Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - biophil

Pages: 1 ... 52 53 54 55 56 57 58 [59]
871
General Discussion / Re: any document for the maching algorithm?
« on: January 31, 2014, 07:12:45 pm »
Yep, I see. I think you've answered to my satisfaction... I guess the incentive to under-represent your willingness to trade still exists in traditional market matching, in a way that it doesn't necessarily in a 1-sided auction.

Obviously, 100% transparency will help a lot too.

So will it be possible for someone to create a client that automatically walks the book an order at a time? One that does traditional market-matching on the client side? I guess it will depend on how the network prioritizes orders - if it does something like "process high bids and low asks first" then client-side market matching would be quite a bit harder.

872
General Discussion / Re: any document for the maching algorithm?
« on: January 31, 2014, 06:13:41 pm »

Rather than present a "technical" justification I will simply present an analogy.  You have 3 players in the market, someone who wants to buy A, someone who wants to sell B, and someone who knows both A and B, C.   

If A and B do not know each other and C knows there is an opportunity to buy from B low, flip, and sell to A for a profit then he will make the market and keep the profit.   If A and B know that C will be doing this then they have incentive to minimize the spread.

Perverse incentives that could result... users must create many small bids/asks to get the best possible price.

Also the process of giving people exactly what they ask for makes validating transactions much easier because I can search for equality rather than >= and thus avoid ambiguous cases.


I think I've figured out my misgivings:

In your example, B has an incentive to raise his ask price; A has an incentive to lower his bid price. That's all well and good, but here's the the problem: B's ask price no longer accurately reflects his true willingness to sell, and A's bid price no longer accurately reflects his true willingness to buy. Since they don't know each others' willingness to exchange, B could raise his ask price above A's bid price and make the market illiquid.

So there's the dangerous perverse incentive: giving people exactly the exchange rate on their order incentivizes them to act like they're less willing to exchange than they really are, and market liquidity suffers.

This is why eBay doesn't charge winning bidders their bid; it charges them the bid of the second-highest bidder, because then bidders do not have an incentive to under-represent their willingness to buy. eBay's goals are essentially the same as BitShares'; i.e., they both want to facilitate the highest possible number of exchanges.

I can't speak to your last point; it could be that the technical aspect of validating transactions wins in the end.

873
General Discussion / Re: any document for the maching algorithm?
« on: January 31, 2014, 02:44:10 pm »
Bytemaster, have you guys done a game-theoretic analysis of your matching? I haven't thought about it yet in enough detail to be sure, but from what I know of auction theory, it seems to me that giving people exactly what they ask for (instead of the best price available) could introduce some unexpected perverse incentives...

On the other hand, the fact that bidders/askers get the best price available if they bid the market price could essentially be an incentive to price stability, which is of huge concern.

Over the weekend I'll try to write up what I'm thinking a little more formally; till then I'd love to hear whatever technical justification of your matching algorithm you have.

874
MemoryCoin / Re: Volunteer Request - UN Petition
« on: January 27, 2014, 06:03:13 am »
Why hasn't this gotten more attention?

Maybe it's just my insane curiosity, but I'd love to know what you have in mind...

875

Here's my take:

A BitUSD is essentially a certificate that says “The BitShares network promises to pay the bearer $1 worth of BitShares upon request.” So suppose at some point a BitUSD is worth 9 BitShares, but $1 is worth 10 BitShares. In other words, BitUSD is trading at a bargain - if I buy 90 BitUSD, I can redeem them for $100 as long as I trust the BitShares network to make good on its promise. So I will go and buy as many of these cheap BitUSD as I can until the price of BitUSD again matches that of fiat-USD.

Good. Only 50% of the answer.   What about when it is overpriced and  how does the network make good? 


The exact opposite. Suppose one BitUSD is worth 11 BitShares when $1 is worth 10 BitShares. Now, BitUSD is trading at a premium to USD: I have to buy 110 BitUSD to get a redemption value of $100, so I will not buy BitUSD. Rather, owners of BitUSD will sell as many of them as it takes to bring the price of BitUSD down to parity with fiat-USD.

876
Here's my take:

A BitUSD is essentially a certificate that says “The BitShares network promises to pay the bearer $1 worth of BitShares upon request.” So suppose at some point a BitUSD is worth 9 BitShares, but $1 is worth 10 BitShares. In other words, BitUSD is trading at a bargain - if I buy 90 BitUSD, I can redeem them for $100 as long as I trust the BitShares network to make good on its promise. So I will go and buy as many of these cheap BitUSD as I can until the price of BitUSD again matches that of fiat-USD.

877
MemoryCoin / Re: IMPORTANT! Help support MMC on Cryptsy
« on: January 21, 2014, 03:26:01 pm »
The problem with Bter is that it doesn't feel trustworthy - their charts don't work, deposits and withdrawals are slow, common wisdom is that their tech support doesn't answer emails, the site just lacks polish...

Ethical question: would it be a bad idea to simply pay Cryptsy to list MMC? Surely a few thousand MMC would be incentive enough for BigVern to fast-track us over at Cryptsy. I'd love to see a discussion on this.

878
MemoryCoin / Re: Cross-Coin Satoshi Dice Idea (Exchange)
« on: January 21, 2014, 12:12:33 am »
Seems very clever - what if you specified an asymmetrical fee? Charge more to convert MMC to BTC than BTC to MMC; remove as many barriers to MMC acquisition as possible.

879
MemoryCoin will be the GEICO of altcoins - a flood of amusing unrelated marketing jingles :)

880
MemoryCoin / Detailed voting results and statistics?
« on: January 19, 2014, 07:28:30 pm »
I'm a newcomer to the MMC community, and I'm excited by the possibilities of this coin. One of the things that has frustrated me, however, is that the voting process seems unnecessarily opaque. Is there a simple way to see election rankings other than manually searching through every vote in the results and adding up address coin totals?

Pages: 1 ... 52 53 54 55 56 57 58 [59]