I am going to make a large modded diplo game. I vaguely know what mods I want, but I figured it wouldn't hurt to ask others. Also need to pick a map.

The mods I want rn are:

Buy Cards
AI's don't play cards
INSS Bonuses
Late Gifts
Essentials
Bomb Card+
Gift Gold Ultra 3
Factions

Where Am I

Edited 5/18/2025 21:59:10

include the "Where Am I?" mod, it can be useful in some cases.

what do you want in your game?

Edited 5/18/2025 20:18:46

The biggest thing would be that everyone starts at peace, to get rid of the PE struggle at the start.
commerce game
probably ffa since custom senareo would take a long time XD
semi-steep increase on unit cost
cards: spy, gift, airlift, bomb, reconnaissance, surveillance, sanction.
Gift cards would be cheap for trading purposes.
will probably decide all the other costs after a map is picked,
not sure if I do or don't want cities yet.
Surrender/boots turn to AI, so things aren't completely free.

the mod "safe start" would be amazing then for you, for X turns players cannot attack eachother

i just realized that their's theirs a filter for mods that work with mega games and no diplo mods show up with it..

The mods that are made available for mega games are selected based on their performance.

You don't want a game to become too slow with all the mods, and diplo mods unfortunately tend to impact the game quite a lot since they need to check every order.

With 400 players you have a lot of orders, so these mods would just run so slow it will crash the game.

factions plays diplos on all players instead, so it would be closer to working right?

yes, but it is still quite slow. if every player is at peace, the mod still needs to play 400 x 399 diplo cards xD

and then we're not talking about storage space. the diplo mods store a lot of data, that scales with the number of players in the game.

thinking about all this, it would be nice to have at least some diplo mod available in mega games. might look into that

Edited 5/19/2025 13:06:45

(i started asking this in the mod maker discord also.)

Is it possible to do something like check all bordering players at start or end of turn then put a max amount of people you can be at peace with to cut maximum amount of checks down? If mods can ckeck statistics, would be able to tell bordering players by if they have a troop count above zero as long as 1 army must stand and light fog is on.

would pay for a working diplo mod on mega games. XD

that honestly could work.
mods cannot access the statistics, those are calculated once you open the statistics menu unfortunately.

Here's a better explanation of the problem and why we don't have a clear answer rn:
Currently the two ways to create diplomacy mods are 1) check every attack whether both players can attack eachother and 2) by playing a diplo card on both players.

both ways have their (dis)advantages. But no one ever has tested whether what works better performance whise. This is not easy, since the performance of (1) mainly depends on the amount of orders that are submitted in total, which does not necessarily scale with the number of players. (2) depends on the number of players alive.

Both approaches are not optimal. They both can scale really quick depending on the game. I can think of one approach that probably scales way less whatever game it is.
Quick layout of the approach:
1. Loop over all territories of the map
For each territory, do the following
2. Check every connected territory
3. If the territories are controlled by different players and they cannot attack eachother, store this combination of players
4. Now we have all combinations of players where we need to play a diplo card.

This approach also needs some testing and speed ups here and there, but in theory this approach does not scale hard when there are lots of players and/or lots of orders

Maybe you could be clever and offload work to clients too? Each client figures out who they can fight and sends this to the server.

Then on the server, the mod could take make sure both players agree or independently verify the state. Then you play diplo cards.

Ofc, higher cost for clients, but it's timeouts during server advance that is the bottleneck here I think. And, as mentioned in Disocrd - UI scales poorly if you have a list with tons of usernames.

Does mean you will live with the tradeoffs ( ie. two non-bordering players both players attacking the same territory will fight) , but we can call stuff like that a feature. By giving clients some work, one do need to keep in mind clients trying to cheat ( by sending wrong info) but I do think it's possible to engineere around that.

Edit: Actually, every client could figure out all bordering players they can see.

[ Already said this in Discord, but since convo is happing in 2 places I will put it here too ]

Your best bet is probably asking Fizzer if he will consider raising the maximum execution time for mega games.

He wasn't against that last time we asked, but that was the first time a modded mega game was made at all, so was still very much at a testing phase.