Simultaneous space race end - how is the winner determined?

[Silverbow]

I've just played a game (HoF VI 3) where I lost space race to Joao... even though we had same arrival turn. Aside from infuriating me beyond all belief, it made me wonder, how the hell does the game determine who wins the spacerace in such case?

Oh, and a sidenote, why the hell is spaceship flight duration 6/7 on quick, when it's 10/12 on normal? Shouldn't it be 5/6?

 

[Verarde]

I've just played a game (HoF VI 3) where I lost space race
Oh, and a sidenote, why the hell is spaceship flight duration 6/7 on quick, when it's 10/12 on normal? Shouldn't it be 5/6?

I can't answer that first question, but anything in Quick mode will happen faster than Normal. Anything in Marathon will take longer than Normal. Takes less beakers to research Meditation in Quick than in Normal for example.

 

[Undecided]

I encountered the same problem today and didn't find any useful information on my admittedly quick search. I would guess its random and you got really unlucky.

In general Quick Speed is two-thirds of Normal Speed plus rounding down/up. (333 turns to 500 turns for example)

 

[obsolete]

It's not random. It goes by fractional turns. Actually, a better way to look at this is TURN-ORDER. These are set at the begining and don't change, unless something interesting happens like a civ gets eliminated, haha.

 

[Undecided]

Interesting, i thought that the human player always was Player 1 as it is the case with wonder building. Thanks for the clarification.

 

[coanda]

It is random.

http://forums.civfanatics.com/showpost.php?p=7895324&postcount=6

 

[obsolete]

Wait wait wait..

I remember the debates going back and forth on the world-wonder thing for years. I believe 2 years ago or so some coders stated on another forum that there was still RANDOMIZATION applied to world wonders on the tie issues.

So I am assuming the last patch tampered around with ties AGAIN?

More than once I've seen code posted here as BTS, when in reality it came from a different flavour of Civ4 and and was quite different.

 

[vranasm]

this actually is easily checked if the code posted by Dan is still the same ... go into your BtS folder, subfolder named Cv...dll, there is source code with which is the dll linked.

then look for CvGame.cpp and then search for the function testVictory()

if you use BULL you have to check BULL's source code of course.

 

[Silverbow]

The same code is present in my CvGame.cpp

So I lost on the last turn due to randomness? 'Awesome' waste of time.

 

[obsolete]

Let's make sure once and for all what the case is now in the last patching.

Turn your SEED to re-randomize, and then reload your last turn a few times to see if it ever changes.

 

[Silverbow]

Let's make sure once and for all what the case is now in the last patching.

Turn your SEED to re-randomize, and then reload your last turn a few times to see if it ever changes.

Do you mean to enable the random seed at reload? Dunno how to do that, with Lock Modified Assets on and all. But if you know how to, I'll gladly try it out.

 

[TheMeInTeam]

Let's make sure once and for all what the case is now in the last patching.

Turn your SEED to re-randomize, and then reload your last turn a few times to see if it ever changes.

Usually, in these space race "ties" the AI just launched 1 turn earlier and the human wasn't informed until the following turn (this is a horrible design failure on the part of firaxis, but that isn't new)

However, people have looked into the code obs. Wonders and standard things use turn order. Victory condition, however, does not. When you complete multiple victory conditions, the one displayed is random. When you and another faction win a victory on the same turn, the winner is also random (unlike most other "ties"). That is true even if you launched a space ship while they win an AP vote; doesn't matter. You get 50/50. Seed tests should confirm it though.

Do you mean to enable the random seed at reload? Dunno how to do that, with Lock Modified Assets on and all. But if you know how to, I'll gladly try it out.

Reload and switch different numbers of citizens onto or off of working hill mines. As each turn a mine has a (very small) chance to discover a resource, mines call on the RNG. You can swap them around to get different RNG values in a chain to be applied.

And yes, that DOES mean the "new random seed on reload" is largely irrelevant, because people can just manipulate the outcomes anyway by doing things that call the RNG up.

 

[r_rolo1]

^^That or put a unit in auto-explore.

To be honest, I find that the no random seed on reload is actually better to manipulate things us, since you can grasp the general RNG output easily in some reloads if you have a lot of units with varied combat odds ... and once you can grasp the next 20-50 "RNG" outputs you can do pretty wacky stuff like choosing the GP you want, manipulate the AI queues or goody hut results or win all the battles. You can't simply do that kind of thing in random seed on reload, atleast on usable time....

 

[Silverbow]

^ I did that mine trick. The game actually had autosave when I was 1 turn from victory. I swapped a citizen to work a mine, and sure enough, I won the spacerace.

Which means I lost to random nonsense. It's really a pain, since I'm looking to complete the HoF Challenge and replaying this crap mapscript is kinda unappealing. OTOH, it would've been a horrible, dead last time anyway (1996). I played like crap, and I chose a very poor civ fit (Celts, obviously hardly looking for good score. Hell, I'm gonna do just that again on retry. Screw score optimization!).

 

[obsolete]

Usually, in these space race "ties" the AI just launched 1 turn earlier and the human wasn't informed until the following turn (this is a horrible design failure on the part of firaxis, but that isn't new)

I think it was so bad you could build a wonder 2 turns after an AI had technically completed it. But I still remember the days where the text notices for stuff never even showed up on the screen until 4 or more turns later, because it was based on some low priority threading.

 

[Jouzou]

Usually, in these space race "ties" the AI just launched 1 turn earlier and the human wasn't informed until the following turn (this is a horrible design failure on the part of firaxis, but that isn't new)

Since the human player moves first, doesn't it mean that every bit of world news actually happened in the previous turn? Or do you mean that a rival launches the SS on, say, turn 200, but the player is informed as late as turn 202?

 

[Quaren]

Since the human player moves first, doesn't it mean that every bit of world news actually happened in the previous turn? Or do you mean that a rival launches the SS on, say, turn 200, but the player is informed as late as turn 202?

Well of course all the news is what the AI did after your last turn so it is the previous turn since your first. The problem with the programing as I have noticed it is that in the case of wonders and stuff like that the AI can finish a wonder and after that you can still choose to build it.
Example:
Turn 200
Your turn: do whatever
AI turn: finish wonder and attack you
Turn 201:
Your turn: get message of attack but can still start constructing/keep building said wonder
AI turn: does whatever
Turn 202:
Your turn: get failgold and message that wonder is built.

I remember plenty of games where I thought I had missed a wonder by 1 turn and reload to finish it 1 turn earlier then I get fail gold but still fail.

 

[TheMeInTeam]

Since the human player moves first, doesn't it mean that every bit of world news actually happened in the previous turn? Or do you mean that a rival launches the SS on, say, turn 200, but the player is informed as late as turn 202?

The issue is that the game doesn't tell you what happened on the previous turn in a timely fashion.

 

[obsolete]

I sent a PVM to Dan to clarify a few issues.

I remember a while ago it became agreed on (finally) that wonders were not based on a random method. But then a couple years ago code snippets were surfaced to show that wonder ties were actually broken by random rolls. But now we are back to no randomization again for wonder ties.

Hmm... I wonder if it's the difference in the SDK or elsewhere.

 

[GenerallyGreat]

Any time I'm looking to rush a wonder I just check if it's been finished in the built wonders screen. If it hasn't I always get it if I whip/chop it that turn. I'd never just trust the fact that I can still build the thing.