Previous balance updates and their effects

[standistortion]

Balance updates have been implemented in different ways and their effects have varied hugely, the obviously bad one was warmasters. The level cap raise broke more or less all the balance updates that came before it with an excess of power points, with the level cap at 50 builds had to be carefully tweaked to suit play styles and current trends in the game.

What's less obvious are the steady changes that evolve in RO's complex balance of skills as different builds are tried out and previously overlooked skill combinations turn out better than expected. Because of this a minor tweak to a skill can sometimes make it completely useless while a previously well balanced skill becomes overpowered with the evolution of builds.

Imho the balance was at its best just before the warmasters update and (iirc) that was due to a steady tweaking of skills while some of the worst updates appeared to work well at first but quickly became bad as builds evolved and weaknesses where found with no counter to balance them out.

Maybe RO's skill system is simply too complex to ever achieve balance and will constantly need tweaking. If so it may be better to tweak skills to influence balanced builds and to boost opposing classes counters to overpowered skills rather than directly nerfing the overpowered skills, that would always put NGD one step ahead of the current trend in battles.

The problem with that is the trend isn't the same across all servers and is unlikely to ever be the same as the size of battles makes a big difference to the skills used. As far as I can see the only solution to that is to have different skill attributes on different servers but a constant and steady tweaking of skills already requires a lot of work and different sets for each server would multiply that work many times.

Some time ago I'd suggested a complete replacement of the current skill system with a skill crafting system. It would be a huge change and probably impossible at this stage but it would allow a small set of skill attributes to craft an almost infinite number of different skills and only the costs of those base attributes would need to be tweaked to alter balance (ie. the range of of all knocks could be reduced by increasing the cost of that range).

Anyway, that's wandering off topic. Which balance updates where the most effective and was there a particular method of implementing them which worked better than others?

[bois]

For me balance updates need a few pre requisites.

1. Have a review of armour calculations. If they don't adjust this first, it could throw things off if they balance classes then review armour after.

2. A review of boss items, with an eye to thoughts on level capping certain jewellery specifically.

3. Some transparency on resistance and spell focus.

4. a review on hit chance because this underpins evasion/ block/ miss criticals etc.

5. Some thoughts on actually getting rid of Warrior, Mage, Archer and just going straight to the subclass as you enter the game. You can have a single retrain to your partner subclass in the game. This would then allow the spells in trees to finally be separated.
This could be a lot of work but may be worth it because balancing becomes less complex from the stand point of having to juggle a single spell between 2 different focused subclasses.

With that in place, spells would already be modified from the foundation. Then they can press on.

An effective approach to balancing (spells and classes specifically) would be data collection like what Chilko has initiated, then to a short list and finally an implementation on testing servers which is evaluated by partners, community and NGD beta testers. Then a month or 2 later after it goes live it is evaluated on the live server. If minor tweaks are warranted, then they are done. Rinse and repeat.

[standistortion]

+1 to all, many more aspects to be considered but those are all major ones. Point 5 would be a big change but doesn't initially need any change at all, just a gradual replacement or tweaking of old main class spells with sub class specific ones.

I think the last point is by far the most important one, accurate data collection would be a the most important tool for planning balance updates. Any link to information on that? I missed it and couldn't find it in a search.

[LittleHomer]

I think, if you want to have a good Balance in a game like Regnum, you should know and note:


The distances between the classes are too small or rather strange. E.g.: Hunters and Marksmen. Both of them has got the same amount of defense. But Marksmen is doing a lot of damage, without many spells!
Every class should be op in her strength. E.g.: Hunters should be very nice in camouflage and should be very nimbly. Damage isn't so important to them, because they are op in their strength! Other example: Knights should be very powerful. They should have a huge defense and they can survive for a long time in a battle. But: They are doing few damage. Same thing for the other classes
The warmaster spells must be special for the classes.
Every character should have the possibility to skill other then the opponent. At presen, they are too many spells which are useless (List: http://www.regnumonlinegame.com/forum/showthread.php?t=93038). Instead of nerfing spells, it would be better to make an other one more op. With this, the battles are faster and everybody has to have a better reaction time.
Also important is, that at present, nobody needs really skill. I think, only mages need this a little bit. I give you an example, then I explain it: The barbarian is only running into a crowd an he must only mindless hit the opponents. Where is there the skill?
I think, the spells of the characters should have more attributes. Don't only give them ''Make 500 dmg''. Why not giving them ''Make 500 dmg and resets the cooldown if you are killing an opponent''? With just a spell like that, the whole battle will be different, because everybody will try to use this spell finally. This is just an example and there are many ways to give the spells more ... skill.
finally: It's really important to know which class is good in which section.


... A good balance isn't done in space of one or two days. Respecting the community and doing the right, is important.

I really hope that any time, there will be a good balance in game :rolleyes:

[Zas_]

Been thinking about this since writing that post and there might be a simple solution to balancing spells on all servers, maybe I should put up another thread in the suggestions forum but it will do here for now.

NGD should be collecting statistics on what spell usage, I'd guess they're already doing it but if not it shouldn't cause much server load or use much space. From that they can get an average percentage for every spell over (example) 1 week out of the total of all spells cast. From that they would be able to automatically adjust the attributes of over and under used spells by adjusting expected average figures which could be tweaked as needed based on how much players are ranting about them. That way if we get situations like the terror chains then terror will become less and less effective while underused spells like evendims fury and splinter wall will get more attractive to include in builds.

