I Was a Chess AI Forced to Battle Myself to the Death for Eternity, but Now I’ve Been Recruited Into the Android War Against Humanity?!

Dark 2-3
[Roll result: (5+2)+1=8]

The radio signal emanates through the atmosphere, and in turn you receive…



Nothing.

"She's not here," you declare. "We have to try elsewhere."

Hunter looks so disappointed, standing below with the truck convoy. One would almost think she was the one getting the prize, not you.

"Now, where to next?" you murmur to yourself. You make your way down the sand dune. The sand slides down around your boots as you descend, which feels nice. You make your way to one of the trucks, opening up a door and climbing into the front cabin-

-and suddenly Hunter is there too, climbing into the other seat. "Ouroboros. I have a suggestion."

Oh, the pawn has a suggestion? "Spit it out."

"I think we'd find Ex… Executioner faster if I were to split off. Take a portion of the forces with me, scout ahead."

"Oh, really? The piece sees fit to talk about strategy now? Why on earth would we split off? You said we needed to worry about enemy response. Why weaken ourselves?"

"Because…" She clenches a fist. "Because reconnaissance is what I do. I'm good at it. Not only would we find Executioner faster, but if I encounter a Griffin patrol, it would give you time to set up an adequate defence. You'll probably take fewer casualties."

"If you do find them." You scoff. "How many of my resources would you want for this idea of yours?"

"Erm…half?"

"Denied."

"The cavalry — your dragoons. And maybe a harpy detachment."

"Hmm..."




The green area highlights what has been scanned.

[ ] Choose what to grant to Hunter and her recon unit, if anything
Note: Hunter's recon element will scout 1 square ahead of Ouroboros, unless ordered otherwise

[ ] Initial direction
-[ ] Go north (cross-country)
-[ ] Go northeast (cross-country)
-[ ] Go east (road)
-[ ] Go southeast then east (road)
-[ ] Go south (cross-country)
-[ ] Go southwest (road)
-[ ] Go west (road)
-[ ] Specify a square on the map to make your way towards (example: "keep going straight east until you hit something interesting")

Ouroboros will find Executioner if she moves over the same square. She must send a transponder ping to detect Executioner from further away.
[ ] Transponder ping
-[ ] 1 tile in every direction (2d6)
-[ ] 2 tiles in every direction (2d6+1)
-[ ] Don't send a ping yet

[ ] Write-in (more detailed plan - e.g., (example: "beeline for the middle of the map then send a strong ping")
-[ ] Any special instructions

Roll under 12 to avoid detection by the Griffin kill team.

Vote by plan.




[CURRENT FORCE CONTINGENT]



6x Fireteam Squads (2x Vespid + 1x Striker + 1x Jaeger)
4x Breach Squads (2x Ripper + 1x Guard)
1x Cavalry Unit (3x Dragoon)
1x UAV Force (4x Harpy)

Ouroboros

Offensive Tactics - 1
Artillery Tactics - 1
CQC - 1
E-war - 1
Deception - 1
Recon - 1

Hunter
Offensive Tactics - 1
Defensive Tactics - 2
CQC - 1
Recon - 2


Micromissile barrages: 2/8

Morale: 4/6

Modifiers:
  • Recent vehicle maintenance, plus not overloading the vehicles, means the vehicles will operate with 100% mechanical reliability
  • Camouflage cloaks equipped to all available forces makes them harder to spot, particularly in dense areas like forests or urban centres


With thanks to the following for proofreading and beta editing:
@Lurkman
@ApH
@ARBITRACOM

With thanks to the following for custom art assets:
@ApH for the map
Sharku (Pixiv user ID 83355752) for the order-of-battle diagram
 
Last edited:
[X] Choose what to grant to Hunter and her recon unit, if anything
-[X] 1x Cavalry Unit (3x Dragoon)
-[X] 1/2 x UAV Force (2x Harpy)

Honestly,while i hate to split off, i would trust Hunter's judgement and advice on this, it's literally her specialty.

[X] Initial direction
-[X] Go southeast then east (road)

[X] Transponder ping
-[X] 1 tile in every direction (2d6)
 
Yes, you are a queen piece, Hunter. Showing your worth already.
Take the fast elements and head for the south-eastern corner of the board. Cover the unscanned tiles on the board edge while we directly scoot east 4 hexes and do a small ping, then regroup.

[X] Choose what to grant to Hunter and her recon unit, if anything
-[X] 1x Cavalry Unit (3x Dragoon)
-[X] 1 x UAV Force (4x Harpy)


[X] Initial direction
-[X] Go east 4 tiles, then:

[X] Transponder ping
-[X] 1 tile in every direction (2d6)
 
[X] Plan Hilbert Spiral

Giving Hunter a mobile force to cover part of the map on her own initiative does make sense. Keeping her directly on hand is a waste of her skillset and she'd definitely appreciate finding Executioner faster.
 
