17-08-2003 07:40									#61
	Support Apolyton, buy Alpha Centauri

BFM, Number Crunchers,

I have loaded the Top 5 Cities & Demographic Scores from Turn 1 to Turn 104 into spreadsheets. But with my family on home leave, I need to do a lot of house chores, and not able to make good charts and analysis this weekend.

So, for all your number crunchers who like to play with the data, there you go :

http://apolyton.net/upload/files/rush/Top5CitiesZip.ZIP

http://apolyton.net/upload/files/ru...graphicsZip.ZIP

	19-08-2003 23:47									#62
	Support Apolyton, buy Civilization 2

Updated: Demographics analysis.

The info below is old info.

quote: Turn 69 Analysis I ran some analyses by changing the Luxury slider and the cities' Labourer Force. This affects some Demographics statistics directly, so we can see how close we really are to the other civs. These are the results: code: GNP MFG FS Prod AR Com pos Sh pos Fd pos C+S+F pos HF AR pos 29+ 3 22+ 3 10+ 5 42+ 3 13 100% 1 28 3 21 3 9 5 41 4 12 96% 1 27 3 20 3 8 7 40 5 11 92% 1 26 3 19 3 39 5 10 88% 1 25 3 18 4 38 6 9 84% 1 24 5 17 4 37 7 8 80% 1 23 5 16 4 36 7 7 76% 2 22 5 15 4 35 7 6 73% 3 21 5 14 4 34 7 5 69% 4 20 5 13 5 33 7 4 65% 5 19 5 12 6 32 7 3 61% 6 18 5 11 6 31 7 17 6 10 7 30 8 16 7 9 8 15 7 14 8 From left to right; Poly in yellow; most important variable in (): GNP (commerce before corruption) The civs are at ??, ??, 25, 25, 22, 18, 17, 15 million MFG Goods (shields after waste) The civs are at ??, ??, 22, 19, 14, 13, 11, 10 shields Family Size (net food) The civs are at ??, ??, ??, ??, 10, 9, 9, ?? net food Productivity (net commerce + net shields + net food) The civs are at ??, ??, 51, 42, 41, 39, 38, 31 units Approval Rating (Happy Faces / %) The civs are at 80, 76, 73, 69, 65, 65, ??, ?? percent (these %s are approximations) Two civs (the '??' in the first four stats) seem really stronger than the rest. I suspect them to be the Netatolians and Webringians. I don't want to meet these guys just yet. If you compare Poly to the #4 civ in Productivity, we're at +9. What causes this difference? Food will not make a huge difference; we're +3 in shields, we're only 5th in commerce. Conclusion: our corruption is significantly lower than the other civ's.

	20-08-2003 22:26									#63
	Suffering from ads?

Updated: Tech Formula (!).

	21-08-2003 08:12									#64
	Support Apolyton, buy Civilization 2

Just curious - where did the new divisor factor of 1.75N replacing 2x(N-1) come from in the tech equation?

	21-08-2003 17:17									#65
	Enter the AD-FREE zone

CFC. See the link to the appropriate thread in the Science part of this thread.

	26-08-2003 12:34									#66
	Enter the AD-FREE zone

Science section updated.

	05-09-2003 14:33									#67
	Support Apolyton, buy Civilization III: Complete

Hi BFM.

We have news from the chat.

It seems as though the formula we have been using for military strength analysis needs some tweaking because we stumbled onto some factors that we had not previously considered. Trip's Military advisor (F3) was telling him we were "Strong" to GCA, while nye's Advisor said we were "Average." In trying to determine why that was so, Trip asked nye if he had fortified all units on the map or not. nye said he hadn't and Trip said that in his version, he did; so there's the difference right there . I inquired if anyone thought terrain made a difference as well, it made sense to me at this point that it would since the game uses general strength numbers to determine the outcome of all combat, so why wouldn't it include the terrain bonuses it uses for combat in the strength analysis as well. Needless to say, it does.

You can move a unit from plains to a mountain tile and go from average to strong; likewise you can go from a hill tile onto a plains tile and go from average to weak (not that we do that, this is just an example.)

