The Daily Click

::. ::. ::.

Post Oekaki

Posted By	Message
Sketchy Cornwall UK Registered 06/11/2004 Points 1971	4th November, 2010 at 16:00:15 - Sorry for the really long post, but I could use some advice. I have what I believe is a good concept for a game, but there's one fairly important part that I'm having trouble figuring out. It's pretty ambitious, so I want to be sure it's somewhat viable before I begin. Basically, it's a hybrid of "Advance Wars" and "Pokémon" - with unique Pokémon-like creatures replacing tanks etc (minimum of 50 unique units). The game would be turn-based, and would play something like this: * Select a unit. * Select where to move it to (movement range is determined by terrain, movement type, and movement speed). * Perform an action. * Repeat for all units... The available actions would depend on the unit, and each unit would have two to choose from (at least one of which would usually be an attack). eg. A turtle-like unit could choose between: * Bite (basic attack against an adjacent unit) * Hide (hide in shell - increasing defense by 50%) eg. A leech-like unit could choose between: * Bite (basic attack against an adjacent unit) * Suck (weaker attack, but restores some lost HP at the same time) Hopefully, I'd be able to design it such that certain groups of unit are very synergistic, so you could form a really strong combo (almost like in collectible card games). Units would also have various other natural abilities - such as bonuses to attack/defense on certain types of terrain, etc. I should just point out that it wouldn't be anywhere near as complicated as the system used in Pokémon. Now here's the problem... The AI for a game like this would be way to difficult for me to code, so I'd intend to make it multiplayer only. That way, by beating another player, you could unlock some of his units - and there'd also be potential for trading units between players. Ideally, you'd only start with a handful of relatively weak units, and then you could unlock more powerful units as you keep playing. But how exactly would this work? How can I make the more rare and powerful units hidden at first, and then unlock them later? I can only think of 3 solutions, and I don't really like any of them... 1. All units are progressively evolved (upgraded) forms of 5-10 basic types. 2. Players start the game with 1 very strong unit, and 5 or so random common units. 3. Some kind of credit system, where players can pay to unlock units. Anyone have any thoughts or suggestions on how to handle unit unlocking? Would anyone actually play this game? n/a
GamesterXIII Registered 04/12/2008 Points 1110	4th November, 2010 at 16:36:56 - You couldn't make some psuedo-AI or some pre-scripted battles that are slightly altered based on how the player acts? I mean . . .it doesn't sound all that impossible if you ask me and you of all people should be able to pull it off. Sorry that this doesn't answer your question at the bottom, but I think you're defeating yourself before you try. Something for singleplayer would be better than nothing. n/a
Jon Lambert Administrator Vaporware Master Registered 19/12/2004 Points 8235	4th November, 2010 at 16:43:31 - If it's going to be multiplayer, then tie the unlocking of new units into that. As you play against people, you could gain ranking points based on how well you did and what units they had. As your ranking points increase, your rank goes up, and the higher your rank is, the more likely you are to find a strong wild unit. This also requires that you implement wild units. "Wild units" would be units that come to you as your rank increases. A person with a certain rank might be in the lobby area (or wherever you are when you aren't fighting) and be approached by a wild unit. When approached (which is random both in what approaches you and when), you can ward it away or you can attempt to tame it. To tame units, you take them into your next battle, and if you fulfill the taming requirements, you tame the unit and it joins your group permanently. If you fail to meet the requirements in a certain amount of battles or time, the unit leaves you. For example, say a person of rank 5 is in the lobby when they are approached by a tiger unit. They can either ward the tiger away, or attempt to tame it. The tiger's taming requirements are that you have it defeat 10 units of any level within 3 battles. The tiger will leave immediately if it is defeated or you lose the battle, or if you fail to meet the taming requirements. Now suppose this person enters one battle and defeats 5 units with the tiger, then enters another battle and kills another 5 units. After this battle the tiger unit will be tamed and stay in your party. Suppose instead, though, that the tiger unit was defeated in the second battle. After this battle, the tiger will leave you and you will have to encounter it again some other time. Sandwich Time!Whoo! JoyCheck & KeyCheck Widgets For easy implementation of customizable joystick and keyboard controls. http://www.create-games.com/download.asp?id=8364
OMC What a goofball Registered 21/05/2007 Points 3516	4th November, 2010 at 17:58:19 - I quite like Jon's idea (he's chock full of them) for unlocking. It would probably have to be online for enough people to play it. As for whether I would, it would depend on how well-designed it was. When I think of Advance Wars-like games, LOTR Tactics comes to mind, which never caught my interest because it seemed too complicated. If you manage to capture the strategic simplicity of Pokemon and Advance Wars, and also create compelling creatures, I'd totally play it. I look forward to seeing this game. Out of curiosity, do you have much experience with coding languages?
