Jump to content

My A18 feedback


~Kevin~

Recommended Posts

In A16 there were infinite loops. Build a stairs with a one block gap and because zombies couldn't leap they would climb the stairs and fall and climb the stairs and fall forever.

 

Used to be?

Some delusions of grandeur mate. :D

This is still very much used today. Granted, its a little bit harder than just 1 block now tho... xD

Link to comment
Share on other sites

Used to be?

Some delusions of grandeur mate. :D

This is still very much used today. Granted, its a little bit harder than just 1 block now tho... xD

 

Did you start reading the thread from that post? There was nothing “used to be” even implied. I was responding to someone who said there are infinite loops in A18. My post was saying they also existed in A16.

 

Also existed not used to exist— Obviously.

 

Don’t start reading a thread mid-conversation and expect to understand what is going on especially if English isn’t your native tongue. You should start at the beginning and read everything. Perhaps you thinking you didn’t need to do that is the true delusion....

Link to comment
Share on other sites

I've come to a conclusion as to why very few players claim AI is not better in A18, granted I could be entirely wrong.

 

These players look at it like this. Since the AI can be exploited in A18 that makes the AI bad. Which.....come on....thats not the way you judge, the question is why is AI better prior A18. there were ways to exploit it in A16 as well as in A18. IMO from all the exploits I know of, there are less and or harder to do exploits in A18. I still would not say oh A18 is better becauce of that alone.

 

No, one must look at all of the AI capabilities and AI performance when comparing between Alpha. I think the few just do not understand how to judge the AI correctly when looking at it from a technical stnd point. I dunno....maybe I'm wrong for viewing it like this. I mean no offense to anyone that thinks prior A18 is better. :)

Link to comment
Share on other sites

Okay. I'm guessing you could force an easy-win horde night against the A16 AI with a "legitimate" build, if wedge-tip skirts aren't enough a dry-moat-and-bridge defense, just leave the bridge down, give the zeds a way out of the moat, that'd make those also work in A18. And i'm'a try that suspension-base idea for sure. I like boring just fine if it keeps me breathing, I like having to think, not having to fight. It's one of the reasons I hanker after buffing the crafting and survival games, make success require more territory than you can easily defend. If you want to advocate for zeds getting (more?) quickly frustrated when they're blocked by other zeds and switching to either random-destruction or wreck-the-nearest-non-terrain-block I'll help, I agree that zeds "should" be able to find that, and my weak-point bunkers maybe "should" need more than their wedge-tip skirts, having to dig a moat so the zeds physically can't reach anything but the killzoned entrance would feel right to me. How about also having the zeds focus on any player-placed blocks on the map if they can't reach the player, so if you build your horde base so they just plain can't get to you, when they go random-destructo they just might head off to your farm and family home 1km down the road?

 

Yep, you could. I built kill cages, stilt bases, took over POIs, and built full on castles and more in A16. The threat was rarely high in SP - if we're measuring by zombies managing to breach a wall - and we got our walls (one block only) breached once in MP. No repairs needed once you hit reinforced concrete and steel. Not to say I didn't enjoy A16 horde nights. I did. But once they got close enough to your base, they rarely *rarely* assaulted from multiple areas. Instead, they congregated to where the player was, spreading out to whatever degree was necessary for the number of zombies on screen.

 

I've come to a conclusion as to why very few players claim AI is not better in A18, granted I could be entirely wrong.

 

These players look at it like this. Since the AI can be exploited in A18 that makes the AI bad. Which.....come on....thats not the way you judge, the question is why is AI better prior A18. there were ways to exploit it in A16 as well as in A18. IMO from all the exploits I know of, there are less and or harder to do exploits in A18. I still would not say oh A18 is better becauce of that alone.

 

No, one must look at all of the AI capabilities and AI performance when comparing between Alpha. I think the few just do not understand how to judge the AI correctly when looking at it from a technical stnd point. I dunno....maybe I'm wrong for viewing it like this. I mean no offense to anyone that thinks prior A18 is better. :)

 

Personally, I don't have much of a 'better' or worse in this case. I like A18's AI more and it fixed some of the truly broken aspects of A16. Things like chairs and other objects were way to easy to use to make A16 zombies useless and their inability to track a path to the player often meant they spun or stood uselessly below the player, making for easy targets.

 

With the upcoming bandits, AI changes were even more desperately needed. I see A18's AI as the beginning. They've already added more randomization to zombie behavior and are looking at some of the 'probably not intended' aspects. This will not only make zombie AI better, but likely the bandits as well.

Link to comment
Share on other sites

I tried an experiment with this A16 running in circles thing, because I've never seen it. Taking the info from the other thread about AI, I used "pole blocks". Basically, centred pole pillars that look like lamp-posts. A block type I've never used in a base which might explain why I have never seen this circling effect.