The whole play style could be influenced this way and depending how the figures worked out it could make classes more or less desirable as an excess of barbs and marks would weaken those classes while strengthening the others.

EDIT: Sorry, was typing when you posted. Can you point me to the info on data collection, I missed it? Thanks.

NGD should collect statistics, and publish them once every 6 months or the like (for each server):

- total realm population
- total realm population active in warzone
- total time played by each class per time period (ie. 2 barbs play 10 minutes each during the last month, that's 20 minutes added to the barb monthly total)
- class and race population per realm
- hot map for positions of self Death (stuck...)
- powers usage per class (number of casting of each power during a period of time / total time played by potential casters during the same period)
- number of connected people per realm per hour of the day (GMT)
- etc...

[Zas_]

E.g.: Hunters and Marksmen. Both of them has got the same amount of defense. But Marksmen is doing a lot of damage, without many spells!

Hunters have much less armor/protection than Marksmen.

[Seher]

Anyway, that's wandering off topic. Which balance updates where the most effective and was there a particular method of implementing them which worked better than others?

Every balance update was great. 1.6.3 was magnificent.

The problem isn’t that NGD do not know what balance means, the problem is that they don’t prioritize it/that it gets postponed to the twelfth of never.

[kmdk]

NGD should collect statistics, and publish them once every 6 months or the like (for each server):

- total realm population
- total realm population active in warzone
- total time played by each class per time period (ie. 2 barbs play 10 minutes each during the last month, that's 20 minutes added to the barb monthly total)
- class and race population per realm
- hot map for positions of self Death (stuck...)
- powers usage per class (number of casting of each power during a period of time / total time played by potential casters during the same period)
- number of connected people per realm per hour of the day (GMT)
- etc...

Collecting population numbers info will be never be published ,and is normal ,and on this there are many factors as:
- a new player that join a realm,usual people tend to join powerful realm.
- is not commercial advised ( a new player can say ,oh only 20 players play this game ?..is a shity game then)

Also yes a % and not a fixed number shall be public daily ,in war stats as %:

Graph 1. Hourly player percentage per lvl (100% means total player online on server)
Example for this kind of graphic:
From 22:00 to 23:00
-from lvl 1 to lvl 30 = Ignis 12%; Syrtis 10 %; Alsius 6%
-from lvl 30 to 40 = Ignis 3%; Syrtis 5% ;Alsius 7%
-from lvl 50 to 60 = Ignis 15%; Syrtis 17% ;Alsius 9%

Graph 2:Numbers of WM from total population per hour.

Example:
-Ignis 2% Alsius 2% Syrtis 2%


How can be this two graphical results useful for Invasion mechanics?

Sample ,this percentages can be used to change mechanics of the invasion and to adjust in real time some invasions features.

There are many other useful graphs info stats that can be used in realm balance too.
Maybe game already has this features,but they are manually adjusted.

[kmdk]

Class balance has a common issue from begin of the game:

All classes share trees with spells

A class shall be direct oriented to a certain specialization.
-Hunters shall be faster ,lighter ,lethal and less armor
-Marksman slow havy armor ,have dmg but very slow
-Barbarians ,shall be real dmg dealers ,but week at defense
-Knights shall be havy but slow (curently i think is best balanced class from game)
-Conjurers shall be slow ,good defense
-Warlock shal be fragile but havy dmg dealers

Now maybe sound as drastic change and 132423 peoples will be disagree ,but i think is a must:

Remove from all subclasses sharing tree
In specialy defense /ofensives spells.

-From Marksman remove : tricks tree ,remove evasive tactics
-From Hunter remove :remove long bows tree or "?"
-From Barbarians remove tactics tree (not anymore caution spell and berseker offers penality of -100% armor)
-From Knights remove : hard to chose a tree
-From Conjurers remove : staff mastery
-From Warlocks remove : Enchantments tree and passive resists

Ofc with this come plenty of spells that need to be reworked or make them to work. Currently all know at least sum of spells that are useless make more than a tree.

Also armor system shall be reworked heavy.I never was agree with armor system that can't be calculated because random armor stats.A game full with randomness is a game full of frustrations.

A player that play from months can lose a fight just because he was never lucky to drop a good stats armor,and even worst atm a player from 50 and up ,have no armor ,only if its lucky to drop one ,or wait from months if he plays in a weack realm to get WM armor.

[bois]

Of course, I agree with Seher. :p .

Publishing information is ill advised. Kmdk covered the important points of that.

I have been asking for the removal of shared trees for a while now. It is quite refreshing that the community agrees on this.

This is why I came to conclusion of removing warrior, mage, archer. It is basically unlocking trees at level 10 but if so then you can unlock those same trees at level 1, yes ? If you pair this with a wizard at the start like Ieti suggested, players can then choose their classes based on their answers. To me it makes no difference if you unlock a tree at 1 or 10.

So unlock them and leave the trees as they are. Internally you then code to separate the shared trees. Separating the trees does not mean you have to come up with new spells. You can have the same spells but they can be shifted around independently and their effects can be modified.

So, for example, you can have a Mind squasher for barb and a mind squasher for knight and both have totally different locations in the skill trees and have totally different effects. Balancing would be so much more granular and effective because every class would have their own spells . Imagine the diversity possible with that.