Ribannah
Fighter Druid
Never mind 
(Hey, how come the discussion has stopped?
)
(Hey, how come the discussion has stopped?
)
Side discussion for the New Scoring System
- Thread starter Aeson
- Start date
hotrod0823
Deity
I have just read this entire thread again and I have question. The modifiers are for the map, civ, and difficulty level but what modifiers were used for GOTM 18??
The map was not a typical standard map and I am trying to grasp how the best dates were determined for GOTM 18. Specifically a diplo victory in the 880's. The base modifiers for Diplo and SS, from what I can tell, are based on when the technology in the most well played game can be researched.
It makes sense that the diplo victory could be 40 turns earlier than the fastest space (10 techs, 4 turns each etc). All the other modifiers are identical for diplo and space.
That being said what are the other modifiers that reflect the tech rate? The number of civs and the land mass, size, type, amount of commerce available. One thing that is a bit troubling is how to balance the necessity to grab land via force to improve tech rate, more cities more citizens more commerce, and still keep the peace and get the votes.
Yes the best players may get the Fission Tech in turn 250, build the UN with a very well timed prebuild but everyone they left alive is furious. Will you be able to win??
The whole reason I even thought about this today is Karasu's data on GOTM16 and 17. Diplo and space have lower scores than the domination, cultural and conquest games. My question is why?? Is it the effect that to win my diplo and space doesn't require a huge empire (effecting the in game score) ? It only requires a faster tech pace than the AI and some well planned prebuilds? Or is it that the modifiers need to be looked at again? I think the data from 2 games isn't enough but from what I can tell so far my diplo win in GOTM18 is a crappy score. Domination and Culture will smoke it. Oh well??
Hotrod
The map was not a typical standard map and I am trying to grasp how the best dates were determined for GOTM 18. Specifically a diplo victory in the 880's. The base modifiers for Diplo and SS, from what I can tell, are based on when the technology in the most well played game can be researched.
It makes sense that the diplo victory could be 40 turns earlier than the fastest space (10 techs, 4 turns each etc). All the other modifiers are identical for diplo and space.
That being said what are the other modifiers that reflect the tech rate? The number of civs and the land mass, size, type, amount of commerce available. One thing that is a bit troubling is how to balance the necessity to grab land via force to improve tech rate, more cities more citizens more commerce, and still keep the peace and get the votes.
Yes the best players may get the Fission Tech in turn 250, build the UN with a very well timed prebuild but everyone they left alive is furious. Will you be able to win??
The whole reason I even thought about this today is Karasu's data on GOTM16 and 17. Diplo and space have lower scores than the domination, cultural and conquest games. My question is why?? Is it the effect that to win my diplo and space doesn't require a huge empire (effecting the in game score) ? It only requires a faster tech pace than the AI and some well planned prebuilds? Or is it that the modifiers need to be looked at again? I think the data from 2 games isn't enough but from what I can tell so far my diplo win in GOTM18 is a crappy score. Domination and Culture will smoke it. Oh well??
Hotrod
- Joined
- Nov 16, 2001
- Messages
- 2,686
GOTM18 was treated as a Large map based on the number of land tiles. Everything else was the same as it normally would be.
Diplo is a victory condition that there just haven't been as many well played games of in the GOTM as most of the other victory conditions. Last month Alexman won the silver with a hybrid early/milked type Diplo game though, and my Diplo game in GOTM14 would have come in 4th if the Jason score was being used. If DaveMcW, Ribannah, or Kemal (among others) were calling a vote 40 turns before their very early Spaceship victories their Diplo score would be right up there with their Spaceship ones (probably not exact, but close). So I'm pretty sure the potential for 10k Diplo games is there.
It's pretty easy to expand to the Domination limit and still get a Diplomatic win. The hard part is if you go for milking (in which case score trumps date), and so to save resources cut all the AI back as much as possible... it tends to make them less likely to vote. Just getting to the domination limit means you can leave 2, maybe 3, voters untouched. As long as you don't break any deals, they really don't care much about how you treated the extinct ones.
I would expect more very good Diplo games to be submitted once people realize the competition in that category is less intense than in the conquest/domination/spaceship.
