snipped...
"Coins are mined continuously. Do you want the last block of the most profitable coin, or the average for the last [time period] of the most profitable coin?"
my response: i am suggesting the current price which made it the "most profitable" at that time we are mining it. (In other words, the reason it was "most profitable" was due to its price in respect to litecoins price.)
It is extremely rare that the choice to mine is based on price. Price is the slowest changing of any of the factors. Normally the main profitability change is difficulty. The price however also will likely not match up with any exchange prices at the time, since it is a huge amount of stuff built into the price, it could vary wildly (not matched to the actual exchange prices).
"What if the "most profitable" coin being mined is only 25% of the pool?"
my response: makes no difference what percentage it is of the pool.
It does make a difference, specifically with smaller (low difficulty) coins have such low expected block times that the propagation time across clusters (eu to usw, use to sea) makes you start seeing a massive amount of wasted time, meaning that its actually
more profitable overall to mine the 2nd/3rd most profitable (raw profitability) coin.
"What about "price/btc"? Time it was mined? Average over the last [period of time]?"
my response: actually that is a good point in a way , the btc/price has no bearing on the first " most profitable" price number until after said "most profitable" coins are exchanged into btc.On waffles pool for instance the " most profitable" is determined to be based off ltc.But in theory the price of btc while mining " most profitable" coin can have a effect on "most profitable" as well. Just how much is the real mystery.
This is incorrect. Profitability is based solely on BTC (nothing about LTC), it is only presented on the front-end as connected to LTC (and then only in the vsLTC stat). Price of BTC (compared to your choice of fiat) has only two effects on profitability (and these are negligible), one is that it might be more effective to mine sha256 instead of another algo (which essentially is not going to happen), and second is that the price of btc affects the amount in fiat you're making that is needed to be converted to cover electrical costs. Everything outside of that, btc should have no effect on mining/profitability calculations.
"You also want the actual exchange rate. Sometimes that doesn't happen for a while (hours). Wait to post until the exchange actually happens?"
my response: paid stats are not posted until after they are actually exchanged any ways, so no change here except the fact that while said "most profitable" coin is mining the pool records the price at which made it "most profitable" and simply displays it when complete. You may ask why this is important, well in my mind it would show transparency of "most profitable" coin from start to end and effectively insure that the pool stays profitable long term. The formula for profitability is very simple in that it is basically income minus expenses, there is no complex formula that needs to be adjusted or tweaked every so often. Now maybe the exchanging part might need adjusting or tweaking, but the overall profitability formula should stay the same.
"You also want the number of coins exchanged, and the realized BTC. What happens if there are a bunch of blocks that are all exchanged at the same time?"
my response: again nothing changes current system on this. Just looking to have displayed the price when "most profitable" and the price it was exchanged for, no matter of when it occurred (unless it occurred like a year later or something crazy like that).
The only downside would be the possible reverse-engineering of our pricing algorithm (exchange splitting, depth calcs, etc). Not a huge deal however. Also, the price listed when we mined a block would not necessarily be the price we expected when we started mining a coin. A large number of our miners (moreso on scrypt than other algos) take a few seconds to change work. So if the price changes, we switch coins, and a block comes in from one of those slow-to-change miners, the price is not very indicative.
"PW's accounting has that all matched up in order to allocate distributions. If you want real-time data, you can't get the actual exchange rate, because it may not happen for a while (takes time for coin to become confirmed, to transfer coins to exchange, and to execute an exchange order)."
my response: well i would hope that it is all real time data actually, as how would it be " most profitable" at that moment we are switching the pool to mine it if it was not real time data?
We pull real-time price data of course. But again, the price we base all of our decisions on has a TON more baked into it than that price.
"PW already publishes a list of recently found blocks on the Stats page. I suppose he can add to that list the spot BTC/coin at the time it was mined. Maybe also add the pool hash rate. Then you can estimate your share of the payout. "
my response: Waffle pool already does a great job of estimating our shares of payout. This proposal is all about pools showing more transparency of "most profitable" coins. You can call it a sorta checks and balances system for all involved(really a win/win for miners as well as pool operator) as usually the pools mining software setup is only as good as it was programmed to perform.
I'm just a bit confused on where the transparency helps. I can't think of a situation where I would actively want to skew the profitability data. The main reason the inputs aren't shown is that they're not that useful to miners, and it keeps pool competition down. It would be extremely easy for me to open-source everything behind WP, but I put a
ton of work into getting to this spot, and would love to profit from my work (selfish I know, but the truth). The only reason anything is hidden on the site is for competitive advantage, not to hide things from miners (compare to other pools, we publish a TON more data for transparency). Many pools it would be trivial for the owner to skim earnings from the pools, and as a way to show that I'm not doing that, I try to publish everything I can that doesn't compromise competitive advantages (for me, and for our miners).
As for the profitability part, I'm confused as to why/how there would be a reason that more transparency would be better. Transparency is there to show that I'm not doing something that would harm miners (to keep me accountable), which I'm all for. But in the case of profitability calculations, I really can't think of a way that I could be doing something in the calculations that would benefit me, and harm miners. In the profitability calculations both sets of interests are aligned (miners/owner). I want the pool to do the best it can in terms of profitability for two reasons, 1) I directly take a %, so higher profitability = higher earnings, and 2) higher profitability looks good, attracting more miners leading to a repeat of #1. Transparency just isn't needed, and only hurts competitive advantage.
If you have a theoretical situation that results in higher earnings for me personally, but hurts the miners in the pool based solely on profitability data that we don't publish, I'd love to hear it (and if it is a legitimate concern, we can discuss getting the data published to alleviate the concern), but until then, I can't see a reason...
