Topic: A way to fix Doge (Read 456 times)

Dogecoin has a problem recently. 

Short version of the story; in Doge there are variable rewards per block.  The variability is based on the hash of the previous block.  Some miners (you could call them "cheaters" but they're just exploiting the rules as written) have caught onto this and are only jumping in and mining when a larger reward is upcoming.  The result is that people mining steadily (you could call them "honest") get lower payouts, the miners who are jumping in only for the higher-reward blocks are getting more than a fair share for the amount of work they do, and the network hash rate is swinging wildly up and down. 

One possible "fix" is to throw out variable rewards entirely and go with a fixed block award.  But that sacrifices a feature of the coin that some folks like. 

Otherwise, the problem here is that the miners jumping in for only the higher-reward blocks *know* which blocks are higher reward while they still have time to mine in those blocks.  It's supposed to be like a lottery, you shouldn't be able to count cards.

A simple way to fix it without sacrificing the feature is to base the variable reward on the low bits of the current block's hash.  Because the block is formed when "enough" of the high bits are zero, mining means using all your hashing power to get the high bits right; If miners try to also select for low bits that yield a larger reward, they'll have to throw out (not claim) most of the blocks they find.  And why would they do that?  The odds of finding a high-reward hash, per minute of hashing, are the same whether working on the current block (after throwing out a potential block) or working on the next block (after claiming the current block), so there is no gain to be had by throwing out a potential block.

Anybody see any logical flaws with this?