Diplo is a victory condition that there just haven't been as many well played games of in the GOTM as most of the other victory conditions. Last month Alexman won the silver with a hybrid early/milked type Diplo game though, and my Diplo game in GOTM14 would have come in 4th if the Jason score was being used. If DaveMcW, Ribannah, or Kemal (among others) were calling a vote 40 turns before their very early Spaceship victories their Diplo score would be right up there with their Spaceship ones (probably not exact, but close). So I'm pretty sure the potential for 10k Diplo games is there.
It's pretty easy to expand to the Domination limit and still get a Diplomatic win. The hard part is if you go for milking (in which case score trumps date), and so to save resources cut all the AI back as much as possible... it tends to make them less likely to vote. Just getting to the domination limit means you can leave 2, maybe 3, voters untouched. As long as you don't break any deals, they really don't care much about how you treated the extinct ones.
I would expect more very good Diplo games to be submitted once people realize the competition in that category is less intense than in the conquest/domination/spaceship.
According to Bamspeedy research, every time we eliminate a civ or raze a city of some other civs (none of their friends, of course), we will get an attitude mark against us too. Enough of those negative attitudes, they will become permanently furious at us and they will never vote for us no matter what. However, it's possible to trick them into voting for us by luring them into an alliance with us against the other remaining civ.
hotrod0823
Deity
Thanks for the response! I am just interested to know if the fastest space victors think they would've won with a vote or if the AI was all furious.
Hotrod
Hotrod
SirPleb
Shaken, not stirred.
For sure! [Edit: chopped a slight GOTM spoiler from here. Suffice to say that although I didn't do this, I do think that passing 10K in GOTM18 with an early diplo win is possible.]
With two exceptions: each war declaration you make, and each city you raze (or abandon) counts for a permanent one point attitude hit with all Civs that know your target. I estimate that the safe limit is around 25 such hits - with that many (and none of the worse hits for breaking deals, and with no large attacks against your friends), it should still be possible to get a vote at the end. Over 30 could even work but it would be dodgy, you'd have to work it out carefully and consider favorite and shunned governments. Subtract 10 from my numbers here if you want to get the vote at the end without being at war and in an alliance against the other contender in the UN vote.
Bamspeedy
CheeseBob
Originally posted by Moonsinger
According to Bamspeedy research, every time we eliminate a civ or raze a city of some other civs (none of their friends, of course), we will get an attitude mark against us too. Enough of those negative attitudes, they will become permanently furious at us and they will never vote for us no matter what. However, it's possible to trick them into voting for us by luring them into an alliance with us against the other remaining civ.
This is mostly correct, but you do not get a penalty for eliminating a civ (unless you razed the city, of course).
DaveMcW
Deity
- Joined
- Oct 8, 2002
- Messages
- 6,489
Code:
Map size modifiers for 'best' date
tin sma sta lar hug
diplomatic -30 -15 0 -15 -30
space race -30 -15 0 -15 -30In my opinion, the modifiers should be zero for all map sizes.
- Joined
- Nov 16, 2001
- Messages
- 2,686
I must not have updated the modifier.xls file or posted the latest changes, sorry. I'll get a completely updated version of the modifiers posted shortly. Input on the modifiers is always welcome, and certainly there is room for improvement.
The GOTM17 and 18 best dates for diplomatic and spacerace were using less drastic modifiers for map size. Both are at -15, -7, 0, -7, -15.
Part of the 'large' modifiers is to account for the number of goody huts. The slowest techs on bigger maps certainly are the Ancient/early Middle Ages ones, but that is somewhat offset by the increased number of techs given out by goody huts both to the player and AI. Then the corruption discrepancy kicks in to make later techs faster (or at least easier to maintain the faster rate). Might still be too big of a modifier, but I do think Large and Huge maps have slightly faster research up to the Modern era when compared to Standard.
Some other slight factors are the increased number of AI's makes it more likely to have all the starting techs represented, and that the corruption difference allows for leaving an extra productive AI (or the other AI's more productive) past the domination limit. The number of AI factor should probably be seperated out and given it's own set of modifiers. Also, the number of land tiles per civ (given default # of civs) is higher the bigger the map. That allows for more peaceful expansion, which helps compensate for a longer expansion necessary to get to a optimum research capacity.
