Steven James - The Rook

I was typing in my scribbles when Lien-hua tapped at the door.

“It’s open,” I called.

“Aina can’t come.” Lien-hua had her computer with her. She sat on the edge of the bed and flipped her laptop open. “The 911 caller was a dead end, but she’s busy evaluating potential ignition systems from the previous fires. She gave me the codes.”

“Good. Give me a minute, then I’ll need them.” I clicked on one of the icons on my screen, and a map appeared with each of the arson sites marked by a small flickering flame. “See the placement of the fires?” I punched a couple of keys, and a series of red lines laced the city, highlighting the streets and highways of San Diego.

I traced one with my finger. “Roads.”

“Oh. Is that what they are?”

“Sorry. I’ll try to stop stating the obvious.” I tipped the screen so she could see it. “Now, if you look at the distribution of the fires, what do you notice?”

She studied the map carefully but shook her head. “Nothing.”

“Neither did I, until…” I uploaded the trolley routes and overlaid them against the map of the fire locations. A couple more keystrokes, and my computer-mapping program calculated the distances between the fires and the nearest trolley depots and flashed the totals in the lower left-hand corner of the screen. “See? The fires all occur within two hundred meters of a trolley stop, usually the Orange Line or the Blue Line.”

She gave a soft gasp of acknowledgment. “He’s starting the fires and then boarding the trolleys to get away.”

“I think so. Let’s check the timing. Look up the trolley schedule for me, and I’ll compare it to the times of the fires.”

She surfed to the San Diego mass transit site and read off the trolley arrival and departure times, and I punched them in. Hit enter.

A crisscross of lines appeared along with a detailed timing chart on the right-hand side of the screen.

“It fits,” I said.

She gave a long, soft whistle. “So, except for last night’s fire, each of the other fires was reported between five minutes prior to a scheduled trolley departure or eight minutes after the trolley left. How long would it take to walk two hundred meters?” she asked.

“I don’t know. Maybe three to five minutes. We also need to factor in the time it takes for the fire to grow, get noticed, and be reported. We can have some officers walk the distances to confirm the timing. But it looks like he didn’t park. He rode.”

“So, if that’s how he’s getting away…” Then it struck her.

“Access codes.”

“Right,” I said. “Every trolley depot is monitored by video surveillance. Let’s see how photogenic our arsonist is.”

I surfed to the city’s mass transit DVA, the Digital Video Archives, and logged in with the access codes Lien-hua had gotten from Aina.

I was annoyed to find that the older footage had been deleted, so only the last six months were available. That left us with only eight fires, and I wasn’t sure if it would be enough, but I downloaded the footage from the trolley depots that correlated to the time and location of those eight fires to give it a try. “This gives me a good chance to try one of the new toys Terry’s designing,” I said. “You know him, right? Terry Manoji? My buddy from the NSA?”

“We’ve met,” she said coolly.

“You don’t like him?”

“He hit on me at a conference once. Told me I reminded him of someone, striking resemblance. Not something you tell a girl.”

“At least he has good taste.”

She didn’t smile.

“Don’t worry. He told me he’s seeing someone. I haven’t met her yet, a woman from the West Coast. Anyway, he sent me this new program he’s designing. Watch this. It’s called CIFER.”

“What does that stand for?”

“Characteristic Inventory for Focused Evaluative Recognition.”

She shook her head. “The government could save millions each year by not spending so much time coming up with acronyms.”

I opened the program and pulled up the file menu. Lien-hua didn’t seem too impressed. “It looks like a facial recognition program.” “More like identity recognition.” I was getting juiced. It felt good to be cranking forward on a case. All I wanted to do was find this guy and then I could spend some quality time with Tessa.

And maybe a little with Lien-hua.

Who knows.

“This program combines facial structure and characteristics, just like traditional facial recognition software, but also evaluates height, weight, pace, posture, nonverbal communication, and spatial movement patterns. Even voice recognition, if we have audio. Plus it amplifies Sagnac Interference to block GPS tracking of the user’s location. It’s for field ops, just in the testing stages right now. This is the only other copy, but from what I’ve seen so far-”

My computer beeped and displayed seven faces in separate panels across the screen.

However, none of them looked anything like each other.

One man appeared to be a transient, another a business executive, the third, a jeans-clad man with a prominent beer gut, and the four other men all looked unique as well. Beard, no beard. Wavy flaxen hair, then black. All were similar height, but that was about it.

“Strike one,” Lien-hua said. “It’s a different person each time.”

“I… I don’t understand,” I mumbled.

“Maybe the program isn’t all it’s cracked up to be.”

Or maybe I’d improperly weighted the importance of some of the signifying trait values. “Wait, let me try something.” I tapped at the keyboard and brought up the similarity index, altered the values.

The same men appeared.

I glanced at the legal pad, then typed in a few default changes to CIFER’s index, and punched enter.

Still the same men.

I was ready to scrap the whole idea, but then I noticed something.

“Hang on, Lien-hua.” I punched a button, and the videos began to play in slow motion, the time counter flashing at the bottom of each image. “There.”

She studied the screen. “What?”

I hit “pause,” then “play” again. “Look.”

“What is it?”

“Just a sec.” I adjusted the sharpness and saturation of the pictures to make them less grainy, rewound to the beginning, and hit

“play” again. She still looked confused.

“Do you want me to play it in-”

“Wait.” She studied the videos carefully. “I got it. Disguises, right?”

“Yes.”

“And pedal supination with his left foot?” she said.

“I think that’s what it’s called… when he rolls his foot outward after each step?”

“Yes,” she replied. “That’s supination. When your foot rolls inward, it’s called pronation.” And then she added, “I pronate.”

“I won’t tell anyone. I promise.”

“How kind.” She reached across my lap and hit “enter” to play the videos again. “Yes,” she said. “His is very pronounced. Maybe from an old injury. A broken ankle, something like that.”

“Wait. Hang on.” I pointed to one of the videos on the bottom of the screen. “This guy is limping on his left, this guy isn’t.”

“You can fake a limp,” she said, “but you can’t fake the way your foot turns when you lift it after taking a step, or maybe you could, but almost no one would think of doing it.” Lien-hua shook her head. “Whew. It may surprise you, but I’m actually tracking with all this. But I have to say, you’re telling me your computer thinks it’s the same guy because of the way his foot turns when he walks?”

“The way he walks, the spatial relationships with the other people at the depots, arrival times compared to the distance of the depots from the fires, facial similarities, arm length, bone structure-”