Page 1 of 1

not expecting the expectedUNQ chains

PostPosted: Fri Jul 23, 2010 10:39 am
by blazer
i'm using this formula from sc00bz cheat sheet
expectedUniqueChains ~ chainCount / (tableWorkFactor / 2 + 1)

chain length: 150 000
my current chain count: 50 000 000
after perfecting: 7 080 082
merges: 42 911 282, 85.84%

odd the numbers don't seem to add up since 50 000 000 - 42 911 282 <> 7 080 082

anyhow we'll skip that part

keyspace: 7.35092E+11
table Work Factor: 10.2
expected Unique Chains: 8 194 836.455

seems like i'm roughly 1 million chains short, is this due to the new reduction function? am i just doing it wrong? is there some new formula to use?

Re: not expecting the expectedUNQ chains

PostPosted: Fri Jul 23, 2010 1:03 pm
by Sc00bz
Having more merged chains is a sign of a non-evenly distributed reduction function and/or hash function.