The Tiny and Small modifiers I think you are right on. The huts work against them, and the research advantage is mostly played out by the mid Middle Ages. Most of the other factors are slight enough that they won't make a difference one way or the other.
The GOTM17 and 18 best dates for diplomatic and spacerace were using less drastic modifiers for map size. Both are at -15, -7, 0, -7, -15.
Part of the 'large' modifiers is to account for the number of goody huts. The slowest techs on bigger maps certainly are the Ancient/early Middle Ages ones, but that is somewhat offset by the increased number of techs given out by goody huts both to the player and AI. Then the corruption discrepancy kicks in to make later techs faster (or at least easier to maintain the faster rate). Might still be too big of a modifier, but I do think Large and Huge maps have slightly faster research up to the Modern era when compared to Standard.
Some other slight factors are the increased number of AI's makes it more likely to have all the starting techs represented, and that the corruption difference allows for leaving an extra productive AI (or the other AI's more productive) past the domination limit. The number of AI factor should probably be seperated out and given it's own set of modifiers. Also, the number of land tiles per civ (given default # of civs) is higher the bigger the map. That allows for more peaceful expansion, which helps compensate for a longer expansion necessary to get to a optimum research capacity.
The Tiny and Small modifiers I think you are right on. The huts work against them, and the research advantage is mostly played out by the mid Middle Ages. Most of the other factors are slight enough that they won't make a difference one way or the other.
- Joined
- Nov 16, 2001
- Messages
- 2,686
I recently did a complete wipe of my hard drive, and seem to not have backed up the Excel modifier file. I'll get one together, but in the meantime, here is the modifier.h file (you can open it in any text editor) used for the latest build of the utility that analyzes the map and outputs the best dates. It's a bit hard to read the modifiers from, sorry.
- Joined
- Nov 16, 2001
- Messages
- 2,686
One other point... on the 'crossing' modifiers, Conquest and Domination sometimes use different settings than each other. This happens when the domination limit is available on one setting, but 1 or more AI are 'farther' away.
If say, a Pangaea map turns out to have 1 AI on a smallish island that can only be safely reached by Astronomy, and all the rest of the land is connected, then the Domination modifier would be the 'Land' one, and the Conquest modifier would be the 'Astronomy' one.
All the other victory conditions use the same crossing setting as Domination.
If say, a Pangaea map turns out to have 1 AI on a smallish island that can only be safely reached by Astronomy, and all the rest of the land is connected, then the Domination modifier would be the 'Land' one, and the Conquest modifier would be the 'Astronomy' one.
All the other victory conditions use the same crossing setting as Domination.
pterrok
Warlord
You should maybe also check the map to see if there are ANY one square islands--it seems that some Civ will ALWAYS settle them and then you need to get up to Marines to take them out for Conquest!
I guess then you'd also have to see if they needed Astronomy to get there in the first place...just gets more and more complicated, doesn't it?
I guess then you'd also have to see if they needed Astronomy to get there in the first place...just gets more and more complicated, doesn't it?
- Joined
- Nov 16, 2001
- Messages
- 2,686
Good point pterrok. Not sure how to account for that though, as setting the best date based on a 1 tile Island would really screw things up if the player beat the AI to it. Could get submissions beating the Conquest date by 100 turns or more.
Probably will just leave it alone and make sure there aren't any 1 tiles islands when the maps are generated. If there was an island like that, the player would just have to go for one of the other victory conditions or 'milk' up to Marines and finish the conquest at that point.
Probably will just leave it alone and make sure there aren't any 1 tiles islands when the maps are generated. If there was an island like that, the player would just have to go for one of the other victory conditions or 'milk' up to Marines and finish the conquest at that point.
I concur! Acording to Bamspeedy's study and my own observation on the last game or two, we notice that the Jason score for Conquest and Domination are always going up a few points all the way to 2050 AD (sometimes, they do drop a few points but they do gain back their losts within a dozen turns). Basically, we don't lose points and we don't really gain anything in delay Conquest or Domination. The only thing we lose is the medal for fast finish (but most of us us aren't playing for medal). Therefore, in the worst case, we can always milk up to Marines then finish them up as you has sugguested.
