When an AI rolls M being Cautious, and its objective is in a straigth line, without any cover in front or to its sides, and its preferred attack is ranged (and cannot block, like an Skeleton archer, and cannot hide because of the LoS), does it go to the enemy and stay at 1" and ends its activaction losing the 2nd action, being more “aggresive” to attack next time? or does it go to a cover, even if the cover is in the opposite way to its objective, and try to hide, being more “cautious”? The rulebook says it would go near and stay there without blocking or hiding in this case, and it would be everything but no cautious…
Hey @DK-dark
The Preferred Attack wouldn’t really be relevant here - only if the model is Aggressive or Cautious.
In this instance, the model would just move as close as possible towards the model it considers an enemy.
Thanks!
Should it engage or stay at 1"?
The only stipulation on movement here would be to move behind cover if it is available. In your example, there is no cover so they would follow the instruction given in the rules -
‘… it will simply complete the maximum move as normal’.
so it would engage… even if they are archers
Yes, they would.
But then, using the Skeleton Archer as example, they would stay engaged (after rolling M and moving to their objective) and the 2nd action would be lost? Even if they have a 2nd attack wich is Melee, the rulebook says they would end the activation if they cannot block or hide after moving, should not they attack with melee?
How would be the case if the adversary just have a ranged attack?
Hey @DK-dark
If they rolled ‘M’, had to move forwards, ended up engaging an enemy model and couldn’t take either a Block or a Hide action then yes, they would have no valid second action and their activation would end.
The flow chart on page 64 for Cautious Move does not grant them a Melee Attack so they would not attack. You only roll once for an Adversary Activation and in your example, they would only get the Move response.
If they just had a Ranged Attack, the Cautious Move would function exactly the same. That is because the response does not mention an attack at any point.
Just a little caveat to tag on to the end of this; in that example; if the skeleton archer was still engaged when it next activated then it would automatically make a Withdraw action, followed by its ranged Attack.
1 Like
But that would be if the Skeleton Archer would not have a melee attack, so that is not correct. In that case, if they roll another M, would be changed into an A, and they would attack with their melee weapon.
The problem here is why they should not attack this way in the first contact? It has no sense, they Move to engage, and they do nothing? They might have another option before end their activation, priorizing blocking and hiding, but not wasting their turn in every M rolled without a cover near them
They act in this way because of the response. The response does not facilitate an Attack Action, so in that case, they would not be allowed to attack.
But if they roll the M in their next activation, then yes, they could attack…
Should not the situation be checked after every action, so the M would be A in the 2nd action like rolling M in their next activation (because engagement)?
Apologies, I’d forgotten that they actually did have a melee attack with respect to what you’re discussing; ignore me on that one.
As it currently stands, yes, the Skeleton Archer would engage then do nothing with its second action and it has been brought up as somewhat illogical. Just to note that the Adversary responses are one of the areas the Developers have said they’ve been looking at for the revised rulebook and this scenario might look a little different when that’s rolled out.
@JimmyW is right on the melee attack thing.
As for the checking, no, the response is rolled for once at the start of an activation. This is because unlike F:WW, a CTA response can provide two action points. Adversaries may be unable to complete both of those actions (such as in the process you have described) but that is no different to a player model - player models do not have to complete both Actions.
yes, I would understand that, this case is easy, it feels odd when you activate a “cautious” adversary, with a preferred ranged weapon, and you move it to engage the enemy (because there is no cover near in the path), and you engage and do nothing (even having the option there with melee, I think they should add this bullet point before end their activation), but with the same response next time, then yes, they attack
Just to also play devil’s advocate, they’re called ‘cautious’ responses for a reason; I always imagine that one as the adversary hesitantly approaching to engage because they haven’t got anywhere to hide and are too close to try and shoot effectively so they move closer then, by the time they activate again, have no choice but to desperately attack in melee.
Ultimately, game mechanics > real logic; it’s not like Skyrim’s skeletons are master tacticians anyway and these sort of seemingly odd responses may well be getting tidied up a bit when the revision comes anyway.
Yeah, it is like when you play a videogame, and the enemies go and suicide at the lava pit… 
I hope not to get a model blocked inside a piece of terrain like a rock or something like that, like back in the videogame 
.
The reason you said is fine, the different here is they have time to perform that attack at first contact.
In any case, I think this paragraph should be reviewed, writing some exceptions or situations.
I am trying to get into the game, and teaching more players to take them into the game too, so when we are playing and we face this situation, I need to do something to get a “refined” feeling with the AI, not giving a “fool” behaviour to the enemy.
In my game with a friend, I took the Skeleton Archer back to take cover, because the rulebook says the cover must be in the path without taking too much deviation (not matching my situation), and then it performed a hide action after getting obscured LoS, so I think I achieve the goal of a cautious model (shadow path) with a bow, trying to shoot in its next activation (you cannot think they are fool or bad “tacticians” for some actions and not for another, they just want to move and shoot an arrow…). To avoid this, you just need to add a point there on the rulebook