wu :: forums - Print Page


    
      
        wu :: forums
        (http://www.ocf.berkeley.edu/~wwu/cgi-bin/yabb/YaBB.cgi)
      

        riddles >> easy >> ESP Puzzles
        
(Message started by: THUDandBLUNDER on Apr 2^nd, 2003, 11:02am)

Title: ESP Puzzles
Post by THUDandBLUNDER on Apr 2^nd, 2003, 11:02am

Here's a couple of ESP puzzles:

http://sprott.physics.wisc.edu/pickover/esp.html

http://mr-31238.mr.valuehost.co.uk/assets/Flash/psychic.swf

And here's something else which is not a trick:

http://intelligence.sergi5.com/

PS Sorry, perhaps these should be in the Easy category.

Title: Re: ESP Puzzles
Post by aero_guy on Apr 2^nd, 2003, 11:18am

number one: [hide]pick your card, then take a look at the others before you click[/hide]

The sick thing is he has a page with a hundred or so people saying how amazed they are and giving explanations like that they think the monitor can read their eye movements. Sad.

I couldn't get the second link to work.

The third one is interesting if poorly coded. I got 50.98% after 254 guesses. You can intuit what patterns it is ascribing to your behavior and then counter them. I have heard of similar coding contests. There is one where you develop a code to win at rock-paper-scissors. If your code is able to generate truly random guesses you will win 50% of the time, but if you can figure out how the other code is generating answers you can come close to 100%. One of the recent winners used, I think, 7th level Cicilian logic (think Princess bride). Since I was able to get about 50%, this code sucks.

Title: Re: ESP Puzzles
Post by towr on Apr 2^nd, 2003, 11:24am

The first two are very easy.

The second is tricky.. It doesn't use tricks like the first two, but it does use a 'trick' so to speak.. I could probably make a program that did the same thing, or perhaps even one that could beat his program.
(It's similar to a rock-paper-scissors/roshambo competition, random play gives 50%, a strategy can give higher results if the other also uses a strategy)

Title: Re: ESP Puzzles
Post by aero_guy on Apr 2^nd, 2003, 11:44am

I think I updated mine while you were writing yours towr.

Title: Re: ESP Puzzles
Post by towr on Apr 2^nd, 2003, 12:22pm

I think I submitted at ?:24 and oyu were finished editing at ?:35

Anyway.. if you have two programs running against each other, and one is using random guesses you can't get much better than 50%, and the only by chance..
The only way to really beat another competitor is if he uses a strategy you recognize and can beat.
And once you have a significant lead, you can use random guesses to keep on top (though I think it is disallowed to use a random guessing program)

In any case, 50% might still be good.

I lost once, and beat the program once.. The second time I tricked it early on, then used a random number generator to keep on top. :P

Title: Re: ESP Puzzles
Post by wowbagger on Apr 2^nd, 2003, 12:22pm

on 04/02/03 at 11:24:44, towr wrote:

The second is tricky.. It doesn't use tricks like the first two, but it does use a 'trick' so to speak..

Hm, what exactly do you mean by "the second [...] doesn't use tricks like the first two"?
I guess you mean the second link. It took me a little while to see the pattern, so to speak.

Or maybe you mean the third link. Well, I won't spend my time figuring out how that one works.

Title: Re: ESP Puzzles
Post by towr on Apr 2^nd, 2003, 12:24pm

hmm.. yes, I meant third..

the second is easy once you see the pattern, think [hide]multiple of 9[/hide]..

Title: Re: ESP Puzzles
Post by THUDandBLUNDER on Apr 2^nd, 2003, 6:55pm

For the predictive algorithm, if you input the first 250 binary digits of pi it scores no more than 48%.

Title: Re: ESP Puzzles
Post by Jeremiah Smith on Apr 3^rd, 2003, 12:11pm

http://www.snopes.com/humor/iftrue/takecard.htm

http://www.snopes.com/humor/iftrue/psychic.htm

Title: Re: ESP Puzzles
Post by Pancake on Apr 7^th, 2003, 10:34am

Funny, I tried the very first one but it doesn't seem to work at all. It has /never/ removed the card I chose.

Title: Re: ESP Puzzles
Post by Pancake on Apr 7^th, 2003, 10:46am

Is that what the entire trick is? Always remove "Black Jack" and watch the person waste 45 minutes trying to figure out why it doesn't work?

Title: Re: ESP Puzzles
Post by Kitty on Apr 7^th, 2003, 10:52am

Huh???????? Wat r u going on about pancake?

Title: Re: ESP Puzzles
Post by Pancake on Apr 7^th, 2003, 11:07am

I don't know. I've tried Cliff's experiment at two websites with my web proxy and without it, yet it always removes the same card.

Title: Re: ESP Puzzles
Post by Pancake on Apr 7^th, 2003, 11:11am

Wow. That trick really sucks. I've got it now. I just wasn't paying attention to the shape at all.

