Are there balancing changes planned for A24?

[coworotel]

This question is especially for the devs: how likely it is that we are going to have any balancing changes in the next A24? I mean particularly unit attributes (attack, armor, etc.). And if we are going to have such changes, is it going to be a few/small changes, or lots of/big changes?

[Stan`]

It is likely. Depends on borg using SVN and uploading his balancing patches to Phabricator mostly.

[coworotel]

@stanislas69 so borg-'s mod may become the vanilla game, or only partially?

[Guest]

Expanding on the premise, is there some kind of feature todo list? Is it in the works or are you planning on just winging it.

[Stan`]

I have my own but I'm trying to push one. Dunno if that'll stick

22 minutes ago, coworotel said:

@stanislas69 so borg-'s mod may become the vanilla game, or only partially?

Well would be nice if it was completely. But it contains more than just balancing it also adds new code.

I also wrote this code So it might be worse.

[Nescio]

It would be nice if there were some clear procedure to get simple gameplay changes implemented. 0 A.D. has greatly improved over the years and a lot of content has been added, but when it comes to gameplay, fundamentally the status quo has hardly changed.

@borg-'s and other mods do much more than just a few balance tweaks and are therefore not straightforward to implement. Overhauling everything at once is not necessarily a good idea. Doing it in small steps is preferable, because then it's easier to explain what has changed and why.

For instance, the last stat tweak of rams was 3 years ago (18735); since then it has been frequently pointed out that rams are still too powerful, yet nothing has changed so far.

[Stan`]

4 hours ago, Nescio said:

It would be nice if there were some clear procedure to get simple gameplay changes implemented. 0 A.D. has greatly improved over the years and a lot of content has been added, but when it comes to gameplay, fundamentally the status quo has hardly changed.

Indeed, it would be very nice to have such a procedure. However, I am not sure how it would work. One would also have to define simple gameplay changes. Just because it's easy to change one value in the root template doesn't mean it is a simple change. It could have big repercussion down the hill. Some might also affect Petra. 

That being said, the design document is still being worked on. It is currently private though. Once it's up to date with the current game, one will be able to submit "patches" just like for the main game for it.

 

[borg-]

5 hours ago, Nescio said:

It would be nice if there were some clear procedure to get simple gameplay changes implemented. 0 A.D. has greatly improved over the years and a lot of content has been added, but when it comes to gameplay, fundamentally the status quo has hardly changed.

@borg-'s and other mods do much more than just a few balance tweaks and are therefore not straightforward to implement. Overhauling everything at once is not necessarily a good idea. Doing it in small steps is preferable, because then it's easier to explain what has changed and why.

For instance, the last stat tweak of rams was 3 years ago (18735); since then it has been frequently pointed out that rams are still too powerful, yet nothing has changed so far.

In fact my mod has few changes other than balancing. Just a few things like: units do not celebrate promotion, numbers of icons on the dashboard, animals can get fat, that kind of simple thing. No changes of risk.

In addition, I did not get any kind of mistake in multiplayer games.

[av93]

1 hour ago, stanislas69 said:

That being said, the design document is still being worked on. It is currently private though. Once it's up to date with the current game, one will be able to submit "patches" just like for the main game for it.

 

What are the plans? Write down everything that is made in the game, and then start changes?

[wowgetoffyourcellphone]

3 hours ago, stanislas69 said:

Indeed, it would be very nice to have such a procedure. However, I am not sure how it would work. One would also have to define simple gameplay changes. Just because it's easy to change one value in the root template doesn't mean it is a simple change. It could have big repercussion down the hill. Some might also affect Petra. 

 

Why so much fear in experimenting? Make a change and if it doesn't have the desired effect then revert that change or change it some more. I swear the team didn't used to be so tepid when it came to this stuff. Who cares if some people howl at the changes? Listen to those who make constructive suggestions and ignore those who don't. Battering Rams have been OP for 3 years, Bolt Shooters have been UP for years. Siege Towers are just weird. People spam Slingers, when really melee infantry is what they should spam (ranged units should support the melee infantry and cavalry should be for raiding or targeted strikes).  Perhaps it's because there hasn't been someone designated to decide these things? 

Edited May 5, 2019 by wowgetoffyourcellphone

[Stan`]

13 minutes ago, wowgetoffyourcellphone said:

Why so much fear in experimenting? Make a change and if it doesn't have the desired effect then revert that change or change it some more. I swear the team didn't used to be so tepid when it came to this stuff. Who cares if some people howl at the changes? Listen to those who make constructive suggestions and ignore those who don't. Battering Rams have been OP for 3 years, Bolt Shooters have been UP for years. Siege Towers are just weird. People spam Slingers, when really melee infantry is what they should spam (ranged units should support the melee infantry and cavalry should be for raiding or targeted strikes).  Perhaps it's because there hasn't been someone designated to decide these things? 

