Zsozso's attempt at fastest transcend on huge map - Page 3

zsozso · February 10, 2003, 07:00

Quote:

Originally posted by jimmytrick
Can the Byron style beat the traditional style without a jungle start?

I'm quite certain it can. I already played it on the Ultimate Builder Challenge scenario - which gives as bad a starting position as you can get without special scenario editing to make it even worse. On that challenge it also performed admirably, much better than the traditional style I also tried there.

Quote:

Of course someone can manufacture an even more friendly map and then what.......

Well, I did not make an ideal map by scenario editor. In fact, if you want to set any kind of standard benchmark, I think the most suitable map for such thing is the only map provided with the game itself - which is what I used.

I certainly picked to most favourable circumstances one can get in a "standard" game. But I think that is part of this challenge, i.e. to know what is the best to play from. You could argue that it is better to start in the uranium flats or the Mount planet for the bonus energy. With a traditional play style it might actually be worth picking the uranium flats.

Quote:

[SIZE=1]Very impressed with your military operations.

Thanks. However, the last one was practically a failure - I eliminated Zak3 at the end, because he wouldn't submit even when he had just 1 pop1 base left surrounded by my troops. I do not understand why he was so stubborn. I did not commit any attrocity...

MariOne · February 10, 2003, 07:09

(follows - & bad chance: 9-minutes X-post... )
I agree that the jungle helps a lot, especially if the winning strategy entails growth and expansion.
(not without reason it has always been rated the most useful Landmark in polls, and was most rated in Bingmann's System Landmarks/Upods pbem setup auctions)

But then also Zak does.

If you object that you should go for it without the Jungle, then you could also attempt to find out the Huge/Transcend transcendence limit for other factions, for instance Believers or Cult...

That would bring us to two "cubes" of variables, instead of one planar table: Diff/Size/Faction w/Jungle, Diff/Size/Faction w/o Jungle.

You could even more finely categorise:
- you don't know where's the Jungle (and you find out it's near)
- you don't know where's the Jungle (and you find out it's far)
- you don't know where's the Jungle (and you transcend without finding out or exploiting it!)
- you start in the Jungle
- you know where's the Jungle and it's near
- you know where's the Jungle and it's far
- you know there's NO Jungle (scenario created map)

This is going to be a bit complex and involuted.

Zsozso set the *generic* HugeTranscend game record (so far).

Then we could keep up the interest by investigating more customised challenges, if you wanna rather compare individual player skills, specific strategies or specific factions potentiality also in relation with the influence of speficic game features like landmarks and why not planetology (ranininess, erosion, oceanness, etc)

zsozso · February 10, 2003, 07:13

Quote:

Originally posted by MariOne
JT, zso showed that it is *possible* to Transcend at huge transced by 2194.
*possible* doesn't mean "anytime" or "always", it *entails* favorable condition.
Transcend-Huge are already TWO restraints on the achievement.

Yup, I'm not even sure I could repeat it even starting from the same point. It was a pretty lucky set of pod pops. In fact I was aiming for a finish 4-5 turns later even in 2181. I think I could achieve that in a repeat game - to transcend within 100 turn. OTOH, an even luckier game could end a few turns earlier. I'd say playing the best possible strategy can get you to finish anytime between 90-100 turns depending on your luck.

In fact if you want to make tougher restrictions and measure skills that way, you could say: you must play Miriam against 6 other Miriams, 90% water random map, tech stagnation, no pod scattering (of course keeping the huge size and transcend difficulty). Now that game would be tough to even transcend ever - I mean that 400 turns compulsory retirement might be too soon to research the tech tree. It would be a rather boring game that's for sure.

MariOne · February 10, 2003, 07:18

We could call that "Huge Faith in Transcendence"...

zsozso · February 10, 2003, 07:24

Quote:

Originally posted by MariOne
If you object that you should go for it without the Jungle, then you could also attempt to find out the Huge/Transcend transcendence limit for other factions, for instance Believers or Cult...

