Dria - TEST

[pooty]

Some of you may have noticed Dria-Randomizer-V22-TEST

I moved node 1 back a bit, in an attempt to even out play. The one side had distinct advantage that 5 to 1 was much shorter than 10 to 2. There's still disparity in 6/9 routes, but the fact that from 5 you could hit 1 directly from the building (with SR, LG, or even accurate tank shots), not to mention Hurricane spamming. Mino can still hit 1, but needs to move more on the hill so its harder, and exposes the mino more, like the other side.

So please as we play it let me know if this is a worthwhile change. There's only been one match on it https://www.omnipotents.com/utstats/mat ... match=8478 (Just Pharizwheel and Terminator). And regular dria is still there.

[pooty]

So it seemed like this was a good change. The winning team won on both sides, and lost the first match.

It definitely opens up the area in front of 5/1 so its more like the other side.

[Judas]

I think we should redo the vehicle loadout for Dria Randomizer to something more similar to what has been done on HSH. Below is a quick sketch of what I had in mind:

Dria is a good map, and there's plenty of reasons why it's popular. However, a lot of the time when I play this map, I see the following problems:
- Slow/ineffective vehicles sitting unused at rear nodes or core
- Players not involved in action driving a slow vehicle from a rear node like node 4 and wandering around
- Not enough or too much anti-air; same for flyers
- Camping with hitscan

Currently, I see a lot of vehicles like a fire tank or rail tank at node 4, and maybe a flyer or manta. It takes ~60 seconds in a ground vehicle to drive to the middle nodes, which are the main attack priority. There's no reason to use those slower vehicles: it's boring and ineffective. Flyers can get anywhere on the map in ~20 seconds, ignoring the terrain completely. A small minority of players camp the flyers and the map outcome is ultimately decided by them with everyone else on the sidelines getting scolded repeatedly by the air pilots: "Why am I alone at this node? Are noobs wasting my flyers?"

Therefore, I think it makes sense to have flyers furthest away to provide some balance in time to attack across all vehicles. My idea here is to try to apply a kind of paper/rock/scissors paradigm: power vehicles win vs anti-air, anti-air wins vs air, air wins vs power vehicles.

[Rhamp]

I second this. I might actually like playing this map again.

This is where the category randomizer would come into play so that each node can have guaranteed types of vehicles. The "kitchensink" randomizer maps by nature are not balanced, so doing this would help with that. Seeing that I've starting learning unreal script, I could dig into the existing randomizer and see if I can make a new version capable of doing this, and also fix the absolute spaghetti that sometimes causes the randomizer to just break and throw random bullshit regardless of how vehicles are weighted.

Another alternative is using math with the existing randomizer. Shit-tier vehicles like badgers and basic scorps would have value of 1, anti air vehicles like hellhound, mantis, centaur, ballista, railtank, etc would have a value of 10. Fliers would have a value of 100. Power tanks and mechs would a value of 1000. Now all you'd have to do is for each node, set the range for every digit of how many vehicles of each category you want. So for example, for node 3 the range would look something like 1010-1021, which would spawn a power tank, an anti-air vehicle, and either another anti-air vehicle or a shit-tier vehicle, guaranteed. Node 5 would look like 30-31, as it would only spawn 3 anti air vehicles. Node 6 would be 10-23, spawning an anti air vehicle, a shit-tier vehicle, and one that is either anti-air or shit-tier. Node 4 would just use the fliersOnly boolean, but would otherwise be 600-6001 (I think it spawns 6 vehicles, but I forget). Node 1/2 would be 10-22, spawning an anti-air + either another anti-air or shit-tier. The bastion is a forced factory, so that isn't counted in the randomizer. The core would be whatever the fuck you want, something like 2222-3542, giving 2-3 power tanks, 2-5 fliers, 2-4 anti-air, and always 2 shit-tier. The only problem would be is if any node/core had 10 or more vehicles, since now you'll have the possibility of carry-over, which if that is the case, separate each category by another digit, so shit-tier would be 1, anti-air would be 100, fliers 10000, power tanks 1000000.

[pooty]

Well most like the current edit of Dria-Randomizer.

I think if you're going to be so specific on vehicle loadouts, just play the non-randomizer version.

And WRT to HSH, its one of my least favorite maps... too many vehicles, devolves into a stalemate vehicle DM, with winning being who can OMFG 1/4 more.

That being said, I think the "Vehicle Category" idea is a good one. Let me think on that if we could code it up. We can do Rhamp's math trick with how the Randomizer does point assignments, but it'd be much cleaner otherwise. Right now the Rando has a "banned" list at each node, might be easier to just add "Node Group Vehicle Lists"

[Judas]

Makes sense. I think I went overboard on specifying loadouts (after all, it's Randomizer!). Thanks for having a look at it, pooty!