I think there's an issue with map size that's completely unaccounted for. When you're going for an extremely fast tech game it's really everything you can do to maintain the 4 turns per tech rate. You have very little opportunity to expand your territory after you've already gained control of your FP and Palace cores. Therefore your territory Firaxis score will be very low, unless you slow down your tech pace, which goes against what you're trying to do.
Now what does this have to do with map size? Well the map size already accounts for tech costs. The smaller the map, the less techs cost. Therefore whether it's a small map or a large map you'll still achieve victory around the same date. That means your Firaxis bonus points for an early finish will be the same. Now the other part of your score is territory/happiness. Happiness is pretty much based on territory, because the more territory you have the more potential happy citizens you have. So if you go for an extremely fast tech pace then your Firaxis score for territory/happiness will be fairly low compared to people going for other victory conditions.
The problem is, the territory/happiness part of your Firaxis score is much smaller on a smaller map than it is on a larger map. While the bonus points for early victory is the same either way. That means a fast tech game on a large map will fall farther behind in score compared to players going for other victory conditions, while on a smaller map they will still remain much closer. This makes it harder to achieve a high Jason score with a fast tech game on a large map than it is on a small map.
I thought of this based on DaveMcW's diplomatic victory in GOTM18. He achieved an extremely fast tech pace, but just didn't acquire enough territory to get a high Jason score. I believe if it had been a smaller map his Jason score would have been much higher.
edit: This actually applies to any game where you're going for an early finish. On a smaller map the early finish bonus is a much larger proportion of your overall Firaxis score. Therefore in a small map you'll want to finish as early as possible, while on a large map it may be more beneficial to delay your victory and milk it more. But it's more apparent with a victory condition that prevents you from maximizing territory, like fast tech games.
Now what does this have to do with map size? Well the map size already accounts for tech costs. The smaller the map, the less techs cost. Therefore whether it's a small map or a large map you'll still achieve victory around the same date. That means your Firaxis bonus points for an early finish will be the same. Now the other part of your score is territory/happiness. Happiness is pretty much based on territory, because the more territory you have the more potential happy citizens you have. So if you go for an extremely fast tech pace then your Firaxis score for territory/happiness will be fairly low compared to people going for other victory conditions.
The problem is, the territory/happiness part of your Firaxis score is much smaller on a smaller map than it is on a larger map. While the bonus points for early victory is the same either way. That means a fast tech game on a large map will fall farther behind in score compared to players going for other victory conditions, while on a smaller map they will still remain much closer. This makes it harder to achieve a high Jason score with a fast tech game on a large map than it is on a small map.
I thought of this based on DaveMcW's diplomatic victory in GOTM18. He achieved an extremely fast tech pace, but just didn't acquire enough territory to get a high Jason score. I believe if it had been a smaller map his Jason score would have been much higher.
edit: This actually applies to any game where you're going for an early finish. On a smaller map the early finish bonus is a much larger proportion of your overall Firaxis score. Therefore in a small map you'll want to finish as early as possible, while on a large map it may be more beneficial to delay your victory and milk it more. But it's more apparent with a victory condition that prevents you from maximizing territory, like fast tech games.
Shillen,
I agree with most of what you said. The best type of victory condition for the Jason Scoring System is different in every game. For example, in the GOTM17, conquest or domination would yield the best possible score. In the GOTM18, cultural victory would yield better than spaceship, etc. IMO, that is the most wonderful thing about the Jason Scoring System is that we don't know what victory is best until we discover the whole worldmap. Therefore, we can't just play the same old boring style over and over again and expect to always end up with the highest possible score.
I agree with most of what you said. The best type of victory condition for the Jason Scoring System is different in every game. For example, in the GOTM17, conquest or domination would yield the best possible score. In the GOTM18, cultural victory would yield better than spaceship, etc. IMO, that is the most wonderful thing about the Jason Scoring System is that we don't know what victory is best until we discover the whole worldmap. Therefore, we can't just play the same old boring style over and over again and expect to always end up with the highest possible score.