That would bring us to two "cubes" of variables, instead of one planar table: Diff/Size/Faction w/Jungle, Diff/Size/Faction w/o Jungle.

You could even more finely categorise:
- you don't know where's the Jungle (and you find out it's near)
- you don't know where's the Jungle (and you find out it's far)
- you don't know where's the Jungle (and you transcend without finding out or exploiting it!)
- you start in the Jungle
- you know where's the Jungle and it's near
- you know where's the Jungle and it's far
- you know there's NO Jungle (scenario created map)

Yes, there are too many variables to set a single "standard". Of course, I'm not claiming that I play the best strategy either. But if you really want to measure strategies against each other, then we must play with no pod scattering and random events turned off. Someone could CMN a map with opponent factions placed deliberately far away, no special landmark but big enough continent to build a good empire on. Any volunteers for such a work ?

MariOne · February 10, 2003, 07:26

OK
Huge
Transcend
Max Ocean
Dry
Low Heights
Abundant worms (Miriam would hardly profit of hunting, make it Avg if you think it harder)
7 Miriams (do we need to make them also Aggressive?)
No pods
No map
Events? (if you're the big dog, negatives are more likely)
Tech Stag
Blind Tech!!!
Don't look first

What else?

zsozso · February 10, 2003, 07:32

I think Blind tech is overkill - it introduces random chance and limits skill effect. Same for events. But making the opponents agressive is good.
I daresay such game is impossible to win by transcendence.

MariOne · February 10, 2003, 07:46

Well, Max Ocean minimizes the need for Scenario factions placement...
Those are the settings, any of us can start a new game on his own and enter them, extreme settings leave room for less variability and make for higher comparable maps...

Capt Dizle · February 10, 2003, 09:16

You guys are nuts. I only meant to point out that the jungle start was a jungle start is a jungle start. It's not like it was a real game.

I am more interested in trying to prove that the Byron method is or is not superior to traditional styles.

zsozso · February 10, 2003, 10:15

Quote:

Originally posted by jimmytrick
I only meant to point out that the jungle start was a jungle start is a jungle start.

JT, you lost me in that sentence...

You seriously want to convince me to play Byron from your start on Hendrik's game ? If I beat you there with Byron style is that proof enough ?

MariOne · February 10, 2003, 10:27

Sorry, I did read the fastest transcendence where David Byron set the 2167 record and gave some hints on his method, but can you refresh me what it all boils down to?

Is it just about building only Cpods and never facilities nor units?
Was there more in it than just that, to characterise it?
How did it relate with the "Borg style"?

I do have attemped, although not earnestly, several strategies stemming from ICS, I'm just not good at labeling them...

Capt Dizle · February 10, 2003, 11:10

zsozso, could you post a MY 2101 save. Although I don't have time now I will want to eventually play jungle bunny myself.

Now, if I remember Byron correctly he advocated no facilities did he not? Whereas you did build some, especially in your SSC.

I was already convinced that a modified Byron strategy is very powerful. Do I believe that a pure no facility Byron strategy is optimum. No. I rather think that a hybrid of the traditional style and Byron where a certain number of core cities with traditional facilities is enhanced with a slew of Byron cities would probably be optimum.

As I thought I said earlier, albeit in different language, transcend time is a function of map, style, skill and luck. Of course any jungle start will be faster than a non jungle start other things being equal. And it would be a piece of cake for even a map hack like me to create a map that would be faster than the one you played.

But you deserve full credit for setting the standard.

If you played Byron from the start Hendrick and I used we would be map neutral but still skill, style and luck would apply. So, what would you be trying to prove? You have already "proven" the best time on huge transcend, at least until someone knocks it off.

I think if we set up a normal map environment, took steps to nullify luck as practicable, even out the skill by having the game played by multiple persons, then we might get a definite answer on style. But that was not your goal and I did not mean to take away anything for the inovative way you approached the task at hand.

