Analysis: The Opening Game Unraveled

INTRO

As most readers here probably know, there is quite the ongoing debate about the Ljosalfar and allegations of imbalance. Finding myself inspired by the smack-talk found in the various Elf threads (much of it originating with yours truly) and tantalized by the chance to do several hours of tedious data collection, I had no choice but to play several test games and tally up the results.

I've now played The Clan led by Jonas, Lanun by Hannah, the Khazad by Kandros, and the Ljosalfar under Arendel. These Civs were chosen based upon comments found here on this forum. “Common Wisdom” agrees these Civs/Leaders have particularly good potential for good starts and strong economies.

METHODOLOGY

I've played these games to turn 200. I took notes during play. No realm was played perfectly, but as things happened, no realm was thrown far off-track by external forces. I tried to minimize micromanagement, but much of it was unavoidable. So occasionally the AI would work a city tile ‘off-plan’ and occasionally it took me a few turns to catch the error. But for the most part, the each Civ developed ‘according to plan’, so to speak.

Each realm was started with boosting production as a very high priority. Commerce production is not the be-all or end-all in a game of Civ, but it is the engine that drives technological development. Rapid tech development is important in the opening game in particular. First-tier techs were ignored after discovery of Ancient Chants and Exploration, afterwards, Education was always researched. Cottages were then erected in as many worked tiles as circumstances allowed. Afterwards, techs were pursued as required, generally with the second goal being discovery of a religion.

Other than that, I tried playing the realms as circumstances demanded. If they needed to build a city to control a chokepoint, that's what they did. Sometimes hindsight made alternatives look better. But I never operated a Civ to artificially boost it’s economy, knowing I was playing to only turn 200 and would not have to deal with a mortgaged future. All effort was made within those parameters to maximize each nation’s economic output, overall, with an emphasis on Commerce.

Games were played on Fractal maps, Large, Temperate, Medium Seas, Emperor skill, Raging Barbarians, Aggressive AI, Epic. I was not particularly picky about map starups, with one exception: Every game was started with a River / Flood Plains near the starting Settler. The operating philosophy here is that there are far too many variables t to directly measure them all. A practical way to evaluate each Civ's strengths is to measure them while near maximum capacity. A skilled player can achieve near-maximum performance, but can never exceed it. So the games were played on maps that offered a promising start for that racial type.

PHASE TWO

The numerical results from each nation cannot be compared directly. Each game was played on a different map, so of course it was easier to grow on some maps than others. What will not change are certain mathematical functions. For example, a Farm on Flood Plain will always yield 4 Food and 1 Commerce (unless you are Lanun) regardless of what else is on the map with that Flood Plain.

I’ve set up a spreadsheet to simulate city growth in FfH. It needs some work, but it will be sufficient (with a lot of eye-squinting) to simulate how each Civ would grow had they played on that map instead. What it won’t do is simulate a lot of micromanaging of city tiles. (Or rather, I won’t eye-squint that much.) But it will reflect trends and it should be good to within a couple percent.

You say Barbs delayed workers on turns 70-80? They wouldn’t for the Clan, so we skip the development gap on the spreadsheet. You say the Khazad don’t have Aecheron next door on this map? That means they would skip this unit build and finish that Elder Council that much sooner. Stuff like that can be modeled, with eye-squinting, close enough, to gauge how the Civs compare regardless of map.

(Sanity Check: This could prove to take up quite a number of hours, a full week of dullness, NOT playing version 2,015, may prove too burdensome. )

PHASE TWO PART TWO
or
WISH I'D THOUGHT OF THAT

Now it just occurred to me today that perhaps a great thing to do would be to design a test map big enough for 4-6 civilizations. Startup zones could be designed to provide nice solid potential for any Civ. Then the whole test could have been run with the single biggest complexity removed: map-to-map differences.

Sigh.

Well, too late for that. This time.

TEST PHILOSOPHY

The first 200 turns on Epic were studied for two reasons. The lesser reason is that tech variety is very limited at this stage, so most every Civ is reacting to the same set of needs/dangers. Tech selections are very similar, with the exception of the chosen religious path. Each Civ is fighting the Barbarians for most of this time with the same units, Warriors and Scouts.

200 turns proved to be a pretty good choice. The Civs start to 'personalize' themsevles around turn 130-190. By turn 200 the early seeds of specialization are just sprouting. The influences from the chosen religion play a strong role in the latter half of this period. But towards turn 200 and beyond, choices start to be made on long-term desires and/or specific strategic challenges encountered in that game. (e.g. Damn it, I want Magic users this game, period!)

Civs develop in different ways, but their success or failure depends upon how well they’ve positioned themselves in the opening game. The biggest reason to end the test at Year 200, therefore, is that the opening game shapes the endgame more deeply and profoundly than any other factor.

These four Civs were chosen to reflect four presumed advantages:

Clan - Barbarian Trait: Can start development right away regardless of the Raging Barbs. Cultured Jonas was considered the better ‘builderish’ leader choice.

Khazad - Dwarven Vault: Dwarves get hindered in early city expansion, but OTOH they have a unique tool in their armory: gold. The Financial leader was selected to maximize their Commercial advantages.

Lanun - Seafaring: These sushi-snarfing surf Sultans get extra from each sea tile, talk about fast population growth. Now, add Hannah’s Financial trait, and Anchors Aweigh!