- Joined
- Nov 16, 2001
- Messages
- 2,686
There are still some issues with most of the modifiers, and they will be refined as much as possible. The more examples of 'fast finishes' we get the better the modifiers should be able to be set.
The point you've brought up is a factor that I've tried to account for. My experience is that the tech rate being capped at 4 per turn techs makes late middle age and industrial conquest feasible while keeping the tech rate at or near maximum capacity. The domination limit should allow the player to leave 2-3 productive AI's to help the tech rate along.
As regards to the mapsize/tech rate/ingame score balance, there are several other factors involved. These are some of the things the modifiers/curve try to account for:
The OCN:Tile ratio doesn't match up throughout the map sizes. The larger the map, the more the ratio climbs in favor of the OCN. This allows for a faster middle and late tech rate (or rather one that is easier to maintain at 4 per turn) the larger the map. This somewhat balances the difficulty of continued expansion on larger maps, as more resources can be dedicated to military than could be otherwise. The effect starts to kick in when the Palace/FP cores have been set up, and so does a pretty good balancing act.
The way the AI handles it's empires also doesn't scale very well through the map sizes. While the OCN:Tile ratio does help it, as the OCN seems to dictate about how large an empire the AI will try for, the AI doesn't seem to handle larger empires anywhere near as efficiently as smaller ones.
The AI military is one of the most effected parts of AI management. They'll do dumb things like send their SOD to take an undefended city on the other side of the world, leaving their cities mostly open for the taking. The larger the map, the more room for error in the AI's military undertakings. PtW seems to have helped a bit in this regard, but not much. Another point on military is that any starting advantage the AI has in units isn't as big a deal on larger maps. On a small map, 5 more units may be a big deal, but as the map gets larger and larger, their overall effect becomes less and less. Overall, I'd say military expansion is easier the more extreme the map settings, relatively speaking. Smaller maps have AI's that just don't have many units/cities, and the player can rather quickly eliminate the competition. Larger maps have AI's that while relatively stronger, have much less focus. A good portion of their units can often be bypassed entirely.
Other factors are the Tile:Civs:OCN ratios, number of civs, number of goody huts, time needed to make contacts, ect.
The more discussion on these points, the better the modifiers can be set. So any points made about them are welcome.
The point you've brought up is a factor that I've tried to account for. My experience is that the tech rate being capped at 4 per turn techs makes late middle age and industrial conquest feasible while keeping the tech rate at or near maximum capacity. The domination limit should allow the player to leave 2-3 productive AI's to help the tech rate along.
As regards to the mapsize/tech rate/ingame score balance, there are several other factors involved. These are some of the things the modifiers/curve try to account for:
The OCN:Tile ratio doesn't match up throughout the map sizes. The larger the map, the more the ratio climbs in favor of the OCN. This allows for a faster middle and late tech rate (or rather one that is easier to maintain at 4 per turn) the larger the map. This somewhat balances the difficulty of continued expansion on larger maps, as more resources can be dedicated to military than could be otherwise. The effect starts to kick in when the Palace/FP cores have been set up, and so does a pretty good balancing act.
The way the AI handles it's empires also doesn't scale very well through the map sizes. While the OCN:Tile ratio does help it, as the OCN seems to dictate about how large an empire the AI will try for, the AI doesn't seem to handle larger empires anywhere near as efficiently as smaller ones.
The AI military is one of the most effected parts of AI management. They'll do dumb things like send their SOD to take an undefended city on the other side of the world, leaving their cities mostly open for the taking. The larger the map, the more room for error in the AI's military undertakings. PtW seems to have helped a bit in this regard, but not much. Another point on military is that any starting advantage the AI has in units isn't as big a deal on larger maps. On a small map, 5 more units may be a big deal, but as the map gets larger and larger, their overall effect becomes less and less. Overall, I'd say military expansion is easier the more extreme the map settings, relatively speaking. Smaller maps have AI's that just don't have many units/cities, and the player can rather quickly eliminate the competition. Larger maps have AI's that while relatively stronger, have much less focus. A good portion of their units can often be bypassed entirely.
Other factors are the Tile:Civs:OCN ratios, number of civs, number of goody huts, time needed to make contacts, ect.
