Listener 4075 - Square bashing by Arden

[Philoctetes]

Hoorah - numbers. Again one must marvel at the construction. There are some fairly obvious approaches, but I have not ventured into the hinterland much. I do hope the required clues give a satisfactory payoff.

[bobbycollins]

Boohoo - numbers

See you all next week, lads & lasses

bc

[midazolam]

This is one of those that you think you have got going and then hit another brick wall, it doesn't help when you go TITs up and make the wrong choice and then have to backtrack (but then I do prefer the "normal" weeks)

[AHearer]

I think this particular game of two halves is one I'll leave alone. I never find the rewards of the final revelation on the numerical puzzles are worth the agony.

[Ruthrobin]

Yes, I am with you all - this is total misery. For me a crossword should be just that and not a cross number (which is what it turns me into!) See you in a week's time.

[cluelessJoe]

Agreed bc, ah & rr ... I've got painting to do, and for once that seems to be the more attractive prospect. I'll maybe have a go at this a bit later when the only alternative is watching paint dry.

[Clamzy]

Somebody's got to support philoctetes so I arrive from the opposite direction. However I do have a query to raise. This puzzle follows a recent trend among numericals in that the clues comprise real words. I am suitably agog at the skill required of the setter to produce such a feat, but what's in it for the solver?

[Mysterons]

I'm with you Philoctetes, this one was great fun. You might find this site useful for working on 1 across:

http://www.naturalnumbers.org/psquares.html

[emcee]

Have solved it all; all looks very plausible and have both words at the bottom BUT is there a typo in the clue for 1 across?

[emcee]

Ignore my last message. Forgot to square it! DUH.

[dr b]

I was able to deduce the 2 digit grid entries and am pretty confident in some of the three digit ones, but my "logic problem" approach is failing as the grid entries get longer - a 6 digit grid entry has a clue answer somewhere between 317 and 999, which is too wide a range to narrow things down too much (knowing that a certain letter can have a value between, say 8 and 20 is not much help).

So my next tack to was set up the whole puzzle in a spreadsheet, so that entering 20 number/letter combinations will show the grid entries for all the clues. I think I have it down to about 5 or 6 letters that can take on any number of values, with the rest fixed or nearly so. Unfortunately there appear to be dozens of potential solutions (in the sense that the solution provides all the squared grid entries with the correct length). Most of these solutions are infeasible as they have 1A with a zero somewhere, which obviously can't work. None of the others I've come up with provide a 1A that will cross properly with the 1 thru 6 downs. So brute force is not working either.

Just as a check that I am not barking up the wrong tree entirely, are others finding 1a to be a negative number?

Must grade papers so I will try not to look at this for the rest of the day....fat chance....

[emcee]

dr b,

I had 1ac < 0 and then remembered it needed to be squared!

[Mysterons]

Scanning the tables given in the final dowload of the above link can help with possibilities for 1 across, once you have the possible options for about half of the intersecting downs. Another useful link for later on in the puzzle is:

http://www.markhorrell.com/tools/factors.asp

[AHearer]

Well, it was raining here today, so despite what I said earlier I kept picking at it and (those of a sensitive disposition look away now) despite what my mother told me it did get better. Overall the construction was very clever, especially the final resolution.

[dr b]

Eureka. That was marginally more fun than grading papers..... now what else can I waste my time with?? Only 5 hours till the EV!

[midazolam]

dr b if you would like a copy of todays IQ (by Dysart) email me at [email protected] (and anyone else who wants to get back to normality)

[bonefixer]

I always manage to solve these (they're the only listeners I attempt) right up until the last bit where you've got to put the final words in.

I've got the grid filled (and I don't think 1ac is a squared negative number, mine was positive) but I can't make a word on the right half of 23ac. Is it something obscure I need a big dictionary for?

[bonefixer]

Thanks Midazolam. I actually tried this after thinking about it (presumably while sleeping) before reading your post and am now amazed that simply switching the only 2 ambiguous letters switches the whole grid around. Before I got the calculator out and had my 'Eureka' moment I thought the switch would give me the same grid by simply negating 1 across and giving the same answer for its square. That was amazing enough. When the grid worked last night I didn't even try the negative possibility as I didn't even consider that there'd be two options for a 12 digit number to fit.

There's probably some really deep number theory in here somewhere - perhaps I'll investigate later. Hats off to the setter though.

[starwalker]

I have always disliked the number puzzles, but a few years ago I forced myself to tackle them. Despite getting the majority of them correct and even winning a prize once, I still have no love for them and sadly this puzzle has not changed that. Unless I missed something, it seems to be too much of a slog using trial and error.
Certainly not one to encourage any newcomers.

[tristram37]

I really don't know why I submit myself to this torture every three months. I'm no mathematician and the puzzles take hours and hours. Having spent most of yesterday on it I have discovered the letter which equates ro 1. Pehaps I will have discovered 2 by Easter!
Tristram37