I see you’re copying @zawy’s charts, but not fully understanding them. (Also, you should credit your sources - he deserves credit for his hard work.)
In the research terminology, a “hash attack” was originally an influx of hashpower done with malicious intent. The modern phenomenon is similar, but in the GPU-mining space, isn’t actually malicious (done to damage the coin or steal value); it’s just done in an effort to be competitive. It’s an effect of people using automated systems to mine whichever coin is more profitable at the time. This way, when the difficulty for a particular coin is low, their systems automatically jump in and mine more coin.
When many people do this at once because of automation, it resembles a “hash attack.” In the research and the code, the defenses are the same - it doesn’t matter why more power comes in suddenly, you just have to find the best way to deal with it.
When this data says 50% “stolen” it does NOT mean 50% of all blocks. It means that during that brief period of high hashpower (the magenta/purple sections), there were more blocks produced than we wanted. (We wanted the amount that make for one block every ten minutes.) This impact is much, much smaller than 50% of all blocks.
Speculation based on false assumptions.
End of March 31, block height: 521227
End of Apr 7, block height: 522348
That’s ( 522348 - 521227 ) = 1121 blocks mined over seven days.
That’s ( 1121 / 7 ) = 160 blocks per day.
That’s (160 * 12.5) = 2000 BTG per day - nowhere near 2700!
The excess was 200 BTG, not 900 BTG… and of those 200 BTG, most of them did not go to the hash attackers. During that excess mining period, everyone who was mining got excess coins.
So, not anywhere near as bad as you say - but still, not perfect. People like h4x and Zawy and others will continue to work to make it better. Remember, our testnet v4 already includes an improved Difficulty Adjustment Algorithm, because work has already been done to make it better, and work will continue.
That said, this problem will never go away entirely; so long as there are at least two coins mineable on a GPU, there will be some miners jumpting back and forth based on profitability. It’s not only natural - it’s also a necessary dynamic to attract more miners when necessary by becoming economically attractive to miners. When things flow smoothly, it’s the balancing effect. When they move too fast, it turns into an “attack.” It cna never go away, it’s a matter of minimizing the negative impact.
This assumption is incorrect. These mined coins are no more or less likely to be sold on the market than any other mined coins. Some miners mine to HODL, others mine to sell. The same is true of miners who use these auto-changing pools to try to get the most coins from their mining.
Sell walls do not crash prices, but I don’t think this is the right thread to go into market dynamics.