Releasing mining rewards over 50 weeks


#143

IF the block reward is shifted btw. both algos than you are right, there won’t be any increase, but i have my doubts this will happen again for several reasons, but i could be wrong with each of them of course:

  • Zooko said that algo-A can not be changed to timelock or whatever because there is an agreement with the investors which are being paid from the Funders reward. Of course this could be eventually splitted btw. both algos somehow.

  • I only see just a prosposal, not any verivied changes for algo-A or algo-B, which at least in my option leaves all doors open to whatever. including that algo-B will create additionally coins within the max. supply of 21M.

  • Lowering the algo-A block reward and later the coming halving will have a serious impact on algo-A, mining operations, whatever… This even wasn’t discussed so far.

  • IF algo-A miners are aware that their block rewards are proportionally lowered as algo-B rewards increase there would be huge resistance in my opinion.

  • Algo-B beginning with 0 or very minimal rewards will make it an even weaker algo as i assume it will be.

Until i see something black on white that it IS decided i will set me bet on additionally issuance of ZEC for algo-B…


#144

I didn’t spot this assertion before. Either it is false, or the alg-B hash rate will definitely be insufficient, or both.

For concreteness, let’s say “a few” is 5 weeks. The claim is that 5 weeks of mining “at first”, i.e. when a miner starts mining on alg-B, is sufficient to match what it would have been without the time lock. Suppose that the difficulty without time locking is such that the miner gets C blocks per week, and the difficulty with time locking is such that the miner gets D blocks per week. After 5 weeks the miner has C*(1+2+3+4+5)/50 = C*0.3 times the full alg-B block reward, whereas without the time lock they would have had D*5 times the full alg-B block reward. If these were equal (for the sake of argument) then C*0.3 = D*5, i.e. C ≈ 16.7*D. Therefore, for equal effort put into mining, it would have to be the case that the difficulty is at least 16.7 times lower with the time lock than without, in order for what Zooko said to be true.


#145

In other words, given the assumptions, time locking could collapse the hashrate by 94%… (regular difficulty is 1, with time lock difficulty is 0.05988024)


#146

I think this is hard to agree with as I don’t know any power utility that accepts ZEC as compensation for power. (Although a power utility that accepts ZEC could be interesting :wink: ) Otherwise I agree with most of your assessments in this thread.

It would most certainly decrease the velocity of money as it would constrain the free money entering the system. Which would in the traditional money system would aid deflation.


#147

Will Harmony mining possibly make Zcash vulnerable to a similar attack and be delisted from Coinbase and Gemini?


#148

@naimed Brazilians can pay their electric utility and internet bill with ZEC using Kamoney


#149

I guess it would have to be a calculation they do about the possible returns of only 50% of those figures versus selling the machines, running them for free is all well and good but with the idea of profitability does it make more sense to offload before? I really dont know (btw I love your posts but some others are so long I dont have time!)
Besides the developers are focused on all of that now, we should be looking more towards NU3!


#150

This has also been expressed on this thread as “alg-B miners are being treated unfairly” since they have to wait for their block reward while alg-A miners don’t.

I see it differently. Yes, it is true that 10 ZEC received over 50 weeks is worth less than 10 ZEC received today. The net present value of the former is lower. For the sake of explanation, let’s say the time-locked reward is worth 9 ZEC in present value. (That is, the average person would be indifferent to having 10 ZEC over 50 weeks to having 9 ZEC right now.)

But miners care about the rate of return on investment (ROI). If their cost to earn 9 ZEC is 0.9 of the cost to earn 10 ZEC, then they’re indifferent. They’re not being treated unfairly! As Zooko has said, mining is an extremely competitive activity with low barriers to entry and exit. What happens is that hashing power (cost) is distributed in proportion to the net present value of the income (reward). So, yes, 9 ZEC is worth less than 10 ZEC. This will lead people to invest fewer resources in mining blocks that are effectively worth only 9 ZEC. But they will invest resources. People will mine the time-locked POW, using about 90% of resources relative to alg-A.

