Epic Six - Sullla's Game

[Sullla]

Having read through all of the commentary, I thought I'd throw out a couple final thoughts before we all move on from this. The controversy generated here over parking a single archer outside an AI capital was quite a surprise to me. It's something that I have done before in the past in previous Always War games (and seen others do in Always War games) without anyone batting an eyelash. Perhaps that's due to the context of this event - a tournament setting rather than a testing environment - or simply the sort of people we attract here, I don't know. I've been choked by humans so many times in MP, it never crossed my mind that sending units to shut down the AI in similar fashion would be viewed as exploitative. However, it's clear that a majority of our community frowns upon this sort of action, so I'll make sure it doesn't happen in future Always War events.

I am a bit disappointed that this one comparatively minor part of my report greatly overshadowed everything else that I wrote about. Not a single person has commented on the merits of attacking early and often versus waiting and teching to superior units, which I felt to be the most important strategic question of this event. A lot of people began their attacks much earlier, yet finished in the 1450-1500 period, suggesting that both paths were equally valid. I'd been hoping for at least a little discussion in that area...

It's understandable that this incident generated a lot of buzz, since I tend to be one of the more vocal members of this community. As Sirian pointed out, I'm also very much on record as being critical of others for various sandbagging efforts in past games. That may well have contributed to some of the comments on report day, in a "what comes around, goes around" fashion. As long as we can all remain civil to one another, which fortunately was the case here, I don't have any problem with others asking as many questions as they see fit. It would be nice not to have quite so many people immediately calling me out for an "exploit" which had never been discussed prior to this game, however.

It was also interesting that several other players immediately beelined units into India's territory, even earlier than I did, yet I saw no one accusing them of exploiting the AI. Sometimes writing extremely detailed reports can come back to bite you in the ass, it seems.

I hope this one game didn't draw too much attention away from the many excellent efforts turned in by our other players. They've been a pleasure to read, even if I didn't have the time to reply to each one individually.

[uberfish]

I am a bit disappointed that this one comparatively minor part of my report greatly overshadowed everything else that I wrote about. Not a single person has commented on the merits of attacking early and often versus waiting and teching to superior units, which I felt to be the most important strategic question of this event. A lot of people began their attacks much earlier, yet finished in the 1450-1500 period, suggesting that both paths were equally valid. I'd been hoping for at least a little discussion in that area...

Well I'm done arguing the CS issue here in deference to Sirian as the event host, since he has clearly made his mind up anyway.

So on to Sulla's question, I tried a mixture of both early and late attacks and lost most of my Iron level units to bad dice rolls which stopped my momentum completely until I teched up. I think both work, but it's better to commit more heavily one way or the other. The lesson I learned was that in AW, when you go on the offensive, you need to bring a greater mass of units to protect yourself against bad RNG outcomes so they have enough to do the job even if you lose a couple 90% fights.

[Sirian]

Well I'm done arguing the CS issue here in deference to Sirian as the event host, since he has clearly made his mind up anyway.

If you have a compelling argument to make, make it, skipping the inferences about your estimation of my openmindedness.

None of your points so far have persuaded me, but if you have more you have not brought out, they should be on the table for consideration.


- Sirian

[Zeviz]

...
None of your points so far have persuaded me, but if you have more you have not brought out, they should be on the table for consideration.
...

Since the discussion is still open, here is the post I was going to make before it was eaten by the forum earlier this morning:

...
No gambit should be so valuable as to force competitors to chase it if they want any shot at winning an event. I see this as no different than letting a 3200BC Warrior Rush rule an event.
...

I agree with this statement. However, I don't think that it applies to Oracle->CS slingshot. Let's take a look at how Oracle was used by "winners" of RB events so far. (I took fastest finishers as "winners" in unscored events. I also haven't looked at Gentle adventures and events whose results haven't been tabulated yet.)

Oracle->Metal Casting
Adv. 2
Adv. 3

Oracle->Monarchy
Epic 3

no Oracle
Adv. 4
Epic 2
Epic 4
Epic 5
Epic 6

unknown
Adv. 5

CS slingshot won 0 events, while Metal Casting slingshot won 2. So if anything, we should be banning the Oracle->Metal Casting route.

This slingshot can give a significant advantage, but so can other approaches. The reason so many people used it in this event is because it was the right choice for builders in this situation. This is just another example of the idea that no choice is perfect for all situations, but there is a perfect choice in any given situation.

The cost of going for CS slingshot isn't the risk of missing the Oracle. The costs are:
- Your first GP is very likely to be a prophet, so no early academy for you.
- You will get Currency much later, and with foreign trade Currency can give you +2 cpt per city, and resource-for-gpt deals, and ability to get cash in deals with AIs. Under normal circumstances, these 3 things combine to give much stronger economic benefit than Beurocracy would.
- You will get Construction much later, so no early Catapults for you.
- On water maps, you are likely to miss out on Colossus-Great Lighthouse combo. (Which was the winning strategy in a couple of events.)

I could go on, but I think that's enough to show that CS slingshot is far from being "The One Right Choice".

PS For a slightly deeper analysis, there are 3 paths through the ancient/classical tech tree:
1. The top path emphasizes economy and is good for builders. (Alphabet, Currency, CoL, and CS)
2. The bottom path emphasizes military and production and is good for fighters. (BW, IW, Metal Casting, Machinery)
3. The middle path emphasizes religion and is initially weaker than the other two. This path misses Alphabet, Currency, and Math, but it also misses Iron Working and Metal Casting. In fact, CS is the first really good thing you get on this path, so CS slingshot is the main thing that makes the middle path viable.

