back, give step-by-step *solutions*, in the same way Dell magazines do for their Logic problems, at least for the harder puzzles.

3) There are dozens of variations. ALL of the Japanese puzzle magazines include many variations in each monthly issue. (They also often contain discussion on advanced solving and creation techniques.) Include them. But don't pretend you invented them unless you actually did.

4) Stop blowing smoke up our tushes by labeling easy puzzles as hard, and anything that actually requires thinking as "diabolical" "killer" "fiendish" "impossible" -- especially while denying that much harder puzzles exist. Where did this start? The crossword world never seemed to do this. They use ratings like "challenger" or "expert" simply "hard" and are usually objectively tied to what sort of answers and clues are used. Do logic problem solvers have self-esteem issues so that we need to be pumped up?

5) Add at least one thing new to the mix. Something so don't we feel that anyone could have cranked out this book in her lunchtime.


Maybe I shouldn't be so harsh. They're all going to lose money on this anyway. They see the trend going up, so they keep increasing their stock of vanilla books. The public will become bored with the same-ole-same-ole. The trend will crest and quickly drop and they'll be stuck with a warehouse full of unsellable books, wiping out their ill-gotten windfall. They'll try to make up their loss by putting out more books that might follow the 5 suggestions above, but it will be too late.

[PaulIQ164]

The thing is, you have to accept that people who like to solve sudokus using fishes and Nishes and colours and chains and what-have-you are in the vast minority. The rest of us are perfectly happy with going only up to 'fiendish' (which I certainly found very challenging at first, and even now they're hardly easy or trivial or anything). It's true that it would be nice if there was a little more to these books, like at least one step-by-step solution to a tough puzzle (I mean, who looks at the solutions anyway? You can tell if you've got it right easily enough), but I don't think the books can stand to get any harder than they currently are really (having said that, an 'all-fiendish' book wouldn't go amiss).

At the end of the day, most people don't want puzzles of the difficulty you describe, and they're happy to buy the books. People who do want puzzles of that difficulty, I'd suggest are likely to know where to find them without buying a book, so would be unlikely to buy one even if there were one. So where's the problem?

[lunababy_moonchild]

Yes, I've seen them too but I didn't bother to look at them, other than in passing. It's like pulp fiction in a way.

I did check out Carol Vorderman's but she didn't have anything to tell me I didn't already know - which disappointed me because my solving skills are not that advanced, imho. Maybe I was expecting too much, I honestly thought she'd be way more advanced than Fiendish.

I think that the British publishing houses have a long way to go to catch up on the Japanese - as I understand it they've been doing this for years - and if this Sudoku craze lasts will have to come up with something more entertaining, especially as more people get more advanced.

Luna

[PaulIQ164]

By the by, since I've never really tried any puzzles any harder than Pappocom ones, can I ask: do the people here who do the really hard puzzles do them in a way that could be solved in a book in any case, that is, without candidate filtering, without more working out than you can write in a grid and around its borders (preferably that is without using an eraser either), without computer help, etc.? I ask because if you don't, that part of the discussion is surely fairly academic, and if you do, gosh, you must be clever.

[Sue De Coq]

Perhaps these books will start to include Sudoku variants as soon as free software to generate such variants appears on the web ...

[tso]

Perhaps these books will start to include Sudoku variants as soon as free software to generate such variants appears on the web ...

Quite a few variants here.

By the by, since I've never really tried any puzzles any harder than Pappocom ones, can I ask: do the people here who do the really hard puzzles do them in a way that could be solved in a book in any case, that is, without candidate filtering, without more working out than you can write in a grid and around its borders (preferably that is without using an eraser either), without computer help, etc.? I ask because if you don't, that part of the discussion is surely fairly academic, and if you do, gosh, you must be clever.

I do all my solving on paper, though I often print grids very large if the puzzle is very-hard. In some cases, if I know in advance the puzzle is ultra hard, I'll use Simple Sudoku to figure out the candidates before printing to save time and avoid clerical errors. This could easily be addressed in book format by printing the harder puzzles larger. As it is, most books print one to a page with lots of white space. They could even choose to print ultra-hard challengers with candidate lists. It's not without precedent. It has long been standard practice to publish word based logic problems with a truth value grid, and it's much more difficult to solve them without one. For example, look here. Sometimes the solver must create the grid, and if the puzzle is particularly tricky, this can be the most difficult part. Word based logic problems are regularly published that are far more difficult -- measured in the time it takes a reasonably experienced solver to finish (not the best way to measure difficulty, but for comparison purposes...) -- than all but a handful of Sudokus that are published in books and newspapers. (The same can be said for Crosswords.)

