1. We have added a Gift Upgrades feature that allows you to gift an account upgrade to another member, just in time for the holiday season. You can see the gift option when going to the Account Upgrades screen, or on any user profile screen.
    Dismiss Notice

'Cause I'm the Taxman

Discussion in 'Civ4 - Succession Games' started by Rusten, Oct 30, 2009.

  1. Silu

    Silu Chieftain

    Joined:
    Jul 13, 2009
    Messages:
    2,964
    Location:
    Finland
    Oh sorry, I didn't notice that Athens wasn't 100% Prophet also :blush: ... Then Vienne first is better without doubt. 2nd shrine isn't that urgent, only 12% worldwide autospread and if we just want missionary cash then Buddhist missionaries easily keep us busy until then.

    5 turns of Caste feels better to me than forfeiting the Clam until regular pop for the Camel culture issue.
     
  2. babybluepants

    babybluepants Chieftain GOTM Staff

    Joined:
    Feb 5, 2009
    Messages:
    4,825
    Location:
    Vancouver
    Missed that completely. :) Forget what I said.
     
  3. RRRaskolnikov

    RRRaskolnikov Goldfish

    Joined:
    Oct 10, 2008
    Messages:
    4,231
    Location:
    Paname
    Sorry for lack of input folks... I am moving of place this week end so I lack time for civ :cringe:.
    Will try to look at the save before Rusten plays...

    @Rusten: please build me an army, please :please: :hammer: :lol:
     
  4. Emperor Peter

    Emperor Peter Chieftain

    Joined:
    Aug 11, 2008
    Messages:
    159
    Location:
    Belgium (GMT+1)
    It's Athens not Sparta that's building library, and it wasn't intended as a placeholder, i wanted culture there but forgot no science variant

    This seems like a good spot for the settler near gergovia:
    1 fish + 3 lake


    we don't have sheep, so we could go there with the vienne settler, or explore west of Isca for a better spot
     
  5. RRRaskolnikov

    RRRaskolnikov Goldfish

    Joined:
    Oct 10, 2008
    Messages:
    4,231
    Location:
    Paname
    @EP: I must agree ;)
    Maybe we can put it a bit more east so we can put another one to work those grass (where the chariot stands on your pic)? Not sure though :dunno:
     
  6. Emperor Peter

    Emperor Peter Chieftain

    Joined:
    Aug 11, 2008
    Messages:
    159
    Location:
    Belgium (GMT+1)
    i was thinking about a city in that area as well, so we could settle like this:


    downside is this uses both settlers, and there's 2 health resources we want to claim. Sheep on the western island, and the eastern one tile island for crabs. We're short on health (this is why i went moai before forge in bibracte).

    We could postpone the war and build 2 more settlers, we have enough ships to move them. I'd settle the fish(white) first, sheep second, no resource (orange) third and crabs last
     
  7. Silu

    Silu Chieftain

    Joined:
    Jul 13, 2009
    Messages:
    2,964
    Location:
    Finland
    The resourceless lake city is by no means urgent... We can postpone that for after the war if we can't afford to build extra Settlers now. If we also postpone the Fish city we don't have to worry about defense in that area that much if Argos can't invade cities by land.

    IMO grab the health resos before the big island cities, they benefit the whole empire as health is a limiting factor with HR and tons of resources kicking in in addition to the previous paragraph :) . At least the Sheep one definitely, Argos is going to contend the Crab soon. Still, IMO take that with the other Settler.
     
  8. Emperor Peter

    Emperor Peter Chieftain

    Joined:
    Aug 11, 2008
    Messages:
    159
    Location:
    Belgium (GMT+1)
    Pericles could easily settle south of Argos, so I'd consider that one more urgent than crab city
     
  9. Silu

    Silu Chieftain

    Joined:
    Jul 13, 2009
    Messages:
    2,964
    Location:
    Finland
    Considering we're gearing up for war and if he's going to settle he will steal a spot of ours in any case (the chariot city if we take the fish site), that might simply save us a settler ;) +1 health will help us a lot more in the short term than the fish site (both give the most important aspect, trade routes) and in the long term the whole archipelago will be ours! :hammer:
     
  10. Rusten

    Rusten Chieftain

    Joined:
    Aug 14, 2007
    Messages:
    3,198
    Location:
    Oslo
    A forge will bring in some extra hammers from chopping the forests too, so I think it's stronger.
    IMO, yes. I will get some courthouses, but I will start a troop build-up too. Maybe we can DoW either at the end of my set or start of the next (and wait for his stack).
    Will try. ;)
     
