Tournament Poll#2

[Natch]

Long poll question, I'll explain.

When you mod a weapon, it will cost differently. Should this delta(change in) cost scale with the number of changes, or grow exponentially with the number of changes?

Example:

I have a Blaster. It costs 2 points, lets say. I mod it, and my formula says it's stronger by a factor of two. Should it cost 4 points, or 5 points?

Keep in mind, this entails several things. My formula is practically designed to be exploited. I'm not weighting one stat change more than another, as things stand. So, this means that one weapon that's said to be better by a factor of two may be worse than another weapon of the same kind which is also said to be better by a factor of two.

Now, having an exponential increase in point screws over the guy who made a crappier design even more. But it ALSO raises the challenge of getting the most out of your weapon alterations. So basically, an exponential point cost will make normal weapons a more viable alternative, and make it more of a challenge to use them effectively.

Basically:

Scaling Cost: Less challenge
Exponential Cost: More challenge

[ChaosTheory]

Exponential = more challenge = good.

Plus, I thought of some . . . interesting exploits just now. Like how to get double (or more) the firepower for a weapon with no price increase and no overall loss of stats. :twisted: Actually, I could get 10 times the firepower with no cost increase, technically. And if I use it the right way . . . even a 100 times better. Beautiful. I love this.

Also, you had better do something to take into acount how different stats affect differnet weapons, such as how deviation is completely different for blasters, beamers, and pulsars.

[Natch]

I've PMed Th15 to ask for the default of all weapons and modules.

He's yet to reply. Also, since YOU don't know my formula, you may not be able to do that as much as you think.

I'm doing things a little differently, as I'm not going to have to account for the differences in beam deviation, blaster deviation, and pulsar deviation.

I'm basically giving you a limit on how MUCH you can mod a weapon stats-regardless of what they are-and see who can use that limit most efficiently.

[ChaosTheory]

th15 also never replied to my question about his section code. I don't think he's had access to a computer since he left for SF.

Anyway, my exploit method is extremely simple and suprisingly effective . . . and I thought of it after I posted, so that my vote has nothing to do with it. Anyway, good luck finding a fairly balanced system. Maybe if you're smart, you'll prevent the exploit I'm thinking of.

[Sponge]

It depends on the attribute, but I'd typically go for a linear increase. Things that aren't increased directly, but instead indirectly, may need to be an exponential increase on top of the linear increase already in use. For example, Doubling a weapon's DPS should make it cost twice as much (linear). However, if in the process of doubling the DPS, the weapon becomes harder to intercept by flak, an additional exponent should be added to the price. This doesn't have to be anything big, just raising the price to the power of 1.1 should be sufficient.

Alternatively, you could just include a linear multiplier that multiplies the price instead of raising it to a power. This would be easier, but not as effective at preventing people from abusing the power of indirect benefits of weapon stat increases. This may be a good thing, though, as it would allow for more strategy.

I also think each weapon needs a separate formula. Granted you could use a universal formula based on percentages, it would be easier to create one formula for each weapon. Why? First, some weapons have different changeable attributes than others. This will make parsing an absolute pain in the ass. Second, some weapons are effected differently by these indirect changes that I mentioned above. Missiles, for example, start to tie up flak as you increase their projectiles. Other weapons do so to a much lesser extent. You will have much more accuracy if you have an individual formula for each weapon, and it wouldn't be much harder.

[Natch]

Okay. Look.

Custom weapon tournies have NEVER been done before. I am going to use a universal formula that limits the strictly numerical changes of each and every stat, regardless of what this stat may be. I know it would be more balanced if I made a separate formula for each weapon, but we are in uncharted territory.

As a logical thinker, I know that changing as little variables as possible is a good thing. We'll keep the formula relatively simple, and then adjust it as I and other people do more and more of these tournaments.

Remember, CmdPrompt is going to code this program. We don't want to put too much work on him to start with. What if we find out that in giving each and every weapon a different formula, we screw up one and a single weapon suddenly becomes OMGWTFBBQ HAX! Now we have to DEBUG THE CODE. AND THAT IS A PAIN. Or, maybe we need an entirely NEW FORMULA. This is a pain too. Also, keep in mind this will be mid-tournament, and so that'll push back the date where we finally accept entries, etc.

I want to avoid this, for now.

Now, I assure you, we will start fine-tuning the formulas and making new ones as time goes on, but we need to keep it as simple as possible with this brand-new frontier we have here. My idea for this tourney is the same as the original custom ship tourney-we give you a limit on points, a few base rules, and then let you go crazy from there. In the first tourney, no one had any idea what would work the best. Nobody.

This is that first tourney. We're giving you a point limit, a few base rules, and then letting you go crazy from there.