Ljosalfar - They are Elves: The love 'em / hate 'em civ of FfH, is their economic potential as strong as their critics say? The one 'must-have' civ in a test such as this.

Generic: Originally, I planned to run a 5th, 'vanilla-ish' Civ as a control. However the Lunan game played out with them acting exactly 'vanilla' for the bulk of the game. I believe the differences Lunan and ‘vanilla’ in the first 200 turns are small enough to be estimated, perhaps, discounted.

WHY? WHY?

In theory, a balanced situation is possible between low-potential/easy-to-optimize Civs and high-potential/hard-to-govern nations. So even if this study demonstrates strong economic advantages for certain Civs, that is not proof the overall Civ is out of balance.

However it is quite the optimistic attitude to assume this sort of balance can be achieved unless each Civ’s performance is well and truly understood. This is not always an obvious or trivial exercise. But once the situation is measured and understood, then root causes can be found. After root causes are understood, successful tweaking of the system is much more likely.

Jonas / Clan through Year 200

Turn 1 notes

A nice map with a couple Flood Plains centrally located. Good deal of desert in areas, but not such that it will ****** early growth. Looks like an easy area to block off with border cities leaving a large area to exploit unmolested.

Turn 200 notes

The Clan was allowed to develop unmolested, of course. Their early development was hampered by the lack of a starting tech, and the Barbarian trait -10% R&D penalty. In practice, this penalty ranged from -15% to -22%, usually -20% give or take a percent. This is due to decimal truncation (truncation, not rounding) done not once on the national aggregate, but separately at each city. This means a city producing 1 beaker will effectively produce 0 beakers, or a 100% penalty. (1 x 90% = 0.9 which when truncated become 0.)

This crushing early technological deficit shaped Clan development. They were despaate for additional Commerce, but, thanks to truncation, delays occured even as Commerce increased. This is because the Capitol started with 10 total Commerce. Raising this to 11 had no effect on R&D, because 11x90% truncated still equals 9. After construction of the 2nd city, the overall R&D efficiency was almost always about 80%, give or take a percent. Truncation leads to a loss of about 20% of R&D production, not the -10% one might expect from reading the Barbarian Trait.

Despite uncovering 13 or 14 Goodie Huts, the Clan learned only 1 new tech from Huts. And that turned out to be essentially useless, as that particular tech was already 95% or more complete. But a lot of angry villagers (Barbarians) were set loose.

In retrospect, the Clan should not have pursued Exploration until after Education. The Clan MUST learn Ancient Chants and then Education. Cottages are desparately needed ASAP by Barbarian-trait civs. Furthermore, the ability to explore among Raging Barbarians more than makes up for the Goblins' slow 1 Movement.

Despite their slow early development, this game would be a boring expansion runaway for the Orcs. They have a huge swath of land to develop unmolested, thanks to the demise of the Mercurans in Year 201. It would be "over" long before borders started to bump together. This is due to the "splashy" benefit to the Barbarian trait. There's no telling how it will help in a given game. The lesson here might be to reconsider using large maps when playing Barbarian realms.

QuickData

6 Cities
19 total population
Fellowship of Leaves discovered, active in 2 cities.

21 Road tiles ... Basic road net 2 turns from complettion
0 Mines
0 Farms
0 Pastures
0 Coastal

Suburbia 16 settlements average age 49 turns

1 Cottages
8 Hamlets ... 1 River, 2 on Cow, 1 Rice, 1 Marble
7 Villages ... 2 River, 2 Flood Plain
0 Towns ... (First Town would appear in Year 210)

2,037 Hammers produced
3,379 Beakers produced yielding 8 techs
32 Gold in Treasury
3,411 Total Commercial output

Army:
9 units of combined 16 base Strength
2 Workers

Hannah / Lanun through Year 200

Turn 1 notes

This is a fabulous, resource-rich starting spot. A River near the capitol, Corn, Marble, Silk in 2 G/F tiles, perfect location for efficient placement of first-ring cities, either coastal and/or riverside. This is an extremely promising starting map.

Turn 200 notes

With Seafaring as the starting tech, the Lanun find themselves in much the same boat as the Clan. They don’t have the R&D penalty, but they do have to invent all four tier-1 techs. The Lanun too should probably set aside Exploration until later, but that is not always possible if Barbarians activity is heavy.

Exploration efforts were given up early in order to hunker down and defend the realm. No pillaging occured, but the lack of Cultural border expansion meant Orcs had to be killed as they entered the nation. Usually they could not be allowed to die attacking the city, lest they pillage that FP/Village instead of storm the walls. Obelisks were started and paused at time, Gold and Hammers being in short supply at times. The game played out for the Lanun very much like it would for non Elf/Dwarf/Cultural civs, with one major exception. City 4 and City 5 were ordered to exploit their Coastal sea tiles for the 2 3 per tile. This maximized their commercial output, a primary goal of this study, while maintaining average city growth rates. City 6 operated in a similar manner, but had only 1 sea tile to work until cultural border expansion in Year 201.

