Random Scenarios

[doublydavebrown]

I was thinking about the longevity of games the other day, don't ask me why, and it occurred to me what really gives a game long lifespan: random scenarios. The way you can just fire up the game, select a few variables, and play something completely new, completely unique, without relying on the community to deliver new scenarios.

So then I got thinking about the Advanced Tactics random game engine. It's a significant improvement over People's Tactics, but I still feel there's a few things lacking. Here's what I think would make it better:

O The ability to select the % of the map/population/cities occupied by each player or unoccupied (if city occupation, then the map occupation follows such that borders are halfway between cities)
O The ability to select starting units per player (requiring one page per player, with one slider per subformation type - the AI would be used to split these up into units)
O The ability to select the distribution of units (3 sliders for borders, cities and other)
O The ability to select alliances/war/peace per player
O The abiltiy to select between large and small island sizes
O The ability to select the amount of starting political points
O The abiltiy to select the size of the starting city (without editing the masterfile)
O The ability to select the technologies owned and available to each player (requiring one page per player, with one slider per technology, where 0 on the slider is already own, the slider is relative cost, and the other end of the slider is cannot own)
O The ability to select a turn limit
O The ability to edit the victory conditions
O The ability to select the amount of starting forts/ports/airstrips (one slider per location type - perhaps sliders for distribution - near cities, along borders)
O Perhaps some generic action cards that are randomly available (e.g. build new city, reinforcements, grow city, area changes side)
O A whole load more master files (WWI, Vietnam, Colonial, Napoleonic, Fantasy, Sci-fi, etc)
O The ability to select the AI production bonus from 0 to 250 % (thanks oufti)
O The ability to select side name, colour and flag for each player
O The ability to make a mirrored map with more than 2 players (thanks Scott_WAR)
O The ability to set the number of political points necessary to declare war (thanks Zaratoughda)
O The ability to set the turn at which war is automatically declared (thanks Zaratoughda)

Before anyone gets upset and thinks I'm criticising Advanced Tactics, that's not what I'm saying: no game designer can think of everything; and no programmer can implement everything a designer thinks of (and from what I understand Vic did both!). Advanced Tactics is a brilliant game. But if, on the offchance, any of these features could be added in a new release, I'm sure it would just be me who'd be very grateful!

Added 26th Jan '09
Added 28th Jan '09
Added 1st Feb '09

[oufti]

I would like to complete the previous list with :

The ability to select the AI production bonus from 0 to 250 %.

I agree Advanced Tactics is a brillant game.

[Joshuatree]

Well I agree with you, the random game engine is quite usefull, espcially combined with the editor... *but* I think the random game engine could indeed use some additions/improvements without going into the editor.
Sort of, make a quick North Africa scenario, press a few buttons and off you go.
Ofcourse, I have high hopes on AT2 [:D]

[Jeffrey H.]

I've come to the conclusion that one of the true limiting factors of a random game is initial placement. Usually, the player with the most favorable initial placement wins.

In some way, it would be neat to come up with some mitigating factor to initial placement, (other than simply generating another random map or "peeking" at the initial map and selecting a location.) 

One idea I came up with is to complement city size with political points, production points or some combination thereof, with a formula for discovery reward based on the discovering players current production base differential and time of discovery.

That way, a player with one city who struggles for many turns to find another city gets some equilibrium against a player who starts in a good position and quickly discovers several large cities. At the time the player discovers his second city, the city strength and or PP bonus is determined so that his production gap vs. the other player is eliminated.

There may be a way to do this within the editor but I think city strength is fixed and cannot be altered except by damage. I'm not very experienced with the editor, (not at all really) but there might be several ways to do this.  

[autarkis1967]

Maybe for AT2 a option for city growth. Base city size on population of the city. Each turn a city/town grows slightly larger. When a city is damaged part of its population is killed as well as infrastructure. This of course would need to be balanced out so that a megacity doesn't become a small village after 1 bombing run, but it would reward keeping your population safe. If this was implemented it would add a whole new bunch of ideas like national manpower for the army. Each turn a percentage of the population is available as manpower. Each time a unit is built it costs so much manpower. If no manpower is available no more units are built.  

[doublydavebrown]

Jeffery H,

I think the first bullet point will go a long way to alleviating the problems of unfair initial placement. For example, in a two player game you could set each player to have 30% of the starting population and then have 40% unoccupied. As there is less unoccupied land, there is less of a chance of early advantage. In fact, you can make it really fair by having all the map already occupied at 50% each.

I feel that your solution, whilst addressing the problem you highlighted more directly, feels "unnatural" or perhaps "unrealistic" (not that a game should ever strive for 100% realism).

What do you think?

[Jeffrey H.]

ORIGINAL: doublydavebrown

Jeffery H,

I think the first bullet point will go a long way to alleviating the problems of unfair initial placement. For example, in a two player game you could set each player to have 30% of the starting population and then have 40% unoccupied. As there is less unoccupied land, there is less of a chance of early advantage. In fact, you can make it really fair by having all the map already occupied at 50% each.

