i'm working on a new opensource project now: "Swarm War
-
happyjack27
- Posts: 1439
- Joined: Wed Jul 14, 2010 5:27 pm
lol. didn't realize that one. pet salt = pet rock.
in the context of the swarm wars game, it makes it so that you can have a very large number of intelligent units on the map at once, all in constant communication with each other, without slowing down the game.
with pet rock, the required processing power grows in proportion to the number of units, rather than to the number of units <i>squared</i>. so the game can handle 100 like it was only 10, and 10,000 like it was only 100. right now it tops out at about 100. so presumably w/pet rock, it'd top out at 10k. meaning it could handle 10 player games where each player has 1,000-large swarms. (i want one of the appeals of the game that you can have just ridiculously enormous swarms. (given sufficient resources, the swarms will evolve and grow in number on their own.))
in a nutshell, it allows me to scale up the game a couple orders of magnitude without increasing hardware requirements.
in other contexts, well, it means essentially the same thing. it gives you the ability to scale particle/field based sims up orders of magnitude on the same hardware.
in the context of the swarm wars game, it makes it so that you can have a very large number of intelligent units on the map at once, all in constant communication with each other, without slowing down the game.
with pet rock, the required processing power grows in proportion to the number of units, rather than to the number of units <i>squared</i>. so the game can handle 100 like it was only 10, and 10,000 like it was only 100. right now it tops out at about 100. so presumably w/pet rock, it'd top out at 10k. meaning it could handle 10 player games where each player has 1,000-large swarms. (i want one of the appeals of the game that you can have just ridiculously enormous swarms. (given sufficient resources, the swarms will evolve and grow in number on their own.))
in a nutshell, it allows me to scale up the game a couple orders of magnitude without increasing hardware requirements.
in other contexts, well, it means essentially the same thing. it gives you the ability to scale particle/field based sims up orders of magnitude on the same hardware.
This sounds like an interesting appraoch to scalable gaming. one of the issues at hand in "Role Play" simluation games is the control of forces. There have been many attempts to build in automation that allows scalable interaction between the "player" or "Player team" and the entities controlled. The issue being a balance between accuaracy and playability. If you think about standard games you may have experienced, a certain level of frustration is caused when playing the game as one tries to direct individual or groups of individual units compared to the entire side. One of the primary issues causing this stress is the management of the time scale verses levels of play.
This is a deep issue in professional gaming when a desire exists to progress the game through a series of events that are single unit or group of unit dependan, but maintain the proper role and invovlement (and interest) levels of the players.
I am interested to follow your development process. Keep on chugging!
This is a deep issue in professional gaming when a desire exists to progress the game through a series of events that are single unit or group of unit dependan, but maintain the proper role and invovlement (and interest) levels of the players.
I am interested to follow your development process. Keep on chugging!
-
happyjack27
- Posts: 1439
- Joined: Wed Jul 14, 2010 5:27 pm
ya, i haven't done much work on the ui (user interface) yet. still working on the game engine and ai. (not so much the ai now - for that i just need to implement the learning algorithm and maybe evolvable learning rates, and its done.) with the ai an unholy union of {geneticly evolved + neural nets + w/temporal difference learning + swarm intelligence}, the units ("organisms") are pretty frickin' autonomous. the challenge of designing the ui won't be so much reducing the player's workload, as actually giving them something to do!
-
happyjack27
- Posts: 1439
- Joined: Wed Jul 14, 2010 5:27 pm
i will say it's at least fun to watch so far. if i make the plants grow and multiply too much, the organisms evolve into pure vegetarians and live forever in peace off of the plentiful food. if the plants don't grow fast enough to supply adequate nutrition, on the other hand, they grow a thirst for blood!
when i make it so they breed based on food supply rather than stay at a fixed population, i imagine it will automatically maintain a better balance, and the dynamics will be more interesting. after all, if your food supply can't support your population, why not just go and take someone else's?
when i make it so they breed based on food supply rather than stay at a fixed population, i imagine it will automatically maintain a better balance, and the dynamics will be more interesting. after all, if your food supply can't support your population, why not just go and take someone else's?