Sketchy Cornwall UK Registered 06/11/2004 Points 1971	4th November, 2010 at 20:24:31 - Thanks guys Gamester: Ideally, I would just make it a single player game. I worry that for an online game to be fun it needs a lot players, and to get a lot players it needs to be fun, and so on... Has there ever been a successful multiplayer online klik game? Also, I know coding an online game is difficult. The AI in "Advance Wars 1" is actually very simple, and I could probably copy it easily enough (AW2+ is more complicated). The problem is the non-attack actions. It's easy to judge the strategic value of attacking an enemy unit (value of unit * damage * etc). It's much more difficult to judge the value of an action that doesn't cause damage - eg. raises defense by 50%, or something like that. Jon: That's an interesting idea. It could work... Thanks. OMC: I never played LOTR:Tactics (never even heard of it TBH) - always a bit skeptical of games based on movies. I doubt complexity will be an issue, just because simpler is easier. I have some ideas for cool creatures, but it's going to be hard to be original when there are already 500 odd pokémon, plus numerous clone games... Anyway, I think I'll start on the basic game mechanics now, and worry about AI and/or multiplayer much later. n/a
Don Luciano Heavy combat pancake Registered 25/10/2006 Points 380	5th November, 2010 at 00:54:37 - It's not that hard to code ai for turn based actions. Basicly you just prioritize skills, If you make all skills groups checks one after another. Heal -if damaged 75% Defense up - chance 50% Some other super skill ... ... Attack - 100% First check if ai has skill heal(if not skip to next) and if damaged use 75% chance that action will be executed. if fails jump to next group... check defense up then Some other super skill And if everything fails attack! Hardest part of this technique is deciding which skills have the priority to be checked first, but it's easy to code in and you can add a lot of conditions to skills to decide if the unit should use it.. Code me a sausage!
Sketchy Cornwall UK Registered 06/11/2004 Points 1971	5th November, 2010 at 13:08:04 - Thanks for that. I know those are just examples, but the actual conditions would have to be much more complicated. eg. Suppose you have a unit with the two skills "Attack" and "Heal". Your unit is very badly damaged and there is a single enemy nearby. It's not enough to just say "I'm badly damaged, therefore I will heal myself" - there are other options: * If the enemy is weak, I could destroy it before it gets the chance to attack me - but then, what if moving to attack brings me within range of another enemy? (ie. It's a trap!) * If the enemy is very powerful, it might still be able to destroy me even after I've healed myself - should I just try to cause as much damage as possible before that happens? Should I try to retreat out of attack range? * Is there another friendly unit that could destroy the enemy for me? Now imagine how much more complicated it would get if you have a skill that lets you heal any adjacent unit... I don't know - maybe dumb AI would be good enough... Edited by Sketchy n/a
Don Luciano Heavy combat pancake Registered 25/10/2006 Points 380	5th November, 2010 at 13:29:38 - I don't have to imagine, i have that skill actually It's not that hard as you might think. you can easily add to check first if there is a weak enemy you can kill off with one strike before checking heal. Facing an opponent that can kill you with one strike even when you heal, would happen extremely rarely... think about it, how much the unit get damaged to heal itself if that oponent can kill it with ex. two strikes? Doesn't make sense unless healing is ultra low. And if it could kill it with one strike it wouldnt get damaged at all to begin with. And retreating is ussually a not a good idea, cause a gameplay where you chase enemies across the map to finish them off is not very interesting. And in terms of tactics retreating doesn't do much good, unless a ranged unit. But you can easily add behaviors, adding a string to a unit ex. tank, healer, ranged - so they have different priorities when checking skills. Code me a sausage!