This game can serve as an adequate test for a 'Vanilla' FfH Civ, thanks to the very early (before turn 10) discovery of Agriculture from a Goodie Hut. This erased the Lanun's penalty for not knowing any Tier-1 techs at game start. The Lanun suffer from bad farming thanks to their starting Seafaring tech. So the Lanun did not build any Farms, but wouldn't a 'vanilla' civ build them? The answer to that is: no. By the time Workers have been built and given time to work, most cities are already Happycapped. They can't grow bigger until new tech are developed, so why build more Food production? So the bottom line is, our 'vanilla' civ would not build Farms at the outset, just as the Lanun did not build Farms.

Now, City 4 and City 5 would be developed differently by 'vanilla'. But those cities were safe from Barbarians, and that means development is predictable. It will be sufficient to assume earlier construction of the 3rd Worker and calculate the differences from there. Therefore, the 5th planned test game will probably not be played.

QuickData

6 Cities
26 total population
Octopus Overlords discovered, active in 5 cities.

23 Road tiles ... Basic road net is complete
1 Mines ... Gold
2 Farms ... Rice, Corn
0 Pastures ... 2 Cow begining construction
1 Plantation ... Silk
7 Coastal ... 6 regular and 1 Clam/FBoat

Suburbia 9 settlements average age 76 turns

0 Cottages
0 Hamlets
9 Villages ... 2 River, 3 Flood Plain, 1 Reagents, 1 Marble
0 Towns ... (First Town would appear on turn 201)

2,304 Hammers produced
5,147 Beakers produced yielding 10 techs
14 Gold in Treasury
5,161 Total Commercial output

Army:
8 units of combined 16 base Strength
3 Workers
1 Fishing Boat

Kandros / Khazad through Year 200

Turn 1 notes

Fabulous starting spot. A river curls through a desert, ceating 13 FP tiles. The first three cities can share these tiles while at the same time screening them from invaders. Hills are in reasonable quantity, the terrain is promising in every direction but the SE, where the rest of the desert lies. Even so, the city started to the SE will have enough nearby Forest and Flood Plain until it grows large. Until then, the tiles wasted by desert couldn't be worked anyway. This is an extremely high-potential starting map.

Turn 200 notes

The realm is small, but tight and super-efficient. Five cities lie in near-perfect efficient X-layout around the Capitol. The tight realm has allowed the army to move to meet any Barb threat. Occasionally the Barbs managed to scare off / delay Workers, and once most of the army was wounded by the time the latest barb surge was thrown off, but never was a city really threatened with loss, nor was a tile plundered. Dwarven +20% city defense and leader Aggressive trait certainly didn't hurt. But any realm can defense against 0exp Orc Spearmen with 5* Warriors. Some development was slowed by having Aecheron next door, but the primary effect was to make the Khazad military by far the most experienced of the test armies.

QuickData

5 Cities
27 total population
Runes of Killmorph discovered, active in 4 cities.

2,266 Hammers produced

5,534 Beakers produced yielding 9 techs
1,535 Gold in Treasury (Dwarven Vault: Full)
7,069 Total Commercial output

17 Road tiles ... Basic road net is complete
6 Mines ... 2 Gold
1 Farms ... Wheat/River
2 Pastures ... 2 Cow
0 Plantation
0 Coastal

Suburbia 12 settlements average age 50 turns

1 Cottages ... 1 Flood Plain
3 Hamlets ... 3 FP
8 Villages ... 8 FP
0 Towns ... (First Town would appear on turn 219)

Army:
10 units of combined 18 base Strength
3 Workers

Arendel / Ljosalfar through Year 200

Turn 1 notes

A solid starting spot. A river crosses a corner of a continent, with 2 Flood Plains where it empties into the sea. Also visible inland are Gold, Cows, Corn and an Oasis. Small patches of desert breakup the continuity of the arable tiles. Some city placements may be 'imperfect'. The starting settler and obvious capitol site are near the coast, which is not as quick a startup condition as a centralized capitol nation. So inland the starting settler goes. A reasonable amount of hills and forests are available, but do not come close to covering the whole area. It’s a very promising startup map for most any Civ, but runs little danger of making the ‘All-Time Best’ startup map list.

Turn 1 notes

The two Elf Scouts explored the entire continent quickly, with a bit of early luck. The Barbarians seemed to appear slowly this game and the Scouts did a lot of exploring before having to enter Caution Mode. This game was also different in that it featured very close placement to the nearest AI civ. That influenced city placement heavilly, as the Elohim beat me to a couple sites. But conversely, our relations were very peaceful, and they screened us against Barbs, so economic development went very smoothly.

Another difference came in tech development. In the other games, I prioritized economic/growth techs. Usually, the choice was pretty obvious. I departed With Ljosalfar by taking Archery after hitting the critical goal of Fellowship of the Leaves. I felt the chance to build a 5 STR unit and it a Hero to boot was too advantageous for an Elf player to pass up.

Towards the end of the test period, I had already completed the planned invention sequence, so I just started going after bypassed, economic-related techs. In actual game-play, a player would probably instead have decided to start specializing, for example, learning Arcane Lore before Year 200 in order to get a jump on building Adepts. The three other test realms did not have the same flexibility in late-game tech choices.

QuickData

7 Cities
35 total population
Fellowship of Leaves discovered, active in 7 cities.

16 Ancient Forests and 26 normal Forests in 69 turns

4,184 Hammers produced

5,663 Beakers produced yielding 9 techs
8 Gold in Treasury
279 Gold spent on unit upgrades
5,950 Total Commercial output

19 Road tiles ... Basic road net 3 tiles from completion. (City 7 only 5 turns old.)
1 Mines ... 1 Gold
1 Farms ... 1 Corn/River
4 Pastures ... 3 Cow, 1 Horse
1 Plantation ... 1 Silk/Forest (Incense complete in Year 202)
1 Coastal ... Coast/Crab (no boat)

Suburbia ... 24 settlements.

