Vel's Strategy Thread, Volume II

[binTravkin]

Someone should sticky this.

NOW!

[Velociryx]

And maybe un-sticky the one that's gone beyond 500 posts?

-=Vel=-

[vmxa1]

Well why not make a thread with links to the I and II and the workshop stuff.

Then they could unsticky I and sticky the index thread. Now you do not have to deal with a moderator as things change and threads will still be easy to find.

[Hermann the Lombard]

No posts in five days?? Horrors!

[Velociryx]

Been making up for it via the Devel's Workshop series tho...

-=Vel=-

[Hermann the Lombard]

True, true, true...but those are specific and this is more general. OTOH those are practical cases and this is more theoretical.

[Helghallen]

Sorry this is a little late in all this. The Settler first discussion got me to thinking since I'm a SMAC baby. I like the idea. Spread like a plague. I tried it. The use of food for both Settlers and workers production seem to slow production. I have decided a warrior first to get a city to pop 2 is more efficient. The loss of pop growth really bothered me and having the city at pop 1 for so long seems to be a definite bad idea.

[Velociryx]

SMAC.

Remember tho, pop 2 is ONLY more efficient if you can work a special that provides a big enough boost to speed the settler production along. If all you have are "regular" tiles or floodplains, then you're not doing yourself any favors by delaying.

-=Vel=-

[Blake]

I've started to think that it's worth growing first if you have "4fop" tiles to work, specifically floodplains (3-0-1), oasis (3-0-2), resources in forests (might be +1 commerce or more rarely +1 food). I don't consider commerce-heavy tiles worth growing for (ie lakes and comm resource on river), unless going for Religion.

Growing to size 2 and working a second floodplain (or comm resource forest) gives about +10% research and uh, +20% worker/settler train rate.

Assuming it's balanced, it makes sense that 3fop tiles (ie forests) are neutral or slightly negative, while 4fop tiles are good or slightly positive. 5fop tiles (Oasis) are probably definitely good to grow for, but I've only ever seen one double-oasis start.

* fop = "Factor of Production", food, hammer, commerce.

[Helghallen]

Originally posted by Velociryx
SMAC.

Remember tho, pop 2 is ONLY more efficient if you can work a special that provides a big enough boost to speed the settler production along. If all you have are "regular" tiles or floodplains, then you're not doing yourself any favors by delaying.

-=Vel=-

That makes good sense....dangit, more crap to think about while I play

[Velociryx]

Where the growth vs. settler argument is concerned, I think that commerce values of tiles are incedental. I mean, sure, if there's a river tile vs. a non-river tile, the obvious choice is to work the river tile for commerce, but this is seldom a factor (ie - it is seldom the case that you have two food specials--say, two corn or wheat--one on the river, and one not. So if growth is what you're after, then you'll pretty much always be working the corn, whether it's got commerce attached to it or not.

And if you're trying to figure which is more efficient turn-wise (settler first or grow first), then that particular consideration only "cares about" food and hammers (since both are used in the production of the settler).

All that to say that when making those kinds of evaluations, I find that I'm making the decision along lines similar to Blake's, 'cept that I'm looking at commerce only incedentally in those cases, and am more interested in answering the following multipart question:

Given the techs I have, or will have, in the timeframe of the settler/growth to size 2, what is the value of tile(s) (food + hammers) I will be working? (ie - I can't count the mined value of a hill if I don't have, or am not researching mining in that timeframe, and I can't count the farmed value of a wheat field or other such special if I don't have, and am not researching agriculture in that timeframe)

If I have the terrain, and I have (or will have) the techs to support the improvement of that terrain, then worker first becomes a no-brainer, because the boost in tile productivity will offset the turns NOT building a settler, so the answer is "worker first."

If I have a seafood special and am coastal, then the answer is fishing + warrior first for growth, or, if I already have fishing, then the answer workboat first.

