High Paying Blocks hopping

[markm]

There are so many chains that have some blocks that pay out way more than other blocks.

Usually when one block is solved you know whether the next one is going to be a high-paying one.

It kind of makes sense the websites that claim to report profitability don't reflect these extra-pay blocks on the instant (though hopefully they do at least average them in) because of the limited time available in which to switch/hop. But do enough of those kinds of chains have slow enough time between blocks that a pool could hop them effectively?

At any given moment, what is the chance that one of the chains with bonus blocks is on one?

(If it is still low, not to worry, as people keep spewing out such chains that chance should keep climbing, yes?)

-MarkM-

[Hydroponica]

In the case of GIL, it is completely random, so, you have the same odds as hitting a 10 coin block, as you do a 50 coin block, every single block. Good luck predicting its profitability, once it hits an exchange

[Eli0t]

In the case of GIL, it is completely random, so, you have the same odds as hitting a 10 coin block, as you do a 50 coin block, every single block. Good luck predicting its profitability, once it hits an exchange

im guessing it will be averaged like Lucky and ARG are currently

[mercSuey]

In the case of GIL, it is completely random, so, you have the same odds as hitting a 10 coin block, as you do a 50 coin block, every single block. Good luck predicting its profitability, once it hits an exchange

im guessing it will be averaged like Lucky and ARG are currently

Exactly.  If it's truly random then the expected value of 25 GIL per block, with 1920 blocks per day will be very accurate.

[Eli0t]

In the case of GIL, it is completely random, so, you have the same odds as hitting a 10 coin block, as you do a 50 coin block, every single block. Good luck predicting its profitability, once it hits an exchange

im guessing it will be averaged like Lucky and ARG are currently

Exactly.  If it's truly random then the expected value of 25 GIL per block, with 1920 blocks per day will be very accurate.

the average will be 30 [/pedant

[mercSuey]

In the case of GIL, it is completely random, so, you have the same odds as hitting a 10 coin block, as you do a 50 coin block, every single block. Good luck predicting its profitability, once it hits an exchange

im guessing it will be averaged like Lucky and ARG are currently

Exactly.  If it's truly random then the expected value of 25 GIL per block, with 1920 blocks per day will be very accurate.

the average will be 30 [/pedant

I thought it was 10 - 40...I don't pay attention to GIL. lol

[markm]

You still know once a block is found whether the next block will be a low paying or high paying block.

The general profitability averaging the randomness of the blocks will presumably be generally known.

Also what seems actually more profitable than following the "most profitable coin to mine" sites/systems is to mine a coin when few people are, because it is easier to mine. if it were at that moment listed as highest profitability all over the web lots of people would be mining it so that regardless of its difficulty making it seem profitable the actual amount of hashing competing for the supposedly profitable block would tend to be out of synch with the actual network hash total (which cannot be recomputed until another block is found as it is based on the speed with which blocks are being found).

So the most profitable approach seems to be when the coin is profitable due to price dump coins you already stockpiled, maybe even enough to bring price back down to a point where the coin no longer is shown as most profitable on those sites all over the web; meanwhile mining easy to mine coins to stockpile them so that you will have coins already on hand to dump when they become "profitable".

That way by the time miners mine new coins thinking they are mining the most profitable one they will no longer be profitable because miners who already stockpiled that coin will be ahead of them in line to sell at the profitable price. Plus that profitable price will be more profitable for them because they mined when it was actually easy (fewer miners to divide the pie among) not when it was merely reported on the web as easy (which causes many more miners to hop there making it actually probably more miners to divvy the pie among at that particular time.).

Whatever the price of a coin, getting the more coins per block blocks rather than the less coins per block blocks should be more profitable than just taking your chances as to which kind of block you get or being stuck with only the lower paying blocks due to bonus-blocks-hopping competitors who make the highest paying blocks the ones that have the most miners to divvy up the chance of actually winning it among.

