Rebalancing Civ4: RtR Mod

[plako]

Highet era Medieval wouldn't be capped. Max there would be 80% (Assuming xml modifier 40 and multiplier 2 and all alive players having the tech). If we assume 24 live players. 1 player known to have that tech would be worth 3.33%.

Looking at this and keeping it consistent capping Medieval to 50% would be in order.

[Fintourist]

I played around with values and for PB18 purpouses Krill's solution works fine although I would drop TECH_COST_TOTAL_KNOWN_TEAM_MODIFIER a bit lower to 15. Actually with constant "iPossibleKnownCount" there is no huge need to stagnate the development of era multiplier because decreasing player amount takes care of it so that boost should not become ridiculous.

For general mod I think we should drop the constant, because we actually want that "% of people knowing the tech" drives the bonus. I got pretty good results for different game sizes with using the previous model but decreasing the growth speed of era multiplier. What I mean:

TECH_COST_TOTAL_KNOWN_TEAM_MODIFIER = 35-40
0 = ERA_ANCIENT
1 = ERA_CLASSICAL
2 = ERA_MEDIEVAL
2.5 = ERA_RENAISSANCE
3 = ERA_INDUSTRIAL
3.5 = ERA_MODERN
4 = ERA_FUTURE

With these setting known-tech bonus plays maybe a bit too big role in the early game of small PBEMs, but overall I think the numbers are in the right ballpark.

Here is the spreadsheet that I sent to Krill if plako or anybody wants to try different variables:
known-tech bonus spreadsheet

[Krill]

Can we just back up a moment? I thought one of the problems that people were complaining about in PBEMs is that there are horrendously large per player bonuses? Is this not the case?

[NobleHelium]

Checked the code and indeed dead people won't be part of the calculation of the known civ bonus. I don't know why I mistakenly thought otherwise.

There's a long-cited CFC post saying dead civs still count and everyone has been citing it since. Someone did look up the code and posted about it in the ISDG, I think it was Seven but I don't remember exactly.

[Fintourist]

Can we just back up a moment? I thought one of the problems that people were complaining about in PBEMs is that there are horrendously large per player bonuses? Is this not the case?

No idea! But it felt a bit weird that if all your 4 opponents in PBEM know the tech you get 32 % bonus and if all your 20 remaining opponents in Pitboss 18 know the tech you get 160 % boost. But I see what you are saying and I'm fine with the logic that the bigger the game the bigger the tech boost for the most backward player (as in big games land differences get bigger too). So if you are an era behind everybody else it's more difficult to catch up in 5-player PBEM than in 15-player Pitboss.

Generally I'm just more afraid of too big than too small bonuses and using the constant as a divider solves this. I'm still voting for dropping the TECH_COST_TOTAL_KNOWN_TEAM_MODIFIER for PB18 a bit. Otherwise I think we'll still see quite a lot of 50-130 % boosts when someone reaches Industrial era.

[Krill]

Can we just back up a moment? I thought one of the problems that people were complaining about in PBEMs is that there are horrendously large per player bonuses? Is this not the case?

No idea! But it felt a bit weird that if all your 4 opponents in PBEM know the tech you get 32 % bonus and if all your 20 remaining opponents in Pitboss 18 know the tech you get 160 % boost. But I see what you are saying and I'm fine with the logic that the bigger the game the bigger the tech boost for the most backward player (as in big games land differences get bigger too). So if you are an era behind everybody else it's more difficult to catch up in 5-player PBEM than in 15-player Pitboss.

TBH I think a lot of the known tech bonus stuff feels weird, but I'm more concerned as well that there are large numbers involved that break stuff than anything else. The largest concern is those small PBEMs that certain values will break the game, and I think that the constant would help limit that problem; that said, play-styles and meta-games in PBEMs are a bit different simply because I don't think that the games last that long, the era boni don't come into effect and so perhaps I'm just been antsy. What sort of values are too large for second and third place players to have in a 5 player PBEM? For example, in the current set up in PB13, in the PBEMs that use the same mod version, the second place player may end up receiving a bonus of 15% from a single person owning a tech (once one of the techs like Corp have been researched). In PB13 it's, what, 7%? Just comparing this to your previous point, the 15% number was implicitly too high. Not saying I agree or disagree with you, just that I'm trying to understand where people think that number should be and why.

My main concern with the constant modifier is it basically returns the tech tree back to base BtS, because those numbers are so small, returning the game back to a straight tech race. But TBH in small games I don't think it's changed away from that much at all, and there are other effects such as the increased tech costs that try to limit that.

Generally I'm just more afraid of too big than too small bonuses and using the constant as a divider solves this. I'm still voting for dropping the TECH_COST_TOTAL_KNOWN_TEAM_MODIFIER for PB18 a bit. Otherwise I think we'll still see quite a lot of 50-130 % boosts when someone reaches Industrial era.

