New Hardest AI Difficulty

I agree the new Ai is boring to play against and impossible to play against for us casual players. Please revert this change or make them “smarter” in another way not like this.
Not fun!

9 Likes

Now the weekend is over, I hope the devs are hot-fixing this issue.

7 Likes

IMHO AI “Hardest” is too difficult for team versus AI games now - I think the resource multiplier is set too high. It feels like a cheat because essentially that’s what it is, and the AI’s just getting too many resources too early.

For casuals, the AI is hitting Imperial Age before the majority of players are getting to Castle thanks to the constant early mass attacks, and the AI is building massive armies with upgrades very early in the game that just aren’t defendable.

I think perhaps the 2x setting should be kept as “Insanity” mode, but there’s now a missing mode between “Hard”, which is way too easy, and the now-unplayable Hardest. It’s a shame that intelligence can’t be increased, but the scaling of 2x resources is just way too high for playability IMHO.

For non-pros I’d say we need more progressive modes between 1 and 2 resource multipliers perhaps. Hard - Harder (1.2) - Even Harder (1.4) - (Hardest (1.6) - Insanity (2.0).

2 Likes

I saw a guy on Reddit defending the new hardest difficulty saying that if you “harass early and constantly” the AI doesn’t know what to do and you can win. I found that hilarious because:

  1. It proves that even in hardest the AI is still extremely dumb if it can’t cope with early harassment and…
  2. Takes away all the fun of thinking strategically because you have to do the same thing every single match.

Age of Noob said it right in his last video, if you are making a cheating AI, you need to make a separate difficulty category for that.

Eventually I would love to see something like the C&C games had where the AI, besides the difficulty had a sub-option for choosing “turtle, guerrilla, rusher, balanced, etc.” or have it randomly chosen for you. That was great and you had to adjust your strategy on the fly to win.

11 Likes

I’m done with the game after these changes. I’m being overrun before I can produce a barraks.

4 Likes

I’ve been trying to play VS multiple Hard AI, but it’s not the same at all. Hard AI is too weak for my tastes :frowning:

6 Likes

here’s my take, undo the 2x resource gather boost for hardest AI, and instead implement a proper handicap feature to lobby settings, that way you can self set how much of an edge you give yourself or AI
to me 2x resource boost suggested that relic took the easy way to increasing the challenge of the AI, the hard way being incorporating a bunch of efficient build orders along side smart decision making (basically what aoe2 DE AI does, and for most part aoe3 DE AI (this one does have a 30% gather boost on extreme and 15% on hardest, hard and below no cheats tho, just logic, and its still good, given the radically unique civs in aoe3))

6 Likes

I just played a few more games against Hardest AI’s and it’s simply not fun anymore to play against it. How this could go through live to the servers is beyond me. Revert this change or create a new AI mode that cheats and reverts the old Hardest AI how it was and continues to improve on it in other ways.

This is not the way!

11 Likes

This has been my experience, it’s just no longer fun to play: you can’t play hard and you can’t play hardest. I’ve had to switch to a different game for now.

4 Likes

LOL, after 1.4 years of waiting, waiting for the AI improvment. I was anticipating a change of logic, like 2000 AOE2 AI to the 2013 AOE2 AI. It needs not be that smart, but at least something like this.
But what we get is a 2x resource gathering rate? :rofl:

7 Likes

i think the devs don’t understand what kind of ai we (the casual, coop, mostly against ai playing) gamers want.

we don’t want an ai which rushes us and bully us with a bigger army in the first minutes. imo the rushes has been good the old way.
we want an ai which can give us some challenge in the later game. which can controll and conter large armies, can siege through walls with towers and castles and not standing with an army at its new build wonder watching my trebuchets destroying it without any reaction.
we want an ai which can build walls and base where i am not able to ride in with some knights and burn all landmarks straight away.

the cheating ai is maybe good for a few pro gamers - but i think they play mostly against real opponents, so i don’t understand what is the idea behind this al?

11 Likes

“harrass early and constantly” is typical answer of someone who doesn’t primarily play against AI (e.g. is mainly playing PvP). It’s almost always some kind of troll or person that doesn’t understand why would someone play RTS against AI as they just keep making stupid remarks… In very few cases it’s someone you could have normal discussion…

I don’t think that we even need cheating difficulty, just make a official tuning pack. Most people will never enjoy playing against these kind of cheating AIs as it will always result in extremely strange game of constant army swarms, exploiting behavior or hours long turtling to make their bonus irrelevant.

Anyway it’s just disappointing as the AI is barely improved from the launch and PvE portion of the game is still ignored, without anything new on horizon… It’s so clear that there is huge portion of playerbase that wants better AI and more PvE content (coop), but they are ignored and when they get something it’s same as punch in the face with the most absurd cheat added to AI…

6 Likes

After more time to think about it, I think there are 2 major problems with the change that has been made.

Before I say what those are, I’m going to say what is NOT a problem. The fact that this AI is hard for most players to beat is not a problem. There is no problem at all in the game containing an AI that a top player can beat, but which will beat even many above average players.

The two things that ARE a problem are:

  1. The size of jump from hard to hardest is far too big and leaves a huge percentage of the player base without a sensible level of AI opponent relative to their own skill. The game needs a more graduated range of difficulties if the top level is going to be so high.

