Giza

MVM Giza b6

UKCS-Alias

Mann vs Machine... or... Mapper vs Meta?
aa
Sep 8, 2008
1,264
816
2nd place in the 2 Skillsets contest layout phase!

An MvM map with control points in gravelpit style. Meaning both A and B are open at once and bots will try to cap them both. The map also features more bombs at once.

Why im trying to make the capture system this hard... because its possible! And ofcourse, new gamemodes being welcome. And i like chalenges in entity work.

Because of the gravelpit style bots are able to use alternative ways to attack you. They are able to ambush you better since when they get past a point, they are behind you. To make it worse, if they capped a point they will use that flank meaning you also have to battle through them to even reach the point.

How the map works
Part 1 (initial phase):
There will be 3 types of bots:
  • Default (bomb carrier)
  • Gatebot A
  • Gatebot B
These bots spread out instantly after the start and try to cap points at the same time. A being on the right side from the bot spawn and B on the left.

Part 2 (A/B phase):
Once A or B has been capped the bots will change behaviour on the point they have capped. They will spawn at that specific gate while the other type still uses their initial spawn. This makes them spread up and walk through the side you normaly walk to the point (they wont use the same path as the other bot type). This means that the point will be less reachable.
However, unreachable it wont be since the spine of the map can still be used to reach the points.
Bombers can spawn at both gates.
Capping progress will be stopped while the bots are disabled (and start fresh as the progress goes down with it).

Part 3 (C phase):
Both A and B have been capped. This moment you will get a 2nd bomb into the game and bots will come from both A and B. The initial spawn is now unused.
Now bots will try to attack C (this is the dish area).

Part 4 (Final/Hatch phase):
Once C is capped there will be 3 bombs in the game and there is no more gatebot activity.

For a more detailed view on the navigating of bots: http://forums.tf2maps.net/showpost.php?p=297113&postcount=29
 
Last edited:

UKCS-Alias

Mann vs Machine... or... Mapper vs Meta?
aa
Sep 8, 2008
1,264
816
Ok, bots now properly attack A and B properly at the start, i have been working on them flanking the other point but atm they walk straight, ill have to fix that. I havent tested C yet because of that. But the hud works:
hud01.jpg

hud02.jpg


And then i also created the final point. It doesnt contain the spawn yet and the bombpit doesnt even result in a loss. but thats because i wasnt finished with the events. I just want to know how it looks.

The new spawn area with bomb pit
2014-01-26_00006.jpg


Total map overview
2014-01-26_00007.jpg


Dish area how it looks now (since the area connects to the final one). To me i need to rotate lights for it so shadows on the dish become better.
2014-01-26_00008.jpg

2014-01-26_00009.jpg
 
Mar 23, 2010
1,870
1,695
cool idea for a gamemode
 

UKCS-Alias

Mann vs Machine... or... Mapper vs Meta?
aa
Sep 8, 2008
1,264
816
Well, even if you were allowed to (i dont know if the contest allows it) i would still do them myself. I like creating waves myself and did that a few times already (see skullcove as example - and the extra popfile on sundown - and some random ones).

Not only that, but im not even sure if you would know how to set it all up. Where in mannhattan bots have 2 settings they need 5 settings in my map to work and those settings are more strict aswel or the bots cant even leave their spawn.

This is the testing popfile i use atm to test the events and nav file on the map:
https://dl.dropboxusercontent.com/u/2070422/mvm_intercept/pre-alpha1/mvm_intercept_pa5.pop

Its a simple wave but it already shows what is required for the gatebots.
- Default allows them to attack both A and B
- Priority means they attack a specific point
- Revert means they will try to cap the bomb

So ofcourse, you can try but i will still make my own popfile. Remember however that the map atm isnt playable yet. Not all of the required entity work has been done and i might change a few things for the popfile.
 

UKCS-Alias

Mann vs Machine... or... Mapper vs Meta?
aa
Sep 8, 2008
1,264
816
And on a full test localy i noticed a few issues:
  • Ugly prerequisites. I was putting them at the wrong spots and had to make them huge because of that. Valve instead just did put them at the spawn of A to make them cap B. This is realy going to help though
  • Broken doors. In my case B seems to autoclose each time. Why it happens im not sure of, its not the reset time since that already was on -1. A and C didnt have this
  • Ignored nav sections
  • pit explosion has the debris too high making it all float.
  • some paths were too short and some were too long. For bomb reset times that can be a problem. Still, i reduced some and lengthened some. They now should be more equal in time
  • Respawn rooms. i was able to walk into them and spawnkill lol
Some are fixed already, some arent. They all still need playtesting however.

