Midlife Crisis? What Midlife Crisis?

Originally posted by mineriva
Kg2 looks ok.

After black moves the Queen White takes the rook. I guess.

Looks like the move to me.

There's some nice problems here http://www.chess.com/blog/NimzoRoy/miscellaneous-endgame-problems if you like that sort of thing. I think endgame problems are the most interesting.

That is their solution.

My first thought was take the Q.
Then I noticed that the K is in check.

Then I realized that there is no legal move for black just prior to this that could lead to this situation where it is white's move.

Or is there? I'm not exactly a great chess player and could easily be overlooking something, but I'm stumped.

Oh for heaven's sake -- I guess Bg2 would qualify as a legal move by black to arrive at this position -- but only if it was at f1. Sorry for interrupting your blog -- and thanks for the puzzle link.

Cheers,
VS

Originally posted by VorpalF2F
Sorry for interrupting your blog

Shame on you for interrupting the seamless, carefully prepared flow of ideas that is my blog.

Last episode of Doctor Who tonight! Ooooooooo!

Then I'm going to have another crack at epsilon exploitability. I must be a able to code it up. I'm meant to be a maths professor ffs.

Originally posted by jbpatzer
Then I'm going to have another crack at epsilon exploitability. I must be a able to code it up. I'm meant to be a maths professor ffs.

Did it! Did it! Hahahahahahaha!

Feeling slightly hysterical.

Congratulations! ♠_thumbsup:

Originally posted by jules97
Congratulations! ♠_thumbsup:

Thanks v much.

I now have a working code that calculates how much player two can exploit player one. I still have to modify it so that I can calculate how much player one can exploit player two which, irritatingly, doesn't just involve swapping ones for twos because of the way the data's stored, but it's not too hard. Then I have to translate it from MATLAB into C++, which will no doubt be harder than I expect. After that, I can try to use it to determine how many iterations I need for a good approximation to equilibrium, and how exploitable my slightly lossy bucketting algorithm is. So far I've looked at the exploitability of solutions in games with small trees and ranges and, for well-converged solutions, it's always been less than one thousandth of the pot. I wasn't quite sure what units to use for this. All the literature on Limit Holdem uses a thousandth of a small bet, but for what I'm doing the only sensible scale for bet size is the pot, so thousandths of a pot it is.

I too know pain. Spent yesterday shoveling dirt for my sister's (bridezilla) engagement party today.

When you talk about the exploitability of solutions is that the nash distance?

Originally posted by jules97
When you talk about the exploitability of solutions is that the nash distance?

It is.

Is it just me who thinks this advert is racist? As a citizen of the great country of UK, I dislike foreigners in general, and the French, with their hairy-legged women and great red wine, in particular, but I thought PokerStrategy was meant to be above this sort of thing. Is there an advert on PokerStrategy.fr that has a picture of Dolly the Sheep wearing a bowler hat, drinking warm beer and saying 'come and play with the English at pokerstars.uk'? I certainly hope so.

Nah, it's not racist, just xenophobic.

The French would never stoop to using stereotypes!

Originally posted by EmanuelC16
Nah, it's not racist, just xenophobic.

Don't be ridiculous. I have no problem with xeno.

Originally posted by grbell
The French would never stoop to using stereotypes!

All comforting and familiar as far as I could see, although MacDonalds is shit in any language.

Originally posted by jbpatzer

Originally posted by jules97
When you talk about the exploitability of solutions is that the nash distance?

It is.

You're probably already aware, or maybe it's not relevant.
But GTROB's aim is to get a nash distance of usually less than 1% and always less than 2% of pot. Perhaps 1/1000 seems a bit large for practical purposes?

He's got another brainteaser up if u hadn't notice. This one is a bit boring though imo...

Originally posted by jules97

Originally posted by jbpatzer

Originally posted by jules97
When you talk about the exploitability of solutions is that the nash distance?

It is.

You're probably already aware, or maybe it's not relevant.
But GTROB's aim is to get a nash distance of usually less than 1% and always less than 2% of pot. Perhaps 1/1000 seems a bit large for practical purposes?

He's got another brainteaser up if u hadn't notice. This one is a bit boring though imo...

That's interesting. For the very restricted problems I've tried so far, I can get within 1% of pot in less than 100 iterations. I'll have to see how many are required for realistic situations.

I'll let my code solve the brainteaser, but I won't post a reply this time. Nobody likes a smartarse. When I posted the solution to the other one I mainly did it to get Mr Sutherland's attention, and we are indeed talking on Skype now. He's been very helpful.

And I noticed that the PS 'play with the French' advert now has la tour Eiffel instead of le moustachioed chat. Does that mean Big Brother is watching me, or is it just Vorpal? Do you think I can get the rake reduced on Stars if I say it's racist? ♠_ugly:

I think you'll need more than some racist accusations to get pokerstars attention. If slander reduced rake on pokerstars, they'd be paying us to play there after the last few months.

so i really just posted to spam these new icon things.
what the hell are these even...
??!?

Feeling disoriented now.

At least I don't need to worry about choosing the right avatar. It's done it automatically for me.

That's got to be my favourite one.

Since JB doesn't reply on Skype, I had to hop in here ... so what's up JB?

Forum changed quite a bit, almost looks modern ♦

It's never going to look anything other than old fashioned as long as I'm still here and the option to remains available.