Average age for all 24 is 80 turns (hard calculation: bare minimum) to 89 (estimated) turns.
Average age for C/H/V only (excluding Towns) is 46 turns for 18 settlements.

4 Cottages ... 4 Ancient Forest (+1 AF complete in 202)
4 Hamlets ... 3 AF, 1 Forest
10 Villages ... 3 AF, 4 F, 1 F/Silk, 2 Flood Plain
6 Towns ... 2 AF, 4 F (Cottage 1 is 151 turns old)

Only test realm to have Town-type settlements (Lanun one turn away from their first).

Army:
13 units of combined 35 base Strength
4 Workers

Only test realm to have units of Base STR > 2
4 Elven Hunters, 1 Elven Archer, and Gilden Silveric

Clan of Embers Decade by Decade

Turn
1 Capitol founded, Exploration begun
10 0 Tech, 1 City, 9 Beak, Exploration in R&D
20 City growth to 2 does not grow economy

30 0 Tech, 1 City, 9 Beak
40 1 Tech, 1 City, 9 Beak, Ancient Chants in R&D
50

60 2 Techs, 2 Cities (Early Tech from G.Hut)
70 2 Tech, 2 Cities, 13 Beak
80

90
100 3 Techs, 3 Cities, Crafting R&D
110 3 Techs 2 Cott, 3 Cities, 15 Beak

120 4 Techs, 3 Cott., 3 Cities, 15 Beak. Hunting in R&D
130 4 Techs, 4 Cott, 4 Cities, 19 Beak
140 5 Techs, 7 Cott, 4 Cities, 25 Beak, Mysticism in R&D

150 5 Techs, 8 Cott, 4 Cities, 27 Beak
160 6 Techs, 8 Cott, 5 Cities, 22 Beak. GK, Relig up, Leaves R&D
170 6 Techs, 9 Cott, 5 Cities, 26 Beak

180 6 Techs, 9 Cott, 6 Cities, 25 Beak, Hunt Lodge up
190 7 Techs, 9 Cott, 6 Cities, 37 Beak, FLeaves invented, in 2 Cities, Mining R&D
200 8 Techs, 9 Cott, 6 Cities, 41 Beak, Gold Mine begun

Lanun Decade by Decade

Turn
1 Capitol founded, Exploration begun
10 1 Tech, G.Hut gives Agriculture
20 Capitol size 2, 11 Beak

30 Exploration nearly done, Arcane Chants next
40 Tech 2 done, Ancient Chants in R&D
50 2 Techs, 1 City

60 3 Techs, 2 Cities, 16 Beak. Education R&D, Silk/Reagents
70 3 Techs, 2 Cities
80 Army = 4 Warriors

90 4 Techs, 0 Cott, 2 Cities, 17 Beak, Mysticism R&D
100 4 Techs, 1 Cott, 2 Cities, 17 Beak
110 4 Techs, 3 Cott, 2 Cities, 22 Beak. Mysticism nearly done

120 5 Techs, 4 Cott, 3 Cities, 26 Beak. Fishing in R&D
130 6 Techs, 6 Cott, 3 Cities, 33 Beak. OO in R&D, 2nd Worker up
140 6 Techs, 7 Cott, 4 Cities, 33 Beak. Coastal tiles working now

150 6 Techs, 9 Cott, 5 Cities, 41 Beak, Army 6 Warriors
160 7 Techs, 9 Cott, 5 Cities, 47 Beak, OO in 2 Cities, Crafting R&D, Farms
170 8 Techs, 9 Cott, 6 Cities, 58 Beak, Mining R&D

180 10 Techs, 9 Cott, 6 Cities, 66 Beak, OO in 2 Cities, Clams!
190 10 Techs, 9 Cott, 6 Cities, 72 Beak, Gold Mine
200 12 Techs, 9 Cott, 6 Cities, 77 Beak, Silk

Khazad Decade by Decade

Turn
1 Capitol founded, Arcane Chants R&D, G.Hut gives Agriculture
10 DV: Low 50g 2 Tech, 1 Cities, 10 Beak
20 DV: Low 50g, 2 Tech, 1 Cities, 11 Beak

30 DV: Low 50g
40 DV: Low 50g, 3 Techs, 1 Cities, 9 Beak, Explore R&D
50 DV: Norm, 3 Techs, 1 Cities, 9 Beak

60 DV: Stkd
70 DV: Abundant 203g, 4 Techs, 1 Cities, 10 Beak, Education R&D
80 DV: Norm, 4 Techs, 2 Cities, 12 Beak

90 DV: Norm, 4 Techs, 2 Cities, 16 Beak
100 DV: Norm, 5 Techs, 2 Cities, 13 Beak, Mining R&D
110 DV: Norm, 5 Techs, 1 Cott, 2 Cities, 15 Beak

120 DV: Stkd 304g, 5 Techs, 2 Cott, 3 Cities, 19 Beak
130 DV: Stkd, 6 Techs, 4 Cott, 3 Cities, 25 Beak, Myst R&D
140 DV: Norm 360g, 6 Techs, 6 Cott, 3 Cities, 37 Beak, Gold Mine

