Gameplay Design Concepts

Acert93

Artist formerly known as Acert93
Legend
I have been tinking with a "Gameplay Design Thread" for a while, even creating some rough posts. The new Fight Night Champion thread dovetailed with one of my drafts so I thought, "Why not just post it?" It is rough but figuring I would never put it on a blog no one would ever read anyhow I might as well post my random ideas here for people to laugh at :LOL: With no further ado, Joshua's first effort: Joshua's Kinect Boxing.

Game: A boxing / workout game for the Xbox 360 Kinect.

Concept: A fun sporting game using pattern based gameplay that passively encourages exercise. The game would have mini-game training sessions, real boxing-oriented workouts, and a full career mode.

Selling Point: Makes good use of Kinect technology as it is similar to the dancing games.
Highly interactive.
An established sport with gameplay mimicking various elements.
Good exercise potential.
Limited game scope could allow excellent visuals on a cheap budget.
Offers both local and online gameplay and a host of mini-games.
Avatar dashboard clothing (robe, gloves, black eye, missing teeth)

Challenges: The Kinect presents a couple challenges for a boxing game.
Tracking: Resolution provided a hurdle for some tracking.
Avatar Integration: The disconnect between human movement and online interaction could be stilted. E.g. As the player throws a punch the avatar may hit the opponent hence while the players arm moves the avatar arm would not.
Latency: 100-150ms and online would add another 50-300ms. Showing the player their exact body movements would appear laggy.

Core Design: Gameplay is similar to a mix of the old school memory game “Simon” and a “pop up shooting gallery” and shared a lot in common with the gesture/posture dancing games.
The player will be presented with a series “targets.” Targets would be small flashes (e.g. translucent arms w/ gloves doing a fast gesture motion w/ a flash for the terminating glove position, so a right handed jab to the face would be a jab with a flash at the face), accompanied by the movement specific sound, that correspond to requested movement and target (so right hand punches may be a small shheeesh sound from the right speaker, left the same sound but lower pitch from the left; hooks would be a swoooosh sound; face targets would be higher pitched than body shots). Successfully completed maneuvers would have visual and audio feedback.

The player will be presented with their targets. Successful series take down the health, stamina, etc of the opponent. Missed series will result in failed punches/blocks. (Multiplayer Dynamic: No health exchange for successful series on both ends as blocks/dodges/bobs/weaves mirror opponent offense. If a player fails to gesture a punch the defender, regardless of success of the block, will not be hit—so no health exchange BUT there is a stamina cost. Landed punches have more heft in MP to reward/penalize. Missed punches and blocks are important as they strongly impact stamina [closing gesture window, the game offers fewer offensive moves]).

“Confidence” bonuses for successful series; player can store up to 3 Bonus Stars. 1 Star = Basic power up; 3 Stars = Super Powerup. Through voice commands the players can request power up use. Players can choose (and unlocked) various abilities. They can be everything from requiring the opponent to string together very difficult defensive moves or series, hit windows being shrunk, very quick gesture sequences, health or stamina boosts, and for 3 Star moves certain unblockables.

Depth: Various player skills (e.g. hard hitters have shorted windows, etc)
Various player styles (offensive, defensive, balance, dancer, etc)
Power ups and Customization (including unlockables and perks ala MW2)
Round # and Length modifiers
Blurred Vision and Muted hearing as players take more and more of a beating
Player Skill: Beginner has fewer moves and presents simpler combo series; Expert presents the full repertoire of movements, punches, and blocks in a dizzying number of patterns to keep the game challenging and fresh
Difficulty: Easy has larger windows for successful movements, a lower penalty for not completing series, allows for missed move(s) in long series while still getting a successful series, and more stamina for energy (i.e. window size) regeneration; Hard has a tighter window for movements, a larger penalty for failed series, requires perfect series completion, and smaller stamina

Replay: Full online mode (as the game’s design is lag resistant)
Career (ala Punch Out story mode) and full boxer customization (appearance, style, power ups, perks)
“Training Workouts” to “prepare” for matches: heavy bag, jumping jacks, speed bag, running, pushups, squats, sparring, combo practice, etc
Tutorials
Mini-games for fun, local multiplayer party games, and to hone skills

I had some additional ideas, e.g. online leaderboards for match-making, and for the gameplay players could select tactic (e.g. defensive, offensive, neutral, counter, etc) every 60 seconds or whatnot and what strategy you chose would impact the sequences. e.g. If you are losing/"tired" and pick offensive while the opponent is good at defense and selects defense your series will be more difficult). As a sequence style game there could be a lot of variables to add gameplay depth and strategy through simple user decisions that have significant impact on the gameplay difficulty. This need not be a shallow game.
 
Will it use MLAA?