Well I guess it started when we started to be very critical on how people do commits. Which in turn lead us to have a very strict review process for anything that isn't art related where everything needs to be as perfect as possible before it's committed. 

Then comes the part where balancing needs experimenting yes, but also experience, and the people who were up to do it in the team went AWOL, or just couldn't stand failing an alpha.

 

[wowgetoffyourcellphone]

16 minutes ago, stanislas69 said:

Well I guess it started when we started to be very critical on how people do commits. Which in turn lead us to have a very strict review process for anything that isn't art related where everything needs to be as perfect as possible before it's committed. 

In your POV what are the pros and cons of this approach?

 

16 minutes ago, stanislas69 said:

 Well I guess it started when we started to be very critical on how people do commits. Which in turn lead us to have a very strict review process for anything that isn't art related where everything needs to be as perfect as possible before it's committed. 

 

I don't think balancing changes are like adding new code or features. It's basically just tweaking stats or enabling things already in the game. Perhaps balance changes could be given fewer restrictions.

[Stan`]

38 minutes ago, wowgetoffyourcellphone said:

In your POV what are the pros and cons of this approach?

Mmmh, in my own POV pros:

1. gameplay mods can strive, if the balancing was perfect, there would be no need for gameplay mods.
2. It also keeps the changelog history clean.

On the cons... 

1. It doesn't showcase the full potential of the game
2. It makes external contribution painful and frustrating, even useless sometimes
3. Makes most contribution about the engine, and not about the game.
4. It makes adding new functionality like adding growing fattening or #3488 very complex, because even if the features is good someone needs to take responsibility for committing it.
5. I can't even think on how hard it would be to get bb #252 and my #2577 in the game...

 

Of course this is my own opinion, and I might be painting a darker painting, than it currently is;

[fatherbushido]

1 hour ago, wowgetoffyourcellphone said:

In your POV what are the pros and cons of this approach?

Everyone working in team know that you have to rely on:

- people

- process

Sometimes you can rely only on people. Sometimes.

[Guest]

13 hours ago, stanislas69 said:

Well I guess it started when we started to be very critical on how people do commits. Which in turn lead us to have a very strict review process for anything that isn't art related where everything needs to be as perfect as possible before it's committed. 

Well, that clearly isn’t working. (Look at the commit log). Might as well try something new.

[Stan`]

30 minutes ago, (-_-) said:

Well, that clearly isn’t working. (Look at the commit log). Might as well try something new.

That's right. So what would you propose ? 

[Nescio]

13 hours ago, stanislas69 said:

gameplay mods can strive, if the balancing was perfect, there would be no need for gameplay mods.

Frankly, I don't quite understand this reasoning. Balance isn't perfect at the moment, never was, and I doubt it'll ever be; nor does it have to. And even if balance were perfect, there would still be plenty of room for gameplay mods.

Anyway, most values in game seem to be rather arbitrary; replacing one arbitrary value with something less arbitrary could be considered an improvement. And if something doesn't work out, it can be reverted or changed again.

27 minutes ago, stanislas69 said:

That's right. So what would you propose ? 

Establish some simple and clear procedure, e.g.:

• identify a problem
• suggest a solution
• make a patch on phabricator and a tiny gameplay mod which contains only the exact change(s)
• launch a couple of test games, for different maps, map sizes, civilizations, number of AI players, etc.
• find other people to discuss, play-test, approve, or review your proposal
• get someone to implement the patch

Only the last step requires a team member, everything else can be done by others. Just let the game evolve.

[Stan`]

2 minutes ago, Nescio said:

Frankly, I don't quite understand this reasoning. Balance isn't perfect at the moment, never was, and I doubt it'll ever be; nor does it have to. And even if balance were perfect, there would still be plenty of room for gameplay mods.

Anyway, most values in game seem to be rather arbitrary; replacing one arbitrary value with something less arbitrary could be considered an improvement. And if something doesn't work out, it can be reverted or changed again.

Well I was looking for pros this is one. 

Well values are always gonna be arbitrary to match a certain vision on how the game play should be, aren't they ?

4 minutes ago, Nescio said:

• identify a problem
• suggest a solution
• make a patch on phabricator and a tiny gameplay mod which contains only the exact change(s)
• launch a couple of test games, for different maps, map sizes, civilizations, number of AI players, etc.
• find other people to discuss, play-test, approve, or review your proposal
• get someone to implement the patch

Why can't you do that right now ? 

4 minutes ago, Nescio said:

Only the last step requires a team member, everything else can be done by others. Just let the game evolve.

How many players are playing SVN ? Because whatever change you make is gonna need to be tested on the latest version, to make sure it's not broken. 

By implement do you mean commit ? 

