My newest milk run

[Elear]

The 'evil sumerian horde' was everywhere! They had cavalry, so if I hid away (like I would do on a beachhead), they would overrun all my Russian land towns! There's 3 places they are came from and I only have 20 cannon/trebuchet. That's no good. I lose cavalry much too fast.

The bottom line is that unlike all the Sid wins you see in the HoF, my only two attack options that would have produced many armies (Sumeria, Babylon) were too advanced for the first enemy. Sort of an odd rut to be stuck in

Look at this, for example:

FIRST WAR LEADERS

Pleb 64k - 13 (10 BC to 600 AD, v. Mongolia)
Moonsinger 81k - 11 (420 to 580 AD, v. Sumeria)
Kuningas 81k - 3 (370 to 470 AD, v. Russia)
Moonsinger 88k - 3 (70 AD to 200 AD, v. Persia)
Elear XXX - 0 (520 to 700 AD, v. Russia)

Ouch! And consider my war was started later than anyone else on that list. You really need a couple armies to get going anywhere (covering your stack, pillaging their resource) In addition, stopping my research enabled me to never get any artilleries. A beeline to replacable parts would have done something, at least. I needed railroads too.

In all the other games, they got railroads quickly to enable the troop floods, production and food bonuses, and support.

I still think it was a rather poor position I was in, despite everything.

This game was start #10.

I went through a few more starts this afternoon. Start #13 looked promising, but at 1000 bc, despite nearly being at Moonsinger's 81k score, having 26 cities, 5 cattle start, etc. I didn't have enough gold and 0 contacts. So I had to give that one up. I was in a strategically bad position too. I've started considering strategically too now. The awkwardness of where I got myself in on #10 was part of what did me in!

-Elear

 

[Elear]

There is... one other thing.

I have noticed a phenomenon on Sid level.

I call it 'Hyperexpansion Effect'. I've given it a lot of thought.

If you play any other level, you can expand expand expand, take over the world, etc. The more cities you get the faster, the better. You get a huge empire, blah blah blah.

However I have developed a theory

Score is all relative on territory and population. Obviously, the domination limit can be reached at similar times. But the actual domination happens at a later date. This actual domination is the point where you can have any milking land you want... the ultima population point. This is a high part of the score.

Let's look into this.

Why, for example, is Pleb's score 64k, and Moonsinger's 88k? etc.?

The answer is quite simple. The ultima population point is reached at 1700 AD, even though after the conquest of Mongolia, Arabia, America, and Zulu he reaches the domination limit, about 1350 AD. 1350 AD his score is nearly 13k.

Moonsinger 88k. At 1350 AD, she has nearly 37k! What happened here!

Well, for one, she went to war at a similar time with Pleb, yes. But Moonsinger's actual domination was about 1150, when she conquered the fertile Babylonian lands. 800 AD was her domination limit reached. Her score is nearly 12k at her domination limit point. This is huge over Pleb's score. Enough that it makes a 25 thousand point difference in final score.

It's all about achieving ultima population. And if your people are happy whilest you do it, great!

Another difference is the luxuries Pleb v. Moonsinger controlled. In 88k game, she had traded for them all by the time she started to war. Pleb had not had them all by any stretch.

In addition, the cleared jungles Moonsinger had gave great population and terrain bonuses!

So how does this all relate back to the 'Hyperexpansion Effect'? Simple. Moonsinger had 35 cities in 88k, versus over 60 on her 81k game. Any graph will show a significantly lower score. She was able to focus on a strong (enough) early military. True, Pleb had a similar amount of cities, perhaps even a few less, but he needed iron as well. Pleb took 600 years for his first war (he was still building, really) and thus he was in Monarchy. On the other hand where Pleb declared war on Arabia (#2), Moonsinger had just eliminated four civilizations (Germany, Persia, Greece, Byzantine) in two directions *including* the biggest AI. She was one turn away from starting a rampage on Sumeria, the other big AI. After that, the domination limit was not far from her grasp. Pleb still had a killer AI to deal with. In addition, Pleb used many resources to raze and replace. Moonsinger took a huge percent more cities than Pleb.

This is not entirely relative however. Pleb had a number of other factors. So let's take a look at Kuningas and Moonsinger 81k.

In both these games, they exceeded 60 cities by 1 AD, and I believe 70 by their first war. Hyper-expansion. They started war both at about 400 AD. This later warring cost them many points, I'm afraid. The only reason they were within 7k points is a good expansion, but that doesn't nearly make up for what they lost. This is why the second Moonsinger game was able to beat both the games. 7 thousand points is a lot.

Another thing is that the earlier you do Sid war, the less infantry you face. This is why early and fast warfare makes up for slower expansion.

My conclusion? It's better to have a somewhat smaller empire and war with knights, early on, on Sid level, and avoid hyperexpansion. Get your war machine going. You will get an output of more points in the long run if you do it right. If you start war early and take too long (Pleb) you will lose everything however. Those first 50 war turns are extremely key. Part of this relates to how fast you can reach and destroy the first enemies core. But that's an entirely different story.

Morale of the story: I would be EXTREMELY surprised to see any score beat Moonsinger's 88k game that got very rapid expansion resulting in large 65+ city empires at 1 AD like those of the 81k games we see. Such a hyper-expansion would not be able to make up the points needed to exceed 88k, and possibly 90k.

