LowtherCastle
Deity
- Joined
- May 18, 2005
- Messages
- 23,756
When do barb units start spawning?
SGOTM 22 - Anti-TSR
- Thread starter AlanH
- Start date
LowtherCastle
Deity
- Joined
- May 18, 2005
- Messages
- 23,756
More thought. More gambits.
1. Gifting a city to AI with GW. We could save a city about to fall to barbs. If AI is creative, we could build a city we want later (jungle ivvories) to spawnbust.
2. IW beeline. Key: 6 AIs bunched on 2/3 of map. When we complete IW we already have a settler with defenders waiting near target #1 to settle iron. Maybe add a stack of chariots to capture target #1 capital. Focus early settling on horses and beakers. Capture target #1, keep good cities will gifting our lousy ones. BSP tried to make this hard, so we do it for that reason. It must be good.
Note: My analysis indicates some AIs probably have lots of forest (healthy demo screen).
3. HBR beeline. Similar to #2 but with HAs.
4. Sailing. Alternative to workers to connect horses, gems, stone without roads, so workers can do other things. Gives us some beakers (TRs + OBs).
5. Settle gems early. Build monument, any needed warriors while settler walks.
6. Defending Horse City. Maybe not so hard. Could settle on plains hill for added archer protection. additional archere for horse tile.
7. LIghtning rod city for barbs strategically settled, hopefully on a hill (don't have save in front of me), gifted to a creative AI without GW, defended with our archer or 2. Need OBs. Located to draw most barbs to it rather than our cities.
The basic idea is to skip GW and focus on crushing AIs asap. How? Above ideas are tidbits to feed into better mixtures of ideas.
1. Gifting a city to AI with GW. We could save a city about to fall to barbs. If AI is creative, we could build a city we want later (jungle ivvories) to spawnbust.
2. IW beeline. Key: 6 AIs bunched on 2/3 of map. When we complete IW we already have a settler with defenders waiting near target #1 to settle iron. Maybe add a stack of chariots to capture target #1 capital. Focus early settling on horses and beakers. Capture target #1, keep good cities will gifting our lousy ones. BSP tried to make this hard, so we do it for that reason. It must be good.
Note: My analysis indicates some AIs probably have lots of forest (healthy demo screen).
3. HBR beeline. Similar to #2 but with HAs.
4. Sailing. Alternative to workers to connect horses, gems, stone without roads, so workers can do other things. Gives us some beakers (TRs + OBs).
5. Settle gems early. Build monument, any needed warriors while settler walks.
6. Defending Horse City. Maybe not so hard. Could settle on plains hill for added archer protection. additional archere for horse tile.
7. LIghtning rod city for barbs strategically settled, hopefully on a hill (don't have save in front of me), gifted to a creative AI without GW, defended with our archer or 2. Need OBs. Located to draw most barbs to it rather than our cities.
The basic idea is to skip GW and focus on crushing AIs asap. How? Above ideas are tidbits to feed into better mixtures of ideas.
So, human barbs when there are 11 cities.
LowtherCastle
Deity
- Joined
- May 18, 2005
- Messages
- 23,756
1. Early settling by us hastens 10 cities.
2. Barb warriors and archers will each spawn at 50% odds.
3. It's safer to move a settler into position pre-10 cities, unless there's a bear lurking.
***Corrected***
No, less than 10.5 -> less than 10 (after truncating) -> 9 or less.
Human barbs spawn at 10 cities.
LowtherCastle
Deity
- Joined
- May 18, 2005
- Messages
- 23,756
WT, you say you're not worried about our tech rate, but you're missing my point. I'm looking at our relative tech rate now, in the early game. You keep saying our exploring warriors will die. Well, that's only because we're not getting Writing soon enough for OBs so our warriors can travel relatively safely from bunched AI to AI.
You also talk about typically not exploring much early and at most with one warrior. Rapid conquest is much more rapid when you know your enemy. Who has copper, et cetera. I see huge value in early exploration. If your GW Gambit sacrifices exploration, I value that as a huge loss. I understand that you, don't, so I wanted to clarify it.
Our warriors might even see into city centers, if they're not much more distant and tell us, via the espy screen, how many hammers they have, whether they have stone, et cetera.
You also talk about typically not exploring much early and at most with one warrior. Rapid conquest is much more rapid when you know your enemy. Who has copper, et cetera. I see huge value in early exploration. If your GW Gambit sacrifices exploration, I value that as a huge loss. I understand that you, don't, so I wanted to clarify it.
