Well, I've finally bit the bullet and decided to post my design pitch, but first a couple things. This project is my senior thesis, due April 2007, so I have about 11 months to work on it. My goal is to become a game character modeller. The main reason I'm posting is that I'd like to get an estimate, from people with more experience, on how much non-art work is involved in all of this. I don't have any real programming experience, so any estimates on the amount of work involved to get the gameplay and AI systems running would be appreciated. I'm working with someone local, but he's been complaining that my goals aren't possible within one year. Personally, I think the actual C++ programming aspect is miniscule compared to the rest of the work (art/sound/scripting), but I'd like some other input. Finally, any critique on the concept and gameplay ideas is welcome. (and I apologize for the lost formatting, did my best to fix it) *********** My goal is to create a single demo level from what would be a full length video game. The game takes place in a fantasy world, with heavy references to Norse and Gaelic mythology in name and theme. The demo will cover a main plot event almost half-way through the game: Sigurd, the protagonist, is the king's only living son and a beloved hero to the Hilgarians. After witnessing atrocities committed by a warring nation, Drystan, Sigurd leads an attack on their capitol, Iolar. He and several of the elite royal guards, the Oshumond, have snuck inside the city through a sewer, with the intent of neutralizing the city's central tower. The tower has a command over the river crossing and bridge, keeping the bulk of the forces from attacking. The costume and architecture is based heavily on medieval, blended with more ancient mythology and culture (based on Germanic/Celtic/Slavic paganism). This is the a simplified version of same technique Tolkien used to craft his Middle-earth; taking existing myths and language, distorting and using them in a consistent manner, synthesizing a cohesive and believable fantasy world. The art style of the game is a 'fantastical realism.' The world will have a gritty, realistic look and feel, usually reserved for WWII shooters instead of fantasy games. This creates an artistic hurdle to overcome: how to present characters and environments that are interesting, instead of just historical. Hopefully a sufficient amount of realism, combined with detailed and interesting decor/ornament and immersive gameplay and sound can overcome this. The game will be using the Source engine, used on games such as the award-winning Half-Life 2. The Source SDK has a robust set of tools, including world and level editors, model viewers, facial animation software, shaders, etc. Models will range from 3000-5000 polygons for the characters. Poly limits of the environments are TBD. High poly models will be made in ZBrush. Low-poly models and normal map generation will be done in Maya. Animation/mocap will be done in 3ds max. Textures will range from 256x256 all the way up to 2048x2048 for the Oshumond, done in Photoshop using both photographic material and digital paint. Textures will be in the Source engine .vmt/.vmf format, models and animations in Source's .mdl/.smd/etc. Nearly everything will be modelled from scratch. The exception are heads and hands (HL2 and other Source art will be used), as well as prop objects if there is not time (taken from Elder Scrolls IV: Oblivion, once a model editor is available). Animations, where new ones are needed, will use mocap. If a seperate animator is brought in, s/he will work on some custom animations. Gameplay will be in the first person. For the demo, there will only be melee combat. The tentative control scheme is WASD for run, LMB for attack, RMB for block, E for use, Q for pickup weapon, tap MMB to push, hold MMB to pick up object, +Shift for walk, +Ctrl for crouch, number keys for switching weapon loadouts (sword+shield, sword2Hand, crossbow, none(picked up)+shield, none), hold Space for squad commands (attack target (LMB), regroup(RMB), stay put(Q), hide(E). The method of attack (thrust, overhead slash, left slash, right slash) is based on mouse movement when the LMB is first pressed (respectively: no movement, up/down, right, left). Likewise, blocking with a weapon will work the same way. Blocking with a shield will block all attacks coming from your field of vision. There will also be counter attacks available; time a LMB press while blocking to disarm the opponent (use your shield to attack), Space to push him back/knock down (use your shield as a ram), Ctrl to strike back (duck and rising attack). Having a shield will make your window of opportunity larger. Each counter-attack will also require a different timing (LMB during contact, Space just after contact, Ctrl just before contact). The game will not have any UI. And because of this, there needs to be a different type of health system. We've decided, especially because of the context-specific WYSIWYG combat, to have 1-hit kills if exposed vitals (head, chest) are struck by a blade or arrow, and an sliding scale of damage. Armour comes in four types: unarmoured (clothes, skin), cloth (leather, quilting, padding), mail, and plate. Cloth armour only acts as a variable, reducing the odds that damage will be struck (and if so, how much damage calculated), and as such it will not be listed below. Plate armour is a more effective armour than mail, especially in regards to arrows and light weapons. A connecting blow to any unarmoured, non-vital area will reduce effectiveness from that limb (reduced attack, block, move speed). A connecting blow to any vital area will result in death. Below is a chart indicating armour effectiveness to different weapon classes. Light weapons are one-handed weapons (daggers, swords, maces, axes, tools). Heavy weapons are two handed (bastard sword, axe, hammer). Light arrows are from shortbows. Heavy arrows are from longbows and crossbows. LightWpn HvyWpn LghtArw HvyArw Unarmoured Hit Hit