I feel that your solution, whilst addressing the problem you highlighted more directly, feels "unnatural" or perhaps "unrealistic" (not that a game should ever strive for 100% realism).

What do you think?

Well, part of the fun for me in random games is the exploration of a shrouded map and the subsequent limited descision making that comes with the limited understanding of the map. Also, your strategies must evolve as you discover more of the map.

So, if I understand your first bullet correctly, you would be essentially pre-allocating which doesn't leave room for discovery on a shrouded map, correct ?

I agree that my proposal might seem a little hoaky or 'rigged'. In addition, I'm trying to think through how such a system could be manipulated for advantage, which is how humans will play it, "the mouse will find the cheese", even if they do so by means we didn't expect.

Also, I'm not sure how to make such a 'bias' system as I propose AI compatible. More thinking required.....

Maybe this idea will be the thing that finally motivates me to get my feet wet with the editor.

[82ndtrooper]

I feel that you folks want to take the random out of the random games.

that is what makes them fun.

I agree with being able to choose your emblem and your tile color but the rest I cant agree with. ( I dont dissagree though)

[Twotribes]

ORIGINAL: Jeffrey H.

I've come to the conclusion that one of the true limiting factors of a random game is initial placement. Usually, the player with the most favorable initial placement wins.

In some way, it would be neat to come up with some mitigating factor to initial placement, (other than simply generating another random map or "peeking" at the initial map and selecting a location.) 

One idea I came up with is to complement city size with political points, production points or some combination thereof, with a formula for discovery reward based on the discovering players current production base differential and time of discovery.

That way, a player with one city who struggles for many turns to find another city gets some equilibrium against a player who starts in a good position and quickly discovers several large cities. At the time the player discovers his second city, the city strength and or PP bonus is determined so that his production gap vs. the other player is eliminated.

There may be a way to do this within the editor but I think city strength is fixed and cannot be altered except by damage. I'm not very experienced with the editor, (not at all really) but there might be several ways to do this.  

If there are only 2 of you use mirror map. Then each of you have the exact same placing.

[Scott_WAR]

I would like the ability to make a mirrored map with more than 2 players.

[Vic]

I am reading this. And i can tell you i am planning to do much more with random games in a future release. But it will take time.

[doublydavebrown]

Jeffery H,

You make a fair point that allocating the map limits the fun of exploration (although sometime you want a quick game and don't want to explore). Perhaps what you're looking for is some kind of option with even distribution of cities, such that each player has exactly the same distance to cities of equivalent size (I'm sure this would be possible with a bit of clever programming). Obviously, the direction and land in-between is still random, so there is still a chance of unfairness (e.g. you may have to cross a mountain range to get to a city whilst the opponent doesn't).

82ndtrooper,

I can see where you're coming from, but I think some of these suggestions add randomness. For example, if you start with a significant starting force, but the distribution is random, then you've added to the randomness. I think these suggestions are about adding to the diversity of random games, rather than removing any of their randomness.

Vic,

Thanks for listening. Anything you do will be warmly welcomed.

[Jeffrey H.]

ORIGINAL: Twotribes

ORIGINAL: Jeffrey H.

I've come to the conclusion that one of the true limiting factors of a random game is initial placement. Usually, the player with the most favorable initial placement wins.

In some way, it would be neat to come up with some mitigating factor to initial placement, (other than simply generating another random map or "peeking" at the initial map and selecting a location.) 

One idea I came up with is to complement city size with political points, production points or some combination thereof, with a formula for discovery reward based on the discovering players current production base differential and time of discovery.

That way, a player with one city who struggles for many turns to find another city gets some equilibrium against a player who starts in a good position and quickly discovers several large cities. At the time the player discovers his second city, the city strength and or PP bonus is determined so that his production gap vs. the other player is eliminated.

There may be a way to do this within the editor but I think city strength is fixed and cannot be altered except by damage. I'm not very experienced with the editor, (not at all really) but there might be several ways to do this.  

If there are only 2 of you use mirror map. Then each of you have the exact same placing.

Good point, I've done this myself once. Worked out pretty good.

[Jeffrey H.]

ORIGINAL: doublydavebrown

Jeffery H,

You make a fair point that allocating the map limits the fun of exploration (although sometime you want a quick game and don't want to explore). Perhaps what you're looking for is some kind of option with even distribution of cities, such that each player has exactly the same distance to cities of equivalent size (I'm sure this would be possible with a bit of clever programming). Obviously, the direction and land in-between is still random, so there is still a chance of unfairness (e.g. you may have to cross a mountain range to get to a city whilst the opponent doesn't).

82ndtrooper,

I can see where you're coming from, but I think some of these suggestions add randomness. For example, if you start with a significant starting force, but the distribution is random, then you've added to the randomness. I think these suggestions are about adding to the diversity of random games, rather than removing any of their randomness.

Vic,

Thanks for listening. Anything you do will be warmly welcomed.

That's another idea, (equidistant or quasi equidistant or some other system of distance to production capacity compensations to city spacing from the orginal start points.) that could work.