Jon Lambert Administrator Vaporware Master Registered 19/12/2004 Points 8235	5th November, 2010 at 13:48:50 - Originally Posted by Don Luciano I don't have to imagine, i have that skill actually It's not that hard as you might think. you can easily add to check first if there is a weak enemy you can kill off with one strike before checking heal. Facing an opponent that can kill you with one strike even when you heal, would happen extremely rarely... think about it, how much the unit get damaged to heal itself if that oponent can kill it with ex. two strikes? Doesn't make sense unless healing is ultra low. And if it could kill it with one strike it wouldnt get damaged at all to begin with. And retreating is ussually a not a good idea, cause a gameplay where you chase enemies across the map to finish them off is not very interesting. And in terms of tactics retreating doesn't do much good, unless a ranged unit. But you can easily add behaviors, adding a string to a unit ex. tank, healer, ranged - so they have different priorities when checking skills. Fire Emblem would like to have a word with you. Extremely rarely? Since when? If it's a low-level unit that you're training, or the enemy is a boss enemy, it's very reasonable to imagine that you might not be able to heal enough, especially if there are multiple damage-dealing strikes per attack. Retreating is not a good idea? Did you just try to say that retreating is a bad idea? If you're going to die and need to recoup, or want to move someone weak out of range in favor of a stronger unit, why wouldn't you? Retreating does all sorts of good: you move out of the way of enemies, you can find time to heal or use some sort of item, or find time to come up with a strategy. Sandwich Time!Whoo! JoyCheck & KeyCheck Widgets For easy implementation of customizable joystick and keyboard controls. http://www.create-games.com/download.asp?id=8364
Sketchy Cornwall UK Registered 06/11/2004 Points 1971	5th November, 2010 at 15:40:09 - I think he just means that retreating is a bad idea in the sense that it will make the game less fun for the player - not that it's an ineffective tactic. However, it is entirely possible for a unit on full health to be destroyed in a single turn - if not by a single powerful unit, then certainly by a group of units. I realize that's not the example I gave before, but it could easily happen, and would still make healing pointless. I guess the standard way of solving this problem, is to make it a battle of brains vs brawn - make the computer player really dumb, and then give him twice as many units to make up for it. It's not ideal though... n/a
alessandroLino I create vaporware Registered 11/03/2009 Points 172	5th November, 2010 at 16:11:07 - Also keep in mind that if the same 'player profile'(or whatever you have) will be available both for single and multiplayer, having dumb ai would make players grind their stuff on single player, and pwnzor others on multiplayer. You could make some way that playing single player wouldn't evolve/give exp to the player. But then almost no one would play alone. But im sure you've got better ideas for that. Also, it will be turn based, which would be playable even with a very slow internet connection, which would make single player pretty much useless if you have a decent amount of players. I'm not against having single player, I'm just pointing that in this kind of game, multiplayer just works better. On the part of unlocking stuff, personally, i'd prefer having a flexible way of training the creatures (like a tiger with 10 atk and 3 def, and other tiger with 3 atk and 10 def), and unlocking moves. Maybe unlocking moves to a 'move bank' of the creature and then you could select the ones before going to battle. Then the moves could be more effective based on a stat or a specific creature. Edited by alessandroLino n/a
Don Luciano Heavy combat pancake Registered 25/10/2006 Points 380	5th November, 2010 at 16:28:05 - Originally Posted by Jon Lambert Originally Posted by Don Luciano I don't have to imagine, i have that skill actually It's not that hard as you might think. you can easily add to check first if there is a weak enemy you can kill off with one strike before checking heal. Facing an opponent that can kill you with one strike even when you heal, would happen extremely rarely... think about it, how much the unit get damaged to heal itself if that oponent can kill it with ex. two strikes? Doesn't make sense unless healing is ultra low. And if it could kill it with one strike it wouldnt get damaged at all to begin with. And retreating is ussually a not a good idea, cause a gameplay where you chase enemies across the map to finish them off is not very interesting. And in terms of tactics retreating doesn't do much good, unless a ranged unit. But you can easily add behaviors, adding a string to a unit ex. tank, healer, ranged - so they have different priorities when checking skills. Fire Emblem would like to have a word with you. Extremely rarely? Since when? If it's a low-level unit that you're training, or the enemy is a boss enemy, it's very reasonable to imagine that you might not be able to heal enough, especially if there are multiple damage-dealing strikes per attack. Retreating is not a good idea? Did you just try to say that retreating is a bad idea? If you're going to die and need to recoup, or want to move someone weak out of range in favor of a stronger unit, why wouldn't you? Retreating does all sorts of good: you move out of the way of enemies, you can find time to heal or use some sort of item, or find time to come up with a strategy. Yes it's extremely rare that a unit which heals itself will be destroyed with one blow, think about it, the damage of the attacker needs to be in the 25% range of the attacked creature for that, with random damage and skills whatnot it's very unlikely. And yes with multiple attacks the unit will still be killed even if it heals itself, but that still has tactical advantage, meaning it's healed you now need more strikes against it to kill it. Offcourse that doesn't help if the last surviving creature is that healer, but you can even condition that very easily with count enemy = 1, dont heal attack and play sound : this is sparta!. I said that ranged should retreat if he has range, and weak units can aswell retreat to give favor, but if you use a system where you cannot attack then move, retreat is actually in most cases the worse case scenario, but i see you are refering a lot to the player actions instead of ai actions. And My post was about AI. Not what player does, player can do whatever the hell he wants. Edited by Don Luciano Code me a sausage!