Once these are done and the map is playable i can make a popfile and start the actualy tests :D

EDIT: Also, i need to add more info nodes. Just to guide a 2nd mapper through the entity work so they can add extra events to triggers. There is alot of events which make use of cases and a math counter and unless they are called a boojum snark i have a doubt that they will know what to bind to what relay without breaking the relays (i have some placeholder ones for them to use which are simply empty and easy to use). But thats fine for later steps.

Bonus screen on edit #3:
entities.jpg
 
Last edited:

UKCS-Alias

Mann vs Machine... or... Mapper vs Meta?
aa
Sep 8, 2008
1,264
816
Ok, i nearly got the map in a state that i can release it. Most of the bugs are removed and most of the optimizing in entity work (also to prevent bugs) has been done. Bots now properly cap each point.

Further for the bot population control i did a major optimization in it. Where originaly i had to use my own settings i now instead am able to use a popfile similar to mannhattan controlling just 2 events per bot. I still need to make them A and B compatible for that (since thats something the mannhattan popfile doesnt have). But that is obviously easier.

To compare:
The old situation: 2 bot types with 5 event settings.
The new situation: 2 bot types with 2 event settings.
The mannhattan situation: 1 bot type with 2 event settings.

So Oswinnin, if you already started its now easier to use the mannhattan ones as base (later i will provide that same template list for both A and B anyway except for the chief versions which to me should be in the popfile itself instead of being a template). although they would still need a few tags on top of the mannhattan ones:
Tag nav_prefer_A
Tag bot_gatebot_A
(or both on B for the cappers of B)

The mannhattan tags are used to cap C. Although those contain nav_prefer_gate1_flank my bots wont use that since it isnt there.
 

UKCS-Alias

Mann vs Machine... or... Mapper vs Meta?
aa
Sep 8, 2008
1,264
816
And the popfile after i converted all non chief classes in valve's gatebot file to be able to work in my map:
https://dl.dropboxusercontent.com/u/2070422/mvm_intercept/robot_intercept.pop

What a hell of a job to do :p took me 30+ minutes

And spawn positions:
Where spawnbot_main_A_0
Where spawnbot_main_A_1
Where spawnbot_main_A_2
Where spawnbot_main_B_0
Where spawnbot_main_B_1
Where spawnbot_main_B_2
Where spawnbot_alt_A_0
Where spawnbot_alt_A_1
Where spawnbot_alt_A_2
Where spawnbot_alt_B_0
Where spawnbot_alt_B_1
Where spawnbot_alt_B_2

0 is before A/B is capped
1 is before C is capped
2 is after C was capped
alt makes them spawn at the top and use the sides in the A/B phase.
 
Last edited:

RubbishyUser

L7: Fancy Member
Feb 17, 2013
414
488
I am super excited for this in the contest. I'm not taking part, but the fact you'll be giving your A-game to make Gravelpit - the best A/D into a Mannhatten style MvM, then after you've spent all your energy, for this to be picked up by some excellent detailer, gives me the willies.
 

UKCS-Alias

Mann vs Machine... or... Mapper vs Meta?
aa
Sep 8, 2008
1,264
816
Download for the first alpha has been added.

There can still be some issues, since the entity work is complex i might have created situations that are either uninteresting, is simply anoying. Give feedback on anything you see that you think doesnt belong there or isnt supposed to happen.

And give the team setup you use. Knowing the meta of the map is important for testing.
 

UKCS-Alias

Mann vs Machine... or... Mapper vs Meta?
aa
Sep 8, 2008
1,264
816
And here are the test results:
  • There is an issue on the left spawn for bots making them stuck when trying to jump down. it can easily take minutes for them to actualy succeed.
  • The left spawn for the bots needs cover to indicate that you cant hit them there yet (this also can fix the nav)
  • There are a few giants that get stunned on capping. This is a popfile bug
  • There is a problem where bots that cap a point get ubered on the point making you unable to prevent the capture. This isnt happening all the time though
  • At the start there seems to lack some options for the scout to get on top of the middle
  • Events dont reset correctly. This makes it possible for gatebots to ignore the capture points or go for the wrong point
  • lack of ammo packs
  • lack of signs
  • The path from the spawn to A and B is quite short making it hard to put in some stronger classes. Unsure if i need to fix this since i can allways make those classes a bomber instead of gatebot.
  • There are some dark areas in the map that need lights
  • Bots can get stuck in the dish on one of the bridges. They try to walk under while they cant.
  • Broken waves. Some are blatently unbalanced and some take ages to spawn all the bots (in wave 2 those 60 heavies if you can kill them as soon as they spawn still take 5+ minutes, with walk times this easily is 10+ minutes)

So there is alot for me to do.
 

