First off, in order for a plan involving bait to work, you need to keep the knowledge that bait exists limited to as few people as possible. You sabotaged your own plan by announcing it publicly.
Second, bait has to be targeted specifically to the culprit's M.O. Random volunteers aren't going to cut it. There are a ton of newcomers in the city. We have to find out what in particular attracts the culprit to a victim, whether it's something about the victim, a time of day, a type of location... Unless we can narrow it down, we have no idea whether the murderer will take the bait, and therefore we're wasting manpower that could be used to run a simple and more effective patrol.
Third, there's been evidence to suggest we're not dealing with a single murderer but an organization. That means your backup is going to find themselves ambushed before the victim's even touched, because their backup will already be circling the perimeter.
First off I've been lead to believe this Network is secure from third parties. If that's not true than yeah, this whole thing it moot.
Second we've been looking at these attack patterns for months, and only very limited MO's can be established from the information we have. Age, gender and appearance seem largely irrelevant, and established fighters have been taken down.
And third yes, we do have reason to believe there's more than one. In fact we have some video evidence that there are likely three of them, of varying degrees of abilities. I can refer you to an earlier post of mine regarding all of that information. For backup to be ambushed, they still have to -know- we're coming. If we assume they outnumber, outmaneuver and outgun us in all scenarios, then we defeat ourselves before we've even begun.
Okay. I'm just going to get this all out right now. I never say anything like this... but thanks. If it weren't for you, I probably wouldn't have gotten better.
I know announcing it is hardly ideal, but I've been through the network and I'm pretty certain it's secure against third parties without severe effort. And if they can put in that kind of effort, none of what I post on here would be safe anyway.
I agree, there are a lot of variables, but I've had no luck narrowing them down, and the longer we delay the more people die. Casting a wide net seems to me to be the best way we can try and catch them if they strike, or maybe spot them when they retreat to wherever it is they go.
That's why the bait needs to be trained at withdrawal and delay, people who can buy themselves some time. And if we have more "ninja" then just the two I've met, what we have -might- be enough.
My lady, I would happy to be the first volunteer. I am a first class swordsman specializing in quick deadly attacks and most importantly the stealth and speed you asked for. My training as a swordsman comes from being in a surprise attack squad which I lead. I am a ventriloquist capable of making my voice come from many different directions at once to confuse my foes. As a skeleton I'm incredibly light weight and do not have much to slow me down, and I am quite difficult to kill. I can move without making a single sound, and if by the slightest chance it was possible for someone to spot me, they'd likely mistake me for some form of nightmare.
Comments 124
Reply
Reply
First off, in order for a plan involving bait to work, you need to keep the knowledge that bait exists limited to as few people as possible. You sabotaged your own plan by announcing it publicly.
Second, bait has to be targeted specifically to the culprit's M.O. Random volunteers aren't going to cut it. There are a ton of newcomers in the city. We have to find out what in particular attracts the culprit to a victim, whether it's something about the victim, a time of day, a type of location... Unless we can narrow it down, we have no idea whether the murderer will take the bait, and therefore we're wasting manpower that could be used to run a simple and more effective patrol.
Third, there's been evidence to suggest we're not dealing with a single murderer but an organization. That means your backup is going to find themselves ambushed before the victim's even touched, because their backup will already be circling the perimeter.
Reply
Second we've been looking at these attack patterns for months, and only very limited MO's can be established from the information we have. Age, gender and appearance seem largely irrelevant, and established fighters have been taken down.
And third yes, we do have reason to believe there's more than one. In fact we have some video evidence that there are likely three of them, of varying degrees of abilities. I can refer you to an earlier post of mine regarding all of that information. For backup to be ambushed, they still have to -know- we're coming. If we assume they outnumber, outmaneuver and outgun us in all scenarios, then we defeat ourselves before we've even begun.
Reply
Reply
Wha...wait. Are you the one who got the transfusion? Yeah, all the time. Nasty side-effect.
Reply
Reply
You're welcome. I'm glad it got to someone who needed it, I was kinda worried SERO might make off with it.
Reply
Reply
You're certainly not the first to say so. Thoughts on how to improve it? Tighten it up? Maybe an entirely new one?
I'm more than open to suggestions.
Reply
There are too many variables. That's the biggest problem. Things should be more controlled, if you're setting a trap.
Reply
I agree, there are a lot of variables, but I've had no luck narrowing them down, and the longer we delay the more people die. Casting a wide net seems to me to be the best way we can try and catch them if they strike, or maybe spot them when they retreat to wherever it is they go.
Reply
Reply
Reply
Reply
Reply
Reply
[Just kind of staring.]
Reply
Would these eyes lie to you? Oh wait, [The skeleton leans back in his chair and laughs] I don't have any eyes. Yohohoho!
Allow me to introduce myself my lady, my name is Brooke the bard, and it's an absolute pleasure to meet you. May I please see your panties?
Reply
This isn't the time for jokes.
Reply
Leave a comment