Galaxy SIII, and it does everything.

OK, just making sure. Hmm... I believe the longest expected loss run is actually -log(np)/log(q). This follows from the statements made at http://mathworld.wolfram.com/Run.html.

Edit: I believe I see where the 1 comes from... and yet it doesn't seem to work.
-log(1*1/2 + 1)/log(1/2) is about 0.5849625007211561814537389439478165087598144076924810, not 0.5 as intuition would suggest. Does the formula converge towards "correctness?"

Please define "expected number of losses in a row." Are you using mean, median, or mode; are you ignoring wins in a row, etc.

I just think it would make the game more "honest," and thus more worthy of study. Maybe one of these weekends I'll make a website for it.

We should introduce BTC into this game. Perhaps a buy-in which is distributed proportionally?
However, I suspect that adding real money to the game will make people more conservative and likely to chose "kill." Therefore I think it should be done like so:

	Kill	Trade
Kill	(-, -)	(+, -)
Trade	(-, +)	(+, +)

For some positive and negative values (which will obviously influence the outcome).

In the case of at least one person having a net negative, everyone's points can be increased by the absolute value of the lowest score:

Player	Raw	Scaled	Payout
A	-1	2	14.3%
B	-3	0	0%
C	2	5	35.7%
D	4	7	50%
Total	2	14	100%

I also have a copy of Yahoo Exploit.exe if someone wants it. It's a 1,030,693 byte file.

I was new once. I have no problem with new people.

Again, I'm "railing on" you because you are trying to "buy trust." That's not how it works.

Oh sure, feel free to give out money. That's welcome. It's the "I gave you 13 bucks so trust me plz" thing that has me worried.

There's no money at risk at all. Also, the negative trust is not a risk either, because your account is so new that you can just throw it away.

"Trust":

• The winner gives you feedback - they'll put 0.0 in "risked BTC"
• The winner adds you to their trust list - I wouldn't trust anyone who would do that
• The winner says "gee thanks" - not a good indicator of trust

If it's #2, I will be making sure that nobody who trusts someone who trusts the winner is on my list.

Here's a guy trying to cheat me out of half of the AC's I own, and he claims that I edited my post!

For reference:

Hey, FRAUD!!
You edited my quote!

I only send 0.01 back to you

I'll send 0.01 back to you
|738551740| (0.1AC)nimda(0.09AC) -> AdamStgBit

Unfortunately this section of the forum allows you to edit and delete posts.

Compare the OBV (orange) with candlewick bottoms (blue):


(Please alert me if the image is not visible)

Ah, smart.


No, thank you I also gave positive feedback with the forum's new trust system

With BTCUSD hanging around $130, you can't expect promotions worth 0.1 BTC to last long.

Thank you, and yes please I suppose sending back to the originating address removes the need to prove anything.

How do I get the hash? This is me:
https://blockchain.info/tx/12612d794b67798729aa00843847b2d975486ade6f8936244079093fb5a1dc4b
and I can prove it with a signature.

When I tried to view source, Chrome asked me to "Confirm form resubmission," which is obviously not what I wanted.

You're making it hard to solve this problem by not telling us the problem, but I think there might be a trustless solution to this. See for example vanitypool, which allows vanity mining for custom addresses, and only the client gets the full private key.

So, given the customer's public key, it's probably possible to generate an address that you do not have control over, but that the customer will have control over once you give them your own part of the secret.

Ask one of the core devs or someone who worked on the vanitypool spec.

Wow, I guess all this thread needed was a bump. The problem is not sponsored slots being left but my time being available.
I'll add available times to the first post. People can then claim them on a first come first serve basis.