[Karyobin]

vast minority

I'm having a bit of trouble conceptualising a vast minority. Unless the majority was incalculably more vast, of course.

[PaulIQ164]

Picky.

[QBasicMac]

Actually, I plan to incorporate "vast minority" into my vocabulary. It is priceless!!

Yep, I was in Europe with no computer, wishing for some puzzles to solve. I found "The Times" Sudoku book 2 for 99.95 Danish Kroner. That's about $16. As was said, it was simply a list of puzzles. I hardly had to use the margins until "difficult", which were still pretty easy.

Well, the community seems to be dividing into those who are facinated by something new who will pay $16 for anything, on the one hand, and really nerd jargon-spouting enthusiasts on the other.

I personally like puzzles that require SimpleSudoku (or my equivalent I wrote in VB) but I also like the ones you can whip out without any notes of any kind.

Mac

[Karyobin]

and really nerd jargon-spouting enthusiasts on the other

...and therein lies the reason behind my post in the 'Killer' thread.

[tso]

Just came back from another bookstore -- over a dozen Sudoku books, half of which I hadn't seen before. Each one contains it's own special selection of bogus information, ridiculous hyperbole and unique set of inapproriate ratings all designed to funnel money into the bank accounts of white men with nothing new to add -- the "vast minority" of which contain even one puzzle that requires anything but the simplest tactics.

For example, Pete Sinden’s books:

"All puzzles are self-created via the application of computer modelling and artificial intelligence and designed solely to create pure brain-teasing Sudoku for the pleasure of humans."

Translation -- "My computer spit these out at random in 12 seconds."


Several books by different "authors" (In what sense are these people "authors"?) make the ridiculous claim that "Sudoku aren't really from Japan, they were invented by Euler..." Again, Euler invented Greaco-Latin Squares (aka Euler Squares), which were based on Latin Squares -- which predate his birth by centuries. There is NO connection between Euler and Sudoku or Number Place. I'm beginning to think there is a racial element -- we want to usurp the Japanese name for the hipness value, but don't want to give them any credit -- no, we'll give that to the dead white guy. Euler is to Sudoku what the Wright Brothers are to Motorcycles.

It's *possible* to publish a book of Sudoku that adds something to the mix. Argus Johnson's puzzle packs are sorted specifically by the type of patterns required to solve -- rather than "Very Very Easy", "Very Easy", "Easy", "Slightly Not-Easy". If he put out a book with 200 hundred content specific puzzles with several chapters on advanced solving, I'd buy it. If someone put out a book that in which 50% of the puzzles were variants, I'd buy that -- I've got a stack of Japanese magazines for that reason. But either of these would require actual thought. The editor would have to actually look at the manuscript. They couldn't just crank them out at will.

Meanwhile, DELL Magazines, who introduced the Number Place puzzles on which Sudoku were based, has come out with a Sudoku magazine of it own. Only $3 USD or 1.63 pounds for 184 puzzles. Or six issues a year for 15 bucks USD. Almost cheap enough to buy a copy. Still no variants, and probably no difficult puzzles -- but not one hyperbole, no goofy lies about the Euler or Japan, no crazy ratings. They rate their puzzles "easy", "medium", "hard" and "challenger".

http://dellmagazines.com/cgi-bin/detail.pl?SiteID=2&ProductID=1051

[dukuso]

> Just came back from another bookstore -- over a dozen Sudoku
> books, half of which I hadn't seen before. Each one contains
> it's own special selection of bogus information, ridiculous
> hyperbole and unique set of inapproriate ratings all designed
> to funnel money into the bank accounts of white men with
> nothing new to add -- the "vast minority" of which contain
> even one puzzle that requires anything but the simplest
> tactics.

do they claim copyright for the sudokus in the book ?