So, upon learning this, I took the liberty to access our units combined strength. With these exact settings we are "Strong" If you unfortify any MI on any grassland tile, from this setup, we go to "Average." This is as of turn 113 - 310 BC.

code:

#  level   unit     tile   Fortified    Off  Def   Comb.(Off & Def)   Multiplied by Level
                                                                     (Reg, Vet, Elite;3,4,5)

1   Vet   Sword     Moun       N         3    4           7             28
1   Reg   Warr      Moun       N         1    2           3              9
1   Vet   NM        Moun       N         2    3.75        5.75          23
1   Reg   Warr      Jung       N         1    1.25        2.25           6.75
1   Reg   Sword     Coas       N         3    2.2         5.2           15.6
1   Vet   Sword     Coas       N         3    2.2         5.2           20.8 
1   Reg   NM        Moun       N         2    6           8             24
9   Vet   MI        Grass      Y         4    2.7         6.7           26.8   (241.2)
2   Vet   MI        Grass      N         4    2.2         6.5           26      (52) 
1   Reg   NM        Hill       Y         2    5.25        7.25          21.75
4   Vet   NM        Grass      Y         2    4.05        6.05          24.2    (96.8)
2   Reg   NM        Grass      Y         2    4.05        6.05          18.15   (36.3)
1   Elite Warr      Grass      N         1    1.10        2.1           10.5
2   Reg   Warr      Hill       N         1    1.5         2.5            7.5    (15)
1   Reg   Warr      Jung       N         1    1.25        2.25           6.75
3   Vet   Horse     Grass      Y         2    1.35        3.35          13.4    (40.2)
1   Reg   MI        Grass      N         4    2.1         6.1           18.3
2   Reg   Gall      Coast      N         1    1.1         2.1            6.3    (12.6)

For a total Apolyton Army Strength of  678.55
I did not count the CAT

Last edited by BigFree on 05-09-2003 at 14:55

	05-09-2003 14:47									#68
	Tired of ads?

Here's something to reference in the future:

	05-09-2003 15:04									#69
	Support Apolyton, buy Civilization: The Boardgame

We can get some EXTREMELY precise estimates of GCA's strength using this.

Once we figure out the average bounds, we should start looking at their cities and what we know they last had in the field vs. CFC.

This we could start making logical assumptions:
At least 1 spear in every city (assume regular... it's better to overestimate than under estimate their possible unit ratios this way). Figure out if cities are on hills or flat land.

Then subtract what we know they have from CFC area... then infer.

Trip has done it already I believe. Roughly 7 vet swordsmen hidden from view. Does this new info affect that at all for total unit stats?

	05-09-2003 20:19									#70
	Support Apolyton, buy Civilization III: Complete

If BFM has not already done so, we still have to calculate the exact formula how F3 strength is derived. It places weight on the number of units, moreso than on their stats. This means that BF's tallying up of all the stats does not give a true reflection of what the F3 screen is measuring.

I will talk with BFM about this and see if I can refine the formula further more over the weekend.

	05-09-2003 21:43									#71
	Put an end to popups!

[Artie Johnson Voice] Verdy Interesting!! [/voice]

The Implecations of this are astounding!! If we want to "lead" someone on, to make them think that they are average compaired to us, all we have to do is unfortify all of our core units and then when they move to attack, the will learn other wise...

E_T

	05-09-2003 22:23									#72
	Enter the AD-FREE zone

Based upon the sim that I have currently updated, when our main stack is unfortified we are average with GCA, and when it is fortified we're strong. This is with GCA having 8 Spears out of sight (half vet) and 8 Swords out of sight (all vet).

However, if I change one of those Swears to Swords (9 Swords and 7 Spears) then we remain average no matter what. With 7 Swords and 9 Spears we remain strong with half our army fortified.

So I think them having 8 Swords and 8 Spears out of sight is a pretty good estimate. They may have a bit less since I can't calculate things when their units fortified or with other defensive bonuses added.

Edit: Update.

Last edited by Trip on 05-09-2003 at 22:50

	05-09-2003 22:58									#73
	Support Apolyton, buy Civilization: The Boardgame

Isn't there a standby order? y?

	06-09-2003 01:47									#74
	Tired of ads?

