Hi,
on 1on1 Ladder games, we mostly have the rule, that
a urich ng or 2 riches/urich-HG on HW is a restart.
A urich ng or a native around 6parsec is a restart, too.
I often have corners which were unbalanced, from the corner of the other player. This is very huge time of waste, cos sometimes we need a huge amount of restarts to get a ~equal start to both.
Is it possible to fix, that there´s no urich ng or 2 riches/urich-HG on HW and no urich ng or a native around 6parsec ?
Is it possible to do, that corners r more balanced ?
... thx ...
Balance Switch
- Lord Brazen
- Site Admin
- Posts:162
- Joined:Mon Jul 18, 2005 10:16 pm
- Location:Toronto, Canada
- Contact:
I was thinking of a map balancing switch myself but I am unsure on what method to implement to balance things. I have several ideas.
I could implement a simple mirroring algorithm where every quadrant has exactly the same systems or a simple flat map where all the systems are the same. These things would not be hard to do but, I personally find such maps boring.
I prefer something more complex. The main problem is that I am unsure what to consider "balanced". How do you rate the systems? You would have to assign some sort of value to each and then equalize the map so that everyone has systems of equal value.
To further complicate things, I think you would have to assign higher value to a good system that is near a home system vs one that is identical but farther away.
The number of planets in a system would definitely need to be factored in. A system with 5 average planets would have more value to me than a system with a single rich planet early in the game. Later in the game I may colonize a rich singleton and move some workers to it.
Also, some may argue that certain systems have more value to certain races. This idea really complicates things. Some people think only in terms of production. For many races, if your empire has planets that only generate 1-2 food per worker, it is fatal.
It is extremely complicated. Its hard to decide where to begin![Confused :?](./images/smilies/icon_confused.gif)
I could implement a simple mirroring algorithm where every quadrant has exactly the same systems or a simple flat map where all the systems are the same. These things would not be hard to do but, I personally find such maps boring.
I prefer something more complex. The main problem is that I am unsure what to consider "balanced". How do you rate the systems? You would have to assign some sort of value to each and then equalize the map so that everyone has systems of equal value.
To further complicate things, I think you would have to assign higher value to a good system that is near a home system vs one that is identical but farther away.
The number of planets in a system would definitely need to be factored in. A system with 5 average planets would have more value to me than a system with a single rich planet early in the game. Later in the game I may colonize a rich singleton and move some workers to it.
Also, some may argue that certain systems have more value to certain races. This idea really complicates things. Some people think only in terms of production. For many races, if your empire has planets that only generate 1-2 food per worker, it is fatal.
It is extremely complicated. Its hard to decide where to begin
![Confused :?](./images/smilies/icon_confused.gif)
"Stars are holes in the sky from which the light of the Infinite shine through." - Confucius.
Hmm sounds good imo.I could implement a simple mirroring algorithm where every quadrant has exactly the same systems
Map is generated Random, like we know it, but map has 4 same quadrants ?
I would be happy with such a switch
![Very Happy :D](./images/smilies/icon_biggrin.gif)
Yea thats a really true. The main problem is that I am unsure what to consider "balanced".
![Sad :(](./images/smilies/icon_sad.gif)
The Start and 6parsec around, is that, what is the most important to balance.
Important is that both get ~equal planets, on HW. Not that the one has 3 huge abd ngs and the other 1 poor med ng and 2 tiny upoors lg.
Maybe a switch to ~equal HWs.
Thats bit hard I guess. Both players should have a ~equal first system and further systems to expand a bit.How do you rate the systems?
~equal = a system with 2-3 small-large poor-rich planets. (not 3 riches ng to 2 poors, thats clear).
I Think so too.To further complicate things, I think you would have to assign higher value to a good system that is near a home system vs one that is identical but farther away.The number of planets in a system would definitely need to be factored in. A system with 5 average planets would have more value to me than a system with a single rich planet early in the game. Later in the game I may colonize a rich singleton and move some workers to it.
Also, some may argue that certain systems have more value to certain races. This idea really complicates things. Some people think only in terms of production. For many races, if your empire has planets that only generate 1-2 food per worker, it is fatal.