When you work on UI, or whoever you get to work on UI, please briefly watch this video: http://www.youtube.com/watch?v=zBHBsBS__1M
The big take-away is when you make a PC version, matching your inputs to the layout of the keyboard is really intuitive, and making hot keys for every function can do work.
The big take-away is when you make a PC version, matching your inputs to the layout of the keyboard is really intuitive, and making hot keys for every function can do work.
-
happyjack27
- Posts: 1439
- Joined: Wed Jul 14, 2010 5:27 pm
i presume the vagary primarily comes from the word "role". what is meant by role is more or less a task or set of tasks to be responsible for. for example:Betruger wrote:Ladajo what you describe in the last couple of posts sounds pretty insightful but is really abstract. Can you give a brief example?
*choosing buildings to construct and placing them on the map
*choosing what army units to build
*telling individual units where to go and/or who to attack / what to build / what to gather
each one of these is a "role" that the game designer can either make the computer do automatically for the player, or make it so the human player has to do it manually.
the game designer has to pick and choose these to try to make the game as "fun" as possible. giving the player too much, or, conversely, too little to do, will frustrate or bore the player, respectively.
what ladajo is implying is that the computer's usually pretty darn good at what ever the game designer makes it do. if the game designer didn't know how to take make a computer player fully exploit the rules, he wouldn't be able to make the rules very non-exploitable; i.e. he wouldn't be able to balance the game very well and thus the game wouldn't be any fun. so because the game designer is necessarily prescient, it's all too easy to make the automated part "too good". thus, either leaving too little manual work for the player, or in the manual work you give them, making them have to compete against computer players who, being fully prescient and able to perform literally billions of calculations a second, is much better at it than any human player could ever hope to be.
ladajo is tacitly suggesting that one can better attack this conundrum by breaking the problem down into "roles", and balancing each role individually, as well as consciously choosing the assortment of roles - which to give the human and which to give the computer. considering, for each role, whether you're under-burdening or over-burdening the player.
Game design is an intensive qualitative process.
In any game a player has a role. His role is dependant on his interactions with the game. This is also primarily driven by what is known as the game mechanism. This is the function integral to any game that the player(s) must do over and over again to progress the game. The challenge is finding a mechnism that, as Happy put it, balances the game and its mechanism(s) against the desired roles for the players.
One of the problems inherant in computer assisted game designs is the desire for technical accuracy in the game that the computer can provide. This in turn goes counter to playability for the player(s). An extremely technically accurate game may provide no chance for the player to ever understand the nuances, or emply them in play efforts.
This phenomena also exists in non-computer assisted gaming. Although during design and development it tends to gain more traction in a computer supported environment.
One of the factors that gets overlooked often in professional gaming is the in effect seperatino of church and state. The church, game design, gets to often embroiled in game development, the state. It is important to clearly seperate the two. In professinoal gaming this drama mostly occurs due to poor definition by the game sponser as to what he needs/wants. The key point: What is the objective of the game? It may not be a player centric objective if the game is to generate data to analyize a process or technology. It may well be player centric objective if the game is intended as an experiential learning event. Most hobby games are of the latter.
When finding the balance for the player(s), it is important to understand that players seek two things. The first is fun, while the second is vested interest. These two things are critical to each other in intent and action. They coexist not so much as ying and yang, but more like ying and ying.
Computer based games of today run high risk in playability. This is a function of automation. It is amazing and incredible for example that in a game like Age of Empires, a player can game as an individual troop in a fight, or a number of other single unit entities in the game, but to be succesful must also be able to "zoom out" and play ruler of the empire while the individual units go about his bidding (or not) as a dependant function on his ability to globally manage them. This introduces an inherant level of complexity to playability that many players can not overcome succesfully. It thus unbalances the co-existance of fun and vested interest to the game as a whole. It is a result of the ability to drill in by an excessive amount to the base level simulations and drive them around individually while losing sight of the full game picture.
