they'd made on the interviewer. Even more, I wanted to see if any of them had anything to do with hyperdrive production. None did, nor had any of my investigators red-flagged any as a potential suspect. With no way to narrow down my search for a hyperdrive connection, I concentrated on the schitz angle. There were about five dozen people with severe schitz tendencies on their medical records in the Swarm. I cut that in half by looking only at males on the theory that the killing was a sex crime. By midafternoon I'd eliminated all but eight of them for having the wrong physical description, for not being on Tiamat when the crime was committed or some other disqualification. I ran a detailed movement analysis on the remainder, tying up my hardware for over an hour. Three were eliminated, none were implicated outright. What to do?
I considered having the remaining five hauled in so I could ask a few questions. I didn't have to haul them in, my desk performs voice stress analysis perfectly well over the screen, but I prefer to talk to a suspect one on one. It makes the interview more personal, raising the stress level and giving the software something to work on. Besides, I like to see the reactions for myself and come to my own conclusions. The computer isn't infallible and neither am I. Using both techniques cuts the error rate.
If it worked I could wrap the case up that afternoon, if it didn't at least I could eliminate those five and get to work finding a new line of investigation. The risk was tipping off the murderer. If one of the suspects bolted, we'd have our man. Then we'd just have to find him. My instincts warned me that we never would. He'd disappear into the Swarm or the mountains down on Wunderland. Maybe in a year or ten the Provopolizei would catch him sniping politicians in Munchen for the Isolationists. The Isolationists would suit k
I considered having the remaining five hauled in so I could ask a few questions. I didn't have to haul them in, my desk performs voice stress analysis perfectly well over the screen, but I prefer to talk to a suspect one on one. It makes the interview more personal, raising the stress level and giving the software something to work on. Besides, I like to see the reactions for myself and come to my own conclusions. The computer isn't infallible and neither am I. Using both techniques cuts the error rate.
If it worked I could wrap the case up that afternoon, if it didn't at least I could eliminate those five and get to work finding a new line of investigation. The risk was tipping off the murderer. If one of the suspects bolted, we'd have our man. Then we'd just have to find him. My instincts warned me that we never would. He'd disappear into the Swarm or the mountains down on Wunderland. Maybe in a year or ten the Provopolizei would catch him sniping politicians in Munchen for the Isolationists. The Isolationists would suit k