GOTM 16 Spoiler

Thanks Peaster - this post has a bunch of random thoughts

Sorry abor misattributing your EC guide - I was actually remembering solo's Early Landing guide - as I recall, that is where we reached agreement in afterplay that Marco's is worth more than was thought, perhaps a much as HG depending on timing. Your EL guide introduced the idea of sending out a task force to the most distant civ ASAP. I don't remember your focus on growth being stated as succinctly as now - I do remember Slow Thinker's game where he built a monster BC civ with Pyramids (heresy at that time!) and put half the theory back to square one! Everybody was afraid to mention it! You perhaps had the same idea independently - but now you certainly have integrated and expanded that insight. For a while we thought that maximizing SSC size in the BC years was a good plan, but you are proving that to be second best strategy.

I think the consensus now is that leaner civs are riskier but get the green star when they work out.

I am sorry you can't open Chofritz's save - as I recall, his 1000BC to 1AD play was spectacular and seemed to me at the time to maximize growth in a similar way to yours. I seem to dimly recall that he was not quite as masterful in the first 2000 years. But I may be mistaken, not for the first time...

I remember that for a long time Smash would build Colossus by hand in his second city, so things have definitely changed.

---

I am interested in roads - when do you connect your cities? I feel that it is a powerful CivII principle to never build roads before 2000 BC without a specific calculable payout. Is a better marker for beginning roadbuilding Trade? Or its precursor?

---

So your early game question is: How valuable is Monarchy in 2850BC instead of 2050 BC? How many cities at 200BC make up for only getting Monarchy a few turns earlier? It's hard to organize how to ask the question in a testable way. Perhaps going backwards from the next status point (Marco's) - some of my green star tries (mostly unsuccessful) got to Marco's too soon - I was not in a position to exploit the information. I found the optimum Marco's time to be about 1100BC for EC. With that in mind, perhaps early monarchy is not so special - it only increases the chance that one will spend beakers to learn a tech the AI will research anyway. So as long as the civ is in shape to get Marco's at 1100BC, maximizing cities is the optimum priority.
 
Sorry abor misattributing your EC guide - I was actually remembering solo's Early Landing guide ...
We should probably discuss EL/EC/GOTM separately. This GOTM16 test was only about GOTM score. I've never made a serious study of EL strategy, so I shouldn't say much about it. I'm vaguely aware of the SlowThinker EL game, and I've played a couple of EL GOTMs successfully, using my own improvised big-civ approach, rather than the ELG approach. IMO a strong "growth player" without many ELG skills will usually beat a average player who has memorized the ELG. But I can't say anything about 100AD Deity EL's... out of my league!

SlowThinker also played an impressive EC game at Poly, with Pyramids + STC + ship chain, which I studied a bit at the time, before I really understood trade. I am a bit less impressed with it now - the Poly guys allowed van rehoming (etc) in their comparison games, which can make trade super-powerful, especially combined with good luck wrt Hides supply/demand (which ST had in spades). I've experimented with trade in several EC/GOTMs since then, without rehoming, and believe that it doesn't usually lead to a great green star game. Combined with fast growth, it IS a good way to play for gold; in fact, that basically describes how I played GOTM 16.

Slowthinker had/has a flair for drama, and an absolute mastery of game mechanics, but unfortunately he didn't codify his ideas and didn't play a lot of games. He also had his share of less-publicized failures. So, it's rather hard to evaluate his theories.

For a while we thought that maximizing SSC size in the BC years was a good plan, but you are proving that to be second best strategy.
Hmmm...for what purpose? GOTM score? IMO Step 1 = enough cities (with Pyr+HG usually); Step 2 = good economy = boats + vans + destination (+LH usually). Step 3 = break the game in republic or democracy. I guess growing the STC is late Step 2. My STC was approx tied with Starlifter and Smash in the BC years in GOTM 16, but I don't always build an STC, or don't rely heavily on it.

I think the consensus now is that leaner civs are riskier but get the green star when they work out.
....
I don't remember your focus on growth being stated as succinctly as now
a) You don't think it's related to the map? Look at "The Gauntlet" (GOTM 58??), which was a very tough map for EC. I won green in approx 1000AD with a VERY big civ. I doubt a lean approach could beat that date, even with extremely good luck. Second fastest was approx 1500AD, IIRC, probably by landing.
...
b) I am not a strong advocate of big growth for EC games (especially on small maps). However, I always recommended that anyone reading my ECG should read about ICS first. Even the lean approach requires fast growth up to a point. Lately, I've played some GOTMs for "green and/or gold" and have included big growth in those.
I am sorry you can't open Chofritz's save - as I recall, his 1000BC to 1AD play was spectacular and seemed to me at the time to maximize growth in a similar way to yours. I seem to dimly recall that he was not quite as masterful in the first 2000 years. But I may be mistaken, not for the first time...
I'm a little surprised to hear so much about Chofritz now. Before I wrote the ECG in approx 2004+, I asked everyone I knew about any EC masters from the past [eg LaFayette?, Smash?, maybe DaveV?], and I don't recall hearing much about Chofritz. I note that he's won 2.5 green stars, but many other players have won many more. (?)

I am interested in roads - when do you connect your cities? I feel that it is a powerful CivII principle to never build roads before 2000 BC without a specific calculable payout. Is a better marker for beginning roadbuilding Trade? Or its precursor?

