View Full Version : Systems Balancing Dev Diary
Archived Post
11-04-2009, 10:00 AM
Are you one of those people who loves crunching numbers in MMO's? Curious which power has the greatest DPS? Do you ever wonder who those people are behind the scenes, making those calculations and balancing everything surrounding the numbers? System Designer Brian Urbanek has written a two-part Developer Diary on Systems Balancing, which includes everything from player progression to combat mechanics. The first part was just released today (http://www.tentonhammer.com/node/76328) on Ten Ton Hammer, so head over and check out the article, then come back here and tell us what you think. Enjoy!
Link to the news article. (http://www.champions-online.com/node/594659)
Archived Post
11-04-2009, 10:19 AM
Greens are still better than purples at 40.
Also, some stats scale badly, he may want to look into that (I'm looking at you Presence for non-Healers.)
Archived Post
11-04-2009, 10:22 AM
Greens are still better than purples at 40.
Also, some stats scale badly, he may want to look into that (I'm looking at you Presence for non-Healers.)
That is why the work is never done
Archived Post
11-04-2009, 10:34 AM
The mafs hurt mah head!
Archived Post
11-04-2009, 11:16 AM
Nice article, very informative.
Archived Post
11-04-2009, 11:53 AM
It was a well written article, both insteresting to read and informative. Now use that info and make the game more balanced >=O
Archived Post
11-04-2009, 11:58 AM
im a bit thick in the brain department, so I found that a bit confusing, but only a bit.
I would like to see tables and graphs to visualize what you are saying.
Im now lev 36 and feel that I have to give all of my stats a good number, with a high con, as I died so often at the start of this game, I still cant get out of the havit of maxing my health as muh as possible, im scared to but all off my eggs (stats) into one basket so to speak so the only way I can see to get a stat score of 200 by lev 40 is the dramactically drop my other stats. What should an archer/supernat drop and what to increase?
What are he best stat sets for each power set, how does mixing and matching sets effect a well rounded hard hitting character?
I would like more info, and some visuals in the next installment, but im glad someone is writing about this subject.
PS Everyone I have spoken to thinks the 'gifts' given from open missions are rubbish, apart from the one given at the prision break in MC, I have never used any, but I have only done about 2/3's of them.
What is the best 'drop' in the game does anyone know????
Archived Post
11-04-2009, 12:21 PM
I <3 this stuff. If I was luckier, I'd totally have his job, I do these types of calculations in my head constantly while I play MMOs, it's my favorite part about games. Seeing where these baselines are, seeing where I can eek out a little bit more to surpass the baseline, seeing how everything balances out, then complaining on the forums when I find something out of whack :D
<3 to you guys for publishing articles like this, they're super fun to read.
Archived Post
11-04-2009, 12:56 PM
Greens are still better than purples at 40.
Also, some stats scale badly, he may want to look into that (I'm looking at you Presence for non-Healers.)
It's a Nemesis.... err.... I mean...
It's a Foxbat plot!
Archived Post
11-04-2009, 01:20 PM
Math sucks. Just play the damn game.
Archived Post
11-04-2009, 02:33 PM
Math sucks. Just play the damn game.
Math is what makes the game fun! Without math it's just Barbie dress-up.
Archived Post
11-04-2009, 04:07 PM
Math is what makes the game fun! Without math it's just Barbie dress-up.
Barbie's been around for 50 years, I guess dress up is fun too. :)
Edited to add: Good article, I look forward to Pt. 2.
Archived Post
11-04-2009, 04:13 PM
Barbie's been around for 50 years, I guess dress up is fun too. :)
Edited to add: Good article, I look forward to Pt. 2.
Yeah, but I don't care about the dress-up as much as the math. I am, after all, a boy. Girls may hate math, but boys love it! Except for Eddie Izzard, he's much more a fan of the dress up ;)
Archived Post
11-04-2009, 04:22 PM
Twenty points of energy equals twenty five points of damage, and players should survive about twenty seconds of trading punches against an enemy of approximately equal power.
Well, the first part of this equation is totally wrong - for good, in my opinion, as that would mean a nerf to pretty much every damaging power that would be the biggest nerf in MMO history ever...
Interestingly, despite this basis being totally wrong in regard to what actually is in the game, the rest of the text is quite well-written, understandable, and (at least for someone like me) absolutely interesting, I love this kind of stuff (weirdly, I end up playing RPGs and MMOs mostly by feel, concept and style and less number-crunching in the end).
The only thing that was really tough was the equations.