Don Luciano Heavy combat pancake Registered 25/10/2006 Points 380	5th November, 2010 at 16:53:53 - And just to elaborate a bit on the healing part, imagine a 1v1 fight, because in mass fights healers are always usefull. Healer HP= 60 Healing amount = 10 - very low Only attacker damage spaning from 35-59 would be able to render healing useless in this case. But as you can easily see that's a very rare case, since having a 1v1 fight beetween a healer and a unit that has such high damage, even if the healer does not heal is pretty much a decided fight from the beggining. Cause if heal spell is 5 times lower than enemy attack it's a useless skill. Or the healer is like 10 lvls weaker than the attacker - meaning that the healer wouldn't do much damage anyway if he decided not to heal. Edited by Don Luciano Code me a sausage!
Sketchy Cornwall UK Registered 06/11/2004 Points 1971	5th November, 2010 at 17:55:39 - 1v1 fights are almost never going to happen. There's a very high chance that the healing unit will already be damaged from previous encounters, so even a weak attack might be enough to finish it off (btw: Attacks dealing 60% damage would not be uncommon). Obviously a healing ability needs to be much weaker than the average attack, or the healer will be virtually indestructible - the healing ability is still useful because you can retreat, and then safely heal yourself before going back to fight some more. Also, you could use the ability quite strategically - use a healer to block a bridge or something, and it forces the enemy to send a powerful force - they can't just gradually wear you down with many weak units. I agree that there can be tactical advantage to surviving an extra attack, even if the unit is eventually destroyed - but sometimes there isn't (even if there are other surviving units). Ideally, the AI would be able to quantify exactly how much advantage is to be gained - which is why I say it's very difficult to code. n/a
Don Luciano Heavy combat pancake Registered 25/10/2006 Points 380	5th November, 2010 at 18:38:08 - "There's a very high chance that the healing unit will already be damaged from previous encounters, so even a weak attack might be enough to finish it off (btw: Attacks dealing 60% damage would not be uncommon)." If the healer is damaged from before, he will probably heal, and a weak attack will not finish him off, and if it will, i mean that's not a weak attack. If dealing 60% damage would not be uncommon, it shouldn't be uncommon to heals be more than 30% making again very rarely a heal will be useless. "Obviously a healing ability needs to be much weaker than the average attack, or the healer will be virtually indestructible" And honestly making the healing ability lower than the average attack i think is a very very bad idea. Because you only need to add cooldown to make healing available lets say every 3 or 4 turns, so healing can't be spammed, and unit is not indestructible, or for any skill for that matter. Or adding something like Mana or Stamina, that recovers each turn to use skills, i mean whatever system you make, enabling units to spam skills is a big mistake. " - the healing ability is still useful because you can retreat, and then safely heal yourself before going back to fight some more." I wouldn't add that, seems you were mentioning attacks doing 60% damage, and now are talking about retreating and healing? I mean since with that high dmg battles will be over in like 2-4 turns, there's isn't much time for ai to retreat and heal some minor damage and then get back into the fight, it's better then to remove healing skill at all cause it will be useless. In my opinion you should give strong healing skills only to support units with medium cooldown. Or low healing skills with short cooldown. And if a strong unit has healing, make a long cooldown, and restrict to only healing itself. "I agree that there can be tactical advantage to surviving an extra attack, even if the unit is eventually destroyed - but sometimes there isn't (even if there are other surviving units). Ideally, the AI would be able to quantify exactly how much advantage is to be gained - which is why I say it's very difficult to code." Yes i agree sometimes it isn't. But then again if in most cases it is useffull, like more than 75% it's actually a very good ai. Because even people make mistakes, and honestly I don't prefer AI to be super smart, cause it's fun when ai is acting dumb. In my game i actually prefered my older dumber ai code, cause now he destroys a lot more of my ships by acting all smart . As long as you cannot exploit the dumbness of the ai i think qualifies as a good AI, no matter if he sometimes does something stupid. Code me a sausage!

Advertisement

Worth A Click