Good! (that you are interested) Maybe we should start a new thread on this. I don't have the final answer, but I generally build short 1-tile roads between my central cities as new setttlers arise from them. In 2000 BC, I might have about 6-7 cities and maybe 2-3 road tiles, usually on shielded grass tiles that the wokers will be using anyway. So, the road is worth approx 20g in extra arrows (based on my theory of interest), plus a little more for its defense value and for speedier growth later on. IMO the cost (2 settler turns) is about the same as the benefit, so the road is not a huge mistake or a huge boon in the 2000BC era. As traffic flow increases, a 1-tile road becomes a much better deal.

So your early game question is: How valuable is Monarchy in 2850BC instead of 2050 BC? How many cities at 200BC make up for only getting Monarchy a few turns earlier? It's hard to organize how to ask the question in a testable way. Perhaps going backwards from the next status point (Marco's) - some of my green star tries (mostly unsuccessful) got to Marco's too soon - I was not in a position to exploit the information. I found the optimum Marco's time to be about 1100BC for EC. With that in mind, perhaps early monarchy is not so special - it only increases the chance that one will spend beakers to learn a tech the AI will research anyway. So as long as the civ is in shape to get Marco's at 1100BC, maximizing cities is the optimum priority.

Monarchy might also deserve a new thread. The question cannot be easily answered mathematically - my prefered approach - but playtesting shouldn't be hard. For example, wait until both player A and B have had monarchy for a few turns, maybe 1800BC or so, and then ask which position is better. IIRC I tried this myself a few times, several years ago. I think an "extreme monarchy strategy" (eg the 3-arrow strategy, with the science bar always at 60) is clearly wrong. Probably "extreme growth" (eg ICS, with good tech choices ) is better than that, but some compromise is ideal. I generally focus on growth (size-one trick, taxes at 60, IRB ASAP, workers mainly on food/shield tiles, etc) until some critical point, usually at approx 4 cities, when my scientists are just starting monarchy. Then I choose the next reasonable oedo year and make that happen by placing a few workers on ocean, building a road, or letting a city remain at size two a bit longer, etc.

My rules of thumb are: Monarchy is late if multiple cities are already rioting (eg approx 7-8 cities at Deity). It is probably earlier than necessary if there are less than 5 cities around to benefit.

1100BC or a bit earlier sounds about right for Marco's for most EC games. I tend to delay it if I don't need map info urgently - for example, in my green/gold games, or on specially designed maps, when AI locations are predictable. I don't value it for tech quite as much as I did in 2004.
 
I just finished replaying GotM 16. Had a score of 4527 in 1823 (turn 243) by spaceship, with a GotM score of 349, which would have been 5th behind the mighty Lucky or 6th behind the mighty Pester. this is an improvement over my original score of 2272 in 1886 by conquest, GotM score of 166 for 12th.
I should say I used the fix to Multigold, that fixes the diplomacy, but you cant use find city. I was able to make an ally of Spain for most of the game and had reasonable relations with the others till I betrayed them.

Also, I was very familiar with the map so had some advantage, but I got everybodies maps so I dont think was a big advantage. I didnt know of map analysis when I originaly played it, so I had no idea where the huts were.

I decided to play without MarcoPolo, but got embassies and maps of all, not sure when but well before 1000 ad

In my original game by 1000 ad, I was just starting the assault on the Celts. In the replay, Zulu and the Celts have been emilated and I have a start on the Greeks, capturing Athens and Sparta.

I purposely didn't build Collossus, as I wanted to see if I could set up a trade system with the enemy city. Madrid built it and I allied with them. Had some good trades but they never built an aquaduct so the trades didnt reach there potential. Therfore, this strategy was a mistake.

On the OEDO year, after defeating all but the Sioux and Spain(1670) I switched to Democracy and started to concentrate on trades. While I wasnt getting pesters $20k per turn, I got more than enough to buy the spaceship parts for a size 4 ship and any improvements, I could possibly want. I have include an excel spreadsheet showing the trades in democracy if anyone is interested!?

My trading was going OK, but I was running out of good demanded items when I discovered Automobile. Then I RB a superhighway & an airport in my SSC and in Athens. Suddenly I had as much money as I could get vans going , rehoming the vans. I shortly thereafter RB superhighway & Airport to a city on island east of capital to give a choice of destinations. During this period I think I was dazzled by all the gold, cause I really reduced building of engineers, which probably cost 4th place.

I have attached a zipfile holding 4 save files and the excel trade file
 

Attachments

  • LVAKUNAreplayGotM16.zip
    71.5 KB · Views: 97
Good work, LV. You doubled your previous score, with a very respectable finish among a crowded field. Also, it's usually harder to get a high GOTM score by spaceship than by conquest.
 
We should probably discuss EL/EC/GOTM separately. This GOTM16 test was only about GOTM score. I've never made a serious study of EL strategy, so I shouldn't say much about it. I'm vaguely aware of the SlowThinker EL game, and I've played a couple of EL GOTMs successfully, using my own improvised big-civ approach, rather than the ELG approach. IMO a strong "growth player" without many ELG skills will usually beat a average player who has memorized the ELG. But I can't say anything about 100AD Deity EL's... out of my league!

Was browsing thru some old threads at Poly today. Apparently, the ELG approach [aka "classical EL" strategy, or just the "lean approach", or "Solo's approach"] grew out of a bunch of comparison games, maybe 6-7 years ago at Poly. Within a couple of years, Zenon, also a strong GOTM player here at CFC, showed that a "robust" approach is usually faster [this opinion is directly from a post by Solo]. I think "robust" essentially means "ICS". As far as I can tell, the SlowThinker game came later, and mainly introduced the Pyramids. So, IMO Zenon should get most of the credit for our current understanding of optimal EL strategy.

It would be nice to have a thread, or a War Academy paper, on the history of Civ2. I'd help, but I started only in 04, after the biggest events and inventions.
 
Top Bottom