Yes, that order is the same as fortify, except that when the unit encounters another Civ's units or Barbs, it wakes up.


Yes, ET, it would be easy to screw with another Civ. Lets say we were in the Industrial age and we had walls in all of our cities/Metros. Inside the cities you get the defense bonus, outside you do not. So, on one turn you move every unit outside of the cities and do not fortify them. Then on the next turn you move them bak in and fortify. You could concievably go from Strong to Weak back to Strong again over the course of one turn only. That would definately serve to confuse the other Civ.

I can see where this tactic would work well against the AI in SP also; where you baited the AI to attack you becasue you were "Weak." Or the opposite where you Fortified on Mountains in order to fool the AI into thinking you are stronger than you really are.


Trip, I have to disagree with your assumption that Civ 3's strength calculator puts more emphasis on unit numbers than their individual stats. The fact that it even counts whether or not a unit is fortified and what type of tile it is on shows what it does count. It may not even count units numbers, only their attributable stats. I'd like to see a SIM where one side has more units, but that their strength, in stats, like I've tallied here, is less than the other side, but is still considered stronger. I do not believe that will ever happen.

	06-09-2003 02:37									#75
	Support Apolyton, pre-order Civilization IV

Can someone check if standby is in fact fortify with wake-up in terms of the strength calculation, or if it might be weaker?

	06-09-2003 02:47									#76
	Got spare money?

I wonder if the other teams even study things like this in order to get ahead? The comparative play styles of CFC and the others so far has shown a bit of haphazardness to their game. I certainly don't want to underestimate what they might be doing, but if they're not playing the game with all their ducks in a row, why would they be studying things like this to try and make a guess about the complement of their enemy forces.

Now, the fact that we pay this much attention to detail is just one reason why Apolyton will eventually win this game.

	06-09-2003 02:54									#77
	Support Apolyton, buy GURPS/ Alpha Centauri

quote: Originally posted by BigFree Trip, I have to disagree with your assumption that Civ 3's strength calculator puts more emphasis on unit numbers than their individual stats. The fact that it even counts whether or not a unit is fortified and what type of tile it is on shows what it does count. It may not even count units numbers, only their attributable stats. I'd like to see a SIM where one side has more units, but that their strength, in stats, like I've tallied here, is less than the other side, but is still considered stronger. I do not believe that will ever happen.

Sorry, but you're wrong. I've already done testing on this a long time ago, but I'll do it again to show you...

Case Study 1:

Twelve 1/1 units are strong vs. four 3/3 units. Thus despite having exact stats on both sides (12/12) the more plentiful 1/1 units are "stronger." Twelve 1/1 units are average vs. five 3/3 units. That makes 12/12 for the 1/1 units and 15/15 for the 3/3 units before they're considered average with each other. This also does not take into account that despite the fact that they're average it does not mean they're EQUAL.

Unfortunately there's no way to know when two militaries are exactly equal (unless you're Soren). However, we do know you can be within ~25% weaker or stronger to still be considered "average." So to make them equal you would need 1 or 2 more of the 3/3 units to reach equality between the two sides. Thus you end up with 12/12 for the 1/1s and 18/18 or 21/21 for the 3/3s. That's a difference of 67-90% simply due to sheer numbers. This clearly shows that unit #s are more important than their stats.

Case Study 2:

It takes at least fourteen 2/2 units to be strong compared with ten 3/1 units. On the other hand, it takes eight 2/2s to be weak to ten 3/1s. This shows that offense is weighted more than defense by the military advisor (as we already knew).

Other testing has yielded that:

Movement does not affect the military advisor's calculations. Thus a Horseman counts as much as an Archer for F3 calculations.

	06-09-2003 08:08									#78
	Support Apolyton, buy GURPS/ Alpha Centauri

Just an idea I haven't tested at all (may do so later) due to not being familiar enough with the editor, but could the rankings be based on the combat chances of success for attack or defence? It seems a little farfetched that, if this were the case, then in Trip's example the 12 1/1s would have a higher ranking than 4 3/3s, but I may check it anyway.

	06-09-2003 10:24									#79
	Support Apolyton buy from Amazon

quote: Originally posted by Trip Case Study 1: Twelve 1/1 units are strong vs. four 3/3 units.