P.S. - A little bit more expansion than was in Moonsinger's 88k game, would not hurt. As long as it isn't hyperexpansion. Also, a higher domination limit.. wouldn't hurt

-Elear

 

[Elear]

Oh and one more thing I forgot.

Because of early war, you get more armies earlier. I was talking earlier about armies being key. They are, and the more you get quickly, the higher your score.

It's as simple as when the apple goes up it must come down.

Obviously Pleb has written a Maximizing Score (milking) article. In this it describes what makes up score. I'm not writing about this. I'm writing about a general set of observations that indirectly (but obviously) ties into the score formula.

 

[bluejay]

Elear: what aggression level are you using? You mentioned everything else in your introductory post. And are you sticking with the same AI civs?

 

[Elear]

Least aggressive AIs.

And I think I will stay will same civs, at least until I'm out of possibilities with my current mapfinder set. This last game was my 10th of 88 possible ones I elected to look at with 4300+ domination limit. I'm at number 37 right now, still looking

 

[nerovats]

My conclusion? It's better to have a somewhat smaller empire and war with knights, early on, on Sid level, and avoid hyperexpansion. Get your war machine going. You will get an output of more points in the long run if you do it right.

Since you can't out-expand the ai at higer levels don't try. Let them build your cities and focus on a strong military. The iro's are very effective for this. Just get 6 cities which will give 10 or even better 15 spt in golden age and after your GA is over you will have more cities then when you would have expanded the peacefull way.

 

[Elear]

Since you can't out-expand the ai at higer levels don't try.

You might be surprised. It's not usually that hard to be one of the biggest, if not the biggest civ period, in the game, even on Sid. If you get a good start like this game, there's no problem. Persia had 45 when I had 66, I believe.

Since the AI doesn't understand prebuilding road spirals or thinking 'ahead', I find their starting settlers and production balances out nicely with my REX and human intelligence (i.e. - settler/worker factories and planning)

However I agree with you on the rest of your point...

Just get 6 cities which will give 10 or even better 15 spt in golden age and after your GA is over you will have more cities then when you would have expanded the peacefull way.

I have many purposes outside my 10 core cities:

--Commerce
--Unit Support
--Boat building
--Harbors for trade routes
--Workers
--Settlers
--Artillery making

So I value a larger empire in general for an easier game, but thus, as I was saying in my article, the idea of expansion through military and a smaller initial growth period and output, is more rapid output.

The difficulty with 'going small' is often what you encounter. Kuningas and Moonsinger 81k had all the resource. Moonsinger 88k (the small, non hyper-expansion) had none as well as jungle/marsh galore. Yet there we have a high score. I think it is fascinating to study from these examples we are presented. Thus, from it, we can learn and grow ourselves, even if the time of these games has long past.

 

[AutomatedTeller]

Part of this depends on the map. On a pangea, the AI will out expand you, i think - I'm not sure there is anything to be done about it. On an arch, it gets slowed down by not being very bright.

 

[Elear]

On a Pangaea, you will get a slightly different result of course. It's quite unpredictable.

Ironically, I consider Archipelago a better breeding ground for KAI.

The reason is simple. You get a full spread on archipelago. One of the reasons it is so fun

I see this usually on archipelago (the full spread):

1) KillerAI
2) VeryStrongAI
3) StrongAI#1
4) StrongAI#2
5) SomewhatStrongAI#1
6) SomewhatStrongAI#2
7) Weaker AI#1
8) Weak AI#2

KillerAI has lots of expansion room, and usually a sub-optimal or average neighbor to crush.

5 to 8 usually get a bit 'crushed' by their stronger 4 neighbors expansions, or stranded on a tiny island.

I find it helpful to attack a somewhat strong class first, followed by a strong, after which one must be prepared to take on the killer AI or at the very least, the #2. However, more success is found from the Killer AI being taken down third, as stated thus previously.

Something that I've probably said before too, is that Moonsinger's 81k game was so unusual. Unlike any other game, I can't classify it the same way.

To sum it up, it did not have a clear killer AI develop, nor an extremely weak civ. Korea and Sumeria were somewhat weaker though, and Byzantine and Russia 'strong' class civs. I'm investigating this... at first glance, it's purely landform. Maya had a big fertile island all to herself. The AIs were paired up well. The biggest landform, 3 AI split extremely evenly. Byzantine, pretty good size. Greece/Babylon, split near half and half on another island. Sumeria and Korea shared another island fairly even too. It wasn't that huge of an island, similar size to Maya isle. Rather than scattered islands for settlements, the landforms were slightly bigger (less offshore islands than I often see).

Weird, but it could very well be simply landform and terrain types that cause this.

-Elear

 

[TheLoneMan]

Nice try, I'm definitely looking forward to your next run.

Your banking was amazing, the amount of gold you had flowing was almost humorous.

 

[Elear]

It's not my banking really

Nothing many here haven't done.

Due to Babylon's extreme richness, I was able to do it to an extreme. In fact, I pushed it almost so far they went from 'strong' to 'superpower'. I forgot the second part of the banking, the beneficiaries .