 

Except that the running in circles happened everywhere without a pole block in sight, again: See

at minute 8:32 as an example

 

I built a standard 10x10 box fort-style base with iron spikes all round, me on top. I then built 8 pole block pillars (each 3 blocks high) at the 8 main compass points round the outside of the base about 5 blocks out from the spikes. I then ran a day 7 horde.

 

Sure enough some zombies began to run in circles around the poles. I've genuinely never seen that. However in their defence, I would add that it was only tiny number of the horde that did this (5 zombies that I definitely saw but not all at the same time) and they stopped circling after a while and joined the main attack against the walls.

 

Being curious I arranged the exact same setup in A18, and the results were far more interesting. A significant number of zombies....perhaps 20, actually attacked the poles instead of the base!! And they did not stop until all the poles on two sides were destroyed. Which took a while as they were steel. Since maxAlive was 24 on the server this is actually very significant, as it left my base being attacked by 4 zombies at a time till all the poles went down. Those poles trivialised the entire horde, as it was basically unable to fully spawn.

 

As long as you add more steel blocks to your base than the zombies can destroy any base is safe. Basically you standing on a big enough steel cube would be good for A16 as well as A18.

 

Fataal has a problem here. If zombies know which blocks are supporting and attack only those, people will complain about structural engineers. If not, they will attack arbitrary blocks and that looks weird. I think you got a great idea there: The AI should prefer blocks that are closer to the player when going into destroy mode. Maybe not all the time, but most of the time

 

Remind me again why A18 AI is better?

 

Incidentally because you can't use arbitrary bases on insane anymore. In A16 a simple one-block deep steel pole wall could hold off a horde on standard difficulty for the whole night, on insane you had to do some repairs. Effectively only player playing insane had interesting horde nights, for normal players base building was trivial.

 

Now zombies are better at breaching walls (for various reasons) and you need 2 block wide pole walls and repair them from time to time even on standard difficulty. And since it is worse on insane you have to optimize much more resulting in many simple base designs not being good enough for the insane horde anymore.

 

My group actually built a basic rectangle ground base melee cage in A18 and it worked on warrior difficulty. In A16 that base would have worked even on insane. THAT is the main difference and the reason why I think AI got better.

Link to comment
Share on other sites

 

Now zombies are better at breaching walls (for various reasons) and you need 2 block wide pole walls and repair them from time to time even on standard difficulty. And since it is worse on insane you have to optimize much more resulting in many simple base designs not being good enough for the insane horde anymore.

Zombies can better break through walls only because one digit of damage in .xml is changed.

AI has nothing to do with it.

Link to comment
Share on other sites

Zombies can better break through walls only because one digit of damage in .xml is changed.

AI has nothing to do with it.

 

That's not true. Part of the pathing code is that if multiple zombies are grouped together they get a damage multiplier and since most will path towards the weakest link in your base armor they will group up against that part of your base and get the multiplier to breach it even faster.

 

In addition zombies are much more likely to pyramid on top of each other and reach higher blocks than they were in previous alphas and so there are times when they can flow over the top of a wall or at the least hit higher up where players may have used weaker blocks thinking that the zombies would only be able to reach so high.

 

Finally, zombies can now for the first time (IIRC since A17.4) damage blocks above them so that placing a closed hatch at the top of a ladder is no longer a guaranteed defense.

 

So....programmed zombie behavior and pathing changes have a lot to do with how much easier they can breach walls and get past defenses.

Link to comment
Share on other sites

1.But what do you think happens when when just increase the block-damage?

2.You say "group". What happens when a group hits the same block in A16?

 

You keep explaining HOW it works. I'm talking about the end result. I don't care HOW it works.

 

(by the way, You explain well, but i'm aware of it. Thank You for this, it may be useful to others!)

Link to comment
Share on other sites

.....

But what happens when you just increase the damage by blocks?

You say "group".

What happens when a group hits the same block in A16?

 

There was no multiplier in A16. It isn't a group hitting the same block, it is that a zombie hitting a block gets a buff to do more damage if he has other zombies pressing against him. That was new in A17 and has been adjusted a few times I believe since then. In A16 a zombie had no buff granted from other zombies nearby.

Link to comment
Share on other sites

There was no multiplier in A16. It isn't a group hitting the same block, it is that a zombie hitting a block gets a buff to do more damage if he has other zombies pressing against him. That was new in A17 and has been adjusted a few times I believe since then. In A16 a zombie had no buff granted from other zombies nearby.

Multiplier? ...that's another achievement...?

What is it for? What purpose does this serve? How does this make the gameplay more interesting?...

Why can't just increase the damage if it's not enough?

