Suggestion: Automatic Mining Template Switching

[ToeKnee]

I have a suggestion for a site improvement: automatic template switching.

Edit: As it stands as of right now, Prohashing mines the most profitable coin and distributes earnings, statically, among the defined coins in the payout menu according to the percentage that a user assigns to a particular coin. The purpose of this post is to "spit-ball" ideas on how to implement changes that would maximize a user's earnings or adopt a coin mining strategy tailored to the user's own goals.

Note: The coins automatically selected will be based upon whether or not the user has a wallet address associated for the particular coin. From within this understanding:

Template 1. "Always Mine & Payout What is Most Profitable":
Based on the user's numerous wallet addresses entered into the payout system, have the system automatically mine the single coin that is most profitable and make payouts in that same coin. Percentage breakdowns will not matter in this template; if a coin is doing well, mine the coin and pay out the same coin--if the user has an address for it. (Note that daily minimum payout thresholds may or may not be reached.) People who use this template are seeking to build a solid foundation of value.

Template 2. "Always Mine the Single Coin That Is Most Profitable *BUT* Payout in Lower Value Coins":
For maximum coinage, consider mining the coin that has the highest value and use this highest value to switch out for the lowest value coins in the user's wallet. This, in effect, would allow users to maximize their coin position. In this manner, the user would be maximizing the number of coins earned. (Note that daily minimum payout thresholds are more likely to be reached.) People who use this template are sinking money into lower performing coins made from higher performing coins with the goal of increasing their position of lower value coins in case the lower value coins "suddenly" have a price spike.

Template 3: User pre-defined percentages #A.

Template 4: User pre-defined percentages #B.

Template 5: User pre-defined percentages #C.

Template 6: Automatic switching between Template #1 and Template #2 based on "Bull Market" or "Bear Market" using the USD price of BTC (+/-) as the trigger, since it seems that most currencies are either themselves "up" or "down" if BTC itself is "up" or "down."

Template 7: User defined switching schedule of Templates #1-#6 based on day of week or time of day, or hourly, or daily, or weekly, or monthly.

This would offer users to enact multiple mining methodologies at the same time and help users to establish their own personal method of coin mining and investing.



Just a suggestion; feedback is appreciated--especially since I cannot see all things or take all things into consideration. Thanks!

[Steve Sokolowski]

Hi Toe,

The reason that the first template isn't implemented is because of network transaction fees. The number of coins that the average miner would end up with would be around 35. Most of those would be unable to be paid because network transaction fees have become absurd at the top of this bubble. Maybe now that a trend reversal (at least in bitcoin) has occurred, the fees will come down as prices decline and demand reduces, and make this option more feasible.

The second option sounds like portfolio targeting - or is that incorrect? I'm currently working on a program to simulate the effectiveness of portfolio targeting compared to the "constant earnings" method. I was amazed at how profitable portfolio targeting ended up being - almost 20% more than what other pools were providing or what constant earnings would provide.

I like the idea of having earnings templates that can be switched between. What would the rationale be of switching depending on bull or bear market? I'm not sure why bitcoin's returning to a bull market would be in any way a better signal than prices of the coins being earned themselves.

-Steve

[ToeKnee]

Yeah, I'm spit-balling.

The main underlying rationale of my post, of course, is to leverage whatever machine learning technologies at Proshashing to maximize either earnings or the number coins (with the hope of future earnings).

The logic of implementation is another story. To be honest, I need to read more on portfolio targeting.

I'm coming over to Prohashing from HiveOS. And I'm still learning how Prohashing works. Heck, I'm still learning how the Crypto market works in general! But I still use HiveOS on all my machines and my HiveOS "flightsheet" points to Prohashing's pool. From my naive point of view, Prohashing is "better" than HiveOS for the simple fact that it appears that I earn "more" USD using Prohashing's Pool when compared against using Hiveon's Pool. (Additionally, the reason I like Prohashing is because of the auto-exchange from ETH to whatever I like according to payout percentages. This makes it possible for me to grab coins while mining using Ethash that would otherwise require me using an exchange to translate ETH to a different coin.)

On HiveOS, I had all my GPU miners configured to run Phoenixminer to mine ETH using the Hiveon Pool. On the Hiveon pool, payouts occur each night only after 0.10 ETH has been mined, regardless of the actual USD translation of ETH. Prohashing, on the other hand, uses a USD trigger. Sticking with the ETH example, Prohashing requires $1k in ETH before sending earnings to a wallet to avoid a transaction fee. Now, sometimes $1k in ETH is 0.50ETH (for example when 1ETH=$2000.00USD), and sometimes, $1k in ETH is 0.45ETH (for example when 1ETH=$2222.22USD), etc. The actual USD value of ETH is always changing, therefore, the number of ETH needed to trigger a payout to a wallet always changes. This is where my mind tangles up a little bit. With HiveOS, payouts occur at a static 0.10ETH whereas on Prohashing, payouts occur at $1k. It appears that Prohashing's model of payment is a little more flexible, especially when the USD price of a particular coin skyrockets.