If I don't have the techs (& am not researching them out the gate), or if the land won't support rapid growth, post-improvement, then the answer is settler first.

(and me personally, if I have hunting, then unless I just have completely awesome terrain, the answer is almost always scout first, but your mileage may vary).

Anything else, and IMO, you're not playing to your maximum advantage.

-=Vel=-

[Helghallen]

As always. Insightful and helpful. Thanks for the detailed thoughts.

[paul57]

Those initial decisions have a huge impact on the future game. Vel, from reading your extensive contributions, I have no doubt that you are in no way suggesting that there is only one right way to start each game. Or even suggesting that the starts you specify above are essential to acheiving the highest possible score in a particular game. However, I find this kind of advise extremely useful to my enjoyment of the game.

Paraphrasing, my understanding is you recommend making the first build of the first city:[list=1][*]Worker, given the terrain & techs to improve it.[*]Workboat, given coastal resource, coastal city & fishing.[*]Warrior (followed by Workboat after researching fishing), given coastal resource & coastal city but not fishing.[*]Scout, given hunting but none of the above applied.[*]Settler, given no other way to gaining an advantage.[/list=1]

Obviously this is based on a fundamental assumption that there is insufficient risk of loosing that city to an opponent (barbarian or otherwise). Interestingly

• Warrior always seems to be "recommended" as the first build (presumably the AI opponents choose it first).
• Strong advise is given by experienced players to always "escorted/protect" settlers all the way to new city sites.

Therefore, do you agree that this assumption is a fundamental way that human players gain an immediate advantage over the AI opponents?

Given my outdated coding background, I am intrigued by the possibility of experts and enthusiasts modifying (improving?) the AI once the SDK comes out. By the way, wasn't the SDK due in January?

Looking forward to your future contributions.

[Velociryx]

Hello Paul! I'd say your summation is spot-on. And you're right....Civ is far too complex to suggest an "always right" approach. The closest I think we can come that is to say that "if optimal growth is your highest priority" then under these conditions, the following are the optimal moves.

Of course, not every strategy will revolve around rapid growth first, so even here, we fall well short of the "always right" claim.

For example...let us say that you really want/need to get Stonehenge early, because your playing a philosophic civ, and an early prophet is pivotal to your plans.

In such an instance, it may make vastly more sense to go with a turn one commerce focus (work a tile that will give you 2-3 cpt, even if it slows/stalls your growth rate in the short term), so you can GET mysticism and get started on the henge early. In such a case, the advice above re: an early growth strategy would not only be less-than-ideal, it could hinder your chances at doing whatever it is you want to do.

Hammer-wise, let us say you're playing Mansa-Musa...it might be in your best interest (say, you find someone quite close to you) to forget growth, and focus exclusively on commerce to speed your way to bronze and archery (for chop and your uu), and then, continue to forget growth, focusing on hammer heavy tiles (forested hills) + chop (make one worker) to speed the production of 6-8 skirmishers. Again, in this case, the optimal growth oriented plays would see your attack force come to fruition several turns later than it otherwise might.

So no...no "always right" answers here....just some "always right under these specific conditions, and for this specific strategy" notes...

-=Vel=-

[Velociryx]

Forgot to answer the question at the bottom of your post. Yes, I think that a human's ability to make one unit fill several roles is instrumental to our ability to jumpstart a lead over the AI.

Specifically, you're right....the commonest advice is always to escort your settlers with warriors (or something), and yet, the "preferred build order" mentioned above sees the warrior built first only a minority of times.

This is because a human can and will effectively use his/her starting warrior (or scout, or whatever) as a guide for that early unit when necessary, thus, that single unit fills two roles....as a starting scout and as a vanguard to protect your investment (settler/worker outside the borders).

As far as I have seen, the AI doesn't do this, playing extremely conservatively where settlers are concerned (I often see them moving with archers, one tile a turn, in the most inefficient manner imaginable....that's cool tho, and our opportunity to gain advantage!

-=Vel=-