People who mined BBQcoin with a CPU core all year last year made more profit from that CPU core for the year than had they directed that CPU core at "the most profitable coin"...

-MarkM-

[mercSuey]

You still know once a block is found whether the next block will be a low paying or high paying block.

The general profitability averaging the randomness of the blocks will presumably be generally known.

Also what seems actually more profitable than following the "most profitable coin to mine" sites/systems is to mine a coin when few people are, because it is easier to mine. if it were at that moment listed as highest profitability all over the web lots of people would be mining it so that regardless of its difficulty making it seem profitable the actual amount of hashing competing for the supposedly profitable block would tend to be out of synch with the actual network hash total (which cannot be recomputed until another block is found as it is based on the speed with which blocks are being found.

So the most profitable approach seems to be when the coin is profitable due to price dump coins you already stockpiled, maybe even enough to bring price back down to a point where the coin no longer is shown as most profitable on those sites all over the web; meanwhile mining easy to mine coins to stockpile them so that you will have coins already on hand to dump when they become "profitable".

That way by the time miners mine new coins thinking they are mining the most profitable one they will no longer be profitable because miners who already stockpiled that coin will be ahead of them in line to sell that the profitable price. Plus that profitable price will be more profitable for them because they mined when it was actually easy 9fewer miners to divide the pie among) not when it was merely reported on the web as easy (which causes many more miners to hop there making it actually probably more miners to divvy the pie among at that particular time.).

Whatever the price of a coin, getting the more coins per block blocks rather than the less coins per block block should be more profitable than just taking your chances as to which kind of block you get or being stuck with only the lower paying blocks due to bonus-blocks-hopping competitors who make the highest paying blocks the ones that have the most miners to divvy up the chance of actually winning it among.

People who mined BBQcoin with a CPU core all year last year made more profit from that CPU core for the year than had they directed that CPU core at "the most profitable coin"...

-MarkM-

MarkM, I'm on matlab working out the change in probabilities right now.  But here is the random block data I have so far, let me know if I'm missing anything.  Plus, only JKC and LKY are on exchanges, correct?  Not having an exchange rate is making it difficult to figure 'profitability'.  But superblock hopping is something I thought of as well, just not many on exchanges yet to motivate me to look into it further.

starcoin:  100/block per minute

BONUS BLOCKS (All Random)
- 1/120 chance 200-800 (every hour)
- 1/1440 chance 2000-8000 (twice per day)
- 1/20,000 chance 10,000-30,000 (once per week)
- 1/250,000 chance 100,000 (once per quarter)

supercoin:

elephantCoin: 20 sec block time
- 50 coins per block, and with following super random blocks:
      - 1/100 chance, a random 100-500 coins block
      - 1/2,000 chance (about twice a day), a random 1000-5000 coins block
      - 1/30,000 chance (about once per week), a random 10000-20000 coins block
  

Diamondcoin:  1/block per minute
      - Every day on average there are randomly:
      - 10 bonus blocks with 2 diamonds each
      - 3 super blocks with 8 diamonds each
      - 1 super+ block with 30 diamonds
      
Junkcoin:  50/block per minute

    - In regular mining (after first 4 bonus days), there will be 1% chance a block will yield triple of the
      normal coins (e.g. in the first 2 years there's 1% chance you get 150 coins per block).
   - There is also 1/10,000 (0.01%) chance that a block will yield 1000 coins. This is valid for all 12 years
      of the mining.

