[Gameplay and Balance]Bad Behavior


#1

Time for a game play thread instead of a bug one, yay!

So I slapped together another video clip I’d like to share and see what you all think.

youtube.com/watch?v=Z7AzQl4U5LE

First off, as you can see my fleet is stacked into one giant pile. The ship design is laughable, I just threw as many far reaching, high damaging weapons as I could on the largest chassis, slapped together a few escorts with anti-fighter weaponry and sent them on their merry way.

Although I’m sure this fleet is easily defeated because of poor ship design, the problem I see is the ‘stacking’. I don’t know if any of you have ever played StarCraft turret defense games, but in the later years some of the popular games were ‘stacked turret D’.

The problem this presents is the ability to concentrate a maximum amount of firepower in a very small area. In the first battle, my fleet is able to decimate enemy capital ships which are relatively close (the ships on the ‘southern’ end of the board) before the ships in the northern section even come into firing range. I don’t have a fleet anymore, I have basically one ship.

Imagine if I took the time to design a real fleet. Well balanced with anti-fighter, defenses, shields, missiles, decoys, and all the trimmings and THEN stacked it into one giant pile? Your fleet, spread across the length of the board like any sensible fleet worried about collision, will get chopped up piece meal by my one giant ‘ship’.

Interestingly enough, I lose the battle anyways despite having a decided advantage from the start(enemy capital ships fleeing out of range after being crippled, leaving their fighters to mop up my defenseless behemoths. This by the way I found to be an amazing piece of AI, thumbs up.)

The second problem… well I lost. Boohoo, guess I need to redesign my terrible fleet. But wait… I decide to just relaunch the same battle, with the same ships, without changing tactics or orders or anything except hitting ‘FIGHT’ again, et voila!! A smashing victory!

Randomness is good, it’s healthy, a little bit of luck should factor into a battle. I should NOT achieve wildly different results from the exact same battle. This severely diminishes competition and leaves the player feeling ‘why bother’. Their best effort and fleet design lost because of factors out of their control, and in this game, the entire battle sequence is out of their control. I can’t stress this enough how bad this is for competitive play or for anyone hoping to take this game as something more serious than a beer and pretzels type games that has a lifespan of 24 hours.

So those are the two problems I have with bad behavior, now for some solutions!

Problem #1:Stacking
Solutions:

  1. Ships are unable to be placed on top of each other during deployment phase.
  2. Ships have penalties to fire (reduced accuracy, slower firing rate) when having to fire in the proximity of allied vessels.
  3. Ships have collision and can ram each other. One giant pile like this will cause all my ships to collide with one another, destroying them all.
  4. Ships exploding cause area of effect damage. Remember master of orion 2, how ships could critically explode, damaging and destroying ships near them, potentially causing a chain reaction as ship after ships causes each other to explode? That would fix stacking in a heartbeat, any one ship in the giant mass could cause every ship to explode in an amazing chain reaction. (That, by the way, would probably look really awesome!)
  5. More weapons that deal Area of Effect damage. I haven’t explored all the weapons this game has to offer so this solution may already exist, but perhaps weapons that deal large area splash damage? Each ship in this giant mass all clustered together is dealt damage by a single weapon because they are too close together.

Problem #2: Wildly random results from exact same battles run repeatedly.

  1. Give the player minimal control over fleet actions during battle. I believe this goes against the core idea of the creator though, and degrades the game into a FPS game of who can micromanage their fleet better and faster. Still, it’s a possible solution.
  2. Reduce some of the RNG(Random Number Generator) in regards to ship behavior. How they decide to move and shoot. Make these things more predictable and more constraining in their parameters. This solutions sounds like ‘dumbing down’ the game though, so I personally don’t much like it.
  3. In competitive play, fleet ability is decided as ‘the best of x games.’ For example, a tournament might be decided by ‘winner is the fleet who wins 3 out of 5 games first.’ A single battle would have to be played out multiple times to see which fleet is superior more often, not decided by a single battle where flukes may occur.

I’d really like to hear your thoughts and input on these topics. I hope you don’t mind the micro-analyzing of this game so soon, I’m just trying to add my own input on how to improve GSB. =)


#2

Some interesting stuff here, that mirrors some of my thoughts during the devlopment of the game.
I thought stacking would be avoided automatically by the ships collission avoidance stuff, but are you breaking this by cunningly using formations? This had not occured to me, and I suspect the best solution will a combination of preventing blatant stacking during deployment combined with explosion splash damage.
I have got very close to implementing splash damage in the past. At the least, I might put in splash damage for ship destruction, at least for frigates and cruisers, in a ‘warp core breach’ style.
It could even add additional strategy to have some modules cause worse warp core breach damage than others. Imagine a suyper-deluxe ultimate power-plant that is super-efficient but goes up like an oil refinery if the ship explodes :smiley:

Great feedback, keep it coming. The replay-randomness issue is a huge pain, and one I really need to scratch my head about the best solution for.