Chance Hit Mail Chance Hit No Hit Plate No Chance No Chance An indie PC game, Mount and Blade, has been my greatest inspiration as far as gameplay goes. AI will be integral to a fulfilling experience. There will be several changes to Source's AI. First, non-combatants must evaluate the situation when they see a hostile and determine a response: attack, freeze, flee). Second, disarmed/knocked down NPCs must decide whether to beg for mercy or continue battle. Third, NPCs who do beg for mercy must decide what to do if granted: honor it, or wait until the player turns around and attack. Fourth, defeated NPCs must writhe on the floor in agony after they are fatally wounded (except if struck in the head). Fifth, a simple commandable squad AI must be made (Source already has a fully fleshed out squad AI). All around, the AI must be tightened up for melee combat, since HL2 and Source games are primarily shooters. 3-5 original tracks will be composed for the demo (low-key, action, epic (custom title theme is optional track)). A volunteer will be sought out when the demo is functioning to a presentable degree (expected by early fall). For the time being, royalty-free music will substitute. Voiceovers will also be sought out at the same time. 8-10 voiceover artists will be needed. Sound will be a combination of original recordings and purchased from sound libraries for some more obscure effects. A dedicated Sound Artist will be sought out later on. Too often in all entertainment, but especially in video games that claim to let the player choose between good and evil, the consequences of their actions are clear-cut and predictable, and present the world in an overly simplified good and evil view. The real world, however, does not function like that. Sigurd's acts do have consequences, and I hope to portray them through the intense reality of next-gen graphics. The difference in feeling between hacking away at the polygon mesh, and strangling the life out of a realistic character is profound. 'Enemies' have become nameless and faceless by design. Whether it is masked Germans in WWII first person shooters or the notorious 'bandits' of adventure games, the 'other side' fails to be presented. It is not often that game plots and gameplay portrays this ambiguity (the only games that spring to mind are the Tactics Ogre and Suikoden series, and Final Fantasy Tactics). Only the consequences of specific events and actions in relation to the character are presented. But they fail to take into consideration the most common activity in those games, killing. How morally absolute is one's quest? Absolute enough to murder a man who has done nothing more than do for his country what you have pledged to do for yours? This level of moral question is not easily portrayed through stylized graphics, and therefore has not, in my opinion, been possible for current generation hardware, which is unable to express nuanced emotion. Realistic graphics drive home the point that this is a very real part of our behavior we must question. A lombax or Jedi Knight swinging a wrench or lightsaber at faceless robots or stormtroopers carries far less moral weight than the 'everyman' soldier who sees it as his clear-cut duty to kill. This moral ambiguity will be represented in the games overall plot, but it is the immediate experience of it that is important, as well as challenging to implement (and really the only instance of the morality aspect in the demo, sans the total game plot). One way to do this is by giving a choice to kill. The route through the level will be mostly linear, and the NPCs set. However, not all of them must be killed. You can grant mercy to a soldier, or not attack a woman that throws herself in your way. The rewards for your actions will be varied. Some characters will sit in silence, some will thank you in tears, some will open up alternate routes or create a diversion to let you sneak past a guardpost. But others will sit in silence until you turn away and attack you, some will thank you in tears and then flee and tell the guards when you leave, and others will try to lure you into a trap. The idea is to question the ideas of mercy and morality, not create a straight-forward 'killing is 'bad'' scenario. Saving someone will get you into trouble as often as it helps you. Indeed, it may be the easiest route to kill every living thing in the game; however, can you bring yourself to do this? Will you kill every man, woman, and child, if there is a chance that they won't get in your way? If the mechanic were a simple 'save life and get help' scenario, there'd be little difference between this act and just offing someone. Either way, you are doing it for the reward, to make your path easier. With a mixed bag of prizes, it eliminates this 'using' of NPCs, and instead asks whether a life is worth anything in and of itself. The actual act of killing will also be 'enhanced.' Enemies will not crumple when they are killed. They will writhe in pain, screaming, crying, on the floor. This includes children. Can you bear to hear the screams of a dying child? Will you coldly put her out of her misery, or watch her life slip away? What if guards are near, and the noise will give it away? Will you also kill the grieving mother? Will you smile at a flaming enemy soldier stumbling towards you, as you realize the physics trap you must set off has children whose screams you can hear, burning to death inside? What will you think when your squadmates scream "For Hilgar!" as they charge and are brought down by a hail of arrows? And then enemy soldiers charge in screaming "For Drystan!" What is it, in just the mere existence of a nation, that they were willing to die for? To value their government's survival more than their own? ************ I thank you if you got all the way through that. Currently the pre-production documents are about 80 pages between story and concept work and management stuff. This is just the pitch, concept, and summary. [Edited by - Professor420 on May 21, 2006 3:18:12 PM]