Our warriors might even see into city centers, if they're not much more distant and tell us, via the espy screen, how many hammers they have, whether they have stone, et cetera.
LowtherCastle
Deity
- Joined
- May 18, 2005
- Messages
- 23,756
What are the basic game elements in this scenario?
1. Exploration
2. Research
3. Infrastructure
4. Goals (conquest, 7 wonders)
5. Raging barbs
6. ?
When I said this scenario is very complicated, what I meant is that every choice we makes helps or hinders one or more of the above. The game constraints do the same.
For example, building the GW appears to sacrifice exploration and somewhat research for infrastructure and raging barbs. Failing at the GW sacrifices exploration and somewhat research and raging barbs for future research (failgold).
That's how I'm thinking in all the variants I outline.
1. Exploration
2. Research
3. Infrastructure
4. Goals (conquest, 7 wonders)
5. Raging barbs
6. ?
When I said this scenario is very complicated, what I meant is that every choice we makes helps or hinders one or more of the above. The game constraints do the same.
For example, building the GW appears to sacrifice exploration and somewhat research for infrastructure and raging barbs. Failing at the GW sacrifices exploration and somewhat research and raging barbs for future research (failgold).
That's how I'm thinking in all the variants I outline.
Mitchum
Deity
I tend to agree with LC on exploration. Knowing what the AI are up to is a huge benefit. We started with 4 warriors and 3 of them are already well on their way to finding at least 1 AI each. That could be huge. Early exploration is a big deal not only to know where the AI are but to have some intel on what they are up to. We've all talked about wanting the 'Mids. Wouldn't it be good to have an idea if one of the AI is building them before we start down that path? The only way to know is to peek into their capital so we can track hammers via the espionage screen. Wouldn't it be great to know where the AI with the GW is located so that we can decide whether to capture it ASAP?
It would also be nice to know if they are reachable without Astronomy or some other crazy tech, right? The sooner we know that the better. In fact, we could be on this large island by ourselves with every AI having its own smaller island. I'd like to know that in ~40 turns (brown number) than ~100 turns (even browner) by pulling all of our warriors back while we aim for the GW.
So, if I had to choose between knowing where all of the AI and fully exploring their lands or knowing for sure that we'd get the GW, it would be about a toss up with the GW being slightly favored. However, given that our odds of getting the GW are anywhere between 75% and 0% depending on whom you ask with most of us thinking the number is closer to 0% than 75%, I'd take the exploration for sure.
If we pull those warriors back now, it's going to be a long time before we'll be able to spare any units to venture out to explore. Plus, the longer we wait to send the troops out to explore, the beefier the unit(s) will have to be as barb warriors, archers, spears and axes start to appear.
It would also be nice to know if they are reachable without Astronomy or some other crazy tech, right? The sooner we know that the better. In fact, we could be on this large island by ourselves with every AI having its own smaller island. I'd like to know that in ~40 turns (brown number) than ~100 turns (even browner) by pulling all of our warriors back while we aim for the GW.
So, if I had to choose between knowing where all of the AI and fully exploring their lands or knowing for sure that we'd get the GW, it would be about a toss up with the GW being slightly favored. However, given that our odds of getting the GW are anywhere between 75% and 0% depending on whom you ask with most of us thinking the number is closer to 0% than 75%, I'd take the exploration for sure.
If we pull those warriors back now, it's going to be a long time before we'll be able to spare any units to venture out to explore. Plus, the longer we wait to send the troops out to explore, the beefier the unit(s) will have to be as barb warriors, archers, spears and axes start to appear.
Are you sure? the formula you posted didn't have a truncation
Is that somehow automatic?
Our unit maintenance is going to be so high that we won't be able to afford explorers...unless we get the GW.
We probably can't get writing before our warriors start to die, and even if we do, then we get to explore one AI then die trying to find another one.
Without GW and with warriors exploring, in order to settle our first city, we will need: 10 new units build...at a minimum.
We could try to leave 1-2 or 3 warriors on suicide exploration duty. 1-2 warriors should be enough or certainly 3 would be able to hold the fort until T41 Archer or GW.
I was only suggesting keeping all 4 home because:
a) I want them for the Rome defender count
b) I think they'll die out in the wild
After GW, we'll be able to get a nice chariot explorer or two.