#3

With the randomness - rewrite your random number generator to take a seed value and to give out the same numbers for everything afterward, maybe?


#4

have a beta-mode readout for the seed used to play the fight… and even an option to set it (to replay stuff). then if there’s a bug, instead of having to describe it/fraps it, we could just email the deployment file and the seed.

obviously, setting the seed would invalidate successful challenges :slight_smile:


#5

Splash damage is a must have! Slap it on some anti-capital torpedoes (Freespace style, for punchy fighters that add even more reason to consider a good PD wall) and some big, shooty kinetic cannons with explosive shells or something, and you have a winner.


#6

Maybe in style as Ackbar in Star Wars - “Concentrate the fire on this Super Star Destroyer” :wink:

Maybe its possible to add a control to define the primary target of the whole fleet?
There will be not much micromanagement that way.


#7

it shouldn’t be hard to make sure the seed is always the same for any specific challenge - or if you want to get clever, have the layout and configuration of both fleets fed into a hashing algorithm to create the seed. It would still be possible to have vastly different outcomes just by changing the position of one ship minutely, but for any specific config you’d get the exact same result every time


#8

I don’t think this would work if the ships are too slow by the time the battle begins their still clustered. Maybe the ships should have hit boxes around them during deployment that prevent overlapping. They dont have to be squares or circles but maybe a polygon shape depending on the ship, this would prevent grouping but still allow you to places near eachother.


#9

Just so you know how the stacking is being done, it isn’t by directly dragging the new ship from the ship selection on to the stack. It won’t let you. You have to drag the ship to some empty space and drop it. Then, pick it up again, and put it inside the stack.


#10

So after figuring out the Challenge system (it seems to be bugged and every single battle by every single person is getting uploaded, hence a GIANT list btw), I thought it’d be fun to fight a mirror match. Same fleet against itself, same formation and orders and positioning and everything.

Results so far?
Left side 2 right side 1 with very different fleet % remaining results…

img406.imageshack.us/i/ss1vwk.jpg/
49% to 6%

img25.imageshack.us/i/ss2apg.jpg/
21% to 6%

img134.imageshack.us/i/ss3s.jpg/
6% to 27%

Those 6 and 20 games were actually pretty close, but being able to go from 49% win to a loss… wow hello variance. My question to you is this. What should happen in a mirror match? Two perfectly identical teams facing off against each other, the absolute epitome of flawless balance. Should every fight be a draw? There’s no skill difference here to offset the two teams, the only difference when these mirror fleets face off is the RNG they get. Which side hits more and which side misses more. Which sides gets the damage applied to more important subsystems and which side spreads their damage too thin across a vessel. This RNG is the only determinant in a mirror match.

So again I ask, in this epitome of flawless balance… what should the outcome be? Should it be a draw every time? Should there be a winner? Since both sides are equal, if there is a chance a side will win, their odds of winning are 50/50. I’m sure if I took a lot more samples, say hundreds or thousands of trials of this mirror combat, left would win half the time and right would take the other half.

Is this what we want? Should we be striving to limit just how much of a margin a fleet wins by, for example every game will be as close as 20% to 6% or closer?

Thoughts?


#11

Keep in mind - you have positive feedback happening with mirror fleets. You can expect an even draw between two fleets of multiple ships, but if you have two fleets with, for example, screen ships to take down incoming torpedoes, obviously if one fleet loses its screen ship early, it’s going to fold up almost instantaneously, even though the two fleets are - as you say - identical. I think this isn’t a balance problem at all.


#12

This is very true, think of it as like the butterfly effect - it doesnt take much randomness for one fleet to have a 1% advantage in ship numbers over the other, and this small advantage can rapidly become amplified.

With 2 very different fleets this effect isnt as prominent, and can in fact become totally reversed. I experimented with a fleet made up of a large number of fast close attack frigates fighting cooperatively to take down enemy capital ships (which, by the way, seems to create the stacking effect regardless of initial layout as all of the frigates converge on the same target). In the initial stages of the battle it looked like i was going to clean up, they very quickly knocked out a few cruisers taking the enemy down to 80% with no losses on my side at all. But then things started to very quickly go the other way as the damage on the frigates which had built up became severe enough that frigates started blowing up left right and centre. Think the enemy wiped me out with 50% total losses in the end.


#13

What would be nice is more modules to reduce the affects of the randomisation. At the moment ships are pretty much at the mercy of the RNG entirely.


#14

A way to avoid the ships spreading out is to not give them any engines at all. Just equip the biggest ships with long range weapons and lots of armor and stack them. Doing this, its way to easy to beat expert on all the challenges.