or just force the AIs into negative gpt spending and see if they start building wealth instead of units
SGOTM7 - Team Peanut
- Thread starter mad-bax
- Start date
zyxy
Warmongering Fool
Very intriguing!
Looks like all games see a linear increase in number of AI units until about 1500BC, after which the PtW games switch to a higher linear increase, while the vanilla civs stay on the same curve
.
Of course linear growth is just what you see in it: the PtW curves in particular could be exponential. Linear is to be expected if the civs are static, but an expanding civ would show (near) exponential growth I think.
So what happened in 1500BC? In my team (Durkz), not that much: we contacted our third civ (Arabs) in 1650BC, the fourth (Persia) in 1300BC. The only other thing that happened in 1500BC is that it was the end of my first play-set (but that would hardly matter I hope
). The slightly nutty people contacted the Arabs in 1650BC, as well as the Persians.
For the vanilla teams: Ivan contact their second civ (Zulu) in 1375BC, Tao their third (Arabs) in 1425BC and fourth (Persia) in 670BC, Xteam meets third civ (Persia) in 1150BC, Arabs get destroyed at some point (too much spam to find out when
).
The slightly later contacts for the vanilla players could mean that the AI's in their games start to build up military later, but it would not explain the huge difference we see here IMO. I am very puzzled....
Question: what is the limit to where the AI will build up troops? Free support, or until all dataslots are filled? Does it depend on its diplomatic status (war/peace)? Could this be different in vanilla/PtW?
Looks like all games see a linear increase in number of AI units until about 1500BC, after which the PtW games switch to a higher linear increase, while the vanilla civs stay on the same curve
.Of course linear growth is just what you see in it: the PtW curves in particular could be exponential. Linear is to be expected if the civs are static, but an expanding civ would show (near) exponential growth I think.
So what happened in 1500BC? In my team (Durkz), not that much: we contacted our third civ (Arabs) in 1650BC, the fourth (Persia) in 1300BC. The only other thing that happened in 1500BC is that it was the end of my first play-set (but that would hardly matter I hope
). The slightly nutty people contacted the Arabs in 1650BC, as well as the Persians.For the vanilla teams: Ivan contact their second civ (Zulu) in 1375BC, Tao their third (Arabs) in 1425BC and fourth (Persia) in 670BC, Xteam meets third civ (Persia) in 1150BC, Arabs get destroyed at some point (too much spam to find out when
).The slightly later contacts for the vanilla players could mean that the AI's in their games start to build up military later, but it would not explain the huge difference we see here IMO. I am very puzzled....
Question: what is the limit to where the AI will build up troops? Free support, or until all dataslots are filled? Does it depend on its diplomatic status (war/peace)? Could this be different in vanilla/PtW?
I think all of that needs testing. If there is a quantitative difference between the two software versions, then that does raise some interesting (read: as GOTM staff, I *really* don't want to deal with this) game comparison issues, doesn't it? Given all the variables, though, and the tiny data set, it's still remotely possible that the patterns seen are just a fluke. That's why I think a bit of more controlled testing is in order.
Renata
Renata
Mathilda
Queen
That sounds like someone who knows what they're talking about could maybe do just that...Renata said:
Mathilda
Queen
I thought you was a graduateRenata said:I only know how to use the C3C editor.
Renata
You didn't skip a lesson did you?
Some more information regarding the Zulu GAs.
Team Peanut: From 550 to 350 to 150 BC, total Zulu unit count goes from 69 to 82 to 89. Zulus have a total of 6 cities; at 150 BC they have converted entirely to infrastructure, 5 Buildings and the HG being built. They are Despotic in 350 BC, and Monarchy in 150 BC.
Team Tao: same time frame, total Zulu unit count goes from 47 to 45 to 41. Zulus have a total of 8 cities; at 150 BC they are building 4 Buildings, 2 Units, HG and Forbidden Palace(!) They are in Anarchy in 350 BC, and a Monarchy in 150 BC.