150 DV: Norm 404g, 7 Techs, 8 Cott, 4 Cities, 46 Beak, GK, Relg adopted
160 DV: Norm 413g, 7 Techs, 8 Cott, 4 Cities, 51 Beak, Wheat Farm
170 DV: Stkd 796g, 8 Techs, 9 Cott, 4 Cities, 17 Beak, Runes in 2 cities

180 DV: Stkd, 9 Techs, 10 Cott, 5 Cities. AnHusb done, BrzWorking R&D
190 DV: Stkd 1,210g 9 Techs, 11 Cott, 5 Cities, 46 Beak
200 VAULT: Full 1,535g 11 Techs, 12 Cott, 5 Cities, 85 Beak, RK in 4 cities

Ljosalfar Decade by Decade

Turn
1 Capitol founded, Ancient Chants R&D
10 1 Tech, 1 City, 11 Beak
20 1 Tech, 1 City, 12 Beak

30 2 Techs, 1 City, 13 Beak, Education R&D
40 2 Techs, 1 City, 14 Beak, Worker under construction
50

60 3 Techs, 1 Cott, 1 City, 15 Beak, Mysticims R&D
70 3 Techs, 2 Cott, 1 City, 16 Beak
80 4 Techs, 4 Cott, 2 City, 17 Beak, GK, Relg, Apprt adopted

90 4 Techs, 5 Cott, 2 City, 20 Beak, Hunting R&D
100 5 Techs, 7 Cott, 3 City, 26 Beak, WotLeaves R&D
110 5 Techs, 8 Cott, 4 City, 26 Beak

120 5 Techs, 9 Cott, 4 City, 28 Beak
130 6 Techs, 12 Cott, 4 City, 31 Beak, Leaves discovered
140 6 Techs, 13 Cott, 5 City, 39 Beak, 2AF, Leaves in 2 cities

150 6 Techs, 16 Cott, 5 City, 44 Beak, 7 AF, Archery R&D since FL
160 7 Techs, 19 Cott, 6 City, 47 Beak, 9 AF, 2 FL Temples up, Agri R&D
170 9 Techs, 21 Cott, 6 City, 64 Beak, 9 AF, G. Silveric born

180 10 Techs, 21 Cott, 6 City, 71 Beak, 11 AF, Corn, Horses, FL 4 cities
190 11 Techs, 21 Cott, 6 City, 67 Beak, 13 AF, Cows, Gold, FL Temple 3
200 12 Techs, 24 Cott, 6 City, 84 Beak, 16 AF, Silk, FL in all cities

AND ALL THIS MEANS?

OK, let’s review. This thread originated in the debate regarding the Ljosalfar’s (and of some other Civs, to a far lesser extent) performance in FfH. Some players feel Ljosalfar is badly out of tune with the rest of the game. Other players counter that this realm is fine, because any economic advantages are certainly offset by military weakness. Some players still claim there is no production advantage to the civ at all.

So this study was launched in with the hope of adding more information for the informed opinion-makers here. There is no way to measure “game balance” directly. And if there were, it impossible to play the game under so many difficulty settings, maps, and such, that ‘balance’ might exist for one batch of settings, but not another. So when one hopes to study this, simplification is necessary. How was this done?

First, study the opening game. The opening game is the one time the various are as similar to one another as ever. With every passing turn they grow more and more ‘different’ But at the beginning they are all going after the same small group of introductory techs. They are much more comparable. And, of course, that the opening game is of critical importance is accepted as axiomatic fact by gamers familiar with so-called 4X games.

Secondly, study the economy. The economy can be measured directly. How a given realm’s army maneuvers and fights a war cannot. So let’s measure what we can, and adjust be the ‘seat of our pants’ where we can’t.

Third, set out to prove a specific point. For this study that point was this: Cottage construction is the rapidest, most-efficient route to economic expansion in FfH, which in turn makes Ljosalfar’s forest/cottage ability incredibly powerful. Why was this point chosen? Because it is the crux of my position from the aforementioned debates, is the answer. The guy who runs the study, gets to design the study.

Fourth, standardize as much as possible. Each civ was launched upon the world with the same General Orders

1> Develop Exploration, so the Army could deploy Move 2+ Warriors against the Raging Barbs, if needed.
2> Discover Education ASAP afterwards
3> Build a Worker to start Cottage construction ASAP
4> Build only Cottages (not even Roads) until all feasible tiles developed.
5> Prioritize R&D to discover a religion ASAP after Education.
6> Avoid hostilities and strive every turn to maximize long-term production.

However the study failed to standardize in the single most important aspect, the map. Different maps were used, not by choice, but because there was presumed to be no alternative. Each realm was given a Large map, with lots of room to expand, and ample River tiles in sight. But there is no question that each realm would perform differently on different maps.

This does not invalidate the study. It does require us to be careful, but it hardly makes learning impossible. We are not expecting to prove precise mathematical predictions. We are not trying to prove that Ljosalfar will produce exactly 1,532 Hammers in the same time Lanun produces 1,000. We are letting each economy rip, as best it can, from a standing start. We are looking for those times when the one realms start to perform differently from the others. And then we hope to explain the root causes for these changes.