I'm sorry to disturb your daydream, but I've been stuck at line five.

"going in through sewers"? "Norse and Celtic mythology"? "Fantastical Realism"?

Man, even if the Romans did have sewers in Rome some 2000 years back, they didn't really take on until late XIXth century in the rest of teh world. Find something else.

As for the AI and the C++ coding, I'm sorry to break the bubble, but they ARE incredibly long. Potentially longer than what you're prepared to go through. And as an artist yourself, I suggest you stick to what you know you can do on your own. Maybe some full length video demo of said scenes in your own style high quality would be better than trying to do everything in low quality, including the AI.

And keep in mind that whatever you intend to present as majoring thesis, your main goal is to answer the all-time puzzling questions that will arise in your professors minds: "so what?" You are supposed to answer everything, and not only in your documentation. You should be able to make it PLAIN!

Its funny that you're stuck on the sewer aspect, its actually pretty insignificant. If you were to point out every inaccuracy or impossibility in game plot/mapping/script, you'd never finish. Some things have to be sacraficed for dramatic and gameplay benefits.

Maybe you can both elaborate on why there is so much programming involved? How hard is the AI programming? Gameplay?

"And keep in mind that whatever you intend to present as majoring thesis, your main goal is to answer the all-time puzzling questions that will arise in your professors minds: "so what?" You are supposed to answer everything, and not only in your documentation. You should be able to make it PLAIN!"
I don't understand this at all, your point or what you mean. I have presented this twice, and both times it was met with massive enthusiasm by fellow students and a jury of faculty and professionals. And I don't exaggerate the 'massive.' I get dozens of compliments and questions after the fact from peers, and my 15 minute presentation (10 min pres 5 min Q&A) went on for a half hour (more than 20 minutes of Q&A) because everyone was so intrigued and animate about the concept.

Well, it seems you have everything well thought out and planned, which means you can reduce your expected devlopement time by a lot. Many people have good ideas but jump in too soon without having everything fleshed out, which takes time later when they run into an unforseen(sp?) dead-end.

As far as the programming time, I don't think 11-months is unreasonable, but you haven't given us very much information. Is your programmer going to be working at it full-time, or just after hours? Are you looking for special effects that would be difficult to integrate easily? You have a good start on the design, but now you need to plan out every single detail, down to every little rock and tree stump that will inhabit your game world. The more detailed you are now, the less chance you have of running into a problem later.

Finally, if the programmer is working fulltime on this, I don't think it would be a big deal because you are already using a well-developed engine. I could probably get a basic First-Person-Shooter done in a month, but without nice effects and fancy models (then again I am a sucky modeler :)). Does your project need to be incredibly finalized, ie. with a menu, installer, credits, and easy to use GUI, or can it just be a sort of tech-demo with the promise of future completion? Also, I am assuming no multiplayer, which would add considerable time to the development process.

A few last notes. If you are going to be a game modeller, I don't think that you he should go with a high-quality movie, as if he does get a job as a game modeler he will be doing models with more reasonable polycounts. Also, it sounds like a reasonable project to me. It will require work and a dedicated programmer, and you will need to work hard to develope all the content, but it is most definitely do-able. Eleven months is a long time.