Why complicate the code if it doesn't solve any problems?

 

 

this is what AI is Now:

Link to comment
Share on other sites

Multiplier? ...that's another achievement...?

What is it for? What purpose does this serve? How does this make the gameplay more interesting?...

Why can't just increase the damage if it's not enough?

Why complicate the code if it doesn't solve any problems?

 

My theory is:

 

If zombies attack distributed over a whole side of a base (like they did in A16), they also do damage to all those blocks. To be really able to break in they would need to do a lot more damage per zombie and would basically break large parts of the base before breaking in at one point. It would mean tedious repair time after each horde night.

 

Instead now the zombies are still relatively weak as singles, but the more they cluster somewhere the better their chances are for a breach. Without doing much harm to the rest of the base. Another advantage is that the player has a small tactical mini-game to prevent or break-up clusters of zombies.

 

A different solution could be that you can make a blueprint/snapshot of a pristine base and later restore it like quest-POIs are restored, just with you footing the material bill (of course). But blueprints aren't in the game yet.

Link to comment
Share on other sites

Multiplier? ...that's another achievement...?

What is it for? What purpose does this serve? How does this make the gameplay more interesting?...

Why can't just increase the damage if it's not enough?

Why complicate the code if it doesn't solve any problems?

 

 

this is what AI is Now:

 

IMO the multiplier serves the purpose of representing what the difference is of one person pushing against a wall vs 100's or 1000's. I consider it acceptable because technical limitations won't and may never let there be 100's or 1000's of zombies on screen *and* without a multiplier the effect of ever increasing pressure on a wall, a breach would never happen. It would stay at 1 zombie hitting the wall (maybe a couple of more if they crowd properly) and only that base damage being done. Steel walls would then make it super easy to keep zombies out and keep the player completely safe - a state that MM, I believe, said is one that they want to make less and less possible.

 

It makes the gameplay more interesting because the harder horde night gets, the less safe the player is. Zombies that can effectively breach walls and overrun a base is, IMO, more interesting than zombies that I can always out repair without really trying too hard.

 

As for why not increase the damage, it's because 1 zom shouldn't have the ability to easily break through a steel wall, the more you add, pressing, pushing, and beating on that spot, the risk of that wall giving way should increase.

 

Lastly, I think it does solve a couple of problems that I've already mentioned above. It wasn't until the change in A17 that I *ever* worried about wall breaches - except in my earliest playthroughs when I was still learning the game. Are there other ways to solve those problems? Sure, TFPs could lower block durability or raise zombo damage individually - but then wood and probably cobblestone would likely become useless in early game, making for a very frustrating early game, especially for people who like to build. Solo zombies with enough damage to break through steel or concrete reliably would, IMO, not be good for the game (except for people who really want that challenge) and would just make early progression a thing that people felt absolutely pressed to speed through because early and mid building materials would have little to no value.

Link to comment
Share on other sites

Yes, i agree that this works ...but this cheap feature, and nothing more - it can be screwed to anything.

But here is normal AI have zed - there is no.

 

here:

look again, and then you can tell me again how you like the new "AI".
Link to comment
Share on other sites

Yesterday i trying setting up animal behavior. And the deer that was ordered to run away from me... why does he look at me for half an hour and then run away??? ..... He's f*cking stupid!!!...

...but this was not before...

Link to comment
Share on other sites

Yes, i agree that this works ...but this cheap feature, and nothing more - it can be screwed to anything.

But here is normal AI have zed - there is no.

 

here:

look again, and then you can tell me again how you like the new "AI".

 

It was cued up to someone rearranging inventories between a motorcycle, backpack, and crate and then driving to the trader. That’s when I stopped watching.

Link to comment
Share on other sites

It was cued up to someone rearranging inventories between a motorcycle, backpack, and crate and then driving to the trader. That’s when I stopped watching.

Damn it! YouTube does not give the correct time code!

Sorry, i swear it worked then.

 

There you can see that pathfinding works only from point to point ignoring the changed situation.

This approach to path programming is, to put it mildly, very rude and useless. It's not good to code like that.

Link to comment
Share on other sites

 

 

 

 

You just go to the point in the video you want stamped. pause it. right click on the video and then click "copy video url at current time". then just paste it on the forums or wherever. :encouragement:

Annotation-2020-02-01-145051.png

Yes, I did exactly the same, it didn't help. This is the first time. The others - everything was right.

This may be due to the long duration of the video...

Thanks :)

Link to comment
Share on other sites

I cannot think of a single factor in a16 that I think is better than 18 - not a single one. The AI is light years better - I played insane, max horde night zeds, always run and dead is dead because there was no real challenge any other way in A16. 18 I play on warrior and 32 zeds - my fort still gets breached - something that essentially never happened in 16. I could easily maintain a wood fort until day 14 too - not gona happen in 18.