I am going to take great pleasure destroying you and two friends of your choosing in FireFight Versus. You will fear my Hot Pink and Fuchsia Covenant Elite!

So it's basically an elite beat agents port then?

I didn't know what the game was so I looked it up:

http://www.youtube.com/watch?v=Ys7fNU7L4EM

I guess if we are going to use a really loose definition of port that would permit, oh, saying Battlefield Bad Company 2 is a port of GoldenEye, then yeah :LOL:

The core concept isn't far off from sequence based gameplay ala a Guitar Hero, Rockband, and QTE in general. Elite Beat Agents mixes sequence based gameplay with stylus tapping instead of button tapping. So yes, in this regards I am taking the general concept of sequence based gameplay and put it into a 3D motion tracking context. So I guess I see where you get the similarities. Of course it isn't just sequence being tracked (i.e. targets) and timing (pace), but the actual motion as well--a left hook to the head would be different from a left jab to the head and a right hook to the body would be different than a right hook to the head.

Maybe I should stipulate I lay no claim to unique ideas? In this case it was an effort to adapt an appealing genre (Boxing) to the limitations of Kinect--1:1 mapping, beyond lag, would create some issues where your game avatar makes contact and the arm stops moving but your continue your movement... how, then, would 1:1 mapping re-merge your avatar with your movements? That would be a pretty big problem to overcome (return to the "fighting position?).
 
I guess if we are going to use a really loose definition of port that would permit, oh, saying Battlefield Bad Company 2 is a port of GoldenEye, then yeah :LOL:

Of course it isn't just sequence being tracked (i.e. targets) and timing (pace), but the actual motion as well--a left hook to the head would be different from a left jab to the head and a right hook to the body would be different than a right hook to the head.

Please don't get offended, I did not mean to be rude, since I think even a direct port of Elite Beat Agents would work well, and would probably be easier to port anyways :) However EBA also tracks motion as you're supposed to follow lines drawn across the screen with your stylus at the correct speed as well, it's not just about tapping, that's why I immediately thought EBA.
 
Last edited by a moderator:
what is the feasability of making this in xna?
 
Last edited by a moderator:
what is the feasability of making this in xna?

I think we'll have to wait and see when XNA gets drivers for Kinect before we can answer that. But assuming that we'd eventually gett them, then it should be possible.
 
Avatar Integration: The disconnect between human movement and online interaction could be stilted. E.g. As the player throws a punch the avatar may hit the opponent hence while the players arm moves the avatar arm would not.
Latency: 100-150ms and online would add another 50-300ms. Showing the player their exact body movements would appear laggy.
This is the major problem (also from actual vids on the net, latency is >150msec)
eg CPU player throws punch (which takes less than 100msec to travel the 1meter), person moves out of the way (yet person is knocked out since the game hasnt yet picked up the player has moved out of the way)
Not the makings of a fun game :)

Theres a reason that all 15 games shown so far are party/fitness games cause latency/accuracy are not so important with this genre.

Any game designs for natal are going to have to take into consideration this issue.
Boxing requires fast reactions/movements thus aint the best fit for natal.
(from left field) How about a wood chopping game?
That should work
 
Boxing requires fast reactions/movements thus aint the best fit for natal.

Exactly. That is why I am suggesting a Boxing-genre game that uses a non-realtime tracking. If I didn't explain it well, the gameplay mechanic that I was trying to outline was one that essentially presented the player with sequences. For example:

Left Jab => Face
Left Jab => Face
Feign
Step => Right
Right Hook => Body
Step = Left

Testing would have to show wether you would have to do the entire series to have it play out or the avatar to begin playing our the sequence (and stop once you fail to do the correct move in a window?)

But I agree that if Kinect as 300ms latency, once you add in the disconnect between your avatar making in-game contact & the realworld player still moving (how do you resync?) a realtime boxing game would have some issues if it is aiming to be serious.

Of course it could be a much simpler realtime model and more generalized. e.g. Require paced, deliberate moves and punches, think very exagerated.

But I think the sequence based gameplay that uses gestures and doesn't try to do realtime 1:1 tracking would work really well and fit the genre (of course it wouldn't be a realtime simulation).
 
eg CPU player throws punch (which takes less than 100msec to travel the 1meter), person moves out of the way (yet person is knocked out since the game hasnt yet picked up the player has moved out of the way)
Not the makings of a fun game

There are some ways around that. Right off the top of my head:

Have the CPU player telegraph their moves; maybe even add a visual que (Punch Out had a ton of these).

As the punch is throw decelerate time to allow player reaction (and the more hurt you are the less it slows down + blur!)

I am sure smart designers can figure ways to account for the technical issues and make a fun gameplay dynamic around it. Only those locked into "Must be 1:1 and not latency tolerant" are going to have major issues. Ditto those who are trying to map traditional FPS controls straight into a shooter--the design needs to be reworked, ala Wii games.
 
