AI, hidden and nearest

Well now my understanding on the procedure has completely flipped again based on that example. Bloody hell.

What is your understanding then?

Iā€™m not sure, it seemingly changes every time I look at this thread :smiling_face_with_tear:

Based on everything discussed prior, I thought the detection/spotting would only take place if the hidden model was the target of an action (not response), i.e. an attack action for the purposes of adversaries unless they one day get a different action that would target an enemy (such as pickpocket etc.). Also that the detection test would take place at the point at which that Action commenced (which would also be the same for non-adversary models) rather than at the start of the Response.

So in the most basic terms; a Response where the ā€˜targetā€™ is hidden doesnā€™t require a successful detection test, only the Actions that require a target (such as an Attack Action) do.

Example of what I thought happens:

  • Aggressive Adversary with ranged attack rolls a Move Response. Target Priority is Nearest. The nearest enemy model happens to be hidden but in clear line of sight.

  • First Action is a Move towards that model (without taking a detection test; an example of the instinctive ā€˜adversaries knowing roughly where the enemy isā€™); Move Actions donā€™t have a ā€˜targetā€™ even though a Move Response does and itā€™s the Action part that determines the need for a detection test. The Adversary therefore makes a Move Action towards the hidden enemy (sprinting if needed).

  • Quick check to see if Target Priority has changed now the Adversary has moved (99% of the time this wouldnā€™t be necessary; Nearest is still likely to be Nearest unless, say a previously blocked LOS enemy was now visible).

  • Second Action is a ranged Attack. As the hidden model is in range of the Adversaryā€™s weapon and it is now specifically an Action (not a Response) that is targeting the enemy, a detection roll is now made (potentially made easier now the Adversary is closer).

  • If successful then the enemy is spotted the Attack Action is carried out against the previously hidden model. If failed then the enemy remains hidden and the Adversary would check if the next Nearest model was eligible for an Attack Action and so on.

2 Likes

Hey everyone,

Weā€™re going to take a good look at this and fold it into the rulebook rewrite weā€™re currently working on. Weā€™re not ignoring you, just to let you know!

1 Like

Thanks for the update; look forward to seeing the revision popping up for a playtest.

I think Gavin and JimmyW sum it up nicely but Iā€™ll add my two cents. I always make them try to attempt to engage the nearest hidden model, regardless of LoS. If they fail that, then fail all other models (which Iā€™ve never had happen), then I would pick the nearest and have them to move to within 1/2" or close as they can get. I do this even if they fail the first time and there are no other good targets. They failed to see the target but they heard it and have moved towards it but cannot engage it. I find that a reasonable compromise that works well in-game.

1 Like