Could it be...

[Winged Monkey]

... that the RNG is actually broken after all!?

I had this weird experience today playing as Julius. I decided to praet rush HC who had a ton of axes in Cuzco. Should be no match for CR-II praets, right? Wrong. 4 praets with 64,5 combat odds + 1 GG with CR-III (yeah, yeah, not the best use of a GG) and 68,1 odds, all died horribly at the walls of Cuzco. Great. The odds of that happening is lowish (0,5%), but as a poker player, I know these things can happen. But I thought it odd enough to reload the game, and wait for the next turn to see if similar results occured. Next turn, exactly the same thing. Turn after, you guessed it, slaughter. This is where it starts to become exceedingly unlikely (0,0048^3 = Uh, LOW!)

What gives?!

 

[Zanttu]

I'm curious to see if someone can explain that...

 

[r_rolo1]

"save random seed on reload" maybe?

 

[6K Man]

If you had the new random seed on reload option selected, yes, it would be very odd/broken. If not, you'll get the same results again if you reload, of course, and I'd chalk it up to bad luck.

I seem to lose a lot more 65%-80% combats in BtS, but I'm calling that happenstance until I've played more.

 

[Refar]

Could you provide the save ?

"Save Random Seed" should not do nothing, as after reloading he waited one or more turns - in those turns some RNumbers should be used up, giving new combat results.

 

[Winged Monkey]

If you had the new random seed on reload option selected, yes, it would be very odd/broken. If not, you'll get the same results again if you reload, of course, and I'd chalk it up to bad luck.

Actually, I'm not sure if I have that setting enabled. That would probably explain it. Though, I waited a turn each time, shouldn't that reset the RNG anyway?

 

[Refar]

Not reset. But if you wait one turn some numbers get used up... so the results should be different if

Loading->Attacking
Loading->Waiting One Turn->Attacking
Loading->Waiting Two Turns->Attacking
etc...

 

[Winged Monkey]

Could you provide the save ?

I'll try to post it a little later (not sitting at the right computer now). Of course, it will probably lead to entirely different results than the ones I posted.

 

[Refar]

I'll try to post it a little later (not sitting at the right computer now). Of course, it will probably lead to entirely different results than the ones I posted.

Possible, but that would mean, that the RNG is Ok, and you just had terrible bad luck.

If however the results can consistantly be reproduced... We would be able to proof that "The truth is out there!"

 

[6K Man]

Actually, I'm not sure if I have that setting enabled. That would probably explain it. Though, I waited a turn each time, shouldn't that reset the RNG anyway?

I don't know how the RNG works. Maybe a new set is generated each turn - in which case your results should have changed. I had thought that if you don't take any actions that use a random number, then the number and subsequent numbers would carry forward.

 

[r_rolo1]

^^Not if you're playing with the random seed preserve. The Civ IV RNG is in fact quasi-random: it picks a number ( called seed ), uses a math process and create a output. That output is used for the game proposes and is the new seed for the next "random" number you need for the game. If you use the random seed preserve option ticked, the game when saved will save the random seed as well, making that every time you load the game you'll have the same random seed, ergo the same random numbers generated. If you do the exact same actions that you did previously you get the same results ( but if you do something diferent ( like putting a scout in auto explore ), results may differ )

 

[Whitedragon]

Things like this happen. How large of a map and how many civs were there?
What speed were you playing?

A smaller map (or running marathon or epic game speeds) can can cause a slower run through what the RNG generates. I'd chalk it up to bad luck.

Threads like this are the main reason I changed my sig.