Luckycoin:  88/block per minute
       Random Super-blocks:
    For the 1st 50000 blocks
    - 5% chances 188 coins/block
    - 1% chances 588 coins/block
    - 0.01% chances 5888 coins/block (so expect 5 such blocks)

    After 50000 blocks
    - 5% chances 2 times the normal coins (i.e. if normal is 88 coins, you get 176 coins)
    - 1% chances 5 times the normal coins
    - 0.01% chance 58 times the normal coins

[gatra]

take a look at this: https://bitcointalk.org/index.php?topic=232741.0

I find the random block reward idea pretty stupid, specially how it's implemented in most sh*t coins. You know in advance the reward, so you could mine only high reward blocks.
In the case of ARG your expected return is 5 per block instead of 3, that's a 66% improvement!
I say it's stupid because not only adds useless variance (supposedly making it more "fun to mine"), and favors reward hopping instead of constant mining, but it also encourages block witholding: if reward depends on previous block, when you find a block you know next's block reward. If it's below average and you know next one is gonna be above average, then you have and advantage by delaying the submission of the new block and starting work for the next block before others. If it's orphaned, you lost a block that was below average, so it doesn't hurt you that much.

[gatra]

At any given moment, what is the chance that one of the chains with bonus blocks is on one?

(If it is still low, not to worry, as people keep spewing out such chains that chance should keep climbing, yes?)

if it depends on the last block's hash (or the height or any other known value) and you know it, you don't need probabilities: you know for sure if next block is a big one or not.
If you want to know the general pobability without knowing last block's hash, then it's constant and it doesn't climb, because each block has a "random" independant probability.

[markm]

Well look on the bright side: they aren't just scamming little old ladies out of their savings by selling scamcoins, they are also scamming miners out of hashing power they could be directing at something else, so if nothing else that crap does help lure away competitors, maybe even in proportion to the sleaziness / scamminess of each individual competitor lured, away from the bread and butter mining of serious miners - miners who are serious about securing a blockchain reliably so services other than scamming / gambling can reasonably be built on it.

-MarkM-

[markm]

At any given moment, what is the chance that one of the chains with bonus blocks is on one?

(If it is still low, not to worry, as people keep spewing out such chains that chance should keep climbing, yes?)

if it depends on the last block's hash (or the height or any other known value) and you know it, you don't need probabilities: you know for sure if next block is a big one or not.
If you want to know the general pobability without knowing last block's hash, then it's constant and it doesn't climb, because each block has a "random" independant probability.

I meant is it unlikely that looping through all such coins checking each to see if its next block will be a high paying one will discover at any given moment at least one such coin, or are there enough now that you will almost certain find, by the tiem you have mined a high paying block of such a chain, that one of the others is already in a "next block is a higher paying one' state so that you already have at least one chain to hop to to find a lucky block?

Maybe it will end up more a matter of how much higher paying, due to typically there being more than one chain whose next block is a lucky block...

-MarkM-

[gatra]

At any given moment, what is the chance that one of the chains with bonus blocks is on one?

(If it is still low, not to worry, as people keep spewing out such chains that chance should keep climbing, yes?)

if it depends on the last block's hash (or the height or any other known value) and you know it, you don't need probabilities: you know for sure if next block is a big one or not.
If you want to know the general pobability without knowing last block's hash, then it's constant and it doesn't climb, because each block has a "random" independant probability.

I meant is it unlikely that looping through all such coins checking each to see if its next block will be a high paying one will discover at any given moment at least one such coin, or are there enough now that you will almost certain find, by the tiem you have mined a high paying block of such a chain, that one of the others is already in a "next block is a higher paying one' state so that you already have at least one chain to hop to to find a lucky block?

Maybe it will end up more a matter of how much higher paying, due to typically there being more than one chain whose next block is a lucky block...

-MarkM-

oh, then it gets tedious to do the math since there are many coins, and each have different block duration times, which are actually geometric distributions. But you don't need to calculate that to implement it. Just hop to the most profitable one and adjust your profitability every time one of the chains has a new block. If 2 chains are in a superblock at the same time, still one will have highest profitability (if tied, flip a coin). If all your chains are in below average blocks, then default to BTC or LTC or whatever.
If you are worried about your ability to quickly switch, you can limit yourself to fewer coins, but I think the calculation and switching can be done very quickly. ARG will be above average 2/5 of the time so I guesstimate you will not be in your default coin most of the time