As for removing randomness, that really isn't the intent. My suggestions are about trying to reduce the effect of initial placement on the game outcome. After playing dozens of these random games, many vs. people, it's very often that initial placement decides the outcome. Less so vs. the AI because it's easier to beat. By placing less emphasis on initial placement, you can put more on gameplay styles.

But you do have a good point, these types of mods will inevitably, (almost by design ?) reduce some of the variability of the games. It's certainly a balancing act to keep the fun intact when making these sorts of changes.

Vic, appreciate your listening in. It's fun for me to extend the basic concepts out a bit and discuss them with others. That's it really, just having some fun with thought experiments. I have been waiting for something to spur me into action WRT tackling the editor, maybe there is something in this discussion that will do it for me.

[Gary Kerr]

One thing that may make the start of the Random games more balanced is to garrison all towns and cities.
 
The garrison would not grow and would be static, but it would force players to build a more balanced military before they simply spread out over the map.
 
The garrison would work best if it would get larger at the target gets larger.
 
A 500 pt value village would have no garrision
 
A 1000 pt value town would have say have a 20 inf garrision
 
A 5000 pt value small city would say have a 50 inf with 10 MG, 10 mortar
 
A 10000 pt value large city would say have a 100 inf with 10 MG, 10 mortar, 2 AT gun, 2 Flak
 
The only trick about this would be how the logistics would work, unit without HQ would lose readness as they don't get supplies. 
 
Just a thought
 
 

[Jeffrey H.]

Someone else pointed out that this is essentially what the "peoples tactics" option does in random games. It loads up neutral cities with some default regime garrison. I haven't tried it yet.
 
Variable garrison size is another interesting idea.

[Mike Parker]

I haven't played this game, but I was reading through it to get an idea and I came across this.
 
I would consider having a toggle to allow random setup to minimize initial setup advantage.
 
Remember I have never played this game so I am proceeding from alot of ignorance, but I think this base idea has some merit.
 
 
First you have to come up with some measure of 'distance' between any two cities.  This would be how difficult it is to move a unit between the two, it cannot be just straight distance it needs to factor in terrain and over the water things.  This isn't horrible and its easily tweaked.
 
Then when a random scenario is made, you would compare each side.  You would then take every neutral city in the game, and assign it a value for each side.  this value us the smallest 'distance' to a city on each side.  Then you just sum these values up (or even take the square root of the sum of the squares, that has some mathematical niceties to it) and the side with the larger is 'disadvantaged' so you can then consider moving cities around (probably only one) so they (it) are closer to the disadvantaged side until your totals are within a nicely defined threshold.
 
This would still keep the randomness, but would help make the initial setup closer to even in most situations.
 

[Jeffrey H.]

That's more or less what we've been describing. What you are missing is the production capacity of the city, (that's also a randomized variable) but that can be incorporated as a factor to your RSS concept. I'm thinking about how I could possbily use the ingame editor to do some of these manipulations. Unfortunately, things like city location are fixed by the map generation process. Changing them would require changes to the game code.
 
It should be possible to load a weighted value for the distance, difficulty of crossing that distance and production strength of each neutral city WRT the initial starting point for each starting regime. Then a leveling process can be initiated to offset some of the differential between the various regimes.  
 
I need to learn more about the editor to see what can be done. I have the feeling this sort of work is outside the capability of the editor.
 

[Vic]

ORIGINAL: Jeffrey H.

That's more or less what we've been describing. What you are missing is the production capacity of the city, (that's also a randomized variable) but that can be incorporated as a factor to your RSS concept. I'm thinking about how I could possbily use the ingame editor to do some of these manipulations. Unfortunately, things like city location are fixed by the map generation process. Changing them would require changes to the game code.

It should be possible to load a weighted value for the distance, difficulty of crossing that distance and production strength of each neutral city WRT the initial starting point for each starting regime. Then a leveling process can be initiated to offset some of the differential between the various regimes.  

I need to learn more about the editor to see what can be done. I have the feeling this sort of work is outside the capability of the editor.

Hi Jeff,

Actually city location is not fixed. Check the Random Towns scenario in the FICTIONAL directory to see how to place a town. Removing is also possible by specifiying -1 to the ExecChangeLocType.

Kind regards,
Vic

[Zaratoughda]

I have been reading this thread and... a lot of suggestions hard to say which would be more important and how difficult to implement each would be.

But, one recent idea of mine that might satisfy a lot of what is being asked here as well as related ideas..... is in the random scenarios being able to select the turn that war breaks out. If set to 0 then things would be as they are now. But, if instead it is set to turn 10..... then nobody would be able to attack others or move into hexes controlled by others until turn 10.

With this..... each side would be able to put an army together before attacking.... whereas now..... at least for me.... I usually attack immediately and that usually means with an army 'under construction'.

So, would be an interesting varient, IMO, to the way things are now... and should not be that difficult to implement.

A related idea... a political point cose to declare war. I believe this is the way things are now in scenarios where sides do not necessarily start at war. So, this one should be easy to implement... and then when the time came in a random scenario, you could see if any of the computer opponents declare war on you and if not, select which one of em you are gonna declare war on.

Z