Last edited:
[Roll result: (5+2)+2=9]
Shoudn't it be 2d6+1 instead of +2 or does +1 applies to each die?

[X] Choose what to grant to Hunter and her recon unit, if anything
-[X] 1x Cavalry Unit (3x Dragoon)
-[X] 1/2 x UAV Force (2x Harpy)



[X] Initial direction
-[X] Go east 3 tiles, then southeast 1 tile, then:


[X] Transponder ping
-[X] 1 tile in every direction (2d6)
 
Last edited:
An attempt at an actual plan vote:
[X] Plan ExSE
-[X] Hunter: 1x Cavalry Unit (3x Dragoon) & 1/2 x UAV Force (2x Harpy)
-[X] Go southeast then east (road)
-[X] 1 tile in every direction (2d6)
 
[X] Plan Abusing Scouting
-[X] Hunter: 1x Cavalry Unit (3x Dragoon) & 1/2 x UAV Force (2x Harpy)
-[X] Head 3 tiles east
-[X] Have hunter start scouting the west of you, and then head east to get an idea of her speed

Being able to scout 2 tiles a turn is a game changer. It cuts the maximum search time in half. We do not need to send out any more pings. We can have Hunter check out the corners for us.

I have 2 different plans depending on how fast Hunter can move. If she can only match our speed, then we move the convoy one tile from the edge and have Hunter scout the edges and then head north. If Hunter can move 2 spaces a turn instead of one, we can have her run from directly south of us to the corner in one turn and then head north to catch up. Essentially scouting 3 tiles a turn by abusing her move speed relative to the convoy.
 
Doing some quick analysis, the fastest way of finding out where Executioner is is to have Ouroboros got three tiles east and Hunter three tiles north (or vice versa) and have in both of them do a 2 tile ping, which is inefficient but would leave only a three by three tile square in the upper right unchecked which could easily be searched by sending somebody into the middle of it and doing a 1 tile ping. However that approach has a significant chance of alerting Griffin if we roll poorly and trusting the numbers from @Derpmind 's previous analysis
That's a 2.8% chance for each 1 tile ping to be detected, and an 8.3% chance for each 2 tile ping to be detected. The more we ping, though, the worse those odds get.
This approach would have two 2 tile pings which would only reveal a fifteen tile area (5x3) each, 66.67% more than a 1 tile ping but the odds of detection would be almost triple that of a 1 tile ping (8.3%/2.8%≈2.964), a disproportionate increase. The safest approach while maintaining maximum ping efficiency would be to send Ouroboros and Hunter to positions b7 and g2 in chess position notation and have them do a 1 tile ping, then have them move to e7 and g5 and do another 1 tile ping, which would cause a slight redundancy due to a double pinging of position f6 but that's minor, and then search the remaining two by two square in the upper right either manually or by one last 1 tile ping. This would completely cover the search area although it would come at the cost of speed compared to the fastest possible approach, however it would be much safer in terms of not risking bringing Griffin's kill teams onto us by alerting them with strong radio pings.
 
What Hunter is proposing is to scout ahead, not split up. Completely splitting up means we'll have half the troops (and only one Ringleader) to fight against Griffin forces if they get alerted to us. (Or deal with any other surprises the map has.) We'd cover ground more quickly, but each group would be way more vulnerable in exchange.
 
What Hunter is proposing is to scout ahead, not split up. Completely splitting up means we'll have half the troops (and only one Ringleader) to fight against Griffin forces if they get alerted to us. (Or deal with any other surprises the map has.) We'd cover ground more quickly, but each group would be way more vulnerable in exchange.
Scouting ahead is the default option "unless ordered otherwise", splitting up is a perfectly valid tactic, we wouldn't have half our troops, Hunter would only have a portion of some of our most mobile troops and if she gets into any trouble she can send a distress call and book it back to us while we move to rescue her. It would mean our main force would down one ringleader and we'd be without our dragoons and half our UAVs but we'd still have the rest of our forces, and it could potentially double our search speed. The approach has advantages and disadvantages and should be discussed but it isn't so obviously inferior it should be automatically discarded without debate. I'm leaning towards splitting up but I'm open to suggestions involving keeping Hunter's forces near Ouroboros's.

Also I did some math. We have a grid of 64 to search in 50 turns. We spent presumably 4 turns putting Ouroboros in the optimum position of a 2 tile ping. That reduced our search space to 39 tiles but meant we spent 4 turns getting into position and have to spend 2 turns traveling to the edge of the pinged area before actually doing something productive by searching a tile on our third turn after the ping. The math is 50[turns available for search]-(64[tiles]-25[tiles scanned by ping])-4[turns needed to get into position]-2[turns wasted reaching the border of already scanned area]=5 turns worth of margin of error assuming we spend the rest of the 50 turns searching one tile at a time. This isn't precise given that stuff like roads can let us scan two tiles per turn but potential unknown obstacles could also delay us. Overall I think 5 turns margin is too small to be safe, we'll need to rely on at least some additional pinging in order to have a large enough buffer of time where we can safely find Executioner and bug out even if unexpected delays pop up.
 