The more discussion on these points, the better the modifiers can be set. So any points made about them are welcome.
Hmm maybe I'm just not good enough yet, but I don't find a late Middle or Industrial Age conquest possible while maintaining a 4 turn per tech pace. Your cities always have more important things to build. You usually can't even achieve the 4 turn per tech rate until the mid-late Middle Age. This is usually about when half your cities have universities. At this point you're finishing up marketplaces and universities in your cities to prepare for the next jump up to harder techs at the start of the Industrial Age. If you're good you just barely get Steam Power in 4 turns. Then you get sanitation and industrialization and need to build factories/hospitals in all your cities. You also need to get ToE and Hoover. Then a mere 40 turns later you're in the Modern Age and you get research labs and the UN. I'd say that's about the first time I have a chance to start building military and it's usually just enough to get a few mech infantries in my cities in case of an AI attack before I launch my ship.
But I admit I'm still learning. Maybe my next attempt at it I'll achieve that 4 turn tech rate more easily and be able to devote some production to military.
But I admit I'm still learning. Maybe my next attempt at it I'll achieve that 4 turn tech rate more easily and be able to devote some production to military.
Ronald
Emperor
Is there a simple way to find out what the maxscore of a given map is? I was trying to understand and apply the formula mentioned at the Jason calculator page, but unfortunately I'm completely lost
Who can help?
Thanks
Ronald
Who can help?
Thanks
Ronald
- Joined
- Nov 16, 2001
- Messages
- 2,686
The MaxScore is determined by the tiles on the map and game settings. All the food is added up on the map and divided by the number of land/coast tiles to give an average food per tile. 2 is subtracted from that as support for the laborer working the tile and the remainder is multiplied by the Domination limit to calculate the number of Specialists the map can support. That is added to the Domination limit * 3 (Laborers = 2, Tile = 1). Then everything is multiplied by the difficulty level to give a max turn score. Then the MaxTurnScore is multiplied by the score modifiers which take into account the mapsize, difficulty, and crossing level. Normally that multiplier is somewhere around 6/10.
So the formula is:
AverageFoodPerTile = TotalFood / TotalLandCoast
Specialists = (AverageFoodPerTile - 2) * DominationLimit
Laborers = DominationLimit
MaxTurnScore = ((Laborers * 2) + Specialists + DominationLimit) * Difficulty
MaxScore = MaxTurnScore * MapModifier
To get MapModifier, add up the relevant score modifiers. Here they are:
Crossing level is at what point in the game all the AI are safely reachable. 'Snaking' is for those maps where the land does just that, where even though all the AI are reachable by land or galley crossings, there are many hops needed.
Mapstat will give a pretty good idea of how much food is on the map but it won't be quite the same as the official number. The utility I use gives the food available 'under' forests (as they can be on Tundra, Plains, or Grassland) and including all bonus food resources.
So the formula is:
AverageFoodPerTile = TotalFood / TotalLandCoast
Specialists = (AverageFoodPerTile - 2) * DominationLimit
Laborers = DominationLimit
MaxTurnScore = ((Laborers * 2) + Specialists + DominationLimit) * Difficulty
MaxScore = MaxTurnScore * MapModifier
To get MapModifier, add up the relevant score modifiers. Here they are:
Code:
// map size score modifiers
Tiny 0.0060
Small 0.0045
Standard 0.0030
Large 0.0015
Huge 0.0000
// difficulty score modifiers
Cheiftain 0.0075
Warlord 0.0060
Regent 0.0045
Monarch 0.0030
Emperor 0.0015
Deity 0.0000
// crossing score modifiers
Land 0.63
Galley 0.62
Galley+Lighthouse 0.60
Snaking 0.572
Astronomy 0.57
Navigation 0.56Crossing level is at what point in the game all the AI are safely reachable. 'Snaking' is for those maps where the land does just that, where even though all the AI are reachable by land or galley crossings, there are many hops needed.
Mapstat will give a pretty good idea of how much food is on the map but it won't be quite the same as the official number. The utility I use gives the food available 'under' forests (as they can be on Tundra, Plains, or Grassland) and including all bonus food resources.