We probably can't get writing before our warriors start to die, and even if we do, then we get to explore one AI then die trying to find another one.
Without GW and with warriors exploring, in order to settle our first city, we will need: 10 new units build...at a minimum.
We could try to leave 1-2 or 3 warriors on suicide exploration duty. 1-2 warriors should be enough or certainly 3 would be able to hold the fort until T41 Archer or GW.
I was only suggesting keeping all 4 home because:
a) I want them for the Rome defender count
b) I think they'll die out in the wild
After GW, we'll be able to get a nice chariot explorer or two.
100% sure. I tested the number of cities in a test game before I even looked into the code.
All numbers in the code are integers and get truncated.
LowtherCastle
Deity
- Joined
- May 18, 2005
- Messages
- 23,756
Getting back to this:
).
Anyway, my basic point is that we already have a lot of evidence that the 1-4 AIs are likely to be building a wonder.
How much of that evidence is conjecture and how much factual?
Factual is that the AIs average hpt for T11 and T12 is 5, that is a total between them of 30-35hpt. The max is 7 and min is 3, so at most 3 AIs are at the minimum of 3hpt, leaving the other 3AIs at 7hpt. Or some other combination. That's factual. Those averages were only slightly lower at 4hpt for turns 5-10. That means at least 24hpt total for 6t. And so on. That's pure factual minimums. The only conjecture is beyond the minimum.
The same goes for the AI fpt on T12 = 8fpt. These AIs are eating gourmet pork, growing portly like good monarchs. They'll produce settlers fast or wonders fast or whatever they want fast. Their teching however is slow to average, except for Darius so they'll soon run out of buildings to build, other than the wonders or settlers, workers and units.
One more factual aspect of the data is the power calculations, meaning whether they have accumulated a power building (rax), built a power unit, grown to an even population (pop2, pop4) or acquried a power tech (hunting, mining, TW, AH, Sailing, archery, BW, et cetera). We know exactly how much these power counts increase for the AI group, turn by turn. Factual data. Because we've known the AIs from early on, we're able to know precisely whether it was a tech, building/unit, or growth. We only speculate on certain things, like whether 3K power was an archer or a Rax, and even that only in certain situations. On T4, for example, there is no doubt that 3 archers were built. Fact.
Because we know exactly how much power was accumulated in the first 12t, we know roughly how many hammers went into power builds. The one exception is T8, where we don't know if an archer was completed for 22h or a rax for 45h. That's a variance of 23h. We also don't know if monuments were completed by the three AIs with Mysticism. (We know for certain that only 3 AIs have Mysticism.) In short, we know with relative certainty that the above hammer estimates are reasonably accurate. We don't really know which AI has exactly how much in his build queue on T12. That's relative speculation.
Do the math. These relatively certain calculations, summed up, are why I'm confident we have relatively 0% chance of completing the GW.
.
I don't know if this splotch of data is even confrontable to you folks, but I'd like to focus on the estimated T12 hammer count for each AI. (scroll to the right) It's saying that 5 out of 6 AIs are past the point of building an archer (cost: 22h). Not so obvious but also in there is that the top four AIs listed have grown while putting those hammers in. That means they're either building a barracks or a wonder. Only other choices would be monument or temple. Hannibal doesn't even have mysticism, so he's building rax, Oracle, or GW (or temple). His hammers are too low for a plains hill stone/marble and he's not working a quarry. He still could be settled on marble or stone, but it seems unlikely he's got a prebuilt quarry at all (WT, you might win that bet.
).Anyway, my basic point is that we already have a lot of evidence that the 1-4 AIs are likely to be building a wonder.
How much of that evidence is conjecture and how much factual?
Factual is that the AIs average hpt for T11 and T12 is 5, that is a total between them of 30-35hpt. The max is 7 and min is 3, so at most 3 AIs are at the minimum of 3hpt, leaving the other 3AIs at 7hpt. Or some other combination. That's factual. Those averages were only slightly lower at 4hpt for turns 5-10. That means at least 24hpt total for 6t. And so on. That's pure factual minimums. The only conjecture is beyond the minimum.
The same goes for the AI fpt on T12 = 8fpt. These AIs are eating gourmet pork, growing portly like good monarchs. They'll produce settlers fast or wonders fast or whatever they want fast. Their teching however is slow to average, except for Darius so they'll soon run out of buildings to build, other than the wonders or settlers, workers and units.
