PC Please DEV's

Not necessarily open for huge exploitation. There would need to be more communication between zombies and more variation on their intelligence.

1. Introduce an action where a zombie does not charge straight from the start (low chance for any zombie to be like that), but walks around seeking alternative entryways (expanding line of sight and not those checking routes used by other Zs), just to signal some zombies in range that there is an alternate path. After a set number of checks it would simply go back to being a straight attacker. Additionally, only non-attacking for a small period of time could shift to a onlooker state.

2. Make Zs reuse previously done checks (i dunno if it is done or not). For example, you have 20 Zs charging from one direction in a big group. Why make 20 pathing checks, when only 1 to 3 could be done for the general vicinity or perhaps for those in front and everyone behind that would walk through where others are, would reuse the same checks.

3. Introduce zombie routing, where Zs would follow other Zs that are in attack mode (special range for that). For example, an onlooker found an alternate route from a side and called out others to the new "entrance", but new Zs running in would not hear him. Yet when they come close enough, they hear vivid Z sounds and first seek them (when there is no direct route to the player or no player in sight).

 
On the other hand, just calculating line-of-sight would reduce the overhead. The zeds would go for the least resistant blocks in visible range. It would only care about the one side, or two if at an angle. Though this method also opens itself to huge exploitation in base designs.
Actually, thinking about this some more, I think you may be right. The code will still have to examine all blocks within a certain radius of the zombie (which is what it already does), but then pathing calculations would only be performed on those ones deemed visible. So, that's less of a load - on the other hand, it then needs to determine whether each block was visible or not (which it doesn't bother to currently do), so that's a little more of a load.

I'm nowhere near enough of a Unity programmer to have any idea of whether the time saved on pathing calculations outweighs the time spent on determine whether to make those calculations on each block.

I do think Khulkhuum's idea of having a Zombie "leader of the pack" makes sense though, since the "followers" could just trail the leader, until the leader died and a new leader was elected from the followers. That might be able to cut down the load somewhat.

 
And this is something I fundamentally disagree with.They should generally try and take the path of least resistance, yes.

BUT there are two things that NEED to be implemented, or this whole system is worse than ever:

a) they must ignore spikes and other traps. ANY of them. Not a single point for traps. They are mindless zombies with no sense of preservation. Traps are not "do not tresspass" signs. They are traps designed to catch the unaware. Be it animals in reallife or zombies in game.

b) if there is no direct line of sight, they should not know what lies behind.
Precisely.

With the present state of the zombies, I won't bother building a base. I actually haven't played since 17.1 because in my opinion, right now, zombies are broken.

 
One way to widen the area of attack would be to make any zombie already standing in a path and hitting a wall count as an obstacle as well. So if a zombie is standing at a wall and hitting on a weak block the rest of the zombies would see that as occupied and look for the next weakest path/block to attack. This would still allow up to 3 zombies hitting the same block (two of them would attack from an angle of 45 degrees) and benefit from the group block damage.

Another idea: All possible paths are searched for exactly as now. But instead of removing all but the weakest path remove only paths that are more costly than 200% of the weakest path. Of these paths take the one with the weakest first obstacle. This would a) make VictoriusIII's szenario work out like he wants, B) add not too much calculations and C) still allow zombies to follow paths the player "handcrafted" for them to follow. And D) still work correctly in POI's

Precisely. With the present state of the zombies, I won't bother building a base. I actually haven't played since 17.1 because in my opinion, right now, zombies are broken.
If you think so. I have made 3 very different horde bases in my last 3 games in A17, all of them worked very well without any exploits (aka what I consider exploits). And they were fun to build and "debug"

 
Last edited by a moderator:
The base design I’ve been using for many versions just so happens to make no point of attack cheaper to use so when blood moon comes, the attack is very spread out & it’s fun to watch the ai work.

I build a reinforced concrete square base 12 blocks by 12 blocks with one door. Outside of the door I build a 6 block by 6 block cage using iron bars with one door to the outside. I surround the entire base with 5 rows of wood spikes with one row of barbed wire against that except in front of the door. Just before blood moon horde starts, I ‘block’ us in by placing 4 barbed wire blocks in the one block wide pathway from the door through the opening of the spikes.

When the horde comes, they spread out evenly along the entire side. The barbed wire slows them for a bit so we can shoot some of them. Once a barbed wire is broken, they quickly all go to that opening and poor into the wood spikes. Once the spikes go down, they go for the concrete walls. By this time, the horde is coming from a different side and the process repeats. Rarely are the walls breached and the killing from the roof or within the cage is quite fun and rewarding. If they do breach the walls, it can be tricky killing those inside. All forges/workbenches &valuable loot is picked up and kept in our backpacks. Horde night is not dreaded, instead we look forward to it.

You just need to experiment a bit to find a design that works then you will enjoy horde nights again.

 
Last edited by a moderator:
Traps are not "do not tresspass" signs. They are traps designed to catch the unaware. Be it animals in reallife or zombies in game.
Um, no. The spikes are not "traps" in the way say a hidden snare trap is. Meant to catch the unaware. They are a in plain sight, very obvious and visible thing. That being said tho the AI we are dealing with for zombies is a bit too smart I'll agree. But, meh, WIP.

 
The base design I’ve been using for many versions just so happens to make no point of attack cheaper to use so when blood moon comes, the attack is very spread out & it’s fun to watch the ai work.
I remember in the early experimentals, my base with a similar design always got hit in the same spot. Every single time, they focused on the same block. I even built a maze around the base that made them come at it from the opposite side, and they still ran across and around to hit that one block.

It was funny as hell. Could have one guy just sitting there repairing the block, and another shooting the crap outta Z's. Later stages we had two guys repairing. They never got through.

 
My base is connected to a "Smettin Wall" and an "Elysian Ring". The zombies are zero threat to it. I snipe every single one in the head and collect every ounce of sweet XP on horde nights. There are no traps to replace or to steal my kills and there is barely and block damage fix at the end of the night. You can build one out of flagstone in one day on the first week and use it for ever. The structures are named after the people who created them who are followers of Grumbul on twitch.

Giving them a complex AI model only makes them easier to exploit. Adding even more complexity to "fix it" is only going to make them even less like a zombie. They should just go directly towards food and smash anything in the way. Even if it means trampling other zombies or walking into a river of fire. The structural engineer zombies are a little much for my suspension of disbelief right now, not to mention being easier to counter.

 
Back
Top