An easy way to see this is to imagine that this dual POW with time-lock has been deployed, and no one is mining on alg-B (the time-locked one). Would you start mining alg-B? Or would you say, “No way am I doing that, it pays only 9 (effective) ZEC, while an alg-A block pays 10!” No, of course not! You would understand that the alg-B difficulty would be extremely low (assuming the two POWs have their own independent difficulties targeting the same block interval, which I believe is the proposal). Of course you would mine alg-B! But others would see the same opportunity and jump in with you, to the point where all of you were making about the same average ROI as alg-A miners, and indeed the same as all other cryptocurrency miners. (Of course, I’m speaking of tendencies; there are all kinds of short term variations.)


#151

The conclusion would appear to contradict the hypothesis, because if all miners end up achieving the same average ROI, why would anyone prefer to mine time-locked coins, encumbered with restrictions? Therefore, to impose this policy on alg-B miners and not alg-A miners can be seen as discriminatory and unfair.

Which is exactly why miners need immediate access to funds so they can better manage risk and take advantage of opportunities in the market.


#152

One thing doesn’t exclude the other. One can sell machines and keep as many as needed … just as another example. It’s not a sell or hold option producers have.

Other than that there are enough “only” mining facilities settles in 2-4 cent per kw/h regions without being a hardware producer. So i think the example i made is absolute valid, no matter some numbers will difference a bit maybe.


#153

Interesting post, in theory, but in practice it won’t be like that in reality.

You miss several factors in my opinion and some of them are very important in my opinion:

  • you nowhere take into consideration the parasite miners that come only on very low difficulty. These mostly won’t care to mine some hours when it’s cheap for them and get rewards whenever. Exactly these will make the loyal real ZEC miner’s life harder

  • auto-algo-profit switching mining pools/services, these are as well coming mostly on low difficutly periods and are not conisdered by nature as a loyal miner but again to put into the category parasite miners.

  • due the above i doubt the algo-B will be extremly low at all, but very weak compared to algo-A only.

  • due the various ETH upgrades (constantinople, Caspar) which fall more or less in the same time period i predict a massive gpu hashpower shifting,especially after caspar release. A good portion as one of the options for this enourmious released free hashpower than will go for sure to ZEC algo-B, be as a parasite miner or more or less loyal miner doesn’t matter. As that hashpower is that huge i have my doubts any gpu will stay profitable than. Time will show if my prediction will come through but having in mind literally millions of gpu’s being forced overnight to look for another mineable gpu coin backups the prediction, enough ETH releases caspar.

  • IF like you write, the difficutly would be extremely low, than things are messed up allready. Extremely low difficutly cries for an attack, especially if it’s on a top coins network. Means, what you think is an opportunity for miners is mostly the biggest nightmare for ZEC having extremely low difficulty. Additionally this would like to huge spikes, see the first points as it attracts the parasite miners anyway.

  • And finally you don’t include volatility at all. Sure 10 ZEC now will be 10 ZEC in 50 weeks, true. But the value of these could be absolutly different than and bevor. Just make a simulation with getting 10 ZEC on 10. January 2018 and 10. January 2019. This simulation isn’t really correct as the rewards are begin released periodicly weekly, but it gives you a good picture about the differences in value and high risk included with volatility.


#154

I’m sorry I didn’t explain clearly enough. That was merely a thought experiment, not a prediction. I was only trying to explain why saying that no one would mine using alg-B (since its payments are delayed) is clearly incorrect, precisely because the alg-B difficulty would never be allowed by the community of miners to become very low! There is a profit opportunity as soon as it just barely begins to drop below its “natural” level.

So, yes, the total amount of real investment (amortized hardware costs plus ongoing costs such as electricity) into alg-B will be slightly less that going into alg-A, but only slightly less – I very roughly suggested maybe 10% less, but it is whatever discount the overall community attaches to being paid over 50 weeks versus immediately, plus, as you correctly point out, the additional volatility risk (including the risk that the value of Zcash will go to zero within 50 weeks).


#155

The average ROI includes (takes into account, is adjusted for) the time-locking of coins. I’m sorry, I don’t know how to explain it more clearly; perhaps my latest reply to @boxalex may help.


#156

Sonya, I think the time-locking part will make it more profitable — except not for “drive-by” miners who switch among coins depending on which one is the most profitable at the moment, or who rent their hashpower to nicehash. I think it will make it more profitable for miners who mine Zcash exclusively.

