Monday, 7 September 2009

(c) Sophie

Maybe yes, Maybe no
Maybe I don't know.

Somehow I thought this was wondrous. Sophie came up with this today.


Class Day Out but it won't count


Class Day is one of the very hardest patiences, and I just got it out. But I did it by replaying my most successful game and Solsuite will not count it for online statistics. Grrr. I think their policy is right. But ... Grrrr.

Any tips? Obviously be very lucky. But beyond that, there were a couple of times I avoided taking a card from a tableau, if the same card was coming up soon in the stock. (Yes, I cheated by looking ahead.) And wait to clear a space until you are happy with what will go in this pile. For example, when a Q is top of the wastepile can be a good choice towards the end, so that it doesn't matter if it gets buried.

By the way, on the redeal, I discovered I had missed the JD the first time round. Which meant that a 10D had not been played early. But this didn't hurt it coming out, and may have even been important to its success.

The game that comes out is 3105127855914.

Saturday, 5 September 2009

I got Idle Aces Out

I got Idle Aces out in Solsuite. This is another one of the hardest games, or at least the least solved, or at least the least solved by people who publish their results on line. The game number I got out is 3021478448855.
This was only my 41st game, so it is indeed not as hard as others (like the awful Japanese Patience.) I used to hate this patience but I think I was being too clever. You can switch between piles you have built up. E.g. if you have got a rising pile up to 7 and a descending pile down to 8, you can switch cards between them to any number you want. But doing this more than a little seems to be counterproductive in the long run.
I did restart a few times until I had a good start. But the good start is more important for the second deal than the first. The deal started something like A A K, which is important because on the redeal it's vital that the first card is a terminal card. So even though I could have played the aces, don't! Then you can use a little bit of cleverness. On the redeal, say I got Q underneath the K, I don't need to play it. Then I perhaps get a J and that needn't be played either. Because eventually the J can be played, then the Q, then the K.
Anyway, I got it out. Don't need to play it again unless I want to.

Friday, 4 September 2009

Taking the credit



One of my students got me last night. We were watching The Spiral Staircase


This is a thriller in which a professor is a suspect. Somebody said that it must be the professor whodunnit. Lars said "No. The professor will take the credit but somebody else will have done all the work."

How to play Japanese Solitaire

Don't.

But if you have to, here are some tips.

Japanese Solitaire is a patience game invented by and released by the Solsuite program. Hilariously, they give the odds as being 1 in 15, but I got it out on my 1400th try. As I write it's the sixth most difficult game of hundreds in the program. Only 12 people have solved it even once and published their result online. That number is now 13, as I got it out last night.

Most of those 1400 tries lasted a second or two. That's one of my tips. Bail out quickly if you need to. Here are my other tips.

  1. You need to be very lucky all the way through, so give up if you are not. If you are not meeting the conditions, feel free to bail out and get another game going.
  2. Look to get about 3 cards out in the first few rows. And ones which don't block you in the future. E.g. if you get two aces out but you have buried one of their twos, don't bother.
  3. At the end of the first deal (i.e. 0 out of 4 redeals) you should have a positive score.
  4. By the end of the third deal (2 of 4 redeals) my cards had crossed over, i.e. the cards going up had hit the same number as the card going down.
  5. I "cheated" on this ghastly game to avoid playing 1400 more. What I mean is, I did a few undos to randomise the next redeal. When I ran through the first time, it didn't come out. Undoing and then choosing NOT to put a card up which would in fact go seems to randomise the next redeal. So even though it's to your disadvantage, the gain can outweigh the pain. The second or third time I did that, it worked.
If you have Solsuite and you want to get this dratted patience out, here's the game number which I know works. 1550537509600. Glad to be of service. (I think Solsuite doesn't let you publish statistics on games you put in the number for, so playing that game won't get you on the leaderboard, sorry.)


Thursday, 3 September 2009

How many trips to the post office does it take to get a passport?

Trips to post office to get a passport for 1 (one) daughter.

1) Pick up form.

Oops, daughter spills sauce on form.

2) Pick up form.

Oops, form is not available from rural post office.

3) Pick up form from St Andrews. In stroke of genius, pick up two copies against repeat of spillage.

4) Wife picks up yet another form from another post office, though technically not really necessary I suppose.

Ok, now get respected colleague with right letters before his name to fill in form and sign photo.

5) Do check and send service at Post Office in St Andrews.

Oops, respected colleague signed outside the area covered by the photo. Yes indeed, the area covered by the photo on the other side of the paper which you can't see. However check and send is civilised. You only pay once and you can go back when you have fixed the problem. So get respected colleague to sign again on the other photo.

6) Do check and send in Cupar.

Oops, second check has to be at the same post office as the first. Why why why why? Ok I'll pay again.

Oops, unsigned photo has to be completely blank, not with remnants of colleague's signature on it.

Scream. Write post.

7) Mythical future trip where I get passport applied for.

Wednesday, 2 September 2009

99.999%

The chance that a random game of Freecell (one of the patience/solitaire games that comes with Windows) is solvable is about 99.999%.
It always gives me great pleasure, surely more than it merits, that this number arises for real. It's not made up or hyperbole. An extensive experimental investigation of a million deals had only 14 not solvable. To five significant digits, that's 99.999%.
99.999% of the time that you come across the number 99.999%, somebody has just said it for effect. But 0.001% of the time, it's the result of real study to find the number. And this is that one in a million time. Ok, one in a hundred thousand time.