[RC5] strategic starting point in the upcoming DES contest

Wouter Janssens wj at vvksm.be
Thu Dec 3 11:36:03 EST 1998


Good point, but wrong. Because here starts the psychology:

If EFF knows that we will use these tactics, shouldn't they pretend to 
start at the very first block, but in fact start elsewhere. If they know 
where we start and how we are dividing the space, they have a 100% chance 
of winning against us. They're faster, so the can stay ahead of us. The 
danger is neither of us will get the price. The advantage of EFF is they 
end first.

EFF will never tell us how they are going to work (supposing they're 
competing with us).

If it's about winning the price, and getting as much money as possible, we 
should work together and divide the space. If that's impossible, we should 
never mention in lists like this one what our plans are.

Wouter Janssens


he mentioned starting in the middle (assuming EFF started at the beginning) 
and working towards both ends, but completly ignoring and conceding the 
first 33% of the keyspace. in this way, they would automatically win all 
contests where the solution was in the first 33%, and we would win the 
other 67%, guaranteed.

i totally agree with this. i think its generally understood that if we 
start at the same place EFF does, we don't have a chance.  i've been 
keeping up with the lists, and from what i've heard and what i can think 
of, this is the best one yet.

If we start at the middle, working 1/6 of the keyspace towards the front, 
and the whole back half, EFF will win if the randomly chosen key falls in 
the first 1/3, but statistically speaking, we can predict victory if we 
follow this path.  The only thing to it is finding out where exactly EFF is 
planning on starting, and then molding this concept to avoid the first 1/3 
of the keyspace that they will compute.

Great idea Steve!

Brass Monkey
Distributer


--
To unsubscribe, send 'unsubscribe rc5' to majordomo at lists.distributed.net
rc5-digest subscribers replace rc5 with rc5-digest



More information about the rc5 mailing list