A switch for equal HWs is a good idea imo, thats a start ...It is extremely complicated. Its hard to decide where to begin
![Wink :wink:](./images/smilies/icon_wink.gif)
thats standard.....(natives should be separeted imho)....a urich ng or 2 riches/urich-HG on HW is a restart.
just this as ladder switch would reduce many restarts.....actually i dont want too balanced maps....so i am not that interested in the others proposals here....
True. Thats the reason.Also, some may argue that certain systems have more value to certain races.
equal
Perhaps create a quadrant and mirror that quadrant *4 rotating as appropriate. This way the maps would be different every time and equal to both or all 4 sides.
Just a thought...
Just a thought...
I propose the following algorithm as a starting point for a "neighborhood" balance switch. LB has seen this, however when I first proposed it he did not have room for the code (this was a while ago, I think he now has more space).
Trash it, or suggest changes as you see fit. This is in the spirit of brainstorming toward a good proposal.
1. Generate the map in the "normal" fashion to use as a starting point.
2. Randomly pick one homeworld system to base the balance process on. The 6 (or more, or less) systems closest to this homeworld constitutes the "base neighborhood".
3. ***** NOTE: THIS STEM MAY NOT BE NECESSARY. SUNDIVER'S MAP LEVELLER PROGRAM HAD ISSUES WITH REMOVING PLANETS AND LEAVING EMPTY ORBITS: fill all empty orbits in the base neighborhood with either a gas giant or asteroid belt.
4. Copy the 6 systems of the base neighborhood into the 6 closest systems for each of the other homeworlds in the game.
The number of systems that constitutes the "neighborhood" may need to be changed. Or, perhaps the switch itself can be used to call it out, for example:
/hood=8
Flaws I see in this algorithm:
1. If Step 3 above is required, then there is no chance of "empty" systems in the neighborhood. Empty systems occasionally play a role in a fleet's strategic moves. I'm sure we've all had the joy of moving a fleet to an un-colonized system deep in the heart of enemy territory with the intent of building a couple of outposts and having several potential targets to attack next turn, only to find that the system we intended to jump off from could not support an outpost...Not a huge flaw, but it could have an impact.
2. This algorithm does nothing to equalize the number of systems that are in range of initial fuel technology. Perhaps it does not matter, though, as it would incorporate a little bit of randomness into the situation.
3. Not a flaw, but there would need to be an upper limit placed on neighborhood size, based on the galaxy size being used. A neighborhood size of, say, 20, Could have a lot of overlap in a large galaxy.
I have no idea if the presence of monsters in the neighborhood would present implementation issues.
There you go, a proposal to love, hate, or modify.
![Smile :)](./images/smilies/icon_smile.gif)
-Gusset
Trash it, or suggest changes as you see fit. This is in the spirit of brainstorming toward a good proposal.
1. Generate the map in the "normal" fashion to use as a starting point.
2. Randomly pick one homeworld system to base the balance process on. The 6 (or more, or less) systems closest to this homeworld constitutes the "base neighborhood".
3. ***** NOTE: THIS STEM MAY NOT BE NECESSARY. SUNDIVER'S MAP LEVELLER PROGRAM HAD ISSUES WITH REMOVING PLANETS AND LEAVING EMPTY ORBITS: fill all empty orbits in the base neighborhood with either a gas giant or asteroid belt.
4. Copy the 6 systems of the base neighborhood into the 6 closest systems for each of the other homeworlds in the game.
The number of systems that constitutes the "neighborhood" may need to be changed. Or, perhaps the switch itself can be used to call it out, for example:
/hood=8
Flaws I see in this algorithm:
1. If Step 3 above is required, then there is no chance of "empty" systems in the neighborhood. Empty systems occasionally play a role in a fleet's strategic moves. I'm sure we've all had the joy of moving a fleet to an un-colonized system deep in the heart of enemy territory with the intent of building a couple of outposts and having several potential targets to attack next turn, only to find that the system we intended to jump off from could not support an outpost...Not a huge flaw, but it could have an impact.
2. This algorithm does nothing to equalize the number of systems that are in range of initial fuel technology. Perhaps it does not matter, though, as it would incorporate a little bit of randomness into the situation.