> For example, Pete Sinden’s books:
>
> "All puzzles are self-created via the application of computer

can a sudoku be "created" as an artwork ?
I'd say, they can only be discovered like other math-entities,
e.g. large primes etc. You won't say these were "created".
I assume the reason is that they want to claim copyright
on sudokus for which "creative work" is necessary in most
countries, while discoveries can't be copyrighted.

> modelling and artificial intelligence and designed solely to
> create pure brain-teasing Sudoku for the pleasure of humans."
>
> Translation -- "My computer spit these out at random in 12
> seconds."

150 per second, currently. With public domain software.
But this can be further optimized.

> Several books by different "authors" (In what sense are these
> people "authors"?) make the ridiculous claim that "Sudoku
> aren't really from Japan, they were invented by Euler..."
> Again, Euler invented Greaco-Latin Squares (aka Euler
> Squares), which were based on Latin Squares -- which predate
> his birth by centuries. There is NO connection between Euler
> and Sudoku or Number Place.

I thought we were through this already. Euler contributed to
latin squares considerably. Probably more than anyone else
before him and centuries after him. Sudokus are very similar
to the problem of completing latin squares with holes.
So there is clearly SOME connection.

> I'm beginning to think there is a
> racial element -- we want to usurp the Japanese name for the
> hipness value, but don't want to give them any credit --

it's generally acknowledged that Nikoli contributed considerably
to discovering sudokus. If "we" wanted to usurp sudoku, we would
give them a different name. So I don't know what you mean here.

> no, we'll give that to the dead white guy. Euler is to Sudoku what
> the Wright Brothers are to Motorcycles.

I don't know about the latter, so I withhold a comment here ;-)
Just wondering, which non-white guy you want to be included
in the history of Sudoku.

> It's *possible* to publish a book of Sudoku that adds
> something to the mix. Argus Johnson's puzzle packs are sorted
> specifically by the type of patterns required to solve --
> rather than "Very Very Easy", "Very Easy", "Easy", "Slightly
> Not-Easy". If he put out a book with 200 hundred content
> specific puzzles with several chapters on advanced solving,
> I'd buy it.

sorting sudokus by difficulty level is even easier (=faster)
for computers than generating them.
Why do you want sudokus printed in book-form, when you can
get them here in internet in computer-readable (and printable) form ?

> If someone put out a book that in which 50% of the
> puzzles were variants, I'd buy that -- I've got a stack of

create one ! It's easy. The variants are not copyright, as I see it.
Collect what you already have, what you find on the web, ask here
and collect what people here might contribute to future new variants ,
and send it to a publisher. And do it, before someone less competent
does.

> Japanese magazines for that reason. But either of these would
> require actual thought. The editor would have to actually look
> at the manuscript. They couldn't just crank them out at will.
>
> Meanwhile, DELL Magazines, who introduced the Number Place
> puzzles on which Sudoku were based, has come out with a Sudoku
> magazine of it own. Only $3 USD or 1.63 pounds for 184
> puzzles. Or six issues a year for 15 bucks USD. Almost cheap
> enough to buy a copy. Still no variants, and probably no
> difficult puzzles -- but not one hyperbole, no goofy lies
> about the Euler or Japan, no crazy ratings. They rate their
> puzzles "easy", "medium", "hard" and "challenger".
>
> http://dellmagazines.com/cgi-bin/detail.pl?SiteID=2&ProductID=1051

[coloin]

I would think that a publisher would pay for the initial publication of the most difficult and prized puzzle of all - a 16er - if it exists, of course.

Guenters analysis of the ways of solving many of the 17ers perhaps would indicate that a 16er [if it exists] would be a "humdinger"

Im still working on it !

Colin

[dukuso]

I would think that a publisher would pay for the initial publication of the most difficult and prized puzzle of all - a 16er - if it exists, of course.

Guenters analysis of the ways of solving many of the 17ers perhaps would indicate that a 16er [if it exists] would be a "humdinger"

Im still working on it !

Colin

I won't support your work, if I knew you intend to keep
the result secret or claim copyright on that sudoku.
Just in principle, not that I'd assume there were a 16 around.

-Guenter

[PaulIQ164]

"vast minority"

You have a funny way of going on about that for someone with a thread about a "variety of variants".