Testnet4 has a hashrate (MA) of 296 TH/s
5 Bitaxe supras would be around 3 TH/s
5 days would be quite lucky
Discussion
If I have 1% of the hashrate shouldn't I expect over a block per day of the ~144 blocks mined?
I stand corrected, you're right.
How do you calculate the hashrate of testnet4? Because what is shown on pools (like testnet.vkbit) and 'getmininginfo' RPC is completely wrong. This is because people mine future blocks with their CPU, so there is a lot of "difficulty 1" blocks, which completely messes up the hashrate calculation.
If you want to calculate the chance of hitting high enough difficulty requirement, you should only use that. e.g. P (probabilty) = 1 / (difficulty * 2^32).
You can get the difficulty from a block that has difficulty > 1 (most blocks on testnet have difficulty "1" because they set the time of the block (nTime) to 'last block + 20 minutes', and then use CPU (or bitaxe or whatever) to immediatly hash that block.
I don't understand how that future thing works. Does that mean I would have equal odds using a cpu as an asic?
testnet3 and 4 have a rule: when the 'nTime' of a block is larger then the nTime of the last block + 20 minutes, the difficulty drops to 1. This means that anyone (cpu/gpu/asic) can easily mine it, because the difficulty is 1.
So you can construct a block with a 'fake' nTime in the future (at least 20 minutes more then the last block), and instantly mine it.
There is a limit to this (else you can keep moving nTime more and more in the future).
The limit is another rule: nTime of block cannot be later then last block + 2 hours.
So it's possible to always mine 5 blocks into the future. This is being done a lot on testnet (because it's an easy way to get some coins without waiting to actually mine it).
You can easily see this on mempool. All the latest block show "Just now", but when you click on them, and view their timestamp, they are often in the future (with a maximum of +2 hours)