Elemental Wars

Rate this contribution

+24

Make the contribution public (lvl 20+)

Comments

45 Comments

9 days ago

I hope you received enough feedback from the community to improve your contribution. Once you have made the necessary updates, publish your contribution!

Mat2099 Unnamed CodinGamer

11 days ago

really cool

N3l Unnamed CodinGamer

15 days ago

Hi, Playtested and went deep into the referee, the statement, and the three bosses while writing a bot for it. It's a solid, well-built game! :) The rock-paper-scissors design is geometrically fair (the three camps are pairwise equidistant in manhattan distance, obstacles are 4-fold symmetric), the capture-then-release ordering is handled correctly, and the referee code is clean and readable. A few suggestions that I think would maybe improve it: - Low leagues feel very swingy. Since the game ends the instant one tribe is fully captured, the 1–2 unit leagues often resolve in 2–6 turns, decided mostly by spawn geometry and who collides first. For a ladder that's a lot of variance --> in the 1-unit league you might play a fixed number of turns and rank by captures, or use best-of-N. - Eliminated players rank last even if they captured the most. So a player who scored 2–3 then got fully imprisoned finishes behind someone on 0. If that isn't intended, keep their capture count for ranking, otherwise it's worth stating explicitly that survival beats scoring. - Lots of 0–0 stalemate draws in League 3. On the larger maps, equal move speed makes a competent elemental nearly impossible to catch without cornering. I see many 200-turn 0–0 games, even boss vs boss. --> Smaller maps for the unit counts, more choke-point obstacles, or an aggression incentive would help. - The highest-leverage idea: make units block movement (right now everyone phases through everyone), that alone adds body-blocking/cornering and would cut the stalemates significantly. A couple of harsh, undocumented "instant loss" edges: - An empty or malformed MSG (e.g. MSG with no text) matches no pattern --> deactivates the player --> loss, - Exceeding the hidden unitsPerPlayer * 3 command cap --> "Too many commands" --> loss. Neither is in the statement. I'd ignore an empty MSG instead of killing the player, and document the command limit. Boss progression: The random --> greedy --> risk-map curve is good! Nice work overall! Upvoted! :)

kosynus Unnamed CodinGamer

15 days ago

Thank you for this deep review! I'll take your suggestions into consideration and get back to you.

Ne-Gok-Sa Unnamed CodinGamer

16 days ago

✅ Approved the contribution. Comment: This multiplayer game is very original and a lot of fun—especially the fact that you can play cooperatively with other players.

Ne-Gok-Sa Unnamed CodinGamer

17 days ago

This multiplayer game is very original and a lot of fun—especially the fact that you can play cooperatively with other players.

kosynus Unnamed CodinGamer

17 days ago

Thank you!

Athos-77 Unnamed CodinGamer

17 days ago

Hello, it's a very impressive work. Well done. I hope it will be approved and make public because I love this idea.

kosynus Unnamed CodinGamer

17 days ago

Thank you!

dcudys Unnamed CodinGamer

25 days ago

✅ Approved the contribution. Comment: This project deserves approval because it offers much more than a simple pathfinding challenge. Players must balance offense and defense, predict opponents' movements, and take advantage of the rock-paper-scissors relationship between the three elemental tribes. The simultaneous capture system creates interesting strategic situations where positioning and timing are critical. The game is easy to understand thanks to its simple rules, yet it provides significant depth for AI development. Participants can explore pathfinding, target prioritization, threat assessment, territory control, and opponent prediction. The presence of three players instead of two also makes the gameplay less predictable and encourages more diverse strategies.

kosynus Unnamed CodinGamer

23 days ago

Thank you!

TheCoderDeyi Unnamed CodinGamer

about 1 month ago

Could this be the new Codingame Challenge? looks pretty well designed, and really nice.

kosynus Unnamed CodinGamer

about 1 month ago

Thank you! I'm glad you like it. This is actually my first contribution built with the SDK, so feedback like this is really encouraging and motivates me to continue.

jacek Unnamed CodinGamer

about 1 month ago

It could use some UI like avatars in square at the side or top so one could know who's who by game screenshot. Also something like MSG a short message, either per elemental, or one next to avatar would be also good.

kosynus Unnamed CodinGamer

about 1 month ago

Yes, already on the roadmap 🙂

blasterpoard Unnamed CodinGamer

about 1 month ago

Simultaneous moves and the possibility of swapping positions with the opponent make captures impossible without luck, dedicating 4+ elementals to the capture, or someone entering a dead-end path while being next to 2+ predators. On top of that, 3+ player free-for-all games are particularly hard to design, and I'm not sure I see too much strategic depth here that doesn't boil down to guessing opponents' moves correctly, even if luck-based captures weren't a problem.

kosynus Unnamed CodinGamer

about 1 month ago

I’m actually quite aligned with your observation. Simultaneous movement makes direct captures harder and increases the prediction aspect. I still think there is some strategic depth through positioning, obstacle control, splitting groups, camp pressure, etc., but I agree the gameplay could benefit from more reliable ways to create tactical pressure. I’m definitely open to ideas and suggestions in that direction !

