Re: Luck test, v 1.00
 

Still, Ulm in there (the classical "Order/Productivity" combo) does pretty well. The initial difference in income with Pan was around 400 gold, and the final difference is of order 10000 gold over 30 turns (30x400=12000). Now, Pan got some nice boosts from random events, but still, Ulm presumably got a better start.

Now all we need is for someone bored to run this kind of test enough times that we'd have an idea of variance http://forum.shrapnelgames.com/images/smilies/happy.gif

Re: Luck test, v 1.00
 

This was a 9 province start, just like the OP.

-Frank

Re: Luck test, v 1.00
 

I am a little surprised at how well luck/growth did. Overall thou, I personally based on Frank's test would take the 44,468 gold everytime. After all thats 10,278 gold extra over just 30 turns, nearly 25% more.

I wonder how Order 3/Luck 3 would do?

Re: Luck test, v 1.00
 

Order and Luck are sort of a bad combination because order reduces event frequency. 25% increase in gold is nothing to sneeze at but it overlooks the fact that luck can do things for you that order can't, like get you items you can't otherwise make, improved chance of heroes, cross breeding, and other stuff not expressed in terms of gold. Sure, order/prod is probably better overall but not so much so that everyone will take order 3/misfortune 2 all the time like in dom2.

Re: Luck test, v 1.00
 

2 players 9 provinces , just hosting

Man/ Manticore/ Dom 10/ Luck +3/ dormant

Ulm/ Manticore/ Dom 10/ Luck -3/ awake +200pts
Ulm only province left by turn 60. Other 8 taken by Indies.

Trn 40

Man
Treasury 35,926 Inc 910 Up 52
Air 98 Water 5 Astral 30 Death 7 Nature 152
27 good events 8 bad 2 mixed 3 national heroes 2 standard heroes with troops.

Ulm
Treasury 23,213 Inc 374 Up 12
Earth 203 Nature 10
8 good events 23 Bad 4 Mixed NO HEROES
4 provinces lost to Knights 1 to Barbarians 1 HELD against Villains(2 heavy inf routed them, very strange)

I suspect hidden sites had a big effect

Re: Luck test, v 1.00
 

Keep in mind that Tien Chi has 3 different luck boosting pretenders to add on top of these figures.

Re: Luck test, v 1.00
 

Remember in the nine-province runs there's variance based on the other 8 provinces. In Trollman's run above, Arco's (order) initial advantage over Pan (luck) is too low, so Pan's initial provinces are better. Even so, it looks like only a 5% advantage so they're very close in outcome after adjustments. So far it's looking like luck is overall coming out better, although the advantage is slight and the startup and variability issues are important and difficult to catch in these tests.

Forrest's run points out a problem with these test runs - misfortunate invasions. Just "next turning" overstates their effect, because you'd actually send an army out to recover it. But, that's a big pain, and managing the reconquest armies makes it harder to make different nations comparable. You'd need similar cost armies at some standard level - probably with similar resource costs as otherwise "same gold" is a very different power level.

I think the money quote is:

Re: Luck test, v 1.00
 

Good test and it's pretty much what I've observed through playing. You're trading gold for gems by taking luck. Gold can get you a lot of things in Dominions 3, but not gems. It actually seems like a pretty balanced tradeoff.

Re: Luck test, v 1.00
 

Another thing I don't like about the original test: growth and production give gold, as well. So you should substract 3 levels of prod and 2 of growth from income (Jotunheim has one level of prod).
That would be, if I remember correctly, 5*2= 10%. You should substract 10%.
----------
I made a quick dirty test, and after 17 turns:

Luck/Growth dominion7 arcoscephale got 130 gems total
Luck/Magic Marverni 178
Luck/Productivity Agartha got 166.

Strangely enough, Luck/Growth Arcoscephale got the most magic items (4), including Horror Helmet and Phoenix Rod.

Agartha got the most +defence events, needs further testing.
I was a little puzzled when I found province defence naturally goes up in all provinces.

Re: Luck test, v 1.00
 

1) You've said that already. I've told that the scales weren't the best possible for testing.

2) When you start with multiple provinces, those provinces start with little PD. It doesn't go up after the game creation without paying for it.