  11. RRRaskolnikov

    RRRaskolnikov Goldfish

    Joined:
    Oct 10, 2008
    Messages:
    4,231
    Location:
    Paname
    Nice chariot spam ;) iirc, I 've never seen so many in 800 ad :lol: (hint: unit maintenance is something)
    Not much to comment, as everything is covered already... I agree that the resourceless site can wait a bit. Agree to trade options too (MC + gold against feudalism with Sury then broker machinery with SB)
    We need to spam cats and take on Stalin as opportunities come...

    Ras

    edit: no beg in my sets Rusten :)
     
  12. Rusten

    Rusten Chieftain

    Joined:
    Aug 14, 2007
    Messages:
    3,198
    Location:
    Oslo
    Did anyone beg/request anything from Sury? If not then I will request a gift after buying feudalism -- it should net us a lot of gold.
     
  13. Silu

    Silu Chieftain

    Joined:
    Jul 13, 2009
    Messages:
    2,964
    Location:
    Finland
    I didn't beg anything at least. In any case you should beg. AFAIK begging "cooldown" is 0.95^TURNS_SINCE_LAST_BEG. Meaning that begging every turn will provide almost equal benefits on average as begging every 20 or 30 turns:

    Once every 20 turns:

    0.95^20 chance of cooldown-induced denial

    Every turn for 20 turns:
    5% chance per turn, so probability of not catching any of 20 of those is (1-0.05)^20 = 0.95^20

    So begging every turn actually yields better results, as there is a 1 - (0.95^20) chance of getting one or more begs through, as opposed to the maximum of one (with the same probability) in the other scenario.

    Of course there aren't proper things to beg every turn but if there is it shouldn't matter if the last begging was 10 or 20 turns ago.

    I'm sure bestsss will let me know if I'm wrong ;)

    EDIT: By begging cooldown I mean chance the beg is refused due to a recent beg. This has no effect on other reasons to refuse begs.
     
  14. Emperor Peter

    Emperor Peter Chieftain

    Joined:
    Aug 11, 2008
    Messages:
    159
    Location:
    Belgium (GMT+1)
    i didn't beg anything either
     
  15. babybluepants

    babybluepants Chieftain GOTM Staff

    Joined:
    Feb 5, 2009
    Messages:
    4,825
    Location:
    Vancouver
  16. bestsss

    bestsss Chieftain

    Joined:
    Jul 12, 2009
    Messages:
    1,955
    @Silu, mm...
    This thing can be >1, here is where it gets increased.

    Code:
    [B]case DIPLOEVENT_ASK_HELP:
    	AI_changeMemoryCount(ePlayer, MEMORY_MADE_DEMAND_RECENT, 1);
    	break;
    [/B]
    case DIPLOEVENT_MADE_DEMAND:
    	if (AI_getMemoryCount(ePlayer, MEMORY_MADE_DEMAND) < 10)
    	{
    		AI_changeMemoryCount(ePlayer, MEMORY_MADE_DEMAND, 1);
    	}
    	AI_changeMemoryCount(ePlayer, MEMORY_MADE_DEMAND_RECENT, 1);
    	break;
    	
    
    The second event ("You made an arrogant demand")is left here (in the snippet) to show that actually on each 'demand' the event gets fired, which is important.

    ...and decrease (each turn)
    Code:
    for (iI = 0; iI < MAX_PLAYERS; iI++)
    	{
    		if (GET_PLAYER((PlayerTypes)iI).isAlive())
    		{
    			for (iJ = 0; iJ < NUM_MEMORY_TYPES; iJ++)
    			{
    				if (AI_getMemoryCount(((PlayerTypes)iI), ((MemoryTypes)iJ)) > 0)
    				{
    					if (GC.getLeaderHeadInfo(getPersonalityType()).getMemoryDecayRand(iJ) > 0)
    					{
    						if (GC.getGameINLINE().getSorenRandNum(GC.getLeaderHeadInfo(getPersonalityType()).getMemoryDecayRand(iJ), "Memory Decay") == 0)
    						{
    							AI_changeMemoryCount(((PlayerTypes)iI), ((MemoryTypes)iJ), -1);
    						}
    					}
    				}
    			}
    		}
    
    Asking every turn can get MEMORY_MADE_DEMAND_RECENT to a very high value and when it's above zero, it only gets that counter increased w/o yeild.