Team Xteam: same time frame, total Zulu unit count goes from 46 to 52 to 50. Zulus have a total of 8 cities; at 150 BC they are building 6 Buildings, GLib, and a Settler. They are Despotic in 350 BC, and a Republic in 150 BC.
Vanilla Zulu are more exploratory, founding additional cities off continent. Since the PTW Zulu in Peanut converted to building city improvements, it seems that Vanilla Zulu wanted around 6 units per city and PTW Zulu about 15. This from only 3 total samples, however.
Looking at the AI being invaded at the 150 BC saves for these 3 teams, they usually have the remaining Wonders being built, but all versions show Unit construction as well. This makes sense if there is a per city limit; with the heavy loss of units they all have unit counts to make up.
If there is a unit per city limit that is different for Vanilla vs PTW, we're likely to see an effect in smaller maps where unit saturation is reached quickly, or in isolationist maps where the AI is constrained in space. One possible side effect is the AI would be more focused on acquiring extra space and cities in Vanilla, which might account for the Zulus being more exploratory for the Tao and Xteam games. And for Xteam's Persia being so aggressive; the AI might want to reduce its Unit count so it enters War more often, and maybe very ineffectually since at first glance it appears that the Vanilla AI's military is choked with obsolete units. I saw very few if any Horsemen in the two Vanilla Zulu military, but several Horsemen built in the PTW Peanut game.
Team Peanut: From 550 to 350 to 150 BC, total Zulu unit count goes from 69 to 82 to 89. Zulus have a total of 6 cities; at 150 BC they have converted entirely to infrastructure, 5 Buildings and the HG being built. They are Despotic in 350 BC, and Monarchy in 150 BC.
Team Tao: same time frame, total Zulu unit count goes from 47 to 45 to 41. Zulus have a total of 8 cities; at 150 BC they are building 4 Buildings, 2 Units, HG and Forbidden Palace(!) They are in Anarchy in 350 BC, and a Monarchy in 150 BC.
Team Xteam: same time frame, total Zulu unit count goes from 46 to 52 to 50. Zulus have a total of 8 cities; at 150 BC they are building 6 Buildings, GLib, and a Settler. They are Despotic in 350 BC, and a Republic in 150 BC.
Vanilla Zulu are more exploratory, founding additional cities off continent. Since the PTW Zulu in Peanut converted to building city improvements, it seems that Vanilla Zulu wanted around 6 units per city and PTW Zulu about 15. This from only 3 total samples, however.
Looking at the AI being invaded at the 150 BC saves for these 3 teams, they usually have the remaining Wonders being built, but all versions show Unit construction as well. This makes sense if there is a per city limit; with the heavy loss of units they all have unit counts to make up.
If there is a unit per city limit that is different for Vanilla vs PTW, we're likely to see an effect in smaller maps where unit saturation is reached quickly, or in isolationist maps where the AI is constrained in space. One possible side effect is the AI would be more focused on acquiring extra space and cities in Vanilla, which might account for the Zulus being more exploratory for the Tao and Xteam games. And for Xteam's Persia being so aggressive; the AI might want to reduce its Unit count so it enters War more often, and maybe very ineffectually since at first glance it appears that the Vanilla AI's military is choked with obsolete units. I saw very few if any Horsemen in the two Vanilla Zulu military, but several Horsemen built in the PTW Peanut game.
Crazy Eskimo
Prince
- Joined
- Apr 14, 2005
- Messages
- 367
Wow. 
I'm glad civ_steve is willing to do all this analysis. Those are some pretty significant differences. I think Renata's right that they need to be explored thoroughly so that we can explain them. The difference in the number of cities and wonder building seem like possible confounding factors. Hopefully, a way to maintain parity between PTW and Vanilla can be found, otherwise the laurels will be less valuable.
Player: I came in first! Woo!
Spectator: But you were the only one playing SGOTM42 in the PTW variant...