111ove Unnamed CodinGamer

about 2 months ago

nice

kosynus Unnamed CodinGamer

about 2 months ago

Thanks!

caparks Unnamed CodinGamer

about 2 months ago

This is very interesting. There's a lot of stuff here to enjoy.. layers and layers of a funion to peel back. Thanks for contributing! Obviously took considerable forethought and planning - the graphics too. This is a very robust project!

kosynus Unnamed CodinGamer

about 2 months ago

Thanks!

eulerscheZahl Unnamed CodinGamer

about 2 months ago

Thanks for the contribution, it looks really interesting. Not sure what a good strategy will look like (capturing some but not all prey helps your predators because they don't have to avoid their predators as much anymore) - I guess we will find out. Skimming through the code, it looks clean and readable, which is nice as well. I suggest adding tooltips with coordinates to each cell. Right now you give the position for each elemental, but the rest has to be done by counting. When you MOVE the same unit id twice, you lose the game. On CodinGame it's common to give a warning and ignore the action, as long as it's syntactically correct. For a perfectly symmetric map you would need a hexagon grid. Not saying that you should do that, it would make catching too hard I think.

kosynus Unnamed CodinGamer

about 2 months ago

Thanks a lot for the feedback, glad you like the concept! You're absolutely right about the strategy aspect: the RPS dynamic with 3 players creates some interesting trade-offs, especially around "how much to capture". I'm curious as well to see what emerges once more players join. Good point about the tooltips. I've added a setting to display the grid coordinates as tooltips, which should make navigation much easier. Regarding duplicate commands on the same unit id, I've updated the behavior to match standards: it now triggers a warning and ignores the extra command instead of losing the game. For the symmetry: yes, a hex grid would be the "perfect" solution mathematically. I intentionally stayed with a square grid to keep things simpler and more readable, and also to avoid making captures too difficult as you mentioned.

5DN1L Unnamed CodinGamer

about 2 months ago

Order of explanation * The overall readability would benefit from reordering the sections for a more logical flow of mechanics: Actions => Movement => Captures => Prisoners. * Reordering would also allow you to remove some redundancy, particularly the first paragraph in the current Movement section. Actions * Please format "x" and "y" in "(x, y)" using <var> tags. You may keep parentheses/comma unformatted or continue to wrap them in <const> if preferred. Movements * Please clarify whether movement is: - 4-directional (N, S, E, W), or - 8-directional (including diagonals). * Please specify a deterministic rule (e.g. fixed direction priority order) when multiple shortest paths exist. Capture * Please specify whether anything happens when prey enters an enemy camp. Game End * "At least one player has all their elementals prisoner.": Needs a wording fix. Suggested rewrite: At least one player has all of their elementals captured as prisoners. Technical Details * "The score of a team is the number of prisoners belonging to its prey team.": Sounds slightly awkward. Suggested rewrite: The score of a team is the number of prisoners it has captured. Inputs * Please state in the following values are integers in the relevant descriptions: - mapSize - elementsCount - elementalId - x - y Inputs / Default stub * Please change "prisoner" from bool to int, as some language stubs don't provide a bool to the player (e.g. Lua, Pascal). Use 0 / 1 instead of false / true. Output * Please format "x" and "y" in "(x, y)" using <var> tags. You may keep parentheses/comma unformatted or continue to wrap them in <const> if preferred. Constraints * Consider clarifying camp distribution: - Does each player always have exactly one camp? - Is the number of camps and elementals guaranteed to be equal for all players? Leagues * Do you plan to introduce additional mechanics in later leagues? For example, the story mentions rescuing elementals; will that be implemented? Referee / Viewer * I'll perform some testing later. Thanks.

kosynus Unnamed CodinGamer

about 2 months ago

Thank you very much for the detailed review! All the points have been addressed: - sections reordered (Actions > Movement > Captures > Prisoners) - redundant parts removed - movement clarified (4 directions + deterministic rule) - capture behavior clarified (including in camps) - Game End and Technical Details reworded - inputs standardized (integers, prisoner as 0/1) - formatting fixed (<var> x, y) - camp distribution clarified Regarding leagues, there is a clear progression already implemented: - League 1: 1 elemental per team (onboarding) - League 2: multiple elementals per team - League 3: prisoner release system Thanks again for the review, I'll be happy to follow up after referee/viewer testing!

leopard-esperanto Unnamed CodinGamer

about 2 months ago

Les règles me semblent claires. Le jeu me plaît. Mais il manque le boss dans l'arène pour passer à la ligue suivante. Du coup je ne peux tester que le niveau de base.

kosynus Unnamed CodinGamer

about 2 months ago

Merci pour ton retour ! Tu as raison il n'y a pas encore de boss dans l'arène car nous ne sommes que 2 joueurs pour l'instant... Il devrait apparaître dès qu'on sera un peu plus nombreux. En attendant, tu peux tester les boss en changeant de ligue en haut à droite. Ils sont bien présents et adaptés aux mécaniques de chaque niveau. N'hésite pas si tu vois autre chose.

kosynus Unnamed CodinGamer

about 2 months ago

Hi everyone, I've just published a Work In Progress version of my multiplayer game Elemental Wars. It's a 3-player tactical game based on a rock-paper-scissors dynamic (Water / Fire / Plant), with simultaneous captures and simple pathfinding-based movement. At this stage, the core mechanics are in place: - movement toward a target (1 step per turn), - simultaneous captures, - prisoners system, - (later leagues) release mechanics. I'm mainly looking for feedback on: - clarity of the rules / statement, - gameplay feel and balance, - edge cases or unclear situations, - overall fun factor Don't hesitate to try it out and share your thoughts even early impressions are very helpful! Thanks

ABOUT US CAREERS PRIVACY FAQ BLOG

The #1 tech hiring platform

SCREENING INTERVIEWING