 

[Nescio]

3 minutes ago, stanislas69 said:

Why can't you do that right now ?  

How many players are playing SVN ? Because whatever change you make is gonna need to be tested on the latest version, to make sure it's not broken. 

By implement do you mean commit ?

Yes, I mean get them committed, and that's currently where the problem is.

E.g. D1761 ( poll and discussion ) or D1762 ( poll and discussion ).

[Stan`]

14 minutes ago, Nescio said:

Yes, I mean get them committed, and that's currently where the problem is.

So yes, I guess the problem lies with the responsibility of committing them. I could commit them right away, but then I would be responsible for fixing it if it broke the game. I would also take every criticism personally (I shouldn't but that's what I do) So the idea behind the design document is to be able to say, this document defines what the game should be, not what I (personally) want, nor what 20 persons on the forum think is good.

I totally understand you point though. You make a valid request, and it makes no sense that it is not being accepted. 

[fatherbushido]

There could also be a commit race which could even result in conflicts (or having a rule like in GO)

 

[Nescio]

45 minutes ago, stanislas69 said:

So yes, I guess the problem lies with the responsibility of committing them. I could commit them right away, but then I would be responsible for fixing it if it broke the game. I would also take every criticism personally (I shouldn't but that's what I do)

If everyone is reluctant to commit anything, then nothing gets changed. It's not about any specific proposal getting accepted, it's about the general fear and lack of gameplay changes; the last one I remember was rP21285, over a year ago.

Furthermore, the primary responsibility for checking, testing, and fixing ought to be with the reviewer(s), not the committer. I know that currently the person who commits something typically reviews it as well, but in person A can propose a patch, person B review it, and person C commit it.

5 minutes ago, fatherbushido said:

There could also be a commit race which could even result in conflicts (or having a rule like in GO)

What do you mean exactly?

45 minutes ago, stanislas69 said:

So the idea behind the design document is to be able to say, this document defines what the game should be

An updated design document is useful for outlining a general vision and listing what's in game (resources, civilizations, structures, units), but it's less relevant for balance and actual stats. Besides, it ought to be a guideline, not a dogma; opinions can change over the years.

Regardless, tweaking can (and should) be done without a design document; e.g.

• rams are too strong vs units; let's give halve their base damage and give them a 2× bonus vs structures
• slingers are too effective; let's give them pierce damage instead of crush
• rome_hero_maximus.json aura is too powerful (structures and units +1 armour globally); let's change it to a 80~m range aura
• wonder food cost doesn't make sense; let's remove it and double the wonder's stone cost
• unit U is considered too weak, it currently has X health; let's increase it to Y; it turns out Y is too much; let's change it to Z
• etc.

45 minutes ago, stanislas69 said:

what 20 persons on the forum think is good. 

Again, if something turns out not to be a good idea, it can be changed and reverted. You never know if it's not tried out. There is no harm in experimenting.

[Stan`]

27 minutes ago, Nescio said:

Again, if something turns out not to be a good idea, it can be changed and reverted. You never know if it's not tried out. There is no harm in experimenting.

Of course, but isn't that the purpose of the mod ? Also given the time span between release, it would mean that balancing could potentially be broken for six months without anyone noticing.

28 minutes ago, Nescio said:

If everyone is reluctant to commit anything, then nothing gets changed. It's not about any specific proposal getting accepted, it's about the general fear and lack of gameplay changes; the last one I remember was rP21285, over a year ago.

Furthermore, the primary responsibility for checking, testing, and fixing ought to be with the reviewer(s), not the committer. I know that currently the person who commits something typically reviews it as well, but in person A can propose a patch, person B review it, and person C commit it

At the time temple was supposed to be in charge of balancing, so it checks out  I haven't seen him since the GDPR Hiatus.

Well the lack of reviews is directly responsible for the lack of commits. 

 

[Guest]

2 hours ago, stanislas69 said:

That's right. So what would you propose ? 

1. The author knowing the code being changed and sufficiently sure the change is correct.

2. The reviewer actually reviewing things rather than just pointlessly adding a checkmark.

3. The author and the reviewer both knowing the code changed. This implies the dev team having two people knowledgeable in each component. 2 people knowing their way around the AI, rmgen etc. This is not far fetched.

4. Interest from both parties

The result would be faster regression-free development.

(I have convinced myself this would never happen since a long time ago. And as such, contributing to such a project was nothing more than waste of time for me.)

Edited May 6, 2019 by Guest

[Stan`]

@(-_-) While it seems really logical, it requires more people than we currently have, and more activity than all team members can possibly give. Where are we supposed to find such people ?

6 minutes ago, (-_-) said:

The reviewer actually reviewing things rather than just pointlessly adding a checkmark.

By checkmark you mean adding inlines, like what I usually do because I don't know the code enough ?