I've spent the time since February doing extensive design, both artistic and gameplay and conceptual and all that stuff. The plan of action has gone through multiple iterations: first it was supposed to be a more generic game in a proprietary engine the programmer is working on, then it was closer to its current concept, then it was being planned for Source, and the actual demo plot/design has been changed a dozen times at least (it started out at the prologue of the game, but to better represent the philosophical aspects, it was moved to the current time), and the scale of it was reduced even more times.

The programmer will be working full-time+ for the summer (until September). My hope is to have all issues resolved and planning resolved by mid-June, when I plan to make a recruitment drive for a couple more people with the art I've been working on (all hi-res bodies for the defenders (that's 36 seperate pieces) are already complete, art production is actually moving just as estimated WITHOUT the double the time I allotted everything). During the year, he plans to work about full-time as well, though this can vary from week to week (but at least 20 hours/week).

The project should be polished, and with so many of the details already done in the engine (ie, minimal engine debugging, most has to do with scripting/textcoding), I don't think its any tall order to do so (this sort of thing is someone I'll be recruiting for). Obviously the art is technically the most important thing, but since this is really a once in a lifetime chance (to work on a game you design for a year and a half), I'd like to fulfill it as much as I can, no sense holding anything back.

As far as the conceptual stuff, I can't say exactly how it will turn out, what the player should do, etc. Take games like Fable, KOTOR, Oblivion. Are these games truly open-ended? Isn't the result of a 'good' deed and 'bad' deed predetermined? Isn't this merely based on our accepted conventions of good and evil and justice, and not allowing the player any exploration of these concepts? If you don't give the player a cookie, it leaves the decision up to his ethics instead of his predetermined plan to play as light or dark. This is the GTA generation we're talking about, though. How far must the envelope be pushed to actually make an emotional impact? I would argue not as far as alot of people think. I'd bet that a vast majority of people play as a 'good' character their first time through KOTOR or Fable. Why is this? Because each situation presents you with two relatively equal choices, and people choose what makes them more comfortable. But when the situation isn't just two choices, when things aren't broken down into right and wrong or good and bad, the player can experience his virtual reality on a non-predetermined level. The player will be deciding the virtue and neccessity of his conduct, as opposed to the developer (or by extension, society).

Quote:
The programmer will be working full-time+ for the summer (until September). My hope is to have all issues resolved and planning resolved by mid-June, when I plan to make a recruitment drive for a couple more people with the art I've been working on (all hi-res bodies for the defenders (that's 36 seperate pieces) are already complete, art production is actually moving just as estimated WITHOUT the double the time I allotted everything). During the year, he plans to work about full-time as well, though this can vary from week to week (but at least 20 hours/week).

It sounds like you have everything under control as far as art goes. If the programmer is working at it full-time it shouldn't be a problem given the eleven month time frame. Although if there is one thing I have learned in programming, its that nothing ever goes exactly as planned and that somewhere something bad will happen. But it seems to me like you have a very nicely laid plan and a reasonable schedule. Good luck with your project!

(PS. Let us know how it turns out! Maybe a link to the demo or something.)

Quote:Original post by Trefall
The Source engine has already built in the crowbar code, it has squad AI, it has hitbox calculation. If you're going to become a game character modeller, why let so much work be put into the programming side of the game? I'd recommend having your programmer replace the crowbar model with melee weapons, some gun model with crowbar, do some minor tweaks to the weapon code so that it feels right (add in a block function to the crowbar code), do the hitbox calculation the way you wanted it (instant death to head, etc), and then let it be with that.

This would let you focus on what you're aspiering to become, and you'd still have a fun gameplay (although maybe not revolutionary) to play with. I just recommend starting simple in the fields you have little experience in, and then build upon them later if you have the resources to do so.

This gets back to what I just said: Obviously the art is technically the most important thing, but since this is really a once in a lifetime chance (to work on a game you design for a year and a half), I'd like to fulfill it as much as I can, no sense holding anything back.
I know it wouldn't be hard to do a simple mod by myself, but I have the ability, and another person with the ability, and the time, and the resources, and the financial security, and everything else that I can realistically hope for. Such a project as you described is really a worst-case scenario, and I'm definately aiming for something much more ambitious. Because like I said, I can't expect to get such an opportunity ever again.

And about the DDS thing, an oversight from one of the many earlier iterations. Fixed (though its pretty insignificant).