Capt Dizle · February 10, 2003, 11:16

And let me qualify that I define normal map environment and "real game" as being a computer generated map. But none are beholden to my definitions for God's sake.

zsozso · February 10, 2003, 12:11

Yes, you are right about the "true" Byron style: no facilities at all, just pods, formers and crawlers - each city having 1 former and 1-2 crawlers to boost production/food as necessary. And I aggree, playing that to such extreme would not be the most efficient strategy. My variant is to play that until you filled the "good" territory available, then build some SPs (like PTS), then return to expanding to the not-so-good territory until you fill that as well.
Finally you can start building infrastructure (in selected cities) and energy park.

Capt Dizle · February 10, 2003, 22:39

Wow! Looking at saves can certainly be enlightening. After reviewing the early saves from zsozso's game, my current game and Hendrick's Morgan game I can clearly state that neither the jungle nor the hybrid Byron play style can fully explain zsozso's record time.

Here is the deal. zsozso got 13 techs in the first 23 turns! Compared to my game in which I got to 13 only by the 49th turn and Hendrick on the 52nd.

The jungle did not play a part in this as zsozso had no more bases or pop than I at that point. I think he had 7 techs in the first 11 turns, that would be six more in the next 12.

Let's see, my 49 minus his 23 gives him a 26 turn advantage. He transcends on 94 +26= 120. Hmm, I can't make 120 in my game without the jungle and all those Zaks and Morgans but I might make 130.

I say the early tech was the ballgame. 13 techs in the first 23 turns!

He had early contact with one of the Zak's otherwise, I am mystified as to how he could have gotten that far that fast. And I would think that two Zak's couldn't have that darn much to trade that early?

Great Caesar's Ghost!

Mongoose · February 10, 2003, 23:50

Aggressive and forunate pod popping, perhaps? You can get almost all the L1 techs from pods.

Capt Dizle · February 11, 2003, 00:00

Yeah. Lucky duck.

Capt Dizle · February 11, 2003, 00:07

In his narrative above zsozso noted that he got one of the Zak's commlink on his 2111 report so that settles that.

I nevertheless unretract my previously retracted statement: "I daresay a sub 100 turn transcend on transcend is impossible outside of some freak event. On huge."

Hendrik · February 11, 2003, 02:24

I must say I was shocked by this amazing time, but JT is right, getting that many techs so fast has a large effect on the game. It took a lot of skill to transcend as fast as you did, but you were still playing under freak conditions, in terms of diplomacy (ideal) starting location (very good) and pod pops, that early forest pop is the only visible one, but 13 techs in 23 turns scarcely seems doable without at least several tech or commlink pops.

I don' t think that early tech can be added on in terms of straight turn advantage. There are very few techs that bring advantages without any costs to implementing them, IE clean reactors make formers cost more, facilities need to be installed, SPs need to be built, etc. Most of the advanced facilities are totally useless to a fledgling empire. I guess you could say that additional techs beyond a certain point in infrastructural development mean diminishing marginal returns... the first few techs make a large difference and don't need a heavy infrastructure to implement (SE related techs, formers, IA) and the like.

Beh a bit of a rant. Anyways zsozso I would be tickled pink if you would play a game on a regular random map using a Byron variant.

Capt Dizle · February 11, 2003, 09:51

Diminishing marginal returns. Yeah kinda, but, for the no infrastructure pardigm lifting restrictions is huge in terms of affect per base. This semi Byron method uses the number of bases as its multiplier rather than facilities in bases.

One example is the effect of Wealth on the center square of a University base. Without this SE in effect you are getting 2 or 3 from the base square and then with it you get +2 (although the datalinks indicates you'll only be getting +1 commerce).

So then your base squares are generating 4 or 5 energy and your adding bases as fast as you can and this has a big effect at a time in the game when research costs are very low. Remember the multiplier of the free network node.

When you get ahead on tech everything can happen faster for you like getting the Empath Guild and soaking up the tech trades and commerce that much sooner.