You could record your own fighting moves and combos in advance, and then choose to which buttons you want to map which combos, and then just play with the regular controller. Maybe more effective would be that you could have the player's movements done in a kind of bullet time, comparable to the fighting sequences in Sherlock Holmes (the Guy Ritchie version).

But your setup could work as well - as a proper 'game', you could do a blow for blow setup where you'll have alternating attack and defend poses taken up by the opponent that you have to react with one of your own movies with. I think that's close to what you're looking at, right? It would be very similar to Dance Central, except that the avatar would perform your real movements, and instead of copying the moves of the dancer online you're responding with counter moves (attack or defense).

Yes, I could definitely see how that could work and heck, even be fun! :) After a round you could offer up a really cool real-time replay as well. After all, not everyone is a real fighter, which made me think of this video I just saw on Gaf:

http://www.youtube.com/watch?v=jA90Lf78n8U#t=2m14s

Your game setup could have a lot more mainstream appeal than this one.
 
Action Adventure Game

Uses analogue triggers for jumping and crouch. Right trigger for jump Left for crouch.

Analogue jumping means no more double jumping. Variable crouch affects attacks. Really shallow jumps can work as dodges.

Attack animation, success and damage, takes into account position and movement of character at button press relative to target. Football (soccer) style shooting mechanic added to attack.

Other than that I was thinking of the main character having a double bladed sword with a sliding grip,


Other idea Genreless Game.
The idea of this is that the game mechanics can be customised as you are playing. Play through the same story in the same world as an RPG or an adventure game or any other suitable genre. choose the camera type top down side on, first person, tracking etc.
 
This is the major problem (also from actual vids on the net, latency is >150msec)
eg CPU player throws punch (which takes less than 100msec to travel the 1meter), person moves out of the way (yet person is knocked out since the game hasnt yet picked up the player has moved out of the way)
Not the makings of a fun game :)

Theres a reason that all 15 games shown so far are party/fitness games cause latency/accuracy are not so important with this genre.

Any game designs for natal are going to have to take into consideration this issue.
Boxing requires fast reactions/movements thus aint the best fit for natal.
(from left field) How about a wood chopping game?
That should work
Aah, well, in your view, no boxing game should work, since even normal controllers can have up to 150ms lag (Resistance: FOM, for instance).
In general, the human body takes more than the time of a punch to start moving, so how does boxing work in the real world? Well, boxers watch their opponents and attempt to figure out how they "telegraph" their moves (Kinda like poker). That's why boxers tend to watch a lot of fights of a prospective opponent, they're trying to spot their "tells".

How could this translate into a video game? Well, you could, for instance, if it were very realistic, put in those tells (opponent usually bunches shoulder muscles just before a jab, for instance). You could add difficulty levels by have the easy levels highlight the "tell" and warn you a specific punch is coming. As you go up in levels, the highlights go away and you learn to read the tells of your opponent yourself.

For the previously mentioned game, you could make the telegraphing a lot more obvious, similar to how you know a note is coming in guitar hero.

If you wanted to, you could easily make an extremely realistic boxing simulator for Kinect or Move, as long as you incorporated mechanisms similar to how real boxers work.
 
Aah, well, in your view, no boxing game should work, since even normal controllers can have up to 150ms lag (Resistance: FOM, for instance).
RFOM is a 30fps game I think (cod6 is 60fsw I think), look at serious driving games theyre 60fps whys that? cause fast reactions are often needed (even though you can see whats a coming from a long way away)
there is a boxing game I know of
http://en.wikipedia.org/wiki/Fight_Night_4
its 60fps, cause 30fps will not work for the genre if you do it seriously

also the way the mind perceives time is not linear
eg say a person clicks on something and action A happens 100msec later, they may percieve it as happening instantaneously (eg the same as 10msec), yet double that time to 200msec and they'll feel it lagging.
This might be a fun experiment to try out, I wouldnt mind coding this up if ppl are interested.
Anyone have any ideas how to test this best?
eg person controls a dot, now they have to keep this dot inside a circle (which is moving around)
the thing is to simulate lag, eg the dot only moves 200msec after the persons pressed the up key or something
Anyone have any better ideas?
 
This is a great thread...I hope some devs read it so that it turns into a wish thread :mrgreen:

I don"t have a fully fledged out concept like Josh, but rather a rough base idea which makes me wonder that no this game dev picks it up:

I like the concept of coop. Coop enhances fun drastically, no doubt about that on my side.
But I am dreaming of a type of "a-symmetric" coop!

You don't know what I mean? Remember Silkworm, the side scroller shooter where you have a Heli and a Jeep, with basically fundamental different behavior (controlls, combat and tactics)...playing coop together in this game is a fundamental experience!