Killing root causes are like killing cancer cells. Zap ‘em when they are small, and they never grow into big problems. I fully agree that this little study would never be presented at FermiLab or CERN. The Unser typing this now, is not the same Unser who started this project 10 days ago. Would I do a lot of things differently? You bet! That’s how hindsight works. In the meantime, there’s lots to learn.

PRE-EMPTIVE FIRE: Answering a few common criticisms.

Lanun:
Many players feel the Lanun would have done far better if only I knew better how to develop them. Well, that may be true, but, once again this is my study and my hypothesis. And that is that Cottages are the better development route. Here’s why: Fishing could be invented about the same time as Education. Then sea tiles can be exploited for 3 easy commerce (Fin leader.) Well, the problem is, there are about only 4 to 6 population points in the entire realm at this time. Only so many sea tiles can be worked if any hammers are needed. But with Cottages, they can be put all over the place. They can be put on Plains (or Marble!) to get hammers. 15 turns later, they improve and generate the same 3 commerce as Coast. 30 turns after that, they generate more. More commerce than Coast, but with hammers. Observe how Lann Cottage production stopped abrupty at 9. From Year 150 Coast tiles were being worked by Cities 4,5 and 6 instead of new Cottages. That was a nice short-term boost towards the end of the test. But in Year 250, Hannah would wish she’d built Cottages even for the coastal cities.

Ljosalfar:
There have been many criticisms of the Ljosalfar results, generally involving a presumed total lack of Barbarian interference. I addresses the Barbarian matter in some detail on a Page 5 message. After going through my notes, I am comfortable that the number of turns lost to barbarian activity is roughly comparable between game to game. Quite frankly, the Raging Barbarians are not so much a threat in FfH, as they are a game of Tetris. (Sorry, too good a line to not use.) Furthermore, the close AI civ also got to two city sites I wanted. Much of the criticism could be filed in this category. Only one side of a two-edged sword was used to cut.

But there are a few legitimate, tangible differences that provided Ljosalfar some small benefits. Their capitol was founded on a Hill/Plains. That situation means the city tile produces a 2nd hammer … very nice for Year 1. Also, the capitol was able to work an Oasis. 3 Food and 2 Commerce is also nice for turn 1. Also the presence of the close AI civ meant for a small advantage no one cited. Each new trade route generated +2 commerce instead of only one. These small advantages do add up and they do compound themselves over time. But OTOH the Lanun map was very nice too. Thanks to Silk in forests and Reagents near a river, they were earning 33% more than “normal” in the early game. And no one has speculated what Ljosalfar might have done on that 13 Flood Plain map. A two-edged sword cuts both ways.

DRUMROLL

So each map came with certain advantages or disadvantages. Despite the best efforts, a lot of variability exists. Were the results anywhere within the same ballpark, I would conclude the test was inconclusive. But the results were nowhere near close. Ljosalfar blew everyone away. Blew everyone away, not in one aspect or two, but in nearly any aspect you care to mention. Military? Ljo. Number of cities? Ljo. Total population points? Development work already done and operating? Hammer production? Projected future growth? Ljo Ljo Ljo Ljo. Only Khazad beat Ljosalfar in any of the major indicators, net commercial output. Everywhere else you look, it is Ljosalfar in a landside.

There is a lot of uncertainty in this study,. But not so much to explain all the massive leads established by Ljosalfar in so many categories. The basic premise has been supported; a realm can develop rapidly by focusing on Cottages, and Ljosalfar is tremendously better at this than anyone else.

Conclusion: Ljosalfar is broken. Fatal flaws exist that must be addressed.

ROOT CAUSES

Here’s how Ljosalfar does it, listed in the order the factors generally occur in the game. I am not suggesting that all or even the majority of these root causes need be eliminated. Some have far more impact than others.

1> Starting Army: Most realms start with a Warrior and a Scout. So starting with two Scouts means starting with an early advantage gathering information and Goodies.

2> Starting Tech: Knowing Exploration means Ljo can go straight for Chants, then Education.

3> Woods Movement/Combat: The Scout pair gets to move faster through the most common hinderance, and fights better in the most desireable end-of-movment terrain. So exploration goes even faster yet.

4> Raging Barbarians Aren’t Very: None move faster than 1 tile/turns, and none will appear before the Capitol can build a defensive Warrior. With fast Woods movement, a 0 exp Elf Warrior often effectively moves as a unit with a Movement promotion.

5> Early Worker Build: Worker builds are often put off until after City two is up. This is because it is hard to defend cities, Workers, and existing improvements against Raging Barbs. And after all, until a few Techs are known, there is nothing for a Worker to DO. But when you can build Cottages in woods, when Education is your 2nd invention, when your 0exp Warriors fight defensively in Forest like they were 3* 2 Move 17exp Warriors in Plains, hell, you start building Cottages in Forests as soon as you can.

6> Elven Workers: When Ljosalfar builds a Worker, it gets an Elven promotion. This means the Worker can enter a Forest and start construction of a Cottage all in one turn. This means the non-Industrious Ljosalfar build Cottages in Forest almost as fast as an Industrious Civ would, if it could. Unlike an Industrious Civ, Ljosalfar does not have to pause in the Forest to build a Road. Not only can Ljosalfar build the most valuable improvement in the most valuable common tile, they even get a mini-Industrious trait bonus while doing so.

