So, the Hardest AI cheats with double resource gather rate.
Are the Hard/Intermediate/Easy AIs the same as before, or have they been upgraded in any way?
Also, the cheat AI (hardest) is going to turn plenty of people off the game. Both games I played today, my team mates (and myself) rage quit because while we can understand losing to an AI that plays too well, no one likes losing to an AI that cheats.
I really think hardest AI should not be tampered with.
Only played it a couple times and fells bit to strong, maybe change it to 1.5 resources. It would be very hard for a causal player to beat it. Best to keep the old hardest ai the same with the new changes and have expert ai that cheats with 1.5 resources
Hard ai is still pretty damn bad. And it seems that ai does not make use of any of the new units. Rus Kremlin ai does not use milita, English builds Wyngard and does not build anything out of it. Same with Abbey. Ottoman lategame seems to be mass archers.
The ai behavior and unit comps are at times really bad. And I also wish that hardest ai was not replaced by an cheating ai. Instead have the cheat ai to be an additional ai. That’d much better.
This is by far the worst change they possibly could have made. AI Improvements Sure, good, great. Making the AI Cheat this hard is just down right stupid. It’s not lazy, it shows a lack of core understanding on the devs part.
Please observe the overwhelming bad response to this change and revert the AI to a non-cheat state.
Just hopped on for a casual game vs the hardest AI only to see them age up impossibly fast, then basically fast castle AND have 70 units on the field. straight up insane.
Perhaps also they should have started with x1.5 resources rather than x2?
Also its probably not something people care about much but it will make the masteries a lot more difficult to complete since the final challenge is a 1v1 against Hardest AI which is now getting cheats.
There is no “non-cheating state” to revert the AI to. Nearly every video game strategy AI cheats in some form or another. There are barely any that don’t.
What I’d prefer is if the devs separated out the difficulty level and the resource multiplier. That way we could specify the AI gets whatever x the resources on any difficulty level, and it’s then optional for people that just prefer the base AI (with all the cheating it currently does, presumably around map vision and whatnot, because it’s way too good at knowing where I put my Outposts).
I don’t know if the AI cheats on vision. It sends a lot of scouts around the map constantly, so that may well be how it always “knows” where you are. Clearly it doesn’t “see” like a human does.
But giving it double resources just means its playing a different game. The old AI was bad because you could exploit it. It couldn’t really harass very well, it couldn’t deal with harassment at all and it would suicide whole armies into Keeps and TCs etc. But now it feels like you really have to exploit those issues, otherwise it will just drown you in troops in 12ish minutes.
I feel like something the devs haven’t understood in doing this is that double the resources gives you much more than double the troops. Both players have to spend resources on:
continuous villager production
Suppose that by 12 minutes you’ve spent 3500 resources on all that, and 1000 resources on army, i.e. you had 4500 total resources collected (it doesn’t matter if these specific numbers aren’t realistic, it’s the general concept that matters). The AI has 9000 resources, so it can spend 5500 on army and still be level on everything else. In practice it will spend some of it on being ahead in other ways, but overall it seems impossible to keep up with its military strength, unless as you mentioned you use cheesy exploits of other aspects of its play that are bad.
I don’t understand why they’ve done this when AoE II DE already showed a much better way to do it - non-cheating AI with a handicap feature.
In my experience it’s very good at routing around Outposts (with Scouts). The range on the emplacement (once I add one) often means I can get some hits in, but never enough.
I agree that upping the resources makes it a lot harder in different ways. I get how it’s seen as cheating.
But the AI would always prefer a deathball (at least on lowest difficulties where I spend most of my time sandboxing). It’s still going to be weak to early aggression, the problem is the window for that is going to be a lot shorter (on the new difficulty). I think the we need to be able to set AI preferences (like aggressive, defensive, eco-only, etc), but that’s a more advanced suggestion that’ll likely take time to implement (if the devs want to do it at all).
I’ve been trying to get some info on AoE II: DE’s AI for some time. Is there a reason it wasn’t replicated in III: DE? I’m assuming it wasn’t, as I’ve only ever seen people talk about II: DE. That could be a reason.
Another could be resource. AFAIK II: DE had community help in improving the AI? Could be wrong there (and maybe that’s something the devs here need to investigate).
I still think removing the resource rate gain from the difficulty, and making it an optional toggle for all difficulties, is a low-cost improvement that would satisfy folks in the short-to-medium term.
Even better if it’s not too difficult would be to allow it to be set to any % in a range, so people can tune the level of challenge, e.g. 100% to 500% in 1% increments. By this I mean 100% = how it was before, 101% is a 1% boost etc, so how it is now would be 200%.
In the long run, though, it’s much better if the AI doesn’t cheat, as you can’t learn much from losing to a cheating AI. You watch the replay to understand what it did better than you, and the answer is nothing, it just cheated.