Perhaps one useful constraint would be the average degree of nodes. Low degree gives us maps with many corridors, while high degree gives us bulky maps. Also perhaps there should be a minimum size for a continent (connected component). For example, we probably don’t want continents with <=5 nodes.
There needs to be rules for player spawns, we don’t really want islands or continents with only one player… (or zero players )
Little change which implies very different strategy :
Your PODs can’t move on an ennemy cell if there is at least one ennemy POD on it, and PODs have a 1-cell range gun. If opponents arrive at the same turn in a cell, fights occurs as now. To chose on which cell firing, it’s the same command as now for moves (if cell is free, you move on it, if not you fire on it with the number of PODs you selected).
The game can be blocked with this rule.
Let’s suppose that each player have enough areas to produce at least one pod each turn.
Let’s suppose again that the only area that can be attacked (different players) is a corridor.
Then each turn, player 1 add pods on this area, same for player 2, and the battle never finish.
Only one continent is good.
How about some neutrals start with pods. These can be chosen randomly or assigned based on neutral production. High neutral productions (>=4) can get 1 pod, the rest get 0. The neutral pods don’t move and just sit there. The rules for fighting against neutral pods is identical to normal fighting rules (majority wins). If there is a draw then you don’t take the neutral zone, as expected. With this rule there will be more strategy at the start where you can simply take any neutral. You will need to think which ones to go for, since you don’t want to waste pods early in the game. You will also need to build your pods in order to take some high production neutrals.
New rule: multiple bases. Each player starts with multiple bases. A new bot is spawned automatically (if there is enough platinum) into one of the bases chosen at random. What do you guys think?
Build expand for 80 platinum
The rule of week 2 (capture the flag) is a bit modified. Pods aren’t created automatically anymore. We have to buy it with command 2.
An expand is built with command 2 -1 position on a zone that belongs to the player.
An expand or a base can’t be built on a platinum zone and are destroyed when another player capture the zone.
So in coding our AI, we have to decide :
- Build pods but long path to reinforce.
- No pods but strategic position near platinum or enemy zone.
Like viraxay’s proposal, neutral HQ Base on map. For capture it, players need to sacrify 3 pods (ennemies or allies) on this area then need last pod for capture it.
- We can build new HQ base on “6 platinium zone”.
- 60 platinium to build HQ base
- The "6 platinum zone " stop production , the player sacrifices that mine.
Exchange rule: At buy phase we sacrifice 2 pods, and put 1 on any own zone.
Fog of war rule: At begining we dont know the number of platinum per zones, we have to visit every zone to reveal it’s platinum.
New rule: Mercenaries
Some groups of neutral PODs are randomly dispatch on the map at the start. The first army that joined the area recruits mercenaries. Since it’s now a 2-players game, mercenaries can be stored in the “pod2” or “pod3”, no matter.
As a ‘bonus’ rule for next week, please remove asymmetric random maps. They are pointless in the leaderboard and irritating in the IDE.
On asymmetric maps, two matches are played against the same opponent with reversed spawns.
Right, but 90% of the time these matches result in one win for each player, since most of the time there is a more and less favorable spawn position.
I’m also curious how this affects TrueSkill calculation. If, for example, my average bot (~35.00) is matched against one of the top players (~45.00) and we each get one win because of the asymmetric spawn, I still come out ahead because my TrueSkill score is lower. This does not seem fair to me.
If you get one win each in an asymmetric game, the game is not fed to TrueSkill, it is simply thrown away by our system… The system is fair…
When the game starts, there are some neutral bases scattered on the map. They can be captured just like other zones. If you own more than one base, you must choose how many drones each base gets on the next turn. Any drones which haven’t been accounted for will be spawned at a random base.
To win the game you either take all the opponent’s bases or hold the most territory when all turns are exhausted.
Thanks for the response. But that just means your servers waste CPU time on these games…
Not every pair of match results in a win/lose for each contestant.