
Scobichevskaya
|
Posted - 2003.09.09 21:47:00 -
[2]
Quote: TomB,
I probably won't be saying anything you don't already know in this post, but I have to say it anyway.
This whole thing reminds me of my early attempts at creating a gravity simulation (I'm a programmer and physics/maths freak). Computing the new state of an object under influence of others and effecting the change immediately would cause unexpected results because the change influences all other objects, which then would influence that original object again... Feedback & Chaotic systems are complex beasts. I managed to circumvent the object update problem by maintaining two states for every object, but that doesn't apply here.
Still, as you are undoubtedly aware, the module balance is a similar feedback system. You cannot change one module without shifting the balance of power of all the others, and too strong nerfs can throw the whole system into a Giant Seesaw of Nerfing +3. In fact, this is happening already. Example, you are nerfing shield hardeners because they are too powerful in the current system, but at the same time you're playing with NEW stacking algorithms which will again affect the hardeners. You're setting yourself up for balancing hell here if you are trying to balance modules for multiple ways of stacking at the same time without having decided on a stacking rule yet (oh, the 75% thing is a bit naive. It allows people to partially cancel out negative effects, as already pointed out by numerous others :).
You're also setting yourself up for documentation hell. If you implement different ways of stacking for certain attributes, combined with the major inconsistencies in displaying the attribute modifiers (1.10x, +10%, etc) and cryptic attribute names, you will probably end up requiring Rocket Science lvl 5 to understand how the system works. In my experience, the average person won't understand why a module giving 10% doesn't really give 10%.. they end up asking questions the experienced people know the answer to but get tired of explaining it all the time. I already have a pocket calculator here to calculate setups. These changes will force me to type longer formulas or even write programs to predict what module X is going to do with module Y.
And like I said before in another thread, too many changes will change the game beyond recognition to those that been here from the start. It's like waking up one day and finding yourself in another world. That's not good.
Note that your balancing changes are also influencing the market system (mods will become uber, or worthless now), and you probably need to change the loot tables to reflect the new balance as well. It is also likely some people are going to be ****ed off to find their expensive blueprints no longer producing the 'good' item they bought the BP for in the first place.
Anyway, good luck, and I sincerely hope you know what you are doing, although to be completely honest, your posts here make me believe you also don't really know how to handle stacking in a decent way. the Chaos test environment isn't widely used by people, and most come there just to test ubersetups and what-if situations. Most people can't afford these ubersetups, and Chaos has no active market and no large battles, piracy, or whatever so it's impossible to see what a new balancing change will do to the game as a whole. So, I come to the conclusion that to truly test a new balance change and stacking algorithm, it would have to be put on live, but then, you will be putting us through another betatest cycle 
Of course I knew I would be paying for continued betatesting this game so it's not really a problem 
Sory for the long post :-)
I agree with Entity. I'm a programmer, too. I mentioned before that the developers should change one module at a time. In order to do a scientific experiment, you need to change just one variable. By changing many things at once, you will have no absolute idea as to how the module change affected the game; there is no standard to compare it to. If you nerf/change many modules you will get complaints now and then will have to nerf/change things again later to rebalance them again and again which will take up a lot more time to balance.
If you have specially hardcoded cases for the stacking of different module types, it can lead to accidental calculations if it not remembered or documented well.
I think the first change should be so that players cannot use more than one of the same type of shield hardeners and leave the shield hardeners at 70%. This will solve the issue of Scorpion tanks.
The second change should be missiles and missile launchers.
After those 2 changes, modify stacking if it is still necessary. Why don't you just have the modules affect the base number so there is no stacking, but there is still less an effect than if they were stacked (this doesn't apply to shield hardeners or energized armor)?
------------- RUS / RUA RUSSIAN Team
|