I think this would actually happen because of the era bonus being based on the lead player opposed to the backwards researcher. I'm not that concerned that some people get a large bonus though, yes it'll almost certainly happen, but I don't think the effect will be catastrophic by any means. Effectively, some people might get to gunpowder and Cuirs a bit faster. That's just about hte only real effect for the people that get the huge numbers, for those players in a position similar to yours in PB13 I don't think it's yhat bad:

It's not too much of a spoiler to say that me and OH are far from leading the PB13, but not playing OCC yet either. There are currently 12 civs left in the game and with the new system we would be using the industrial era modifier. With this era system we would pretty much always be getting 50-67 % known-tech bonus boost (3-4 civs know the tech) and even 100 % to some relevant techs (when 6 players know the tech) (and up to 183 % if we backfill something that everyone has). While not perhaps totally game-breaking, I think that this would simply be a bit too much free beakers and not fair to tech leaders.

Numbers would change to 24-32%, from "50-67% known-tech bonus boost (3-4 civs know the tech)", 48% "even 100 % to some relevant techs (when 6 players know the tech)". So basically it's halved, even with the known tech number changed to 20, dropping it to 15 is making it close to irrelevant IMO.

[Fintourist]

Numbers would change to 24-32%, from "50-67% known-tech bonus boost (3-4 civs know the tech)", 48% "even 100 % to some relevant techs (when 6 players know the tech)". So basically it's halved, even with the known tech number changed to 20, dropping it to 15 is making it close to irrelevant IMO.

Yes, for PB13 known tech number 20 would work, but now we are talking about ~double the amount of players, which makes me believe that there will be ~double the amount of players left in the game at t190 (as if we ever get that far ). Because we are using the constant divider this will also lead into doubling of known-tech bonus meaning that we are expecting same kind of figures as I described in the quoted post (when using "20" as a value).

Basically I'm saying that in this exceptionally large game it would be justified to drop the known tech number further.

[Krill]

Numbers would change to 24-32%, from "50-67% known-tech bonus boost (3-4 civs know the tech)", 48% "even 100 % to some relevant techs (when 6 players know the tech)". So basically it's halved, even with the known tech number changed to 20, dropping it to 15 is making it close to irrelevant IMO.

Yes, for PB13 known tech number 20 would work, but now we are talking about ~double the amount of players, which makes me believe that there will be ~double the amount of players left in the game at t190 (as if we ever get that far ). Because we are using the constant divider this will also lead into doubling of known-tech bonus meaning that we are expecting same kind of figures as I described in the quoted post (when using "20" as a value).

Basically I'm saying that in this exceptionally large game it would be justified to drop the known tech number further.

The number still has to be high enough for the other players though, not just middle of the pack ones. And frankly, civs in your position getting that sort of bonus isn't something as a player I'd care all that much about: the difference between 15 and 20 is 2% of your bpt, we're talking about 20 beakers per turn per player that has a tech if you make 1kbpt, when researching techs that cost from 2K beakers (Education) up to, what is Rifling, 5.5K beakers? I don't think it's a big enough difference to go crazy over.

[Fintourist]

the difference between 15 and 20 is 2% of your bpt, we're talking about 20 beakers per turn per player that has a tech if you make 1kbpt

Take Renaissance era and 10 people knowing the tech, difference is 15 % in the multiplier (45 % -> 60 %), which is 150 beakers.. If only 5 people know the tech the difference is 75 beakers.

I don't think it's a big enough difference to go crazy over.

Yeah, in the end anything goes.

[plako]

Talked with Krill and we decided to make a code that would work decently in PB18 and forget all the other games for now.

Basics:
•iPossibleKnownCount = 10 (constant)
•TECH_COST_TOTAL_KNOWN_TEAM_MODIFIER = 20 (from xml)
•era costs of:
0 = ERA_ANCIENT
1 = ERA_CLASSICAL
2 = ERA_MEDIEVAL
3 = ERA_RENAISSANCE
4 = ERA_INDUSTRIAL
4 = ERA_MODERN
4 = ERA_FUTURE
•caps:
- up to renessaince 50%
- From Industrialism 75%

These mean that ~10 people needs to know the tech to get full bonus.

implementation:

Code:

    iPossibleKnownCount = 10;

    if (iPossibleKnownCount > 0)
    {
        int currentHighestEra = GC.getGameINLINE().getCurrentHighestEra();
        iModifier += std::min(4, currentHighestEra) * ((GC.getDefineINT("TECH_COST_TOTAL_KNOWN_TEAM_MODIFIER") * iKnownCount) / iPossibleKnownCount);
        if (currentHighestEra<4)
            iModifier = std::min(150, iModifier);
        else
            iModifier = std::min(175, iModifier);
    }