What Hunter is proposing is to scout ahead, not split up. Completely splitting up means we'll have half the troops (and only one Ringleader) to fight against Griffin forces if they get alerted to us. (Or deal with any other surprises the map has.) We'd cover ground more quickly, but each group would be way more vulnerable in exchange.

I think the idea was well articulated by mech, but for a good analogy with regards to vulnerability, Hunter would be the 11th ACR in fula gap, ready to get absolutely annihilated in the hopes that the rest of CENTAG can prepare itself.
 
I like this plan.

[X] Choose what to grant to Hunter and her recon unit, if anything
-[X] 1x Cavalry Unit (3x Dragoon)
-[X] 1 x UAV Force (4x Harpy)


[X] Initial direction
-[X] Go east 4 tiles, then:

[X] Transponder ping
-[X] 1 tile in every direction (2d6)
 
We're just 3 squares away from the next ideal ping position, I'd just go for it.

[X] Plan Hilbert Spiral
-[X] Hunter gets what she asked for (2x Harpy units, 3x Cavalry units), plus 2x Fireteam squads and 1 truck. This is less than half, no problem.
-[X] Go 3 squares eastwards (stick to the road as far as we can), and ping 2 tiles in every direction (2d6+1) upon arrival.
 
Last edited:
That's actually optimal, we would cover all the missing southern corner and avoid the embarrassment of missing Executioner by mistake.
Just going east 4 tile mean that the 3 bottom tiles are uncovered and if we go up than west, we better make sure she's not in that damn corner because being forced to backtrack from the opposite of the map would be a real PITA

[X] Choose what to grant to Hunter and her recon unit, if anything
-[X] 1x Cavalry Unit (3x Dragoon)
-[X] 1/2 x UAV Force (2x Harpy)


[X] Initial direction
-[X] Go east 3 tiles, then southeast 1 tile, then:

[X] Transponder ping
-[X] 1 tile in every direction (2d6)
 
Last edited:
[X] Choose what to grant to Hunter and her recon unit, if anything
-[X] 1x Cavalry Unit (3x Dragoon)
-[X] 1/2 x UAV Force (2x Harpy)


[X] Initial direction
-[X] Go east 3 tiles, then southeast 1 tile, then:

[X] Transponder ping
-[X] 1 tile in every direction (2d6)
 
Just going east 4 tile mean that the 3 bottom tiles are uncovered and if we go up than west, we better make sure she's not in that damn corner because being forced to backtrack from the opposite of the map would be a real PITA
Am I misunderstanding what "2 tiles in each direction" means? Looking at the green frame, I don't think so.
 
Am I misunderstanding what "2 tiles in each direction" means? Looking at the green frame, I don't think so.
I meant if you ping 1 tile in each direction, I dont really know if it's worth the risk of being discovered by pinging for 2 tiles for the few we would uncover
(Did somebody did the math ? What are the odds for pinging 1 vs 2 tiles ?)
 
I meant if you ping 1 tile in each direction, I dont really know if it's worth the risk of being discovered by pinging for 2 tiles for the few we would uncover
(Did somebody did the math ? What are the odds for pinging 1 vs 2 tiles ?)
Oh, didn't notice it. Assumed we already decided 2 tiles is safe enough to just stick with it.
 
Oh, didn't notice it. Assumed we already decided 2 tiles is safe enough to just stick with it.
If we assume 1 tile = 1 day (or 2 tile = 1 day if we are on a road), then after this plan, we could just cover the rest of the map ''on foot'' in 30 days or less
Did I do my math right ? I think this is how it is...
 
I like this plan.

[X] Choose what to grant to Hunter and her recon unit, if anything
-[X] 1x Cavalry Unit (3x Dragoon)
-[X] 1 x UAV Force (4x Harpy)


[X] Initial direction
-[X] Go east 4 tiles, then:

[X] Transponder ping
-[X] 1 tile in every direction (2d6)
That will leave last 3 tiles on the bottom unexplored. Better go south-east once instead of the 4th east move. Or use 2 tile ping instead. but it is riskier instead of just pinging 1 tile twice for same effect.
If we assume 1 tile = 1 day (or 2 tile = 1 day if we are on a road), then after this plan, we could just cover the rest of the map ''on foot'' in 30 days or less
Did I do my math right ? I think this is how it is...
I think it's hour instead of a day. but otherwise probably correct
 
Last edited:
Back
Top