3. Not a flaw, but there would need to be an upper limit placed on neighborhood size, based on the galaxy size being used. A neighborhood size of, say, 20, Could have a lot of overlap in a large galaxy.
I have no idea if the presence of monsters in the neighborhood would present implementation issues.
There you go, a proposal to love, hate, or modify.
![Smile :)](./images/smilies/icon_smile.gif)
-Gusset
- Lord Brazen
- Site Admin
- Posts:162
- Joined:Mon Jul 18, 2005 10:16 pm
- Location:Toronto, Canada
- Contact:
Re: equal
I have thought of this simple mirroring algorithm but I prefer a less predictable universe. It would be very simple to code.Perhaps create a quadrant and mirror that quadrant *4 rotating as appropriate. This way the maps would be different every time and equal to both or all 4 sides.
The main problem with this is that your opponents would know where your homeworld is and where your best systems are. They would know exactly what you have and where to attack first.
"Stars are holes in the sky from which the light of the Infinite shine through." - Confucius.
- Lord Brazen
- Site Admin
- Posts:162
- Joined:Mon Jul 18, 2005 10:16 pm
- Location:Toronto, Canada
- Contact:
I do have a lot more space nowwhen I first proposed it he did not have room for the code (this was a while ago, I think he now has more space).
![Smile :)](./images/smilies/icon_smile.gif)
I like your idea and I am thinking of something along the same lines. Here is more or less what I was thinking:Trash it, or suggest changes as you see fit. This is in the spirit of brainstorming toward a good proposal.
1. Generate the map in the "normal" fashion to use as a starting point.
2. Randomly pick one homeworld system to base the balance process on. The 6 (or more, or less) systems closest to this homeworld constitutes the "base neighborhood".
3. ***** NOTE: THIS STEM MAY NOT BE NECESSARY. SUNDIVER'S MAP LEVELLER PROGRAM HAD ISSUES WITH REMOVING PLANETS AND LEAVING EMPTY ORBITS: fill all empty orbits in the base neighborhood with either a gas giant or asteroid belt.
4. Copy the 6 systems of the base neighborhood into the 6 closest systems for each of the other homeworlds in the game.
1. The galaxy is generated as normal (except monsters are not placed yet).
2. The map is scanned and a table of stars is generated composed of all stars within 6 parsecs of a homesystem (excluding the homesystem themselves).
3. The planets (from the stars collected in step 2) are removed from their systems and added to another table (called the planets table). Planets will include gas giants and asteroids.
4. The planets in the planets table are assigned a merit value for each player. The merit of a planet depends on the players race picks.
5. The planets are then divided up between the players equally based on the merit values of the planets.
6. The planets are then re-distributed to the stars collected in step 2 evenly (each system gets a planet from the table then each system gets another until they are all gone).
7. Monster systems are generated as normal.
I would expect this to have some problems. It would obviously have some problems on a small map...I'm still thinking about it.
"Stars are holes in the sky from which the light of the Infinite shine through." - Confucius.
- Lord Brazen
- Site Admin
- Posts:162
- Joined:Mon Jul 18, 2005 10:16 pm
- Location:Toronto, Canada
- Contact:
Not yet. This is where the hangup liesThat sounds like a very complicated task to define.
Do you have the beginnings of a formula or concepts in mind?
![Razz :P](./images/smilies/icon_razz.gif)
![Razz :P](./images/smilies/icon_razz.gif)
![Razz :P](./images/smilies/icon_razz.gif)
I am open to suggestions. I would like to see the following factors in the calculation:
1. Food production per worker factoring in race/planet G, race bonuses, race moral, government...etc. If all you have is a bunch of UR barrens, you may as well be dead. Farmers can't build ships.
2. Net production per worker factoring in race/planet G, polution, race bonuses, race moral, government...etc.
3. Planets worth as a research colony factoring in race/planet G, race moral, race bonuses...etc.
4. And of course the max population.
You should have equal footing regardless of your race picks (within reason). Your planets should make up for some of your races weaknesses (within reason).
The problem with balance currently as I see it is, if you pick the right race and you get the perfect planets for that race, you will win. The next game you could pick a different race, get the same planets, and not even have a chance at winning or vice versa. It is often just dumb luck.