    It look all the leaders have MEMORY_MADE_DEMAND_RECENT = 20, so it's 5% chance to 'forget' one request per turn.
     
  17. Silu

    Silu Chieftain

    Joined:
    Jul 13, 2009
    Messages:
    2,964
    Location:
    Finland
    Oh, OK. Then it's very very dumb to ask every turn :p I wonder what the optimal beg pattern is. Every 20 turns, and if one is refused, wait 40 turns until the next one? I doubt it's that simple though, since the refusal can be due to other things. Wonder if that problem is too complex to solve in closed form and should do some simulations instead.

    Asking every 14 turns might actually be better, since the probability of forgetting the last beg would be 1-(0.95^14) ~= 51% so constant asking shouldn't make the counter diverge(?). Dunno if the wait time should be incremented if a beg is refused. Might have to do some math :p
     
  18. Silu

    Silu Chieftain

    Joined:
    Jul 13, 2009
    Messages:
    2,964
    Location:
    Finland
    Just for kicks, I simulated some hardcore begging with slightly different parameters. Turns out I was quite optimistic about the 14 turns. I altered the base refusal chance (chance to refuse for other reasons than recent begging) and the interval. All numbers done as averages over 100 "games" with 10000 "turns" for each 6 category. The number given is the amount of "successful begs" in a single "game" ":p" Peaks:

    Base refusal 0%:

    Always same interval
    Interval: 35 Avg successes 206.77

    1.5x interval after refusal
    Interval: 29 (43) Avg successes 219.03

    Double interval after refusal
    Interval: 27 (54) Avg successes 220.55


    Base refusal 20%:

    Always same interval
    Interval: 35 Avg successes 164.85

    1.5x interval after refusal
    Interval: 27 (40) Avg successes 172.1

    Double interval after refusal
    Interval: 22 (44) Avg successes 170.85


    I think the 20% scenario is more realistic (though of course far from the actual scenario still). That would indicate that the optimal way would be the 1.5x interval style, though wouldn't matter much as long as the interval is well chosen. In the light of this test constant begging with 20 turn intervals is a very bad idea when compared to ~30.


    Raw data:
    Spoiler wall of text hits reader for 7214 damage :
    Baserefusal 0%:

    Interval: 25 Avg successes 155.75
    Interval: 26 Avg successes 164.25
    Interval: 27 Avg successes 179.47
    Interval: 28 Avg successes 185.68
    Interval: 29 Avg successes 194.74
    Interval: 30 Avg successes 200.36
    Interval: 31 Avg successes 203.1
    Interval: 32 Avg successes 203.24
    Interval: 33 Avg successes 204.56
    Interval: 34 Avg successes 204.45
    Interval: 35 Avg successes 206.77
    Interval: 36 Avg successes 204.57
    Interval: 37 Avg successes 203.78
    Interval: 38 Avg successes 203.93
    Interval: 39 Avg successes 203.87
    Interval: 40 Avg successes 202.73
    Interval: 41 Avg successes 197.9
    Interval: 42 Avg successes 197.2
    Interval: 43 Avg successes 193.29
    Interval: 44 Avg successes 193.47
    Interval: 45 Avg successes 189.82
    Interval: 46 Avg successes 187.8
    Interval: 47 Avg successes 185.39
    Interval: 48 Avg successes 182.22
    Interval: 49 Avg successes 180.55

    Baserefusal 0%, 1.5x interval after refusal:

    Interval: 10 Avg successes 3.14
    Interval: 11 Avg successes 3.32
    Interval: 12 Avg successes 7.09
    Interval: 13 Avg successes 13.26
    Interval: 14 Avg successes 43.89
    Interval: 15 Avg successes 69.38
    Interval: 16 Avg successes 111.01
    Interval: 17 Avg successes 129.82
    Interval: 18 Avg successes 158.58
    Interval: 19 Avg successes 166.12
    Interval: 20 Avg successes 184.92
    Interval: 21 Avg successes 194.85
    Interval: 22 Avg successes 202.37
    Interval: 23 Avg successes 206.58
    Interval: 24 Avg successes 211.22
    Interval: 25 Avg successes 214.1
    Interval: 26 Avg successes 217.75
    Interval: 27 Avg successes 217.08
    Interval: 28 Avg successes 215.36
    Interval: 29 Avg successes 219.03
    Interval: 30 Avg successes 217.58
    Interval: 31 Avg successes 215.93
    Interval: 32 Avg successes 213.54
    Interval: 33 Avg successes 214.43
    Interval: 34 Avg successes 212.1
    Interval: 35 Avg successes 209.29
    Interval: 36 Avg successes 208.96
    Interval: 37 Avg successes 206.44
    Interval: 38 Avg successes 203.71
    Interval: 39 Avg successes 200.27
    Interval: 40 Avg successes 198.54
    Interval: 41 Avg successes 197.18
    Interval: 42 Avg successes 195.04
    Interval: 43 Avg successes 191.81
    Interval: 44 Avg successes 188.67
    Interval: 45 Avg successes 187.39
    Interval: 46 Avg successes 184.32
    Interval: 47 Avg successes 181.12
    Interval: 48 Avg successes 180.08
    Interval: 49 Avg successes 175.62

    Baserefusal 0%, double interval after refusal:

    Interval: 10 Avg successes 22.48
    Interval: 11 Avg successes 55.84
    Interval: 12 Avg successes 98.23
    Interval: 13 Avg successes 128.2
    Interval: 14 Avg successes 153.99
    Interval: 15 Avg successes 173.08
    Interval: 16 Avg successes 186.32
    Interval: 17 Avg successes 193.34
    Interval: 18 Avg successes 202.94
    Interval: 19 Avg successes 208.76
    Interval: 20 Avg successes 212.36
    Interval: 21 Avg successes 218.78
    Interval: 22 Avg successes 219.11
    Interval: 23 Avg successes 220.41
    Interval: 24 Avg successes 219.12
    Interval: 25 Avg successes 220.31
    Interval: 26 Avg successes 219.19
    Interval: 27 Avg successes 220.55
    Interval: 28 Avg successes 218.01
    Interval: 29 Avg successes 213.93
    Interval: 30 Avg successes 213.77
    Interval: 31 Avg successes 214.49
    Interval: 32 Avg successes 210.21
    Interval: 33 Avg successes 211.49
    Interval: 34 Avg successes 206.25
    Interval: 35 Avg successes 203.9
    Interval: 36 Avg successes 204.09
    Interval: 37 Avg successes 199.34
    Interval: 38 Avg successes 197.71
    Interval: 39 Avg successes 194.91
    Interval: 40 Avg successes 192.74
    Interval: 41 Avg successes 190.18
    Interval: 42 Avg successes 188.33
    Interval: 43 Avg successes 185.82
    Interval: 44 Avg successes 184.93
    Interval: 45 Avg successes 181.28
    Interval: 46 Avg successes 178.64
    Interval: 47 Avg successes 177.2
    Interval: 48 Avg successes 175.23
    Interval: 49 Avg successes 173.46

    Baserefusal 20%:

    Interval: 25 Avg successes 122.82
    Interval: 26 Avg successes 131.81
    Interval: 27 Avg successes 141.04
    Interval: 28 Avg successes 148.65
    Interval: 29 Avg successes 155.57
    Interval: 30 Avg successes 159.95
    Interval: 31 Avg successes 160.78
    Interval: 32 Avg successes 164.49
    Interval: 33 Avg successes 163.03
    Interval: 34 Avg successes 164.18
    Interval: 35 Avg successes 164.85
    Interval: 36 Avg successes 164.5
    Interval: 37 Avg successes 162.57
    Interval: 38 Avg successes 162.48
    Interval: 39 Avg successes 162.18
    Interval: 40 Avg successes 158.82
    Interval: 41 Avg successes 158.81
    Interval: 42 Avg successes 156.55
    Interval: 43 Avg successes 155.77
    Interval: 44 Avg successes 153.85
    Interval: 45 Avg successes 151.91
    Interval: 46 Avg successes 149.22
    Interval: 47 Avg successes 147.13
    Interval: 48 Avg successes 145.89
    Interval: 49 Avg successes 144.12

    Baserefusal 20%, 1.5x interval after refusal:

    Interval: 10 Avg successes 2.36
    Interval: 11 Avg successes 3.19
    Interval: 12 Avg successes 6.9
    Interval: 13 Avg successes 11.35
    Interval: 14 Avg successes 37.6
    Interval: 15 Avg successes 60.47
    Interval: 16 Avg successes 93.67
    Interval: 17 Avg successes 109.82
    Interval: 18 Avg successes 128.55
    Interval: 19 Avg successes 139.35
    Interval: 20 Avg successes 152.69
    Interval: 21 Avg successes 151.29
    Interval: 22 Avg successes 164.28
    Interval: 23 Avg successes 167.84
    Interval: 24 Avg successes 167.62
    Interval: 25 Avg successes 169.9
    Interval: 26 Avg successes 171.02
    Interval: 27 Avg successes 172.1
    Interval: 28 Avg successes 170.85
    Interval: 29 Avg successes 168.74
    Interval: 30 Avg successes 168.63
    Interval: 31 Avg successes 166.79
    Interval: 32 Avg successes 167.68
    Interval: 33 Avg successes 163.42
    Interval: 34 Avg successes 162.44
    Interval: 35 Avg successes 162.46
    Interval: 36 Avg successes 159.76
    Interval: 37 Avg successes 157.35
    Interval: 38 Avg successes 155.6
    Interval: 39 Avg successes 153.45
    Interval: 40 Avg successes 152.3
    Interval: 41 Avg successes 148.3
    Interval: 42 Avg successes 146.92
    Interval: 43 Avg successes 143.19
    Interval: 44 Avg successes 141.4
    Interval: 45 Avg successes 141.29
    Interval: 46 Avg successes 137.44
    Interval: 47 Avg successes 136.33
    Interval: 48 Avg successes 134.35
    Interval: 49 Avg successes 133.66

    Baserefusal 20%, double interval after refusal:

    Interval: 10 Avg successes 19.67
    Interval: 11 Avg successes 50.26
    Interval: 12 Avg successes 84.01
    Interval: 13 Avg successes 108.98
    Interval: 14 Avg successes 125.48
    Interval: 15 Avg successes 140.9
    Interval: 16 Avg successes 152.53
    Interval: 17 Avg successes 158.39
    Interval: 18 Avg successes 161.24
    Interval: 19 Avg successes 164.63
    Interval: 20 Avg successes 166.71
    Interval: 21 Avg successes 170.24
    Interval: 22 Avg successes 170.85
    Interval: 23 Avg successes 169.76
    Interval: 24 Avg successes 168.04
    Interval: 25 Avg successes 166.89
    Interval: 26 Avg successes 167.21
    Interval: 27 Avg successes 167.21
    Interval: 28 Avg successes 163.93
    Interval: 29 Avg successes 159.57
    Interval: 30 Avg successes 159.74
    Interval: 31 Avg successes 156.2
    Interval: 32 Avg successes 154.31
    Interval: 33 Avg successes 152.42
    Interval: 34 Avg successes 151.09
    Interval: 35 Avg successes 147.66
    Interval: 36 Avg successes 144.86
    Interval: 37 Avg successes 144.27
    Interval: 38 Avg successes 142.93
    Interval: 39 Avg successes 139.72
    Interval: 40 Avg successes 137.5
    Interval: 41 Avg successes 137.47
    Interval: 42 Avg successes 133.6
    Interval: 43 Avg successes 131.3
    Interval: 44 Avg successes 128.48
    Interval: 45 Avg successes 127.85
    Interval: 46 Avg successes 126.92
    Interval: 47 Avg successes 123.7
    Interval: 48 Avg successes 123.85
    Interval: 49 Avg successes 121.92

     
  19. bestsss

    bestsss Chieftain

    Joined:
    Jul 12, 2009
    Messages:
    1,955
    Silu,
    Remember that the max value one can ask depends on:
    * the current turn
    * power ratio
    * previously given stuff (negative)
    So asking too often even successful wont yeild a lot. I think it's much better to ask a few times for bigger stuff, if you can properly calculate the max the AI will give in.
     
  20. Silu

    Silu Chieftain

    Joined:
    Jul 13, 2009
    Messages:
    2,964
    Location:
    Finland
    Yeah I remember :) . I didn't mean one should ask with a constant interval for the whole game, rather shorter "bursts" when they start having stuff worth asking but the previously given stuff won't spoil things yet - I just calculated long sets of turns to eliminate variance.

    As calculating the max value the AI will give gets pretty hard I think it's better to ask a bit less and repeatedly around the intervals above. After all getting refused due to asking too much is a "waste of time" in this sense (vaguely simulated by the base refusal rate) as the "previously given stuff" penalty is a function of the total value they have given instead of the total number of successful begs.
     

Share This Page