Let's suppose that the price of ETH is rising while the price of DOGE is falling. I would like to use my Ethash GPU miners to mine ETH in order to increase a payout in DOGE. (Template #2). Now if both ETH and DOGE were to rise in price, then I'd use my Ethash miners (ETH) to buy a different coin that happened to be falling or lowest in USD price in order to maximize position. Similarly, if the price of ETH was falling, but another Ethash coin was rising, then I'd use the stronger coin to buy the weaker one. Maybe this template can be called "Stronger Buys Weaker" or "Trend Mining." Perhaps this is what is already happening at Prohashing? Maybe this is the logic behind portfolio mining? Spitballing.

We all know that ETH mining is going away soon and people will want options to point all that unused hashing power. Most people will just say, "ah, screw it," and point their hashing power at Nicehash and run NHoS because it's just plain easier. (Prohashing should develop its own Linux operating system!) It seems that other people are going to switch to ConFlux, VTC, RVN, or ETC when direct ETH mining stops. If people don't want to use Nicehash and they wanted to mine RVN directly, for example, they might have to re-configure their GPU bios, adjust power levels, and use a new miner that uses a new algo, etc. I think RVN uses Kapow and that algo uses more power than Ethash. That's a lot of hassle, to be honest--and I think the avoidance of hassle will drive people to Nicehash. It's not impossible to reprogram your GPU, but it would require that an individual learn the ins-and-outs of a mining a new coin after all they have known is Ethash.

As far as the BTC trigger (+/-) goes, that was just my first thought to capture a signal indicator that tends to predict the rest of the market. If coding could allow the monitoring of all coins, then the (+/-) of each coin could be its own trigger. I was thinking in terms of making the code simple.

Pretty much this is a stream of consciousness post that I made in order to help me clarify my own thoughts. As more people chime in and as questions are posed, I'll be able to clarify my thoughts further. But at the end of the day, the motivation is to make as much coin as possible.

[spauk]

We all know that ETH mining is going away soon and people will want options to point all that unused hashing power.

Why would ethereum mining be going away? My opinion is it's not going away any time soon.

[ToeKnee]

ETH 2.0 will be proof of stake (owning the coin), no longer proof of work (mining the coin).

[Steve Sokolowski]

Ethereum has been saying that it's going to move to proof-of-stake for some time now, but that transition keeps getting delayed. I don't think that's going to happen for two years, at least. And if so, that would be good. The Ethereum developers, like the Bitcoin Core developers, have been distracted by all sorts of unnecessary features - the network functions fine with proof-of-work. Meanwhile, it costs $35 to send one Tether payout to one customer because sharding and similar capacity limit issues have not been addressed.

The dollar-based threshold is simply because we initially developed the system in 2013 to use dollars as the backend accounting method. Other pools started with a primary coin, whereas we designed the system to support mining of any coin and paying of any coin. Therefore, defining payout thresholds in anything other than dollars doesn't make sense to us.

I'd love to develop an operating system, or more reasonably, one of those automatic algorithm switching miners. However, we determined a while ago that those miners are unlikely to ever become profitable because that would require adding support for hundreds of algorithms. I don't understand how pools that support so many algorithms ever turn a profit. It can take more than two weeks to get a single algorithm implemented and tested, and that doesn't include all the research into the parameters and which coins the algorithm supports. The other reason not to develop an algorithm-switching miner is that one algorithm has been the most profitable for some time now, so that sort of software would have been a wasted investment.

I wrote code yesterday to compare portfolio targeting of a number of coins against straight earnings for the past year. Once I get the guide written to demonstrate the findings, then let's continue this conversation to see what other earnings templates can be inferred from those results.

[spauk]

I like the idea of payout templates, but mining templates would probably not work without making a mining OS or algo-switching miner. Lately I've been checking the status page a few times a day because the most profitable algorithm for my GTX 960 rig switches between neoscrypt and equihash 144_5. I think the reason for that is the prices go up and down on the exchanges prohashing uses, or something's going on with the coin networks. So most of that is out of prohashing's control.
I don't know how much I would actually use payout templates though, I usually choose payouts in the same few coins that I like (usually the coins with high value or potential for higher value, and I also like the coins with lowest network transaction fees) and I only adjust the percentages.
I think awesomeminer might work with prohashing, but I've never been able to set up auto algo switching based on profitability, and I think they only let you manage one rig for free. So it seems easier for me to just check profitability manually and switch algo if needed, luckily there's only two options.
I did some more reading about ETH2.0 and it does sound like they want to stop PoW eventually, but I haven't found an exact date or block height when PoW ends. I guess it's whenever they are satisfied with their work on the ETH2.0 shards and merging Eth1 into shards to keep smart contracts, which could be a month or years away. I also read something about a difficulty time bomb that was planned back in 2017, I haven't heard anything more about it. I guess I just have a hard time wrapping my head around the term "soon" used with unknowns.