I noticed something in the rules. On page 212 it says about reactive ice "...untill the decker destroys or deceives it, or otherwise convinces it to go away."

What I can't tell is how one decieves it. Reactive ice tends to do it's thing after the trigger step. Probe for example. Once triggered, the systems suddenly has a huge amount of dice to increase it detection of you. The only way to prevent this is to attack it. There is no current rule for trying to throw a fake password at it or some other authorization.

I have been thinking about make a system test that would let the decker to try just that. But I don't want it to be easy. Combat should be the easier but noisier way. I feel that a really hot decker should be able to do a run without combat in some cases. Real ghost in the machine.

Here's my question. Has anybody else done a house rule that simular? Or does anybody have a good argument why there shouldn't be one?

Couple of different answers to this issue which we have actually addressed in the Idiot's Guide to the Matrix thread.

Deception of Reactive IC is possible by just dropping into Cybercombat and using the Evade Detection Maneuver (or using it multiple times). There's a pretty good example on page 8 of the above thread where the decker does just that to some Scout IC. However this assumes you've detected the triggered Reactive IC in the first place.

On the other hand given the description of Probe (and Scout), especifically the bit about detecting "operations performed by unauthorized programs", I've would assume that possessing a previously Validated Account means you have a Valid ID to throw at the IC which will "turn it off".

the wording smell like some leftovers from sr2 (pre vr2.0) as then you used the deception util to fool access ice to let you past (access ice was about hte same as a password lock).

im with synner on the fact that if you have say a valid user account you can trow that to the ice and get away with some commands (check matrix, mutch info on accounts there)...