You can pretty much tell within the first 30 turns if you have a chance of winning or not. This causes a lot of early game drop outs.
"Stars are holes in the sky from which the light of the Infinite shine through." - Confucius.
Well, but this is part of the game, isnt it?"The problem with balance currently as I see it is, if you pick the right race and you get the perfect planets for that race, you will win. The next game you could pick a different race, get the same planets, and not even have a chance at winning or vice versa. It is often just dumb luck."
Taking aqua is risky. Thats why it is just 5 points.
Sub is more secure and therefore 6.
When I understand your proposal correctly, and you want that the racepicks influence your neighborhood, I have no idea how you will make it "fair".
Some guy who prefers sub will claim that you will give aqua too many wets...and so on.
Balancing hws is a nice idea imho. Anyone can easily understand your "balancing".
Balancing neighborhood. Difficult. And Racepicks should have no influence IMHO.
The "mirror maps". I agree LB, they are boring - nevertheless I heard that some guys would like them. If there is demand....hmmm...prolly a good option for your patch then.
True, maybe a small terran/gaia with 1 or 2 med barren, would be a avrg food system.1. Food production per worker factoring in race/planet G, race bonuses, race moral, government...etc. If all you have is a bunch of UR barrens, you may as well be dead. Farmers can't build ships.
Hmmm minerals which depends on racepicks, isn´t a good idea imo, so HG-Player gets UrichHGs or not ? Dic gets better minerals than a uni race ?2. Net production per worker factoring in race/planet G, polution, race bonuses, race moral, government...etc.
A feudal Player gets artis, cos of research maulus ?3. Planets worth as a research colony factoring in race/planet G, race moral, race bonuses...etc.
I play often subaqua, and have 8-20pop often on each planet, all nonsubaqua players get more terrans and arid planets ?4. And of course the max population.
The main pick which depends on luck is Aqua, but aqua alone is often foold by map, so its really luck. Maybe we should slightly stabilize the wets on map, so that there r not so huge differents with wets in corner. In 65% of Maps aqua cant compete (compete with unitol), in 15% it can compete and in 10% aqua is just the best race u could choose, cos of tons of wets.The problem with balance currently as I see it is, if you pick the right race and you get the perfect planets for that race, you will win. The next game you could pick a different race, get the same planets, and not even have a chance at winning or vice versa. It is often just dumb luck.
To prevent from "dumb luck" u can take sub or tol races, which have more pop on alle planets. Or make a combo of sub and aqua
![Very Happy :D](./images/smilies/icon_biggrin.gif)
Thats why we need to balance our neighbourhood, but not depends on racepicks imo !!!You can pretty much tell within the first 30 turns if you have a chance of winning or not. This causes a lot of early game drop outs.
What I think and prefer about a balanced neighbourhood, look "Posted: Sun Jul 31, 2005 7:03 pm" in this Thread.
- Lord Brazen
- Site Admin
- Posts:162
- Joined:Mon Jul 18, 2005 10:16 pm
- Location:Toronto, Canada
- Contact:
Balancing neighborhood. Difficult. And Racepicks should have no influence IMHO.
I haven't really stated my thoughts clearly I guess. I see from what you both have stated, you seem to have read that I think all races should be equal.thats why we need to balance our neighbourhood, but not depends on racepicks imo !!!
What I think with respect to racepicks and balance is this. If player A spends 7 points on race picks that improve his empire developement and player B spends 10 on the same, then player B's empire should be able to be better. It should not be so much of a gamble.
No...and...NOHmmm minerals which depends on racepicks, isn´t a good idea imo, so HG-Player gets UrichHGs or not ? Dic gets better minerals than a uni race ?
This is exactly my point. What people are complaining about when it comes to balance is that it is a gamble. High cost picks should yield less risk. The map is pretty random right now. I think you should at least get bare minimum for your race to survive.Taking aqua is risky. Thats why it is just 5 points.
Sub is more secure and therefore 6.
I think picks and pick costs should be weighted in when it comes to map generation. There should be some randomness to things but not as much as there is.
Last edited by Lord Brazen on Fri Aug 05, 2005 11:19 am, edited 1 time in total.
Who is online
Users browsing this forum: No registered users and 93 guests