3-Way battle simulator - Page 3

Kipper · 2012-03-16, 07:05 AM

Well I originally started building it just because I wanted to see what the difference between two identical maps with all things equal might be if one of the maps had fixed footholds and the other one didn't. I have a theory of what might happen as explained in the other thread, but I thought a bit of simple scientific testing would back it up.

Problem is, everything took WAAAAAY longer than I expected :P

I would have loved to make a version that has more variables that you can switch up and down, such as who starts where and with what, populations, resource values, some hexes having bases which are desirable, and many of the other ideas people have come up with here.... some of them I think I could do in time, some of them I'm not sure about.

If anyone wants to expand on the idea, maybe work together or modify whats there - then great, I'd be interested in seeing a really neat simulator where we could all play out different scenarios and see what happened too.

Don't forget though - good science is about eliminating variables if you want to purely test one idea against another, and thats what it was about! (except I didn't get to testing the actual other idea!!)

Today though, I think I'd better do some paid work

I've stripped the code out of the site and zipped it up as a standalone,which is attached to this post - feel free to take it, bend it, break it, reassemble it, and see what you can come up with..... (I make no apologies for the state of the code, it is commented, but since I made the whole thing up as I went along, there may be a couple of WTFs in there.... You take it as it comes!)

Shogun · 2012-03-16, 07:22 AM

Originally Posted by Coreldan

It's a tricky situation. Cos overall if the footholds will stay always the same, NC are very unlikely to fight next to the hexes near TR foothold. That's a lot of handcrafted area going to waste most of the time.

that´s my point. in ps1 it wasn´t a big concern if you were always fighting over the same base, because all the other bases were exactly the same. and even then it sucked sometimes when a stalemate would last for several days.

but in ps2, i expect every part of the map to be unique. and so i want to fight in all those unique places instead of just the same base over and over again.

sure, the concept of homeconts/homehexes is nice, too. but diversity on fighting terrain is on a higher priority for me. and since the devs worked so hard on every inch of terrain it should all be used.

Snipefrag · 2012-03-16, 08:16 AM

Great idea mate, keep up the good work.

Soliex · 2012-03-16, 09:01 AM

Good job on the simulator, all it really makes me want to do though is play planetside 2 and help those little blue hexes take over everything.

basti · 2012-03-16, 09:50 AM

Obviously the sim is quite basic right now, but the basic idea is actually quite brilliant.

If you would expand on it, you could replicate Every hex grid of each continent, and fill all the nessesary data in (type of hex, ressources, etc etc etc), pretty much a full replication of the map. Then, hook it up with the games data, means use it to record whatever happens on a continent, and build a database that way.

Sooner or later, you have enough information to really simulate possible outcomes. Would be quite a lot of work, but if you add the ability to use this map to plan a strategy ahead (and maybe even share it with others), it could be one hell of a powerful tool.

Imagine 5 Big outfits (30 each) teaming up for a night. Each outfit picks a target behind enemy lines, and points out how they plan to advance. Given all the collected data from the past, its possible to simulate a likley outcome of the attempt, allowing the outfits to prepare for this possibility (think about TR having much faster and stronger resecure teams going around than NC, while VS are the 5 outfits).
Would be a great lot of work, but damn i would love to see this. I would actually pay for it!

Shogun · 2012-03-16, 09:53 AM

and if planetside 2 doesn´t meet our expectations, we will just stick with this simulation instead ;-)

DxC · 2012-03-16, 10:13 AM

I liked it very cool to watch and fiddle with it Made me think of that map back when Planetside 1 was out where it would show a update on the fronts used to watch that like a hawk at collage :P

Miir · 2012-03-16, 10:49 AM

Very cool stuff!

You should make it so it's more interactive. So you can choose a side and select the hexes you want to attack and see how the other sides deal with it.

Kipper · 2012-03-16, 11:03 AM

Originally Posted by basti

Sooner or later, you have enough information to really simulate possible outcomes. Would be quite a lot of work, but if you add the ability to use this map to plan a strategy ahead (and maybe even share it with others), it could be one hell of a powerful tool.

I plan to have a live tactical map, game feeds permitting - of each server/map so that you can keep up with the war from your desk at work, or just have it on your tablet while you're playing...

