Guild Wars 2

https://forum-en.gw2archive.eu/forum/professions/necromancer/Minion-AI-Can-I-have-a-new-one-plz

The one that got me was this:

Bhawb.7408:

The only change I’d make to the minion AAs is to have a slight range increase, and allow them to attack and move at the same time. It is completely nonsensical that in a game where players can move and cast player summoned-AI has to stand still.

The problem is their pathing, in many cases ANet’s god awful terrain mapping, and the apparent ability to get stuck on the smallest pebble and be unable to move. There isn’t any real reason the pathing should be an issue now when it wasn’t an issue at launch (they ballsed it up when trying to reduce minion over-aggression). And increasing their ability to move over Z-axis differences would remove the pebble problems.

He’s basically asking for his profession’s (optional) utility skills to be better than the ranger’s class mechanics. Funny none the less.

Well, on his defense, if that AI system was improved, it would also improve our pets.

It’s still funny as they stated improvements to the ‘AI’ is mostly limited to band aid fixes as working on it is too risky as it could break the game.

I think the problem is that pets/minions share 99% of the code for regular mobs you fight in PvE. Pets are basically literally the same as PvE enemies.

If they improved the AI, PvE enemies would be able to move and attack at the same time as well and that would greatly change the game, since currenly you can avoid most damage simply by moving around and kiting them. They want you to move around and be active rather than standing still, so they want to keep it this way.

Splitting the code (duplicating much of it) would cause them a lot of headaches and be prone to bugs, so they’re not going to touch that.

Basicly they should have just invested MUCH more time into making genuinely good AI in this game in the first place. It would solve most of our pet problems, help Minionmancers, and make the PvE in the game more engaging that the “stack and full DPS” situation we have. Would have probably saved them time in the long run since they wouldnt have to keep going back and adding in so many crazy mechanics and rules to PvE content to try and stop people doing that.

The word basically gets abused much in this thread.

AI can eat up considerable amounts of processing time. The better you want your AI, the more processing time it is quite likely going to eat up. You can’t just magically have better AI without paying a cost.

Paying more processing time could present other issues.

It is not nearly as simple as many people seem to think.

Then they shouldn’t have based any of the profession around the use of such limited mechanics.

Even if Necro minion AI improve 100×... its still going to worse than Ranger pet AI.

Pathfinding gets heavier performance-wise the more accurate it is.

I don’t know how many players are online at a time, but multiply that with 3-5 and you get a good respresentation of how many enemies are getting their paths calculated. I think the problem here is that Anet probably uses the same pathfinding for everything. That is, pets have the same pathfinding as other enemies. In that case, it would be very difficult to split them and make a separate pethfinding for pets (and minions and such).

Also, I believe the pathfinding was perfect in the early days of the game, when it still wasn’t quite released. During beta tests (or even alpha), Anet realised that the pathfinding would be too heavy on the servers for the high amount of players so they reduced it’s update time. A little thing that supports this is something anyone can try. Go to a fairly empty map and strafe to the left/right. The pet will adjust to your pace and actually run in front of you. This could’ve been used for blocking projectiles with the pet but their reaction time isn’t good unfortunately.