One more factual aspect of the data is the power calculations, meaning whether they have accumulated a power building (rax), built a power unit, grown to an even population (pop2, pop4) or acquried a power tech (hunting, mining, TW, AH, Sailing, archery, BW, et cetera). We know exactly how much these power counts increase for the AI group, turn by turn. Factual data. Because we've known the AIs from early on, we're able to know precisely whether it was a tech, building/unit, or growth. We only speculate on certain things, like whether 3K power was an archer or a Rax, and even that only in certain situations. On T4, for example, there is no doubt that 3 archers were built. Fact.
Because we know exactly how much power was accumulated in the first 12t, we know roughly how many hammers went into power builds. The one exception is T8, where we don't know if an archer was completed for 22h or a rax for 45h. That's a variance of 23h. We also don't know if monuments were completed by the three AIs with Mysticism. (We know for certain that only 3 AIs have Mysticism.) In short, we know with relative certainty that the above hammer estimates are reasonably accurate. We don't really know which AI has exactly how much in his build queue on T12. That's relative speculation.
Do the math. These relatively certain calculations, summed up, are why I'm confident we have relatively 0% chance of completing the GW.
.
LowtherCastle
Deity
- Joined
- May 18, 2005
- Messages
- 23,756
THis is why I listed the basic scenario elements above. Your statement here is correct only relative to certain variants, but you don't say which ones, you just think them (giving you the benefit of the doubt
). In a variant that gets early writing or focuses heavily on beakers, we might get writing in time. Also, our current exploration indicates that the AIs are relatively bunched, so it might not be so hard to go from one to the other relatively safely. INcidentally, analysis indicates that about 3 or so are inland and the others might not be.I'd prefer to let 1-2 warriors continue exploring. Our scout last game survived to the end, iirc. I wouldn't mind killing them off after they're done, if we can't afford them. Cheap price to pay for the invaluable knowledge. Knowledge is power.
With the archer strategy we could get by with 4-6 archers for Rome (for 2-3 cities) and a couple for horses, possibly. I'd have to study the map again. Gems shouldn't require more than a warrior for the chokepoint and I forget what for defogging farther south. Doesn't even need a garrison.
We could settle our second city with 4 warriors or 2 warriors and 2 archers for Rome garrison.
.
LowtherCastle
Deity
- Joined
- May 18, 2005
- Messages
- 23,756
In any case, I think we can all agree this crap is complicated, especiallly without GW. That's what makes GW so tantalizing.
I love that you can do...and volunteer to do...and seem to enjoy digging up...all that data, LC. great work.
I worry that we're outsmarting ourselves and could end up being the only team without GW. I've said how I think the downside is very minimal if we take a shot at it. We'd research a tech that we probably need anyway. We keep some warriors home to defend/spawnbust (which is probably what was intended.) After all, we weren't given 4 scouts.
If GW goes before T30 as has been predicted, then the hammers -> gold will be less than 2 archers worth. Not a big setback.
Even tho everyone has pretty much said I can just go ahead with what I choose, I'm not comfortable playing without more people that see it as a worthy effort with acceptable downside. You guys clearly have the majority, so if you want to play on with Archery, I'll accept the team direction and get on board with "plan B"...I just can't do the turnset.
I worry that we're outsmarting ourselves and could end up being the only team without GW. I've said how I think the downside is very minimal if we take a shot at it. We'd research a tech that we probably need anyway. We keep some warriors home to defend/spawnbust (which is probably what was intended.) After all, we weren't given 4 scouts.
If GW goes before T30 as has been predicted, then the hammers -> gold will be less than 2 archers worth. Not a big setback.
Even tho everyone has pretty much said I can just go ahead with what I choose, I'm not comfortable playing without more people that see it as a worthy effort with acceptable downside. You guys clearly have the majority, so if you want to play on with Archery, I'll accept the team direction and get on board with "plan B"...I just can't do the turnset.
fyi, last game, our scout died almost immediately, right near Shaka's capital. We did not send a replacement or learn where Willy was, but still managed to wisely wipe him out first without early scouting knowledge.
Mitchum
Deity
And that, my friend, is proof that scouting is not only unnecessary but is in fact BAD for you!
da_Vinci
Gypsy Prince
Busy weekend, air show at Andrews AFB on Sat and nephew's 21st birthday on Sunday.
Now that I have caught up on the debate ...