I also think it will make the network more secure for all users (despite the reduction in “hashpower-security-factor” that Daira is helpfully investigating.)


#157

Still a lot of Thinking Zooko, remains the question, how ?
I rather saw you writing , I know for sure the time locking will make it more …
Then there would be any kind of proof of this.
So it are stilll asumptions, I think that’s tricky, since I can’t find any on this subject anywhere, where it’s explained how the miners will get more profit if they stay.

For me to get back in mining, I like to have a solid roadplan, I want my descision based on facts ( a lot of facts, for example price, no one can forsee that, so I am aware of the fact that not all facts are solid or stabil )
And for the scale of my mining , power is a biggie, so I have to be sure if I get out, my losses are doable.
If that’s an unknown, I do not restart my rigs again.


#158

Obviously miners do not like it, so hashrate will go down which makes it much more profitable for the “stakers”. Security is determined by the resulting stakers/(rental market) hashrate ratio. As it goes from 0.51 to 0.49, your risk goes from 0% to 100%. If hashrate drops 50%, an ASIC miner with 10% of the current hashrate will find a double spending attack attractive if the rental market can provide 15% of the existing hashrate (to get the 50% in a half-difficulty scenario). It doesn’t even require that much because some of those percentages are being shifted from the public chain to the private chain.


#159

To be clear, this is more profitable for “stakers” who are people who have their own equipment, but the total number of stakers goes down as most of them do not like the proposal. They go to other coins or put their equipment on the rental market. So both the numerator and denominator get worse for security. There’s no such thing as a relevant “cost to attack” because double spends have a lot more to gain than the costs. ETC and BTG attacks show this. Nicehash double-spending attacks are on the larger coins with higher “cost to attacks”.

The relevant security metric is the probability distribution of (ZEC “stakers” HR)/(POW rental market HR) ratio going below 50%.


#160

‪My view: ‬
A. too risky for Zcash Company to modify coin properties or issuance schedule
A1) investor perception of certainty of supply
A2) regulation of securities (legal/equitable splits, restrictive covenants, debt etc) compared with money (immediately redeemable)
A3) fungibility of coins an essential property
B. Practical loyalty incentives not created. Market routes around restrictions via secondary market pricing risk appropriately (time value of money).
C. Increasing cost base (see B) will cause
C1) centralisation
C2) decreased hash/security
D. “central planning” initiatives perceived as regulation contrary to ideology of community.

In sum you’re proposing restriction to compel behaviour.

My proposal is a seperate legal entity/market participant to incentivise loyalty via profit motive.

A community owned mining pool that charges market rates to aggregate hash on Zcash. Either decreasing fees or dividend payment from fees back to miners, provided a certain time period hashing. Time based per GPU/ASIC unit not hash based. Ideally community funded token model implemented with smart contract/s.

This would allow lots of opportunity for financial/game theory engineering but without the risk to the core protocol, coin and company.

I would love it if a representative from Zcash company (Zooko?) addressed each of the objections I’ve summarised to convince the community of their proposal.

As an aside it seems to be the view that “controversial” proposals are healthy for community. Whereas I think they waste too much attention and valuable resources and most of the “controversy” would be obviated with better communication.


#161

Wouldn’t full-time Zcash miners like myself be more incentivized to mine another coin and auto-convert to zec instead of mining long term?

Let’s say zec is not the most profitable coin to mine:

In this case, auto-converting to zec will get you more zec on an exchange and make them an immediately liquid asset, rather than having to wait the full 50 weeks for a lesser amount.

This seems like a big problem in the incentive structure. It could both increase the price of zec while also decreasing the alg-b security. We want to increase both. :slight_smile:


#162

This is absolutely wrong and unrealistic, for two reasons:

  1. The expected investment of resources is not linear in the value of revenue. At most it’s linear in the value of profit, and even a small reduction in effective revenue can completely wipe out the expected profit for a given subset of miners (bearing in mind that this already applies to miners in some geographical regions, and others are marginal).
  2. It’s completely unrealistic to suppose that the value of the proposed time-locked reward is 90% of an unlocked reward, as opposed to 50% or less. You’re probably failing to account for price volatility risk.