UKCS-Alias

Mann vs Machine... or... Mapper vs Meta?
aa
Sep 8, 2008
1,264
816
I know, i was just summing the things up so i dont forget to fix it.

I have gatebot templates copied from valve, thos obviously have those tags. However, their normal giant bots dont forcing me to manualy give them the tag. And forgetting that is easy to do.

To note, that spawn bug (that they spawn 1 at a time with huge intervals) is due to WaitBetweenSpawnsAfterDeath property. Even if you set it like this:

TotalCount 20
MaxActive 10
SpawnCount 10
WaitBetweenSpawnsAfterDeath 20
WaitBeforeStarting 0


They will still spawn 1 at a time. And in the above example where between the 2 groups you get 20 seconds you suddenly end up with 19x20 seconds instead.
 

UKCS-Alias

Mann vs Machine... or... Mapper vs Meta?
aa
Sep 8, 2008
1,264
816
And on A2 i found other issues. Since the map didnt break at spawning the waves i was able to test it better:

  • Currently you have to guess what they are going to attack. There is no indication on the point they attack unless they walk out of the top spawn. Possibly adding some sort of alarms there could help
  • For engies its a pain to move the base all the time. Especialy if they are at a stronger position. It can be considered a tradeoff but also could mean that there should be a little bit more options.
  • A and B are capped too fast giving no real chance on defending.
  • Once A or B is capped bots seem to forget that they also need to attack C. Instead they go for the bombpit and do nothing there.
  • Once A or B is capped people seem to stick to that cp and ignore the original spawn. I migth have to put a text hint for that
  • Heights at C are anoying for many to use effectively. I probably am going to lower the sections next it.
  • There lacks alot of cover on C. The roof area doesnt feel anything useful
  • Walk distances to reach the point in C are also big, for this im thinking to add a brige to reach it faster
  • Money is able to scatter in this map making the scout job hard to do. The mission realy needs to have money to compensate for mistakes.
  • It seems that when waves are more focussed on 1 point that the other one quickly gets ignored completely. While if they have to focus on both they drop major defense lines. Probably because captures happen fast.

A good test which shows gameplay flaws which i have to solve. I dont have the feeling that the whole system is wrong, its just that atm defending is too hard which breaks the system that i ment to have where people had to move each time. Because they simply cap too fast you cant even move away from the points.
 

UKCS-Alias

Mann vs Machine... or... Mapper vs Meta?
aa
Sep 8, 2008
1,264
816
And i reworked the area at C:
  • I added alot of cover
  • I lowered the higher section from 512 units to about 320
  • Changed the bot route from A (lower path) and used the old path to make it connect to higher C (gatebots and bombers will avoid it). Im thinking on adding a gate that closes when A is capped
  • From the higher C section there now goes a bridge to the cappoint
  • The whole side area (which was only used as bot flank) is removed. Instead the upper bots now will use that new bridge instead
  • The staircase area on the upper section is now removed and its now the corner with the window. This is what i suspect to become the ideal teleporter spot
  • The approach route from the spawn/hatch is now lowered to match the C height

2014-02-03_00001.jpg

2014-02-03_00002.jpg

2014-02-03_00003.jpg

2014-02-03_00004.jpg
 

UKCS-Alias

Mann vs Machine... or... Mapper vs Meta?
aa
Sep 8, 2008
1,264
816
Based on the test of today a3 has been released! This is the first well playable version of the map.

Note that waves are between advanced and expert difficulty. On an unexperienced team these are far too hard and most likely will end up getting steamrolled quite a bit.
 

Snowbat

L4: Comfortable Member
Apr 23, 2013
165
74
We played yesterday and it was pretty good. Lots of action. But it's gonna need some work becaus eright now it feels kinda messy. Bots are running all over the place and don't seem to consistently go for A or B then C.
I think this is a bug right?

I noticed how money is an issue. It's very hard to get A+, meaning there's gonna be a shortage of cash on a map that would actually need you to buy more upgrades than usual. I for one think running speed could greatly help a player get from A to B faster, since it's a big and complex map.
So maybe it's a good idea to have the robots drop a little more money?

It's good to have 2 cappoints, but make sure your waves are built so that it still becomes somewhat possible to defend both early in the wave. If you really split the robots 50/50 from A to B, it's gonna be hard for a team to prevent capping early on. Better would be to send the bulk of the robots to one CP, and a smaller portion to the other.

Other than that, the difficulty was good. Comparable to Two Cities and that's good. Not sure what class combo is best, but with the Medic is was pretty good yesterday