The University is really the best fast transcend faction you know. Morgan can sure rush infrastructure but the benefit from some techs is not strictly tied to infrastructure as zsozso has painfully taught us.

Getting tech sooner means being able to build SP sooner and gosh it's not like it will interfere with his infrstructure builds is it. PTS, PEG with VW to keep all happy. And WP and ME, none of this stuff is dependent on having infrastructure for benfits.

MariOne · February 13, 2003, 08:49

JT, one reason for the undocumented +1 Energy in basetile is Governorship (which thus gives more than just the documented +1 for each Treaty/Pact in that base).
DilithiumDad taught me this very recently (I had always taken for granted the datalinks without checking...)

You now report that there is an undocumented +1Energy in basetile for Wealth (or did you mean for a particular leve of SE Economy induced by Wealth?).

I am positive about the cause I quoted, I verified the effect by comparing a pre- and post-election save, all the rest unchanged (explicitly, SE unchanged).

Are you sure you didn't mistook the Gov'ship effect for the SE cause, or are you actually speaking of a different undocumented effect, which could even cumulate with the one I report???

Capt Dizle · February 13, 2003, 10:25

As you know I am not a math head Mose.

Precontact, non governor Zak running simple survival gets one energy per base square, or two with HQ or tanks. Econ rating of zero.

Precontact, non governor Zak running free market survival gets two energy per base square, or three with HQ or tanks. Econ rating of +2. This is consistent with what the datalinks says "+1 energy per square"

Precontact, non governor Zak running free market- wealth gets four energy per base square, or five with HQ or tanks. Econ rating of +3. This is inconsistent with the datalinks entry for +3 econ which states "+1 energy per square +1 commerce rating". It is actually +2 energy per base square that you are getting and you will keep it after you get commerce.

Also, I do believe you get an additional +1 in the base square with the governorship.

And I think you get a nice boost with golden age of about +2 energy per base square but that I can't be sure about. There may be some other stuff going on with that.

Base square +1
+Tanks +1= +2
+FM +1 =+3
+Wealth +2=5
+Governorship +1=6
Golden Age +2=8

Shows up as 8+ but I think 8 or anything higher shows as 8+.

So you can see that with all of these factors in place zsozso was making out like a bandit with his little bases on the base squares alone not to mention that with the projects he had he was getting a size three base, free node, free energy bank, free hologram. And so it goes.

Capt Dizle · February 13, 2003, 10:34

mathematically, there is probably a correlation between the GA induced increase in base square energy and the +10 psych allocation trick you use in the early game Mose.

By going 40-10-50 I was actually increasing the labs I was making because the +2 per base meant +1 energy to the same 50% energy allocation. The credit income lost by going to 40% was also mitigated by the increased base energy. And of course +2 growth without planned!

I digress, your early game 10 pysh allocation is probably doing the same sort of energy boost but all the numbers are fractions, so small as not to show up except in the final tech per turn where you do gain because, again, the cost is low in the first place. So I'd say that your little trick is kind of linked in the math sense.

MariOne · February 14, 2003, 07:59

To be fair, "my" little trick I actually learned it from cousLee, or JAMiAM, or Bingmann, can't recall.
To my knowledge, it helps forcing the game splitting some energy in a different way it would do by default, without actually allocating any to psych in most cases...

I must say that I only ever see it effective with a 30%Econ 10%Psych 60%Labs.
For instance (just an example, actual figures could be different), with 4 net energy you have 2ec-2labs at 50-0-50. You might find to be still stuck in 2-2 after shifting to 40-0-60 or even to 30-0-70. Usually the 30-10-60 allocation does the trick to split your energy 1-3, despite this means leaving your labs allocation unchanged, or even decreasing it!

But it only works to squeeze the best labs out of your energy in the early game, inevitably at the cost of some ec.
A -1Eff could be enough to make the trick uneffective, as a too good Eff also makes it moot.

And I never saw any money benefit from it, i.e going 60-10-30 does not give any extra ec over those the game already is able to allocate by default with 60-40 or 70-30.