Player: Silence! I am a genius! (/zim)
On a side and unrelated note, how is parity kept for other factors like the lack of MDI in Vanilla? We used MDI fairly extensively, but I would imagine it would change our offensive mix if we knew that Knights were our only 4 attack unit in the Medieval times. Or is it not a big enough factor?
I'm glad civ_steve is willing to do all this analysis. Those are some pretty significant differences. I think Renata's right that they need to be explored thoroughly so that we can explain them. The difference in the number of cities and wonder building seem like possible confounding factors. Hopefully, a way to maintain parity between PTW and Vanilla can be found, otherwise the laurels will be less valuable.
Player: I came in first! Woo!
Spectator: But you were the only one playing SGOTM42 in the PTW variant...
Player: Silence! I am a genius! (/zim)
On a side and unrelated note, how is parity kept for other factors like the lack of MDI in Vanilla? We used MDI fairly extensively, but I would imagine it would change our offensive mix if we knew that Knights were our only 4 attack unit in the Medieval times. Or is it not a big enough factor?
Vanilla's been modded for SGOTM and GOTM games. The missing units and civs were duplicated* and added to the vanilla game.
Except for Mongolia, since the Keshik couldn't be duplicated in vanilla. For Mongolia, both PTW and vanilla were changed.
Renata
Except for Mongolia, since the Keshik couldn't be duplicated in vanilla. For Mongolia, both PTW and vanilla were changed.
Renata
Crazy Eskimo
Prince
- Joined
- Apr 14, 2005
- Messages
- 367
Many thanks.
Crazy Eskimo
Prince
- Joined
- Apr 14, 2005
- Messages
- 367
Well, with Ivan at 800 ad and just about to invade the Persia island, I feel pretty safe in congratulating the rest of Team Peanut on our 2nd place victory. Good job everyone!
Edit: We're the third fastest overall, just ahead of Grumpy who had a finish of 870 AD. I looked at the threads of the two Conquests teams that are still playing and it's unlikely that they will finish by 840. Great job, fellow Peanuts!
Edit: We're the third fastest overall, just ahead of Grumpy who had a finish of 870 AD. I looked at the threads of the two Conquests teams that are still playing and it's unlikely that they will finish by 840. Great job, fellow Peanuts!
Well it looks like we will end up being 3rd overall, with just the fastest C3C and Vanilla teams ending ahead of us. And I believe AlanH said an additional Award for PTW will be handed out due to the differences in AI behavior between PTW and Vanilla! 
Definitely a great game, Peanuts!
To continue, the SGOTM8 thread for signing up is open. I've signed up and I hope we can play together as a team for it. It looks like most of the action will take place in the Modern Era, due to the restrictions on America whom we will be playing. Lots of research action, and diplomacy, leading up to a late beat-them-up (or so it seems.) We still need a final Spoiler entry for SGOTM7 (were you up for that Peanut? I can, or another volunteer, otherwise.) Once our spoiler is in, I was going to put some links to the discussion about the Vanilla and PTW differences, so that's the business left to close this out.
Definitely a great game, Peanuts!To continue, the SGOTM8 thread for signing up is open. I've signed up and I hope we can play together as a team for it. It looks like most of the action will take place in the Modern Era, due to the restrictions on America whom we will be playing. Lots of research action, and diplomacy, leading up to a late beat-them-up (or so it seems.) We still need a final Spoiler entry for SGOTM7 (were you up for that Peanut? I can, or another volunteer, otherwise.) Once our spoiler is in, I was going to put some links to the discussion about the Vanilla and PTW differences, so that's the business left to close this out.
Mathilda
Queen
Hi Nutters!
Sgotm08 definately sounds different again - I like these variants.
Since the issue between ptw and vanilla looks a long way from being solved I was wondering if we might play C3C this time?
Any thoughts?
Sgotm08 definately sounds different again - I like these variants.
Since the issue between ptw and vanilla looks a long way from being solved I was wondering if we might play C3C this time?
Any thoughts?