Wow, you guys play this in some 5th dimension, and my poor 3D mind is struggling to get wrapped around an extra axis or two ...
It is as if you have sucked the brain out of the AI like some ugly brain bug from Starship Troopers. I play much more simply, with much less clairvoyance about such things as AI intentions and resources 12 turns in. I am also not generally a test game simmer (not enough time). So I don't think I am the guy to play until that stuff matters less.
I do have a couple thoughts about process for getting to consensus.
It seems we have a few scenarios: GW success, GW try and fail, No GW warrior ring and No GW archer defense. So four outcomes to put on a decision tree.
Decision analysis requires assigning each outcome a probability, and a utility (value). The smart choice is which product is larger, at least in decision tree analysis. Gambits, of course, violate this logic.
Do we have a sense of how good, and how probable, each of the four outcomes is?
Aside from LC's laudable goal of the perfect game, we are competing against the other teams, not against BSP. So which decision is likely to give us a big edge, or keep us even, or leave us lagging?
Is GW a big edge vs. no GW? Or are there no GW alternatives that compensate for not having the wonder?
Is GW fail catastrophic, in terms of the opportunity cost of the chase? Is there enough failgold to mitigate that opportunity cost?
I can probably manage to ask the questions in three dimensions, but I think the answers have to come from you guys in the fifth dimension ...
3D dV
Now that I have caught up on the debate ...
Wow, you guys play this in some 5th dimension, and my poor 3D mind is struggling to get wrapped around an extra axis or two ... It is as if you have sucked the brain out of the AI like some ugly brain bug from Starship Troopers. I play much more simply, with much less clairvoyance about such things as AI intentions and resources 12 turns in. I am also not generally a test game simmer (not enough time). So I don't think I am the guy to play until that stuff matters less.
I do have a couple thoughts about process for getting to consensus.
It seems we have a few scenarios: GW success, GW try and fail, No GW warrior ring and No GW archer defense. So four outcomes to put on a decision tree.
Decision analysis requires assigning each outcome a probability, and a utility (value). The smart choice is which product is larger, at least in decision tree analysis. Gambits, of course, violate this logic.
Do we have a sense of how good, and how probable, each of the four outcomes is?
Aside from LC's laudable goal of the perfect game, we are competing against the other teams, not against BSP. So which decision is likely to give us a big edge, or keep us even, or leave us lagging?
Is GW a big edge vs. no GW? Or are there no GW alternatives that compensate for not having the wonder?
Is GW fail catastrophic, in terms of the opportunity cost of the chase? Is there enough failgold to mitigate that opportunity cost?
I can probably manage to ask the questions in three dimensions, but I think the answers have to come from you guys in the fifth dimension ...
3D dV
Mitchum
Deity
I think this is where I disagree. If I agreed with this statement, I'd be all over the GW Gambit.
I think that the down side is a lot more than "very minimal" if we take a shot at it and fail. The first handful of tech choices are very important and each needs to provide some short-term, tangible benefit or else we're better off picking something else. We would have spent several turns researching a tech that we otherwise wouldn't need for ~50 turns (brown number).
Researching Masonry before Mysticism (a tech we've all agreed we will need very soon) loses out on the 20% bonus beakers which may amount to 1 to 2 turns of research. Not a huge deal but has ripple affects throughout the early game.
Archers are way more effective than warriors at keeping the barb hordes away from our cities. The sooner we can put hammers into archers rather than warriors, the better. Plus, in this scenario with warriors not qualifying for Rome MP duty after BW (a tech we want in the short term), they are even weaker.
Here is my fear of the worst thing that can happen. We start the GW and get it about half done when the first wave of barbarians start to enter our culture. We've got 2 or 3 warriors and there are archers baring down on us. What do we do? Without the ability to chop or whip and without Archery, we end up hunkering down in Rome while our mines and pasture get pillaged. We end up losing the GW (either because our hammer output goes down or we have to start building warriors/archers instead), we have no improvements and we are sitting at one city having to dig our way out of a MASSIVE hole. We could even lose Rome (and the game).
Maybe I'm overly afraid of the barbs but the fact that in the real game we've seen no AI culture (unlike the test game), we will have more than our fair share of barbs and they are likely to be worse than what we've seen in the test game.
Mitchum
Deity
Huh... I had you pegged as a 2D guy. I'm much happier to know that you're at least qualified to take out our trash now!
Similar threads