As for simulating ahead from the live data ... wow, that would be cool, but given that this just a basic random AI, and the game will be governed entirely by player actions - how possible do you think that would be?

I mean, so many variables - an AI like this would obviously go for the hex with the most resources because it only captures one hex per turn, whereas players might organise to assault 4 hexes each with 25% of the resources to minimise their risk, split the enemy, and increase the chances of capturing something rather than all or nothing, etc.

basti · 2012-03-16, 06:03 PM

Originally Posted by Kipper

I plan to have a live tactical map, game feeds permitting - of each server/map so that you can keep up with the war from your desk at work, or just have it on your tablet while you're playing...

As for simulating ahead from the live data ... wow, that would be cool, but given that this just a basic random AI, and the game will be governed entirely by player actions - how possible do you think that would be?

I mean, so many variables - an AI like this would obviously go for the hex with the most resources because it only captures one hex per turn, whereas players might organise to assault 4 hexes each with 25% of the resources to minimise their risk, split the enemy, and increase the chances of capturing something rather than all or nothing, etc.

Well, the current "one hex per turn" would just get removed right away, because the game doesnt work that way.

To cut it short, the entire thing would try to simulate what players would do in any given situation, and the whole assumption is based on whatever players did in the past.

Thinking about this somewhat more, getting something like this in a working state may require quite a few people for quite a long time. So put it on the list, at the very end. If you dont have anything else to do for your stats, do the AI. Keeps ya busy for ages!

Kipper · 2012-03-16, 06:23 PM

Originally Posted by basti

If you dont have anything else to do for your stats, do the AI. Keeps ya busy for ages!

I think the lack of coherent strategy that zerg mentality is with a constant player churn and people doing their own thing in smallish groups, weighted random probably covers it.

I'd just need to add weightings for valuable hexes, and a choice 'per turn' of attack/defend (maybe multiple attack/defend actions based on population).

It could get quite complex but quite cool at the same time.

Maybe if the devs/PR people have a 'map week' so we can see more of what things are gonna look like, tactically.....

NewSith · 2012-03-16, 06:38 PM

Kipper, can I start a flamewar here about prolonged stalemates and boredom, Figgy style? I mean your sim shows perfectly what's gonna happen most of the time...

Kipper · 2012-03-16, 06:43 PM

Originally Posted by NewSith

Kipper, can I start a flamewar here about prolonged stalemates and boredom, Figgy style? I mean your sim shows perfectly what's gonna happen most of the time...

Well... It doesn't account for multiple attacks, or failed attacks but population being equal, you could expect those to be fairly equal too over time - so even without fixed footholds, it would appear that the battle is going to mostly be a stalemate over the same hexes from the same direction. I hope to be wrong.

**Malorn** · 2012-03-16, 06:51 PM

Originally Posted by NewSith

Kipper, can I start a flamewar here about prolonged stalemates and boredom, Figgy style? I mean your sim shows perfectly what's gonna happen most of the time...

No, it doesn't, because it isn't an accurate model.

I hope it becomes a more accurate model, that would be interesting. I'm always available for ideas on how to help do that if Kipper or anyone else wants. I gave some in my first reply.

Skitrel had an idea about making it Risk-like where troops are deployed along the front and you use dice rolls for battle outcomes and as populations change it does-redeployment.

Any accurate modeling would also have to take resources into consideration. It could be simple - each empire needs each of 4 resources. So depending on whether and how much of each resource it controls it will change behavior accordingly. Like if it doesn't have resource Foo it will find the nearest supply of it and allocate more troops to that location. Likewise if the enemy is encroaching on one of their precious resources they would allocate troops to defend it. Resource denial is another tactic that could be added.

All in all I think a RISK-like approach with adding in resources would give us a much more interesting simulation. Afterall, Higby has stated several times that resources are important and we will change our behavior based on them. So any simulation must account for that or it will not accurately reflect PS2 battles.

basti · 2012-03-16, 06:51 PM

The current version pretty much shows what a big group of random people that only work together moment to moment would do. Means, right now its a zerg simulator.

Once you add a single outfit doing stuff, everything changes drasticly.

Yet, there is truth in your words. Most of the time, the battle will be around a certain area of hexes. Its one of those many things that make me dislike the entire uncapturable foothold idea, as this is a direct result of said idea.

Beta shall come, no way to be fully certain how battles could evolve, without seeing it myself.