I am not finding that. Under what circumstances?

	06-09-2003 15:09									#80
	Lose 30 kilos (of popups)

Ok Trip I ran a few tests, by all means they are not conclusive, but they do seem to support my position.

I set up two differnent scenarios, actual .bix files created using the Editor.

In scenario #1 I have Rome versus Greece.
Rome has 100 Regular Warriors and one city producing wealth. Greece has the same exact thing. The terrain was all plains. I played this scenario in HotSeat mode so I could check stats for both sides.

Here's the results:
Rome with 100 Regular Warriors(1-1-1) is "Average" to Greece when Greece has 100 Regular Warriors.

Rome stays "Average" with Greece until I disband 21 Roman Warriors, the Rome goes to "Weak" versus Greece. So:
Rome with 79 Warriors is "Weak" to Greece when Greece has 100 Warriors.

I checked this conversely and Greece was "Strong" to Rome at the same point that Rome became "Weak" to Greece.


Now to check to see whether or not unit numbers, not their individual stats counted, or a measure of both, I used Babylon and Greece for Scenario #2. This scenario was set up the same as #1 except as noted.

Here's the results:

Babylon with 50 Regular Bowmen(2-2-1) is "Average" to Greece when Greece has 100 Regular Warriors.

Babylon stays "Average" with Greece until I disband 21 Greek Warriors, then Babylon goes to "Strong" versus Greece. So:
Babylon with 50 Regular Bowmen is "Strong" to Greece when Greece has 79 Regular Warriors.

I checked this conversely and Greece was "Weak" to Babylon at the same point that Babylon became "Strong" to Greece.


Now, the way that I tallied things up above puts things like this:

100 Regular Warriors have a strength of 600 [3 X (100 X (1+1))] and 50 Regular Bowmen have a strength of 600 [3 X (50 X (2+2))]

So, 50 Regular Bowmen do indeed have the same strength, as calculated by the F3 screen as 100 Regular Warriors.

The number of units do not count for more in the F3 screen. It's the amount of the stats for each unit that count towards it.

These results seem to also put off the 25% figure that we were previously using; as in the window of "Average" was 75%-125%. The window seems to be 21% or 79%-121%. At a 20% difference, the Advisor from the F3 window did not change, it does change at a difference of 21%, according to my method of tallying unit strengths.

I did not test differing terrains, I can do that at some other time.

	06-09-2003 19:28									#81
	Support Apolyton, buy GURPS/ Alpha Centauri

Did you try Fortifiying them, Man!?!

You know, If you make a scenario, you can always send it to someone else to have them do the testing. It's a fair devision of labor.... They can also do the spreadsheets, too.

E_T

	18-09-2003 17:21									#82
	Support Apolyton, buy GURPS/ Alpha Centauri

quote: Originally posted by BigFree Ok Trip I ran a few tests, by all means they are not conclusive, but they do seem to support my position.

Stop! Before you go any further, please read Army Strength Formula Test Area first.

I did a lot of testing there, and it seems I just need to adjust the formula to account for defense bonuses and fortify status. I won't do that in that particular thread, so we alone benefit from our findings.

quote: Babylon with 50 Regular Bowmen(2-2-1) is "Average" to Greece when Greece has 100 Regular Warriors.

It seems that the fact that a Bowman is a UU may play a role here. Better use a 'neutral' (modded) 2-2-1 unit.

quote: These results seem to also put off the 25% figure that we were previously using; as in the window of "Average" was 75%-125%. The window seems to be 21% or 79%-121%.

It never was 75%-125%. It has been determined as 80%-125% (both boundaries included) all along.

I will run more tests when I find the time.

edit: there is also the factor of "Army Support Cost". 1 city supporting 100 units is different from 1 city supporting 50 (stronger) units.

	19-09-2003 00:30									#83
	Enter the AD-FREE zone

BFM, that's why I tested 100 Warriors against 100 Warriors. The results of strength allow exactly the same conclusion of my formula, which is a rather simple one. If Had not run the test on regular units as well I might be inclined to agree with you that a UU might change the results. I might even consider the fact that 50 units against 100 units could effect the strength factors; but in my testing, it shows it does not.