Title: Re: ESP Puzzles
Post by aero_guy on Apr 8^th, 2003, 9:50am

What, you were picking the black jack rather than the jack of clubs for example?

Title: Re: ESP Puzzles
Post by Jeremiah Smith on Apr 9^th, 2003, 11:58pm

[hide]The way the "remove your card" trick works, as explained at www .snopes.com, is that NONE of the cards in the first set of six appear in the second set of five; only cards that appear similar. Since people will only concentrate on remembering the one card they chose, they don't notice that the other cards from the first set aren't there either.[/hide]

By the way, hide tags won't hide imbedded links...

Title: Re: ESP Puzzles
Post by aero_guy on Apr 10^th, 2003, 1:46am

You can, with a little work, make them (http://www.ocf.berkeley.edu/~wwu/cgi-bin/yabb/YaBB.cgi?board=riddles_suggestions;action=display;num=1043138466).

Title: Re: ESP Puzzles
Post by sergi on Aug 1^st, 2003, 4:29am

hello world
about http://intelligence.sergi5.com/
the original idea was to beat the human brain random generator and not another program. as you can see in the statistic page, the delay between each key is important and if typed very quickly, scores are very high.

Title: Re: ESP Puzzles
Post by wowbagger on Aug 1^st, 2003, 5:43am

Er, how is that supposed to work?
I tried typing 0's and 1's in the long input field, but it just says "You have no tryed [sic] to play!" and "You have not typed enough keys!".
Am I supposed to type the numbers one-by-one? And is there another way to play than to click the "view..." button?
Oh, wait. Is this some queer javascript stuff? Then it's probably my browser again...

I think there's room for improvement regarding the presentation. :)

Title: Re: ESP Puzzles
Post by wowbagger on Aug 1^st, 2003, 5:47am

Wow! Upon closing the window, I got this in a new window:

Quote:

You have played during:1059740955 seconds
Number of key pressed:9
Average time by key pressed:117748995.0000000000seconds
The computer have predicted:undefined% of your thoughts.

I-win-gif

That's the way to guarantee a win: Consider undefined as a win and don't define anything! :D

Title: Re: ESP Puzzles
Post by James Fingas on Aug 1^st, 2003, 9:25am

I sent Cliff this email:
Your ESP experiment is tantalizing! I was so astounded the first 6 times I tried it that I had to figure out how it worked. So I tried the experiment again this way: open the page with the five cards and take a screenshot (without looking at the cards at all). Now open the next page BEFORE DECIDING ON A CARD. Now look at the screenshot and pick one of the five cards, and it has been removed! To be even more sure, try picking another one, and IT HAS BEEN REMOVED TOO. No matter which of the five you pick, THEY HAVE ALL BEEN REMOVED AS SOON AS YOU PICK THEM! But I know how you do it. Because of the law of causality, your computer can't predict my actions, but YOU WAIT UNTIL THE SECOND PAGE IS DISPLAYED TO COLLAPSE THE QUANTUM STATE OF THE OBSERVER/OBSERVATION! -- James Fingas

Title: Re: ESP Puzzles
Post by James Fingas on Aug 1^st, 2003, 9:37am

I beat the "intelligent" computer without using a random number generator. You have to play for at least 205 digits. You just have to have a good idea of what it thinks you're going to do. That being said, I didn't beat it by much... I wonder how hard it would be to write a computer program that could beat it by a large margin?

Title: Re: ESP Puzzles
Post by BNC on Aug 5^th, 2003, 12:09am

ROFLMAO ! :D

on 08/01/03 at 09:25:26, James Fingas wrote:

Title: Re: ESP Puzzles
Post by Sir Col on Aug 20^th, 2003, 4:15am

May I invite people to try out a 'mind reading' experiment that I put together:
http://mathschallenge.net/index.php?section=rec&ref=general/experiment_2&type=general

Title: Re: ESP Puzzles
Post by wowbagger on Aug 20^th, 2003, 4:28am

on 08/20/03 at 04:15:12, Sir Col wrote:

May I invite people to try out a 'mind reading' experiment that I put together:
http://mathschallenge.net/index.php?section=rec&ref=general/experiment_2&type=general

That's nice.
I particularly like the "concentrate hard" part. ;)

Title: Re: ESP Puzzles
Post by towr on Aug 20^th, 2003, 5:16am

If a year had 13 months it wouldn't always work..
And as I recall for some cultures it does..

Title: Re: ESP Puzzles
Post by wowbagger on Aug 20^th, 2003, 5:27am

You're right, towr. Looks like Sir Col's mental abilities have a cultural upper bound. ;)
The idea could of course be extended. However, it wouldn't be good style to claim to read minds and nevertheless ask how many months there are in the subject's calendar before reading their mind.

Title: Re: ESP Puzzles
Post by towr on Aug 20^th, 2003, 8:03am

it's also not really good style to claim you can read peoples mind, then ask them to give you certain information that clearly encodes what they're trying to 'read' :P
(If it were lossy encoding that'd be one thing, but in any encoding used in such a trick you just know it's not lossy)