Wait, wait...some of you might say! Just play a standard coop RPG game, the likes of Baldurs Gate or Sacred 2 more recently: one can choose a Babarian and the other a Mage - et voila, there you have the asymmetric coop experience! True to some extend, but....

What I am dreaming of is an action adventure game with story fundamentally based on the concept of asymmetric coop - something like: Sherlock Holmes and Dr. Watson in a modern setting or even a space opera?!


Some ingredients I would put in:

- of course asymmetric controlls/combat/fighting tactics...I know that this is difficult, as you have basically design two games in some sense which should even match somehow

- story telling could rely on both characters, which can sometimes branch out (extending the Gears of War idea, where you sometimes have to fight different pathes)...where you can get an update by your coop buddy once you met again at the same place (this would drastically increase replayability, as gamers would *love* to experience both sides of the story)

-both characters have to work together to solve puzzles and find hidden stuff (extend the idea of LBP, where you need sometimes a certain amount of coop lpayers to solve stuff, this could lead in very interesting unique funny solving stuff)


I know that it is only a rough weird basic concept, but just imagen asymmetric coop and its potential for entertainment...
 
- of course asymmetric controlls/combat/fighting tactics...I know that this is difficult, as you have basically design two games in some sense which should even match somehow

-both characters have to work together to solve puzzles and find hidden stuff (extend the idea of LBP, where you need sometimes a certain amount of coop lpayers to solve stuff, this could lead in very interesting unique funny solving stuff)
A problem with that is when your game becomes dependent on coop play, you're drastically reducing your potential market. Whereas designing sophisticated coop gameplay alongside solo gameplay is a lot of effort with questionable returns.

One to look out for though is Snowblind Studio's War in the North, with 3 players cooperative gameplay where supposedly teamwork is essential with each character playing a role, and AI taking up the place of coop buddies when none are to hand. Coop gameplay was a big talking point on the Snowblind forum and it looks like they've prioritised their game towards that.
 
A problem with that is when your game becomes dependent on coop play, you're drastically reducing your potential market. Whereas designing sophisticated coop gameplay alongside solo gameplay is a lot of effort with questionable returns.

One to look out for though is Snowblind Studio's War in the North, with 3 players cooperative gameplay where supposedly teamwork is essential with each character playing a role, and AI taking up the place of coop buddies when none are to hand. Coop gameplay was a big talking point on the Snowblind forum and it looks like they've prioritised their game towards that.

Yeah, that is the problem of course, if you want to design it properly...the game would not work in SP mode...so this will stay a dream...


Snowblind Studio's War in the North??
Never heard of this...gonna check this out-thanks for the hint!
 
A problem with that is when your game becomes dependent on coop play, you're drastically reducing your potential market. Whereas designing sophisticated coop gameplay alongside solo gameplay is a lot of effort with questionable returns.

One to look out for though is Snowblind Studio's War in the North, with 3 players cooperative gameplay where supposedly teamwork is essential with each character playing a role, and AI taking up the place of coop buddies when none are to hand. Coop gameplay was a big talking point on the Snowblind forum and it looks like they've prioritised their game towards that.

Well, that's part of the answer though, isn't it? It means you need a serious investment into AI. This is not easy (see Resident Evil 5), but I'm fairly sure that it can be made to work with some trial and error and a bit of half-decent research. I think for instance for Resident Evil 5, you could make the behaviour of the AI partner configurable at least during the test phase, so that people can tweak behaviour according to their preference for the given situations in the game.

You can then use the data that worked for solving the problems in the game to inform the AI behaviour as well as find out what parameters are important. Then you can always decide whether or not you want the player to ask the AI to do certain things a certain way, or make it dynamic based on the gameplay that the player exhibits and what matched that gameplay best in the test phase data.

This could be an interesting way to make motion controlled games that are a little more involving - having to work together that way makes for great party-gaming I think. For Kinect, one person could be doing driving a boat, the other could be dragging people into the boat out of the water in a kind of rescue type game, etc (you could do this with the other controllers as well of course).

In the case of Move, I was thinking that a Descent type of game would be great for either one or two players, where you would use one controller for all the movement types (which basically correspond to just about everything that one Move controller can do), and the other for shooting, controlling robot arms etc. You could play that with one player holding two Move controllers, two players holding one each, two players online, etc.
 
I think Billy Idols idea could work. Resident Evil online while not having different mechanics for each player, had people teaming up to beat levels. Just takes match making like any online game. I do remember waiting around quite a bit in LBP to get 4 people into a level but that maybe because single player was an option and the multitude of other levels a distraction from them wanting to play the particular level you needed four people for. Once we got four people together we had a blast going through the levels, even the infamous wheel of death.
 
Back
Top