 

And yes, exploiting the AI was just as bad then too.

Link to comment
Share on other sites

Played a little bit of 18.3. I liked 18.3, but it was still not as good as 16.

 

I would like to say I was very disappointed in 17 and 18 is a step in the right direction.

 

However, Alpha 16 was an almost perfect game. All we needed were some bigger bosses that could do more damage to bases. It seemed the difficulty leveled out after a certain point. Rather than fixing those small issues, the developers gave the game an entire facelift.

 

1000% agreed

 

 

No doubt many replies pointing out the +16k all the time with A18, but I definitely think A16 is a better game, all things considered

Link to comment
Share on other sites

A16 AI had a very small chance for spinning zombies (less than 1 in 20) but had no super obvious exploits. The randomness of the AI in A16 is something I miss.

 

Also, spam carafting and auto mining were fixed in A16.

 

AI was simply better in A16 - I have 1000 hours and never had spinning zombies

 

The structural specialists with magic weak spot finding powers in A18 are not my cup of tea - jumping is even pushing it IMO

Link to comment
Share on other sites

Sure, no problem...

 

1) In alphas prior to A17, zombie behavior was much more random- I know it wasn't but it appeared so. Since they did not automatically and unconditionally go for the weak blocks, the player could NOT build infinite ramps to exploit the AI. In A17 and A18 the player can use about 20 blocks to build a ramp and the AI will follow an infinite conga path around that ramp forever, and thus becomes absolutely zero threat. The blocks can be made of wood and will never be attacked so you can do this from day 1. Game is now over. Surely exploiting AI with an infinite ramp is not the intended way to play the game??? Now go look at every youtube video called "best horde base" or similar. Every one exploiting the A17 AI, without exception.

 

2) After A17 AI change, players can now sit out horde night on top of a converted POI and never ever worry about it. This is akin to the exploiting ramps described in point (1). Because the AI homes in on the weakest blocks, they tend to almost never attack the supporting blocks and will likely just be attacking all the insignificant blocks inside the POI (once the wall is breached). In A16 and prior there was a very high chance for the POI that the player was sitting in to be raised to the ground by the random nature of the horde attacks on supporting blocks - I've seen it happen dozens of times. This is more or less impossible in A17 and later. Again, surely this is not the way the game is intended to be played?

 

 

So let's assume the player is not going to exploit the AI with an infinite path or impregnable POI, yes? Here's still more reasons A16 and prior was better...

 

3) In A16 and previous, the AI would attempt to path to location where the player was. If it couldn't get there it would revert to attacking the closest block that was blocking it's path to the target player. This actually felt like what a "real" zombie would do and greatly increased immersion. It felt random! Build a 10-block wide wall between zombies and player, and zombies will attack all those blocks because the player is moving about shooting stuff. Watching an A17 "brain-dead" zombie carefully thread a path around your spike traps, following the clear path where spikes have already been cleared, and smartly heading straight for the single weakest block in your wall between it and the player is just stupid, and not how "real" zombies would behave.

 

4) In A16 and previous because of the random nature of how zombies behaved, there were many more base designs viable than there are now. The AI switch in A17 actually took a ton of super-fun base designs off the table. The changes to AI in A17 is what gives TFP their reputation of "no fun allowed". Any change to the game that reduces player choice is a bad one in my view. If a player wanted to build a base deep underground and never deal with zombies at all - SO WHAT? That's his business.

 

I just cannot understand how anyone could think that the AI in A17 and A18 is better. I'm still waiting on a well thought out and articulated argument as to why it is better. To summarise the A17 AI:

 

a) It is super easy to exploit. Even without infinite ramps you can predict which block they will go for and 100% influence which side of your base they will attack, and where, rendering defense trivial.

b) It removes a lot of fun base designs - like classic forts

c) Zombies should not be structural engineers or dodge their way through gaps between traps, it ruins immersion.

d) A17 zombies *cannot* raze POIs beyond simple tier 1s; A16 and earlier used to regularly do so.

 

I agree with this. We’re not all going to agree on things, but I think this is well said for people like myself who liked the old AI better

Link to comment
Share on other sites

Overall, enjoy improvements of game and play it happily. Been doing this for a solid 4+ years now? Every big alpha that comes out, we roll a new world and play for hours again.

 

As part of a small group of people that play I am/was the crafter. The major skill overhaul has kind of stolen my thunder. My group mates still do what they do, just takes them a bit longer now to get where they need to go skill wise. The more advanced they get, the less they need me, the less fun it is for me. I no longer have my niche, and have to branch into things more than I did in A17 than I care to. I just want to stay behind and build, and not waste my labors on fighting.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...