I posit that many of the existing wildly successful games of today are single player first person focused because it is more playable to the average gamer. They can run around with a sword a chop up goblins and zombies all day long. It is fun, and they want to chop up more zombies than the other guy and see spectacular immediate gratification as body parts fly. There has been a struggle for "strategic" and "operational" type games for market share because of this. This exists not only in hobby gaming, but also professional gaming.
I think that by not clearly establishing the point of a game up front in the design phase, identifying the mechanism that supports that point, and then implementing these ideas during the development phase is what leads to a lack of fun and vested interest for the players in "bad" games. There is more to the discussion then the points I raise, but I think they are the effective center of gravity for gaming.
Computers are fantastic aides to gaming, but in general are misused and fall short of potential. After all, a game is about players, while a sim is about data crunch. If you want to do more sim than game, or game than sim, then stick to the point. But of course, you must know what your point is first.
In any game a player has a role. His role is dependant on his interactions with the game. This is also primarily driven by what is known as the game mechanism. This is the function integral to any game that the player(s) must do over and over again to progress the game. The challenge is finding a mechnism that, as Happy put it, balances the game and its mechanism(s) against the desired roles for the players.
One of the problems inherant in computer assisted game designs is the desire for technical accuracy in the game that the computer can provide. This in turn goes counter to playability for the player(s). An extremely technically accurate game may provide no chance for the player to ever understand the nuances, or emply them in play efforts.
This phenomena also exists in non-computer assisted gaming. Although during design and development it tends to gain more traction in a computer supported environment.
One of the factors that gets overlooked often in professional gaming is the in effect seperatino of church and state. The church, game design, gets to often embroiled in game development, the state. It is important to clearly seperate the two. In professinoal gaming this drama mostly occurs due to poor definition by the game sponser as to what he needs/wants. The key point: What is the objective of the game? It may not be a player centric objective if the game is to generate data to analyize a process or technology. It may well be player centric objective if the game is intended as an experiential learning event. Most hobby games are of the latter.
When finding the balance for the player(s), it is important to understand that players seek two things. The first is fun, while the second is vested interest. These two things are critical to each other in intent and action. They coexist not so much as ying and yang, but more like ying and ying.
Computer based games of today run high risk in playability. This is a function of automation. It is amazing and incredible for example that in a game like Age of Empires, a player can game as an individual troop in a fight, or a number of other single unit entities in the game, but to be succesful must also be able to "zoom out" and play ruler of the empire while the individual units go about his bidding (or not) as a dependant function on his ability to globally manage them. This introduces an inherant level of complexity to playability that many players can not overcome succesfully. It thus unbalances the co-existance of fun and vested interest to the game as a whole. It is a result of the ability to drill in by an excessive amount to the base level simulations and drive them around individually while losing sight of the full game picture.
I posit that many of the existing wildly successful games of today are single player first person focused because it is more playable to the average gamer. They can run around with a sword a chop up goblins and zombies all day long. It is fun, and they want to chop up more zombies than the other guy and see spectacular immediate gratification as body parts fly. There has been a struggle for "strategic" and "operational" type games for market share because of this. This exists not only in hobby gaming, but also professional gaming.
I think that by not clearly establishing the point of a game up front in the design phase, identifying the mechanism that supports that point, and then implementing these ideas during the development phase is what leads to a lack of fun and vested interest for the players in "bad" games. There is more to the discussion then the points I raise, but I think they are the effective center of gravity for gaming.
Computers are fantastic aides to gaming, but in general are misused and fall short of potential. After all, a game is about players, while a sim is about data crunch. If you want to do more sim than game, or game than sim, then stick to the point. But of course, you must know what your point is first.