Also, who says I'm gonna use parsing?

[antisocialmunky]

/me 1/4ths Megatac reload rate.

[Natch]

Ok, 50/50 so far with the vote.

I'd like more opinions please.

[Sponge]

Okay, just my $.02 (well, $.02 more than I've already given ). CmdPrompt says he will be writing a PHP script. I don't know how familiar you are with PHP, but it's not a language that you use to write programs, in a conventional sense. It's a scripting language used to generate dynamic web pages, such as this forum. In the sense that CmdPrompt will be using it, it will effectively be a web-based calculator program. Parsing is fairly necessary, because, if I'm envisioning this correctly, the user will get a data string from Ship Builder for their weapon, and paste it into a field in a form. They'll then select their weapon.

The script will parse this string, locating each important bit. It will then take the values, and compare them against the standard values for the weapon. If you use a single universal formula, it will base its calculations off of % change, and probably multiply the initial cost of the weapon by the % increase + the % decrease of the stats.

Still using a universal formula, you could insert constant multipliers that would be distributed to specific attributes that were changed, if you wanted to weigh one attribute higher than another.

This is how I would see the script working. CmdPrompt may have other ideas, but in all honesty, it's a simple parsing/dividing/multiplying script with if statements thrown in to distinguish between weapon types, which is necessary if the script is to determine the starting values for a weapon so that it can find the % change.

Because it needs these if statements already, it would be very easy to incorporate custom formulas for each weapon. Here's a quick pseudo code example using a universal formula. All values are made up, hence the beauty of pseudo code .

Code: Select all

if ($weapon == "beamer") {
$basefirerate = 5;
$basedamage = 3;
$baseclipsize = 1;
etc.
<insert formula and calculations>
echo "$pointvalue";
}

if ($weapon == "blaster") {
$basefirerate = 8;
$basedamage = 4;
$baseclipsize = 6;
etc.
<insert formula and calculations>
echo "$pointvalue";
}

As you can probably see, it wouldn't be difficult to replace the formula in each of those examples with a specific formula for that weapon. It would be a simple matter of coming up with and balancing all the formulas on paper, as opposed to a major programming difficulty. Granted, it would be more work than just copy-pasting one formula, it's still not a big effort.

However, using a style similar to what I typed out above is not necessary. If you really wanted to use a universal formula, with no intent on ever changing it in the future, your if statements would only need to include the variables for a weapon's initial values. The script could then take those values and plug them into the universal equation, which would then only need to be typed once. Advantages of this? Less code. Does this decrease processing time? Not noticeably- all the calculations are done server side, so your computer speed or internet connection speed play a very small role. It's a much less modifiable system, and as such would be harder to incorporate individualized equations in the future. It's a lot easier, however, to modify the single universal formula.

Again, this is your tournament, and so these decisions are ultimately up to you. I'm just throwing the idea of individual equations out there, and letting you know that it's no problem for PHP to handle.

[Natch]

Oh, that's what parsing is. Then I'll prolly be using it. But yeah, what I plan to do is not weight anything, to see what needs to be weighted the next time around.

Also, you gave me a few ideas, thanks.

[Normandy]

This discussion still alive?

The following attributes need linear scaling:
Damage
Fire Rate
Turn Speed
Weapon Arc
Beam Deviation (Beams)

The following attributes need a small curve:
Range
Clipsize (The curve will be determined by Fire Rate)
Burst (The curve will be determined by Clip Size)
Weapon Scaling
Missile Turn Speed (Missiles)

The following attributes need exponential scaling:
Reload Time
Deviation
Beam Duration (Beams)

Those are the attributes I can name off the top of my head.

[TheBlueEcho]

Well at least you covered the basics. I think the only 'Basic' stat not mentioned is Hp, and the not-so-basic Bullet Speed.

[Natch]

Oh no, I'm doing it a tad differently. It'll either be all exponential increases or all scaling increases. This is because we're not quite sure what stats are more easily abusable. This is the very first tourney with moddable weapons.

We need to keep it as simple and consistant as possible until we have actual data to support the weighting of stat mods.

[Squishy]

So lets say someone is allowed a maximum modification of 100%, spread across any stat.


Increasing damage by 100%
Weapon is 200% effective than normal.

OR

Increasing damage by 50%, and decrease reload time by 50% (doubling output of shots).
Weapon is 300% effective than normal.

OR

Increasing damage by 50%, and increase range by 50%.
Weapon is 225% effective than normal.



The system MUST account for the final statistical improvements to a weapon when compared to the default.

[Natch]

*Facepalm*

You people say these things as if you already know how I'm going to calculate the stat modifications. The thing is, you don't.