<base value> + <base increment> * <number of increment steps> * (1 + <inflation rate>) ^ <number of increment steps>
is a lot harder to read than b+i*n*(1+r)^n and then explaining what the symbols mean. Especially hard to read with FOUR line breaks within the equation. Ouch.
Archived Post
11-04-2009, 05:39 PM
I put the equation into Excel. I guess to do not totally understand the factor "increment step." I assumed that would be either an arbitrary value or a value based on level. If it's an arbitrary level then I end up with a linear progress from level 1 to 40. If its a function of level then we'll have exponential growth of the stat value, however, i could not easily find a solution that ran from a base stat value 5 to 200 from level 1 to 40 with a given value of 76 at level 16.
Additional things to consider for the next article is the quality controls that were put into the system of equations. It sounds as though they are designing for the "average" or mean stat values. What was found to be the standard deviation during testing? Was a cap put on the control limits with in the system to meet a specified set of stat ranges?
Was factorial testing performed with the system of equations to identify how significant each factor is? I suspect it was because I remember beta testing without equipment and I did find it to be a significant factor in attack and defense. What other factors were considered during development? Character attributes and gear attributes were mentioned in this last article.
Great article. It can help many engineers or aspiring engineers get a better understanding of what it takes to develop a game.
Archived Post
11-04-2009, 09:01 PM
Barbie's been around for 50 years, I guess dress up is fun too. :)
Edited to add: Good article, I look forward to Pt. 2.
Math is what makes the game play right, without it it would be like dressing up barbie to get slaughtered by your little brothers power ranger.
I still question the calculations though, it seems the greens and purples don't really make sense. As someone said Greens typically is better than purples. In a game were super stats is importaint I don't understand why a purple with 10 here 20 there in 4 places is better than a green that has like 70 in one area you could just stack on super stats to boost your strength and other areas. I guess this is where the art part of the math fails because it seems they are assuming people will spread out their stats more than a person who stacks points over his super stats to get the maximun strength he needs. As the game has shown, a player is much better stacking his super stats better than someone who thinly spreads out his stats evenly.
I would like to see in part two on how they would expect a player who plays at higher levels utilize only purples to dominate a person who uses only two stats/superstats as it seems that is what he is assuming.
Archived Post
11-04-2009, 10:05 PM
As the game has shown, a player is much better stacking his super stats better than someone who thinly spreads out his stats evenly.
I'm not sure if that's correct. I suspect it depends on the superstats and powersets. A Might build that only stats for CON and STR is going to do a tonne of base damage, but will struggle to crit, heal more slowly, build energy less efficiently, etc. Whether this outweights the additional damage is an open question.
I would like to see in part two on how they would expect a player who plays at higher levels utilize only purples to dominate a person who uses only two stats/superstats as it seems that is what he is assuming.
I'd like to see this too, particularly comparing different types of two stat/superstat builds vs more "balanced" builds.
Archived Post
11-05-2009, 01:23 AM
Are you one of those people who loves crunching numbers in MMO's? Curious which power has the greatest DPS?]
First part was a total yes. Then I remembered I'm playing Champions Online. You ask us an obviously great question that the game itself has no answer to. Give us the "finer details" of our powers already :P
Archived Post
11-05-2009, 02:48 AM
Math is what makes the game fun! Without math it's just Barbie dress-up.
And? What's your POINT?
Archived Post
11-05-2009, 03:48 PM
Love this article, but IMO the DPS stuff should be left only to those determined enough to test different setups on their own.
Releasing hard numbers on DPS potential may just push most of the FoTM noobs right onto that particular framework, and before you know it half the server is ego bas - I mean - whatever the numbers say :D. Of course some will realize that DPS is just one facet of an MMO, but others will surely be blinded by the numbers.
IMO Everyone should feel that their particular build has enough DPS for them to succeed, even if its not in the top tier set. Nobody should be be put off just because the numbers say what they are doing is not ideal.
Archived Post
11-05-2009, 03:59 PM
Releasing hard numbers on DPS potential may just push most of the FoTM noobs right onto that particular framework
You realize that power optimizers generally work out their own hard numbers, since released data is usually inadequate to actually describe performance in-game?
Archived Post
11-06-2009, 07:55 PM
So this is the guy, huh? I feel like I know so much about him already. :D
Which of his calculations resulted in these frequent 75% nerfs? Was he off by a factor of 4 beforehand, off by a factor of 4 now, or just off by a factor of 2 both before and after? :o
For my part, I'm just glad he's working in CO's game design and not designing bridges for a living. :p