  2. The nature of how it has been made hard to beat, i.e. 2x resource cheating, forces a single playstyle against it of extreme early aggression to try to prevent it from ever getting started on truly having double the player’s resources. Even though Beasty beat it, he said while playing that you’re going to have no chance against it if you let it get going. This basically removes almost the entirety of the game’s mechanics from relevance as it becomes all about using one of a tiny number of options for the earliest possible aggression, e.g. english villager rush, or scouts harassing at collection points. All the game’s mechanics of age landmarks, techs, units, counter units, pale into insignificance compared to the importance of shutting down the AI’s resource collection ASAP, which in turn needs one skill, that of microing units at enemy collection points while also managing your base. The devs have said in the past that they want the player to “feel like a general”, yet this removes any strategy or decision making skills from the game, it’s entirely about that one micro/multitasking skill. This is not a long term interesting challenge for players.

9 Likes

I agree on big armies,battle-late game stuff which i desire the most but in order to reach that state,both AI and human player needs to reach a status quo.

AI needs to rush with an army size equivalent of a fairly skilled human player can realistically make to weaken, not to destroy so early because of this resource business.So if devs hadn’t put this cheating stuff,one of the core problems why the AI was too easy and boring in previous hardest difficulty would be already solved.They nailed the timings of the early rush comparable to AOE 2 DE Extreme AI.It’s a fundamental aspect of RTS to stop snowballing.

If they revert to previous timings we’ll return back to the cycle of turtle-build walls/keeps then snowball taking 7 AI’s without breaking a sweat and complain non stop in forums, reddit etc about how bad,dumb,easy and boring it is.Investing time and money to impove the AI would be so pointless as snowballing is the cancer that can’t be cured in all of the strategy games known to mankind.

Also AI needs to respond human player’s rush but as of now AOE2DE AI is exactly as horrible as IV at handling early rushes in 1v1 so relic gets a free pass from critism here but if you add one or more AI’s in AOE2DE they handle early rushes decently by building walls,towers and building an army that tries to counter you which aoe4 severely lacks.

‐‐-----------------------------------------
(It’s so thrilling to expect another ban for being ‘too combative’ in this post.Yay!)

2 Likes

Nah it’s just hard to do.

People have been constantly asking for a harder AI (myself included). They’ve made small improvements in most patches, but at that rate it was going to be a long time before it was decent.

So they did this 2x resources to make it hard since people have been requesting a challenging AI every single patch since the game came out.

Obviously they should have made that a new level of AI and not created a huge gap between hard and hardest.

They also made the Easiest AI less aggressive in this patch because people have been requesting that as well.

They’re trying, although I must say some of the silly things I see the AI do seem easy to fix in theory,

I agree in general, especially about the jump from Hard to Hardest, but this part ends up being true about most sufficiently difficult AI even if it doesn’t really cheat. The AoE2 Hardest AI can easily be beaten in this same manner even though most players couldn’t beat it otherwise.

3 Likes

And Beasty is right. If devs really want to die on this AI cheating hill, it needs to at least be better implemented instead of this laziest approach ever.

Have the AI gather at regular rate in Dark (I see it ages up very fast even in Hard), 1.2x in Feudal, 1.5x in Castle and 2x in Imperial. That way at least you have a fighting chance that doesn’t involve rushing with all you’ve got to cripple the economy and keep it that way. Better yet, make different multipliers apply to different resources. It’s very different to get a trash tsunami than it is to get knights, elephants and hand cannoneers.

To clarify, I think the AI needs to be better and I don’t support altering resource multipliers, but if that’s the only choice we’re going to get, something like that should not be too complicated to implement.

2 Likes

Agreed! 2x gather is an unfair advantage. Especially in dark age… that’s an insane boost that makes them snowball way quicker than humans can, then they harass you and not only do they keep snowballing, but you’re stopping your econ as they’re harassing you.

Correct their logic, have them do some build orders, so many things are fixable without necessarily actually making them super smart. Like, how many times have i just put an outpost in a random place, then AI spends the whole game just sending vills one at a time passing close to it just to have it die by my outpost… THE WHOLE GAME!!

just fix THAT ONE thing and you just boosted them a bunch

2 Likes

People seem to keep forgetting but there were a bunch more changes than that.

The thing folks don’t like is the resource boost, but let’s not confuse the two. It’s a shame that it’s overshadowing general improvements to the AI, but it is what it is.

Like I’ve said before, a good short term improvement would be to decouple the resource boost and enable it for any difficulty. This seems to be similar to what others are suggesting as a “handicap system”.

Longer-term, I’d like AI strategies. Defensive, offensive, etc. It’d be more work but would enable all of us to really get a customisable experience (even if some strategies aren’t clever and cause the AI to make subpar choices).

4 Likes

mm we got a thread that basically shows that ai changes failed to fix its issues on difficult scale and balance of it. now that you said about the resource boost change 's feedback means that this patch was a failure and shows why many ppl are leaving according to steamcharts which lowered its avg number of players recently from the 12k+ to 9k in a non weekend day .

2 Likes

My family and I played against the AI. It’s fun for us.

Our favorite AI was the “hardest”, but now with the change it is too difficult for us to enjoy. I didn’t understand this change, to create more difficulty it was enough just to add a level to more, with the name of “insane”.

Now there is a huge gap between the “hard” and the "hardest. The “hard” is too easy for us and the “hardest” is insane for us.

And now? The game got boring for us against the AI after they made that change.

Will the developers go back to “hardest” as it was before and just add an insane level??? It would be perfect for us if it were done this way.

7 Likes