[Rik Meleet]

The "attack early and often" Vs "attack when powerful" difference is too difficult and too subtle for part-time players (like me) to analyse and to draw conclusions from. An easy to see (and announced in the intro of your report) tactic as Archer-parking is something I can immediately see the consequences off.
All 3 tactics have their advantages and their opportunity cost and their real cost. Even after reading (virtualy) all reports I still have no clue whether attacking early or attacing late is more efficient. I do now know that there are ways to lock an AI in. It doesn't fit my play-style so I won't use it unless I really really need it (although it'll probably be too late then). If anyone wants to use it - their choice, not mine.

I did notice that you had more units in your land than I had. I asked (in post 7) whether that had to do with our differences in scouting. I'd still like to know your (and other people's) opinion on that. That'll teach me (and perhaps others) whether it is better in AW to stay close to home to delay contact(s) or to scout the whole earth when it is still possible for the quickest contact(s).

[Sirian]

I am a bit disappointed that this one comparatively minor part of my report greatly overshadowed everything else that I wrote about.

I'm sure that players who took heat over pursuing score in a scored event felt similarly during Epic Two. That is not to discount your disappointment at all. I've been there, including deciding not to finish my Epic Thirty-Six report after catching too much flak over the scenario design from people who didn't even play the event. My own Epic Six report is sitting parked in limbo right now, because the back and forth that broke out here is compelling my attention.

The thing is, this isn't about your choke move any more. Sometimes problems and tensions build over time and erupt at a flash point. I had hoped that with Civ4 and my level of patience and restraint and caution in regard to this area of rule-making that it would be easier and less controversial, less adversarial than it was in Civ3, less burdensome on me personally, but that is now coming in to question. Just the mention of needing to clean up some of the most obvious problems has brought out strong objections. There are those arguing against all boundaries as futile -- a logic fatal to our cause if embraced. We have diametrically opposing views between those who want absolute clarity and precision from central authority and those who want absolute freedom and self-responsibility from decentralized authority. The only people who are speaking up are those who have an objection to this or that, so that the dialogue moves from one dispute to another, centering only on negatives, heating up along the way.


I was hoping to see at least the initial stirrings of consensus. No sign of that on the horizon at the moment. All of our active voices here have been relatively recent arrivals. The near total silence from our oldest members, the ones who've been through the tribulations of the Civ3-rule-making, is not what I was expecting. Only those who contribute to the process will be heard, people. Please keep that in mind as I work to move us forward.


- Sirian

[Zeviz]

It seems that in this case early attack gave similar results to late attack and constant attack. In my game an early offensive push choked after taking 2 cities, but I was able to gradually build up and start a constant attack that lasted from 8th to 16th century, giving me similar finish date to those who did no early attacking at all.

[Kronoz]

I could go on, but I think that's enough to show that CS slingshot is far from being "The One Right Choice".

I agree with you, Zeviz, I dont play the CS-slingshot for the same reasons you mention.


Appreciate your reports, Sullla, as always. And I appreciate very much that your webside slips through the firewall at work, have killed quite a few hours on your reports

[T-hawk]

the merits of attacking early and often versus waiting and teching to superior units, which I felt to be the most important strategic question of this event. A lot of people began their attacks much earlier, yet finished in the 1450-1500 period, suggesting that both paths were equally valid.

OK, here's some thoughts on that. Attacking vs economy was not mutually exclusive at all, thanks to the map. Both the capital and floodplains-copper site had tons of flood plains (with rivers). Those sites really can't develop into strong hammer producers, but it's a no-brainer to cottage them. And two large, well-cottaged cities really are enough to support a solid research economy, especially in Bureaucracy for those who got there quickly. So anyone who tried at all to develop their economy found it running solidly. I don't recall any reports with serious economic difficulty against the AIs, save Sooooo who ignored it entirely.

The next three to five cities would be your military producers. Starting later, they couldn't build economic buildings and develop cottages fast enough to matter, but for military it's just build a barracks and go.They'd take about the same time to acquire whether you built them yourself or conquered them, thus factoring out the variable of when you started attacking.

So almost regardless of your game plan, you had guns and you had butter, and your sources of each couldn't be readily converted to the other. You'd have to go to Sooooo's extreme game plan to get a result substantially different from the 1500 AD timeframe.

My point is this: it all comes from the map configuration. In the aftermath of Epic Five, Sirian outlined how he intentionally set up the map to be bereft of good commerce sites, and correctly expected a lot of crashed economies as a result. But this game proved that the presence of even just a couple good commerce sites (and flood plains + river is as good as they come) will naturally lead to strong economies all around.

[Zeviz]

...
I was hoping to see at least the initial stirrings of consensus. No sign of that on the horizon at the moment. All of our active voices here have been relatively recent arrivals. The near total silence from our oldest members, the ones who've been through the tribulations of the Civ3-rule-making, is not what I was expecting. Only those who contribute to the process will be heard, people. Please keep that in mind as I work to move us forward.
...

I think the consensus is that RB is a great place and we'll be happy to be here whatever the outcome of these discussions is.

Whatever the conclusion of the arguments about rules or no rules, CS slingshot or no CS slingshot, archer park or no archer park, we are going to continue enjoying these events.

PS And just to clarify, the reason I was asking for exact rules is that I was hoping they would reduce the number of such arguments. However, if your experience has shown otherwise, exact rules might not be a good idea.