7> COTTAGES: After running this study, I am of the opinion that a good deal of work is needed on the tech tree. One huge problem is the ability to build Cottages so early on. Cottages are the most powerful of the improvements, in they add very much value to a tile, and that they can ultimately be built on most land tiles. Note I list Cottages as a generic problem. I suspect the game as a whole would benefit from making Cottages a later invention.

8> FOREST COTTAGES: I have no confidence that this concept can be made to work in FfH, except, perhaps, as a late-game invention somewhat on a par with Lumbermills. The problems are, lack of tradeoffs, spamming, and benefits disproportionate to similar game functions.

a. Tradeoffs: One tile produces moderate Hammers, excellent Commerce, and almost always feeds itself, sometimes more. The player never has to weigh priorities.

b. Spamming: Put a Cottage in every Forest, as fast as possible. End of planning process

c. Disproportionate benefits: It is fine for a special resource tile to generate substantial Food, Hammers, and Commerce in one tile. It is not proportionate to be able to do this in virtually any tile. It is doubly disproportionate to be able to build this improvement in the same 6 turns it takes to move into and build a Cottage on Grassland.

9> GOD KING: The impact of Cottages was expected. But the impact of God King was not. Flood Plains are great for a Capitol, because they can be cottaged-up early. When the Capitol reaches its Happycap, excess Food is not needed, except, 3 Food per tile is still good for production when building Settlers and Workers. Plus the Cottage Commerce is collected all that time, of course. But Ljosalfar is not working 3 Food tiles, they are working a Forested Cottages. The Hammers from the Forest gain +50% from God King. But God King does not help other realms build these units (to the same extent) because other realm Capitols cannot afford to work 4 Forest tiles with no Commerce. In the test game, Ljosalfar was producing a Settler in 13 turns. Other realms took 26 turns. Ljosalfar built a Settler and a Hero (Gilden Silveric) in the same time (13 turns each) in the same time other Civs built just a Settler. After Ancient Forests crop up, this improves even more for Ljosalfar. God King comes into play around turn 80, and after that growth skyrockets.

[Edit: Addendum to God King. Ljosalfar was able/compelled to steal a march on the competitors in another way. Not much has been said about the way religion spread through the realm. On average, one city picked up the religon by random spread. All others were via Zaealot/Thane/Disciple.

Three realms used their "free" religious unit to move to the Capitol and spread the word ASAP. I have not stressed how critical this moment is. This is when the initial national Happycap backlog breaks. So ... Ljosalfar found their size 4 capitol under God King so valuable for this sort of W/S production, plus, the advanced condition of their City 2, resulted in Ljo using their free unit to spread the word to City 2. No other realm would probably be able to let a non-Capitol grow up first, But in this case City 2 was also a good hammer producer. City 2 and City 4 (the Holy City) built half-priced Temples, where every other civ had to reply upon a single Temple. Don't think for a minute Cultural/Spiritual is not a powerful builder civ. Then both produced Disciples to spread the word. Here is another speed in benefit of Ljo, at least their builder version ... the speed at which the religion moves through the land. And this effect, too, is related to God King.

10> TRAIL TO LEAVES: Not only is Fellowship of Leaves the ‘Natural’ religion for Ljosalfar, the way there contains Hunting. So as Ljosalfar works diligently to Way of the Leaves, they get as a by-product the ability to build tier-2 units.

11> ARCHERY: Only the Elves can afford to interrupt economic development to get an extra tier-2 Tech at this point. Since inventing it brings a Hero, invent it they usually do.

12> ANCIENT FORESTS: And for desert, Forest tiles that improve themselves. Not strictly a Ljosalfar issue, it becomes one when combined with Forest Cottages. The extra Food / Ancient Forest is not trivial. It can turn a 1 Food/turn backwater town into another city growing at a normal rate. But usually, it means turning a normal-grower into a fast-grower, or helps churn out a quick Worker. The cherry on top of desert is their ability to spawn a free tier-4 strength unit at times. Now, Ancients are not a huge factor in the early game, but they do act to make bad dynamics worse.

OTHER THAN THAT, GOT ANY COMPLAINTS?

Damn straight. In addition to unanticipated synergies and obscure mathematical implications, there is one flat out mistake in Ljosalfar. Their tech tree costs are unlike the other civs. Khazad, Lanun, Clan of Embers each pay 337, 540, and 1080 beakers for Tier-1, -2, and –3 Tech respectively. Ljosalfar pays 315, 504, and1008 beakers…why?

(My guess is typo+tired proofreader eyes.)

RECCOMENDATIONS

I don’t have answers for all of the issues. And what suggestions I have are still in the prototype stage. Frankly, I have spent 10 times as much time on busywork as I have thinking on the data. But there are a few examples of low-hanging fruit I think we can knck off right away. These are listed in the order from easiest/most likely/soonest to forget about it Unser.

1> Fix the Tech costs

2> Start with 1 Warrior 1 Scout or even 2 Warriors. There is no reason for Ljosalfar to start with two Scouts, there is ample reason for them not to.

3> Two Words: Gilden Silveric

4> Return the 1STR 2 Move unit to Raging Barbarians. I hated Barbarian Scouts in Vanilla Civ. 2 Move means the unit can enter a tile and pillage it the same turn. This makes defense require more units, maneuver outside cites, and makes Workers very hard to protect. This is needed for all Civs in general, IMO, not just Ljosalfar.