I think the issues that we saw in SGOTM7 are lessened with lower level games. SGOTM8 will be at Emperor or Monarch, so it will take longer for the AI to reach whatever unit level they think they need. I think map size is standard, and that helps, too; SGOTM7 was small giving the AI fewer total cities. However, SGOTM8 will be archipelago, so there might be some isolation issues.
It would be nice to keep the number of PTW/Vanilla teams higher than lower. However, I have no problems playing C3C if that's what the team would like.
It would be nice to keep the number of PTW/Vanilla teams higher than lower. However, I have no problems playing C3C if that's what the team would like.
Peanut
A Slightly Nutty Flavour
Sorry comrades - Peanut strikes again ! We had a bout of winter illnesses in my two weeks of spare time, which killed my chance of finishing the final spoiler installment. Sorry particularly to you CS with all the analysis you have done. I should have posted earlier to let you all know. And I was looking forward to writing it as well.
I have been busier than a one-armed wallpaper hanger for the last three weeks with the very last practical component of my course with another three to go, so I won't be able to start playing until mid-August. Feel free to use my name and nail me up as the figurehead if you want, or otherwise just reflect reality and rename the team as Team CS or Team Mathilda or another of you illustrious Peanut crew members. When I can I will call by admire your progress.
I have been busier than a one-armed wallpaper hanger for the last three weeks with the very last practical component of my course with another three to go, so I won't be able to start playing until mid-August. Feel free to use my name and nail me up as the figurehead if you want, or otherwise just reflect reality and rename the team as Team CS or Team Mathilda or another of you illustrious Peanut crew members. When I can I will call by admire your progress.
Mathilda
Queen
Oh Peanut, such a shame.
Hope to see you around and back for sgotm09
Hope to see you around and back for sgotm09
Peanut: sorry to hear about the winter illness (hard to imagine - it's been over 100 degrees F for about 2 weeks here!). I can put together a final spoiler, and then post about the analysis.
Team Peanut certainly has more character to it as a name . This is going to be a loooong game (other then launching spaceships, I've had no experience in the modern era, and I've never dealt with nukes either.) In my reading of your post you still wish to play on the team, just not until the 'practical components' of your course are completed. If you're MIA for a couple weeks that should be fine IMO; that might be 3 rounds, perhaps middle Middle Ages (we will be dependent on the AI to make contact, so it might take a lot longer!)
Team Peanut certainly has more character to it as a name
. This is going to be a loooong game (other then launching spaceships, I've had no experience in the modern era, and I've never dealt with nukes either.) In my reading of your post you still wish to play on the team, just not until the 'practical components' of your course are completed. If you're MIA for a couple weeks that should be fine IMO; that might be 3 rounds, perhaps middle Middle Ages (we will be dependent on the AI to make contact, so it might take a lot longer!)
Peanut
A Slightly Nutty Flavour
Thanks CS. That about sums it up. If I can rejoin in about 3 weeks that would be great - it looks like an interesting scenario !
BTW after 20 years in engineering I am converting to secondary school teaching. How's THAT for nutty, eh ?
BTW after 20 years in engineering I am converting to secondary school teaching. How's THAT for nutty, eh ?
That sounds great! I don't know about Australia, but our high schools are always short in their supply of experienced math, science and engineering teachers. To compound the issue, the education bureaucrats seem intent on dumbing down these types of classes so they can graduate more students making them look better. (And the average teacher might have failed these classes themselves and have no aptitude for these subjects so how good can they be at teaching these classes?) Good for you, and good luck in this career change!
If knew we were going to have tons of Ancient and Middle Age action, I'd say 3 weeks is too long to be out. With this scenario, I don't have a real issue if you wanted to join in at the later date. You should post in the sign-up thread, and if anyone has a big objection they should state it.
If knew we were going to have tons of Ancient and Middle Age action, I'd say 3 weeks is too long to be out. With this scenario, I don't have a real issue if you wanted to join in at the later date. You should post in the sign-up thread, and if anyone has a big objection they should state it.
Similar threads