One thing I found in my testing that was important to note is that you need to test in either Hotseat or Turnbased mode and not allow the AI to control one of the Civ's.

Even though I ran only two tests, they were simple and stream lined. I stand behind the results and my conclusion 100%. I welcome further testing by other's. I'm sure Trip did some more testing on this as well.

	20-09-2003 18:24									#84
	Increase the size of your Attachments

Updated: Science, Contact Analysis, Civ-specific Abilities.

Old info:

quote: CDG's City Development Update: Turn 66, Jun 13th, 2003. From the Turn Reports, we can see the development of their cities: Elucidia Turn 02 - 1 - founded ... (building 2 settlers in the meantime) ... Turn 49 - 4 - discovered Turn 54 - 2 - built a settler Turn 55 - 3 - growth Turn 62 - 4 - growth Turn 65 - 2 - built a settler Turn 66 - 2 GoatCity Turn 26 - 1 - founded & discovered Turn 36 - 2 - growth Turn 56 - 3 - growth Turn ?? - 1 - built a settler Turn 66 - 1 Zarigueya Turn 35 - 1 - founded & discovered Turn 45 - 2 - growth Turn 66 - 2 Sheherazade Turn 63 - 1 - founded & discovered Turn 66 - 1 Cuido de Alcohol Turn 65 - 1 - founded & discovered Turn 66 - 1 If CDG has built a worker, they may have at Elucidia before turn 49. Otherwise, they still have their initial 1.

	23-09-2003 21:32									#85
	Support Apolyton, buy Galactic Civilizations

Updated Contacts and Science for CGN contact.

	27-09-2003 15:08									#86
	Enter the AD-FREE zone

There's been some debate about whether the "Fortified" status of units and Terrain bonuses affect their strength for the purposes of Army Strength calculation in the F3 screen.

I finally found some time to run some tests.

There are two things to test: "Fortified" status, and Terrain bonuses, however, if they both have an effect, it's likely that it's a combined one (since both factors determine how well the unit will defend in case of combat).

"Fortified" status

I set up the following test scenario:

On an island with only mountains, I have X units against Y of the same units of another civ.

• The first test is with warriors: X=11, Y=14. 11/14 is 79%, which is below the boundary of 80%, so it should show "weak", and it does.
  
• Now, I fortify all my warriors. I expect a change to "average", but it still shows "weak".
  
• Then, I replace all warriors with spearmen. Again, 79% = weak, but the difference is now, that if I fortify 2 spearmen, it changes from "weak" to "average".
  
• Test repeated with 14 -vs- 18 spearmen. Needs 4 fortified ones on the weaker side to change from "weak" to "average".
  
• Test repeated with spearmen who have been modded to have a Defense value of 3 (it's now a hoplite without being UU). No I need only 3 fortified ones to change to "average".
  

Last edited by BigFurryMonster on 27-09-2003 at 21:22

	27-09-2003 19:24									#87
	Inflate your Upload Space

For the terrain bonuses check. you also need to see if being in a town, city and metro on a hill has any effect.

E_T

	28-09-2003 13:32									#88
	Suffering from ads?

It occurred to me yesterday that the Fortify Bonus might simply add 25% to the Defense Value of a unit...

Edit: Not quite. The 25% bonus seems to work OK for units with a defense of 4, but for other values, it's off by some margin. Since 25% of 4 is an integer value, this may be due to rounding, perhaps...

Last edited by BigFurryMonster on 28-09-2003 at 19:49

	28-09-2003 20:01									#89
	Support Apolyton, buy Civilization 2

BFM, would you please use another colour for your edits besides banana yellow? Yes, I know it represents the Great Banana, but it’s very difficult to read, especially on a grey background. I usually end up skipping the edits because it’s too much of a bother to squint at them to see what they say. It’s times like these I wish we worshipped the Great Apple or the Great Eggplant instead.

Last edited by pikesfan on 28-09-2003 at 20:06

	28-09-2003 20:44									#90
	Support Apolyton, buy GURPS/ Alpha Centauri

I just highlight all of BFM's text. It's easier to read the yellow when it turns blue under highlight