5> Different starting Tech for Ljosalfar. They are hampered less than usual by the lack of 2 Move Warriors thanks to their speed in Forest, so they probably could survive the early loss. Crafting, perhaps, or Agriculture, but under no circumstances Arcane Chants. Perhaps combine this with #6?

6> Remove Elven promotion, Replace with Elven Tech, a la Seafaring. This corrects unintended situations such as the Infernals being able to build Elven Cottages with captured Elven Workers. The premise of Elven Cottages is that Elves live in them. A creature born to rend and destroy is not going to live in harmony with its surroundings just because someone passed along the deed to some hut shaped like a mushroom. (I hear this might cause Tech trading problems? Dunno, throwing it out there.)

7> God King: Forbid Ljosalfar from this Civic. I have hopes for this concept. It by itself might go a long way to “fixing” things. Also nice is that is it transitory … it addresses the problem in the early game, but goes away before the endgame. And really, are not all Elves haugh…accomplished creatures themselves, every one of them of phenomenally long life? Seems like these would be the last People to buy into a God-King. Say, who’s our King? It’s Jim – Why, he’s been alive a thousand years! Oh yeah Jim, I used to date his babysitter. Little brat, used to sip my wine when I was working on her bra. But it’s good to see some young blood in office. So if you ask me from a role-playing perspective this one is right up ElfAlley.

8> Remove Elven Cottages. Just bite the bullet and do it. Do it now when it’s easy. Look, Elves are supposed to revere the wilderness. The WILDerness. The Elf cities were special and hidden. You walked into a forest and you saw a forest. Not the east coast of the USA but with more trees. But in the interim let’s keep Elven everything else, Farms, plantations and the like. Combined with #7, we might have a winner. At the very worse, it acts like a half-way change. Elves can build Farms in forests, and get 3 commerce from them under Aristocracy. Then we are back to Food+Hammers+3Commerce / tile, which is not so differentfrom the status quo. So that provides an incentive for them to deveop and settle into Aristocracy In the end this might be a problem too, because Aristocracy does not take much longer to reach than God King. But at least Farms do not improve over time like Cottages. Aristocracy eventually grows obsolete, and the phenomenon worksonly when Arisotcracy is up. So I say eliminate the big problem (Cottages) and lets see if the other potential problems are or are not. Thus some “flavor” can be retained if we’re lucky.

9> Ancient Forests: Two changes, first, no improvements can be in an AF, nor can one grow in a forest tile with and existing improvement. Yes, this incudes Roads. Ancient Forest sould be sacred, eh? And second, Ancient Forest should be giving out an extra Commerce. What is Commerce used to create? Technology. How do you get Technology? Thinking. What’s another word for thinking? Meditation. Where do tree-huggers meditate? Ancient Forests. Give AF a +1 Commerce, and that’s a sweet enough desert for me.

10> Truely a proto-idea; I am tossing it out there in a proofreading edit. In addition to GK restrictions, Ljo shuld be barred from some of the anti-warweariness Civics. Thus can we create a situation, I hope, where Ljo becomes vulnerable to major headaches from unwanted wars. The most challenging/interesting game I played involved massive war weariness being countered by the Guardian of Nature civic (and everything else I could scrape up.) Anyway, I see possibilities here to transform the Ljo from mutant mega-producer into something fun and full of flava.

11> Cottages: This is beyond the scope of this study but I want it in this list. I am of strong hunch that Cottages and some other techs, perhaps, should be moved back in the Tech Tree. I see Cottages as a big flaw as is. Now for something completely different.

THE END

Game and note files have been atached to the cooresponding Test Results page. This replay page was intended as a collection point for allfiles. It iwll no go unused.


Last Reserved Page

Go nuts.

I had been looking forward to reading this. I cant wait to see your analysis.

La-nun, La-nun, say it with me, La-nun... they are not from the moon.

Is there a reason you're focusing on the Clan of Embers, the Lanun, the Khazad and the Ljosalfar?

Deathling said:

Is there a reason you're focusing on the Clan of Embers, the Lanun, the Khazad and the Ljosalfar?

Click to expand...

Probably because they have the characteristics that may or may not alter the play of the early game.

Clan - Barb
Lanun - +1 Food & a Financial Leader
Khazad - Dwarven Vaults
Ljosalfar - Elves... and trees

How about Hippus and their raiders?

Sureshot said:

La-nun, La-nun, say it with me, La-nun... they are not from the moon.

Click to expand...

Thank you. I was beginning to think that they had something to do with the lunar seas, what with all this Lunan stuff. A quick check of the Civlopedia proved differently.

Good luck on those writeups, Unser. You're probably far enough into it now that just restarting it all on .15 would be impractical, but hopefully that means that you don't have to forgo playing .15 for too much longer. Courage!

Deathling said:

How about Hippus and their raiders?

Click to expand...

The raider is a military unit that can be easily leveraged into early game advantage... but it just makes it easier to make early war, and already semi-viable strategy. I think what he's trying to do here is look at builder starts, which the Hippus shouldn't have IMO.

A test with Lanun that has only 1 sea resource and 7 sea tiles used?

i must say your strategy for Lanun seems unbelievably lacking. By turn 200 if thats the situation i had id consider it a grand failure and learn from it lol

wheres the kurioates?
i really wanted a comparison of that sprawling trait