...Essentially all it does is make it so if somebody drops in and starts mining during particularly easy difficulty times, their reward is cut in half, just for that one difficulty cycle. If they stay and continue mining, they would get full rewards from the next block forward.
I assume the network checks blocks submitted by a node during the last difficulty cycle(s), in order to determime loyalty. How does this keep miners from hopping on pools during low diffs, since the user and network sides of the pool node are completely separate?
If people want to leave mining the coin during easy difficulty times, that means the profitabilty must be really bad, because the people who had been mining are receiving twice as many coins as people who want to drop in and start mining. Then we have a legitimate issue with the coin having lost a huge amount of market appeal. However, it is not very likely that someone who has been mining the coin long-term would stop mining it while they are earning twice as many coins as people who start mining - plus, if they do, the difficulty level would adjust to one step easier yet, to compensate, through existing mechanisms, so there is no systemic problem. You do make a good point - the halving of rewards should be extended to subsequent difficulty cycles, until the difficulty eventually goes up, not just be limited to the one easier difficulty cycle after a hard cycle.
On a more general note, once the mechanism in place, the coin hopping pools would pretty quickly stop including Catcoins in its calculations, out of economic necessity, because they are after "profits" and they would not be profitable enough, receiving only half of the rewards, by jumping into mining Catcoins during easy difficulty times. And by nature, they won't stay through a high difficulty time, to qualify to receive full rewards during the next difficulty period. So reward paid out would approach 100% going to loyal miners (defined as any miner willing to mine the coin for at least two continuous difficulty cycles, including a difficult cycle). Which would dry out the supply of cheap coin to the dumpers. Almost all of the coins would end up in the wallets of people who believe in Catcoins and employ a strategy of mine-and-hold. This should lead to price increases, simply because there is no longer a stream of cheap coins being sent continuously to coin dumpers / coin hoppers.
Keep in mind also, this mechanism is designed to make itself unnecessary to be invoked except in rare cases. Once it is known that it is futile to try to profit by mining the coin temporarily when the difficulty level gets easy, in a "jump in, get rich quickly, jump out" sort of a way, the difficulty fluctuations should become more modest, which should cause this mechanism not to trigger very often. The goal is to have stability in the mining difficulty curves such that the reward mechanism simply never engages at all. It is best seen as a fail-safe only, so it can handle situations such as when a big miner decides to jump in without being careful, and drives difficulty up, that it does not open up an opportunity for a bunch of profiteers to jump in when they leave, cause the difficulty to go way up, which is then followed by the difficulty level becoming ridiculously easy (which in medical analogy is like suffering insulin shock for diabetics, or drug withdrawal symptoms for someone addicted to drugs). The high temporary hashing, is kind of like drugs - causes a temporary sense of the coin's importance, but is followed by a crash and massive sell-off which is the withdrawal symptoms.
and from another post:
You're assuming their strategies are indeed predictable. Not to mention that if a block reward is all of a sudden half of what it should be, people are going to ask serious questions about the legitimacy of our coin, even if the mechanism is well documented and transparent. Unless you can address envy's concerns above, and make the mechanism reliable. It's a non-starter I think. Further, all the tools we have to monitor hashrate seem to be inaccurate, including the wallet itself. we saw plenty of proof supporting this in the last two spikes, network hash was nowhere near what it should have been. How can you predict an emergency mechanism off of unreliable information. Yesterday we were showing 212 Mhash on one tool and 400 on another, and neither seemed to be right.
No, there is no such assuming anything - it is narrowly targeted to people who engage in profit-maximization. It would take away the one-and-only factor which would cause a profit-maximizing strategist to drop in on the coin while the difficulty is easy, and leave when the difficulty gets difficult. Anyone who would drop into mining the coin for any reason not related to immediate profitability maximizing - it would not affect them at all - and for purposes of this design - and I submit any other design - no mechanism can counteract the normal fluctuations in the network hashrate caused by people randomly choosing to mine or not mine a coin, who do not respond to incentive signals. But I would submit that for all practical purposes, people jump into mining a coin during easy difficulty times to try to maximize profits. Otherwise, the huge influx of mining power during easy difficulty times would seem to be an amazing, unexplainable coincidence (unless you want to offer a plausible explanation other than profitability why hashrates seem to soar during easy difficulty times).
I agree such a mechanism should not be implemented unless and until it has been tested in the lab to be reliable. I believe this can be implemented in about 30-50 lines of code and is not that complex. Furthermore, I believe once this is seen out in the marketplace and has the effect I predict for it (and anyone can validate this prediction, I believe it is logical), other coins will simply copy the code, and it will become standard for all the coins. Then it will just become part of knowing how cryptocurrency works - and the new standard will be that you choose a coin based on its longer term potential, and park your mining equipment on it for weeks or months at a time, hold the coins, and reevealuate once in a while. This is healthier for the cryptocurrency ecosystem, much more so than the mine for a few seconds and dump it on the exchange modality that prevails today. This kind of change is of course revolutionary, and I would not expect it to take hold very easily without meeting a lot of opposition - and this is why I did not propose it lightly, and why I would be willing to invest some effort in answering reasonable questions about it - any kind of radical proposal deserves a long contemplation period before being implemented where all of the pros and cons should be vetted carefully.
I agree that tools to measure hashrate is unreliable because they are based on estimations based on how often block solutions are submitted into the network. However, the mechanism proposed is not based on estimating hashrates - it is based on difficulty levels - which are absolutely reliable. It is so reliable that the entire network has 100% consensus on it every time there is a change. So there is no problem with indeterminism, when it comes to whether or not a given difficulty cycle would fall under the criteria of exceptional circumstances, where a miner without recent mining history, would have their rewards cut in half. The entire network can have 100% consensus on this, for the same exact reason 100% of the network has consensus about what the current difficulty level is (and what the previous difficulty level was).
