 |
What's the best tactic for this, and how good is it? I did some testing. It turns out AI units act differently from player units; the AI will refocus its attacks even if the original target is in range, maybe based in part on damage dealt by units, while player units will continue to attack the same target until they lose sight of it or their pathing is blocked: Player units don't automatically switch targets This means that typical results against the AI don't reflect the same tactic used against a human opponent. If you're short on time, you can skip most of the following, since it's about quantitative results against AI scorpions. This is just Feudal-age tactics being used in post-Imperial games. I tried modeling a scene from a deathmatch tournament: (I don't try to pick on expert players, just show that a tactic is not one you learn from standard tutorials.) First, the straightforward approach: End result after previous image, 23 out of 28 scorpions remaining Many different attack methods were tried, with no micro after the initial setup, except sometimes to order the units acting as diversions to attack. This was to simulate the intensity of a death match game. A frequent problem was that the main group of paladins would move around for a bit, either due to user error (right-clicking instead of attack-move) or maybe they were trying to reach a unit in the middle or back. The methods tested included moving a few units around to the back and having them attack then (after the usual slight delay that allows them to get flattened by an onager shot), or splitting the main group of paladins up into two or three groups on the front and sides. No method was found to offer any substantial improvements over the first one tried, which was patrolling two paladins on the sides and back of the scorpions on stand ground or no-attack stance: Two or three waypoints for the front paladins so they don't get too close 14 out of 28 scorpions and 3 out of 4 siege onagers left in first test Typical result for the straightforward approach was 21 scorpions left (7 destroyed). Typical for a patrolling strategy was 17 scorpions left (11 destroyed). Replacing the siege onagers with scorpions led to ~2 more scorpions being destroyed. It seemed once paladins started attacking, scorpions at the back would focus on those paladins instead of the ones patrolling, decreasing the strategy's effectiveness against AI-controlled scorpions. Surrounding the scorpions did not significantly increase effectiveness since scorpions just shot at the paladins opposite from them, whereas if the paladins are grouped up, the nearby scorpions have no useful targets due to minimum range. This somewhat balanced the increased splash damage from being grouped up. Similarly, ordering the diversions to attack meant they were doing damage, but also taking more damage. The paladins managed to get a single victory, thought I had to order the two patrollers to attack in the third frame after all others were dead: In the 'straightforward' results image, only 22 seconds had elapsed before the last paladin died, which included the time it took to close the distance. Their survival for much longer here might be from quickly getting within minimum range of onagers. Here are some cool, but pointless, images of paladins avoiding fiery projectiles, from different tests: https://imgur.com/a/88YLljY TL;DR: If you use feudal-age tactics against units with no ballistics in a post-Imperial Deathmatch game, you won't be able to beat 28 scorpions and 4 siege onagers with 12 paladins if you're fighting the AI. If you're fighting a human player, however, who is busy setting up new buildings and spamming units on singlequeue and doesn't try to micro their scorpions at all, then you can kill them without losing a single paladin. [link] [comments] |
from newest submissions : aoe2 https://ift.tt/2PULXKw
No comments :
Post a Comment