[tor-talk] Why adding Guard-Exit (EE) node in Tor yeild more catch probability than adding guard and exit node separately?

[Matthew Finkel]

On Tue, Aug 05, 2014 at 10:40:17AM +0900, saurav dahal wrote:
> I am trying to observe the catch probability i.e. probability that a client
> selects my added guard as well as exit node while making a circuit.
> 
> First I inserted certain number of guard nodes and certain number of exit
> nodes in Tor network and performed the simulation in Shadow simulator.
> After completion of simulation, I calculated the catch probability.
> 
> Then again I added the same number of Guard-Exit (EE) node and perform the
> simulation and calculated the probability.
> 
> I found that catch probability for EE nodes are much higher than those of
> guard and exit node separately.
> 
> Could anyone please explain why this happened?

Hi Saurav,

Can you clarify what you mean by Guard-Exit nodes? If I understand
correctly, you ran a simulation where you had x nodes which had the
Guard flag and y exit nodes, then you ran another simulation where
you had (x+y) exit nodes which also had the Guard flag? Is this
correct?

Thanks,
Matt