Status as of Friday, October 24

[Steve Sokolowski]

Lots of interesting stuff going on.

• We got merge mining last night. Chris verified that it works, and it will be in the next release. After that release, we should see a profitability rise of 12-15%.
• I also completed the corrections to the concurrency errors, which Chris is testing now. We'll either release them and merge mining on Sunday, or on the morning of November 1.
• A few customers asked to be paid 1% in a number of different coins. That's an interesting investment strategy that's like the way the venture capitalists around bitcoins throw their money around. Get a little of each coin, and even if 50% of them fail, perhaps one altcoin will rise by a factor of 100, turning a great investment. Unfortunately, that caused issues with the payouts this morning when the excessive auditing I programmed aborted the payouts when a daemon failed to respond. The issue should be fixed now. Remember that payouts that are under the network transaction fee don't get sent, so that you don't lose everything. In general, payouts where the balance is under 4 cents won't process.
• Look at Coinwarz's profitability rankings. Notice that there is a huge bank of coins that have a profitability of 10 cents (it may be less by now). The only one above that was BBQCoin, which I have never been able to get running (and apparently nobody else has either). These markets are very thin, and my theory is that we have reduced or even eliminated volatility on Cryptsy's selection of scrypt coins. As soon as the price of one of these coins rises, the algorithm mines and sells, making it impossible for speculators to win with coins that have a very low volume.
• This behavior is not what we anticipated; in fact, we thought that the markets had more volume and that this wouldn't happen because we would need a huge hashrate. Instead, you'll notice that we have been mining litecoins for large portions of each day now, since we extract the value from new coins we add and then their profitability falls down to litecoin profitability. Most of the time, we mine litecoins, breaking off of them when a buy order is placed in one of the markets. This may also be why Clevermining's profitability has fallen to 97% of LTC even though they merge mine. Whether related to this or not, they delisted themselves from poolpicker.eu, probably to draw less attention to their poor payouts over the past few weeks.
• Why is this different than what happened before? Middlecoin, for example, would mine large numbers of coins and then sell them all at once, crashing the markets. People would have time to add buy and sell orders between each time Middlecoin would switch coins, so the volatility between the coins would never equalize.
• I suspect that there is a way to make money off this situation, at least until we implement Poloniex trading, by looking at the price of coins on other exchanges and selling them when they rise above Cryptsy's price, since those exchanges will be much more volatile. It isn't our business model to deal in arbitrage, but if I were an entrepreneur who cared about that, I'm sure there's money to be made.

[cryptorific]

This behavior is not what we anticipated

Really? This is exactly what happened with sha256 coins too, Peercoin has been around for a very long time but it's profitability is just under bitcoins because as soon as it becomes more profitable than bitcoin, people jump on it and push it back down and now we're left with just a few sha coins which are mostly merged mined because otherwise they would have died. Eventually the scrypt coin market will shake out the the elacoins of the world and leave handful of viable coins (ltc, doge, ftc) and the new coins will have to use alternative POW (x11, x13, etc), until multipools are developed for those.

The issue should be fixed now.

Did LTC's payout fail for the same reason?

[Steve Sokolowski]

LTC payouts failed because the script was over-paranoid, and set to fail all payouts if there was an error with any coin. I changed the script to only fail each coin if any payout in that coin failed. That means that the USDe daemon no longer fails the LTC payouts and they processed successfully a few hours ago.

I think this situation is different, both than the other scrypt coins and the SHA-256 coins, because of the improvements in algorithms. In SHA-256 coins, a trading algorithm simply wasn't necessary because there are so few of them. A single person can easily manually review the available coins and perform manual trades to equalize the coins.

In scrypt coins, that isn't possible because there are so many coins, and the market depth is so low in all of them. Therefore, some sort of automated algorithm is necessary. But until now, most pools seem to have focused on creating a switching algorithm and then performing manual sales. That was suboptimal because it's not possible to predict market movements. By the time those pools would manually sell the coins, they would find out that they mined too many and crashed the market, or the bottom falls out before they sell at all.

I predict that the result, however, will be the opposite of what you describe. Assuming that we can continue to attract miners as we add more and more coins and exchanges, stability will come to all of the coins. Coin developers hate multipools because inferior algorithms result in huge dumps that crash the price below a sustainable level. That then encourages other multipools to abandon that coin and its blockchain doesn't advance, leaving it susceptible to 51% attacks.

But that doesn't happen with real-time trading. Instead, the difficulty and price adjust slowly, keeping the coin alive and, since we aren't attacking these coins, providing security from 51% attacks. I think that what will happen is that good-faith developers won't hate us, because we will reduce volatility in their coins. The only people who dislike that are developers who create premined coins they plan to sell later and get rich. If people want those coins, then they will set their payout proportions to them, and we won't sell the coins as they are mined. If we can get huge hashrates someday, price will be based upon demand, not upon speculation and gambling.