Does this game use some sort of "seed" value?

[unmerged(715229)]

What I mean by that question is:

Say if you start a game and you save at 1.1.1936 and now use that save as a start for your games, the result from it is, that all those games are similar and have some tendencies that appear in all those games from this save.

As opposed to using a real fresh start for every game, then that (the above mentioned) is not the case and these games will each play out _completly_ different.

True?
Not true?

 

[grumphie]

as far as i know, no. same goes for all other PI games. unless im severly mistaken in both memeory and understanding of mechanics.

 

[themousemaster]

The exact same start can go dramatically off the rails at any point.

All of my NATCHI playthroughs are off the same savegame I made after reorganizing the OOB once, and they all have a vary high variance.

 

[unmerged(715229)]

I know that the game changes even if you use a save game start but I have also noticed some certain events that appeared only in one save game and the resulting playthroughs from it (e.g. --> german AI doing sealion, me playing USSR) and not in all others I started. But other influences might have led to that.
My experience is limited. So far I might have played 40 to 50 games (still a freshman I guess).
Thanx for the answers.

 

[themousemaster]

I know that the game changes even if you use a save game start but I have also noticed some certain events that appeared only in one save game and the resulting playthroughs from it (e.g. --> german AI doing sealion, me playing USSR) and not in all others I started. But other influences might have led to that.
My experience is limited. So far I might have played 40 to 50 games (still a freshman I guess).
Thanx for the answers.

If you would like to confirm the non-existence for yourself, all Save files are just plain text. You can have a look in one (though they are like multi-MB huge), and you will see that there really isn't a "seed" present.

 

[unmerged(715229)]

I did dive a bit into save file editing. Its hard to really analyse one complete. But youre right I couldnt find something. ;-)

 

[dz_duck]

The seed has to come from SOMEWHERE. If it's not in the save file, then it could come from the system clock (different game every time), timing of clicks on the main menu (different game every time), or maybe a checksum of the save file (same game every time). Of course, once you as the player do something different, even by one hour, then the game will diverge.

 

[grumphie]

The seed has to come from SOMEWHERE. If it's not in the save file, then it could come from the system clock (different game every time), timing of clicks on the main menu (different game every time), or maybe a checksum of the save file (same game every time). Of course, once you as the player do something different, even by one hour, then the game will diverge.

doesnt work with a seed. random seeds work well in situations that are always the same. say, generating a map.

HOI3 is far too complex to be decided by a seed. the AI would be absolutly murdered by the player whenever they did something even remotely unexpected, as seeds are really inflexible.

basicly, the game is not scripted. sea lion happening is a toss of the coin depending on what the AI and player did. if the player is the UK and sends all his fleets away, sea lion becomes a whole lot more liekly. but also things that are much less obvious - let's say, the player is mexico, and goes beat up the USA. japan can then be more aggressive, the UK might send mroe leets and troops to asia to try and contain the japanese there, and sea lion becomes more likely.

 

[dz_duck]

doesnt work with a seed. random seeds work well in situations that are always the same. say, generating a map.

HOI3 is far too complex to be decided by a seed. the AI would be absolutly murdered by the player whenever they did something even remotely unexpected, as seeds are really inflexible.

basicly, the game is not scripted. sea lion happening is a toss of the coin depending on what the AI and player did. if the player is the UK and sends all his fleets away, sea lion becomes a whole lot more liekly. but also things that are much less obvious - let's say, the player is mexico, and goes beat up the USA. japan can then be more aggressive, the UK might send mroe leets and troops to asia to try and contain the japanese there, and sea lion becomes more likely.

I think we are saying the same thing in different ways. Of course the non-random AI has a lot to do with things. But, its choices are driven by random events, even something as simple as, did that brigade lose 452 or 453 men in that battle? There are many, many random number draws in HoI3, so the random number generator has to get a seed from somewhere.

 

[themousemaster]

I think their may be a disconnect here with what people are defining as a "seed".

Yes, events, %-rolls, and the like are randomized. No, there is no fixed-seed-value-at-gamestart-that-carries-through-to-end.

 

[unmerged(715229)]

The reason why I brought this up is:

Sometimes I see a game going in a direction from the very start.
Example:
Like in a certain playthrough i have a really hard time trading ressources.
And if I start over from a save that will again come up maybe slightly less or more but nevertheless.

But if I choose to restart from fresh that effect is gone.

I am talking about 1. Jan 1936 only so theres not much time in which I could have done anything different to explain that behaviour. :huh:

This is just to make clear why I uttered that question in the first place. :mellow:

Edit: I do not count this as anything bad in gameterms, I am just generally interested in game mechanics.

 

[Lynx190]

Both here and previously in EU3 it's always felt like the AI uses diplomacy (including trades) earlier than usual after loading a game rather than starting a fresh game. But I don't believe it produces identical results; reloading definitely can trigger the AI to change its behavior (countries start shuffling units around on the border right after a load) but not necessarily in the same way.