The Big Weapon Study [unfinished]

Printable View

Show 30 post(s) from this thread on one page

13-11-07, 20:20
CMaster

The Big Weapon Study [unfinished]

I was going to write a full, scientific paper style article for this. I even got started, but then I realised it was silly for what is clearly an unfinished, incomplete study.

So instead, you get a brief overview, and an invitation to carry on the project yourself.

For the record:
All damage figures are based on the raw (unresisted) value given from the character.log files - so the accuaracy is dependent on that being correct. All shots were performed at full lock, from point-blank range to the chest hit-box of a fully buffed, fully healed PPU (this ensures that the "blacking out" of damage zones and damage modifiers weren't involved). Based on the results of this study, different damage %s can be used to adjust all damage to an equivilant weapon.
Every shot is done without an ammo mod. This means two things: one, there is an assumption that ammo mods always increase damage by the same amount and secondly that APU and Melee appear relativley more powerful than they actually are.
The poin blank range exagerates the strength of rayguns and makes fusion firles appear weaker. Also, the damage figures for burst weapons assume every shot hits - even stood still at poin blank range this is far from guaranteed.
The devourer adds all DoT to it's calculation.
All frequencies should be the maximum frequency acheviable (cap)
Edit: Oh, and none of the weapons tested displayed random damage - all were completely consistent (in the logs at least).

Thanks to Brammers, Jini and everyone else who helped. This is far from complete - if anybody wants to complete it, be my guest. However the point - weapon TL and damage are barley related - should be pretty plainly apparent.

The Spreadsheet
Google documents version of the spreadsheet

Graphical representation of data:
http://cmaster.linxsoft.co.uk/images...eaponstudy.PNG
13-11-07, 21:47
SorkZmok

The idea is nice but there is a lot more to factor in.
Classes, runspeed, range, achievable setups and what else not. Without all this the numbers are worthless.

And you seriously need to work on those mad excel graphic skills. :lol:
13-11-07, 22:50
Riddle

very nice, I would like to comment further but i'm on a time slot tonight so need to log in for my fix :p

however.........less time in Excel........Get yer ass in game m8!!
13-11-07, 23:16
nabbl

nice study actually...

and smork, what has the weapon study to do with classes? i dont think u should compare tanks with apus and pes with spy just with that kind of excel sheet.

its more like u can see that the TL has NOTHING to do with the damage which is made at the end.

and thats really making me depressed :(
14-11-07, 09:34
jini

Quote:

Originally Posted by nabbl

nice study actually...

and smork, what has the weapon study to do with classes? i dont think u should compare tanks with apus and pes with spy just with that kind of excel sheet.

its more like u can see that the TL has NOTHING to do with the damage which is made at the end.

and thats really making me depressed :(

I agree with Sork on this.
It's not just a matter of "absolute" damage.
Many other factors affects these equations. For example: we dont know how fast reticle closes on a tank and on a spy do we? What affects this factor, because this is the most important shit. I dont care if I have a huge canon with super power but I do care to have an easy weapon that can aim easy and shoots and hits following my particular style. And this is where the balance project is real hard to materialize. We, simply have to trust KK on this, because since they are the makers of the program they see the whole picture in a way we can't. From my experience so far, I can safely say that all classes, except the APU are very well balanced.

Another example is the ionic pistol. This pistol really does great damage, and there is nothing you can do to avoid it. But its so difficult to handle that its essentially useless in pvp: how many low tech pes have you seen using only this weapon?

Seeing just the damage of a gun is like watching a room through the keywhole, trying to imagine the whole picture in that room :p
Classes are balanced and the game is fine. Just populations are low nowadays
14-11-07, 10:13
Glok

Quote:

Originally Posted by jini

Classes are balanced and the game is fine. Just populations are low nowadays

Will the real jini please stand up?

Erm, that graph makes rifles look superior to any other class of weapons hands down, which isn't the case. Factor in some percentage scale that approximates the effect of reticle close time and runspeed (of the weapon user) on the weapons and we could really assess the true DoT capability of the various weapons. The problem with that though is actual player skill, so the percentage scale would be an average of the player skill of all runners. That's a bit impossible I think, a few runners will always fall out of the percentage scale, higher or lower. You could just assume a z=-2 to z=2 scale and truncate the extremes. A running tally of duel results with the participation of most runners and the full weapon range with appropriate setups for requirements would be the only way to do this. I'm not sure it's worth it... (actually it almost certainly isn't)
14-11-07, 10:13
nabbl

Quote:

Originally Posted by jini

Classes are balanced and the game is fine. Just populations are low nowadays

LOOOOOOOOOOOOOOL

are you wearing your pink sunglasses? Oo
14-11-07, 11:58
jini

Quote:

Originally Posted by nabbl

LOOOOOOOOOOOOOOL

are you wearing your pink sunglasses? Oo

You are a nib
As far as I can play with my spy and compete head to head against tanks in OP fights with some, not a lot, ppu support then the two ends are balanced Nabbl. End of discussion. And this also goes for PEs as well. Only exception comes with APUs, which I have already said
14-11-07, 12:24
zii

I never understand your graphs. The x and y axis never denote what the values represent.
14-11-07, 12:57
nabbl

Quote:

Originally Posted by zii

I never understand your graphs. The x and y axis never denote what the values represent.

x = TL
y = Damage

to jini:

fuckin 90% of all the weapons arent even usable. and u call it balanced?

did u ever see bugging lowtech speed tanks in op fights? i can kill everybody with my overpowered ion shotgun cannon, spies dont have a chance even with shields on them, im fast as hell, nobody hits me and i have so much constitution that im always down the sync before im dieing.

that is NOT BALANCED AT ALL...

and its not only the tanks, what about Riggers? Riggers can easily level down there in the ceres labs and cap themselves, make money, rare parts and woc discs in short periods of time.

that is NOT BALANCED AT ALL...

the weapons are NOT balanced, the classes are NOT balanced and youre wondering y all the people left from terra and quit the game? -.-
14-11-07, 13:34
William Antrim

Quote:

Originally Posted by nabbl

x = TL
y = Damage

to jini:

fuckin 90% of all the weapons arent even usable. and u call it balanced?

did u ever see bugging lowtech speed tanks in op fights? i can kill everybody with my overpowered ion shotgun cannon, spies dont have a chance even with shields on them, im fast as hell, nobody hits me and i have so much constitution that im always down the sync before im dieing.

that is NOT BALANCED AT ALL...

and its not only the tanks, what about Riggers? Riggers can easily level down there in the ceres labs and cap themselves, make money, rare parts and woc discs in short periods of time.

that is NOT BALANCED AT ALL...

the weapons are NOT balanced, the classes are NOT balanced and youre wondering y all the people left from terra and quit the game? -.-

Im bloody glad youre still here, because Jini if we left balance down to you; well lets just say you'd probably break the scales mate. Cmaster, thank you for the study, or at least thank you for starting it. Maths is the only REAL way to balance anything even remotely fairly.

Jini get over yourself, you were never the best pvper in this game, and youre still not. Not by a long chalk. So dont go bringing those "I can kill such and such" arguments into the game to make yourself sound so much better. Its not about what "you" can do, its about what the game does.
14-11-07, 13:40
L0KI

Nabbl, I completely agree with your comments.

It's sad that only a VERY small percentage of weapons are viable in PvP, considering the sheer amount of high end weapons there are.

Balance needs to go hand in hand with variation, and customisation. I dont want to be a clone of every other player on the server, I want to use the weapon I like the look or feel of the most, and I want to be suprised by tanks pulling different weapons on me.
14-11-07, 13:41
Brammers

Hmm, been looking at the results.

If all the classes had the same con setup, then 2 characters who start a fight with the same TL weapon, same type of damage ammo would be expected to die at the same time. That would be perfect balance, but different con setups, ammo etc this doesn't happen ingame.

Now if you factored in con setups for the different classes and adjust the damage of the weapons. (Very difficult to do, I admit) then it should be the same situation, an APU vs Spy should still both die at the same time. However, this doesn't happen in game.

Even if you take the results for TL vs Damage, it's not a smooth line. (Have a look at the figures in the XLS file, note the ion Shotgun Cannon for example) There are a few weapons that just deal way too much damage per minute compared with other weapons in their class.

Conclusion: NC2.2 is not balanced.
14-11-07, 16:05
CMaster

Quote:

Originally Posted by zii

I never understand your graphs. The x and y axis never denote what the values represent.

Valid complaint, fixed.

About the apparent weakness of pistols: Note the abscence of any pistols people actually use (xbow, ceres, freeman, BoH). These would probably climb higher - anybody want to go get the data.

Another note - all damages are normalised to as if the player is getting 100% damage on a weapon. In reality of course, it is harder to acheive the same damage %age on a higher TL weapon, although the differences in attainable damage % are minimal.

Does this data tell you everything - no, of course not. There are other pros and cons to weapons aside from pure damage. The classes have different effects. But it does show a fundamental flaw in the balancing process.
14-11-07, 16:38
$ir Mafia

jini big man, jini big balls..ugh!
jini crushing skull of his dog with stone, ugh!

damn, try to show youre over 18 sometimes, will make you
some friends.

Nabbl isn�t a noob at all and he�s damn right in this case
(wow, did i really say that ^^)

NC 2.2 = unbalanced as hell

Even with PPU support on a Spy you aren�t even CLOSE to an unsupported
Tank, 4 good hits from him and you are wasted and he�s on maybe 50% HLT.

this whole thing isnt balanced, its a mess

Show 30 post(s) from this thread on one page