|
Title: Selfish mining paper math analysis Post by: y-c-c on November 07, 2013, 11:13:23 AM I was going to post this in the main thread (https://bitcointalk.org/index.php?topic=324413.0) about the selfish mining paper (i.e. "Bitcoin is doooomed" paper) but realized as a newbie I can't post anywhere! Posting here instead.
Basically I saw a lot of people criticizing the paper for the wrong reasons (e.g. the paper actually doesn't require the new block to propagate faster than the honest block to work), while not really going through the math and derivations in a rigorous way. Well fear not! Turned out the math was junk as well. Here's the reddit post I made about it: (http://www.reddit.com/r/Bitcoin/comments/1q3dna/bitcoin_broken_paper_flaw_analysis_with_math_ie/) Basically one key equation in the formulation (Σpk + p0' = 1) is wrong since it should be Σpk = 1 instead. This has an effect of invalidating all future equations the authors used for calculating the percentages. Haven't done the math yet but I suspect the actual required number for the attack will be much higher than the original 25/33%. Title: Re: Selfish mining paper math analysis Post by: MaxBTC1 on November 07, 2013, 12:54:11 PM Nice find, the paper has been partially ridiculed already. Update us on the maths if and when
|