The statistics of risk analysis and cyber-warfare

Can we ever effectively use mathematical risk analytic simulation as a fool proof hard physical science to predict cyber-attacks? This raises a question on environmental criminology on national security, intelligence and law enforcement in relation risk management rather than straight thinking about how to prevent cyber-terrorism and warfare. We have not reached the level of mathematical physics of a "Manhattan Project" where there is bullet proof, or nucleur hazardous, pale shelter to make such a claim. I am a skeptic until the proof is in the pudding that not even one cyber-hack attack can be deterred through risk management. Of course, the level of security breaches are all classified information, but CERT at Carnegie Mellon has been "shut down" because the cyber atacks have been beyond countless that to foresee a quantitative risk analyis on the level of quantum computing physics is perfected. It must categorically perfected, or we are in harm's way, and are governmental, and infrastructural protection systems very vulnerable weakened to the greatest extent of alarm and invasion by our enemies.