Another Arbitrary BCT Blog

sujaychand.me

[isthisprogress]

[isthisprogress]

Drunken Tavern Brawler Final Submission 

[isthisprogress]

13837076 - Sujay Chand - Studio 6 Contextual and Conceptual Statements.pdf

[isthisprogress]

Drunken Tavern Brawler (Week 12)

I Should Really Stop

Fight Stick Test

[isthisprogress]

Drunken Tavern Brawler - Busy (Week 11)

I did more test and jammed out the presentation with my group.

[isthisprogress]

Drunken Tavern Brawler - Fight Stick Test (Week 12)

Just out of curiosity I test the project with a Fight Stick and it was alright, I may have to slow down the punches a bit.

[isthisprogress]

Drunken Tavern Brawler - I Should Really Stop (Week 12)

I made the cursor stick to the ground, not because it added anything to the project. It was just something @marlosanmiguel kept asking for. Are you proud of me now??????? 

*SlowlyCriesRememeberingAllTheWritingLeft

[isthisprogress]

Drunken Tavern Brawler (Week 10)

Game Update

October GDM and Losing Myself

Control Scheme Analysis

[isthisprogress]

Drunken Tavern Brawler - Reflective Statement (ARDN612)

Throughout the semester, I worked on “Drunken Tavern Brawler” for Studio 6 and ARDN612. Hence, this was a 45 points project. Harry and Ryan worked alongside me, but only dedicated 15 points each. Yet, I can proudly say this was the best project I have been apart of; the outcome speaks louder than anything else. The project taught me many skill and lessons in game development and team communication. Nevertheless, getting to this outcome was a very interesting journey.

During my last semester break, I spent a significant amount of time playing with and exploring Unreal Engine 4 as a tool. While scouring YouTube, I came across a tutorial on how to make a “Gang Beasts” like fighter (Medel Design, 2017). From here I started experimenting with physics assets. Then made this:

After making that I thought I could make a comedic ragdoll fighter, but while talking to a friend; she mentioned she was drunk and that gave me an idea for a Snapchat caption “Drunken Bar Fighter The Video Game”. This snap got a lot of positive reception, which motivated me to make the project a reality.

Initially, the project started off as a Gang Beasts(Boneloaf, 2014) clone. But, through play testing I realised that I wanted something a bit more original and different. Thus, I began to pivot toward a 3D arena brawler and want to make it emphasise the competitive; yet goofy nature I saw the players exhibit. Hence, I began to looking at mainstream product in this genre like Super Smash Bros(Bandai Namco & Sora Ltd, 2014); as well as hidden gems like Shrek SuperSlam(7 Studios & Shaba Games, 2005). Coincidentally, Sherk SuperSlam ended up being a critical factor in making the decisions and morphing the project, in the early stages.

Through the experience, I gained a lot of conceptual and project management skills. One of the most useful skills I gained from these departments was learning how to listen to and judge feedback. I realised that the most important thing to listen too is “Why are they saying what they are saying?”, understanding why something is being said will help one determine what is the problem and if it is essential. I could share many examples, where players were asking for specific mechanics or animations edits; which would have only been polish. However, upon reflection the best example I can think of was controlling the level of drunkenness or having that “unique aesthetic”. From face value both suggestions look like they will add a lot of value to the project. But, upon evaluation I noticed that they are not essential to the programs functionality and enter the territory of feature creeping. Moreover, controlling the drunken level would require significant programming done to the ragdoll and physics; currently I am using default values. Then will have to break the skeletal meshes into separate models, re-rig and animate all of those limbs. This entire element may take up to 6 weeks to do, with two working on it. The time span includes no research and accounts for code documentation to crystal clear and bugs to be non-existent. Speaking realistically this would strip development time by at least two-thirds, if we are lucky and a single problem is never faced. Then again there may be a way to fake it, but it would require a lot of time to develop. Basically, if we had succeeded in doing this the final product would have been an utter mess and the actually “game” part would have been half-assed.

On the technical side many skills and principals were learnt or underwent significant improvement. Firstly, I became a lot more aware of the system architecture of Unreal Engine 4. Where, it became easy to figure out how to expose certain variables. Secondly, I learnt a lot about Physics Assets, hierarchical rig and layer blends by bones in a skeletal meshes; where I now understand how to animate specific limbs and send messages efficiently down a chain. Thirdly, using a minimum area within a rectangle to control a dynamic camera smoothly (More documentaion: https://www.geometrictools.com/Documentation/MinimumAreaRectangle.pdf) (Eberly, 2015). Lastly, I have a much better understanding on judging task length.

During the project there were many occasions where I lost track of what I wanted to make. There were many occasions where I started turning this into a technical fighter and neglecting the intended audience. The reason for this was that I got more interested in looking at frame data and combo chains. Then again when one is passionate they can easily get carried away and lose themselves in their work. I could blame Tekken 7 (Bandai Namco Studios, 2015) for these issues, however it did assist in teaching me how to chain combos and structure fighting mechanics and make certain actions punishable, which led to me nerfing the jump.

In conclusion, I do feel like I cheated by doing very little research into drunken behaviour. But, at the end of the day; the fighter part had more importance and essential to the game experience. The drunkenness just acts as a selling point. However, focusing on the fighter part allowed for the project to get done; which gave me a project to look forward to working on after graduation.

References:

Medel Design. (2017, June 27). 😝 GANG BEASTS Fight Style - UE4 Tutorial + PROJECT DOWNLOAD [Video file]. Retrieved from https://www.youtube.com/watch?v=VV4kLV8DifI&t=2s&ab_channel=MedelDesign

Boneloaf. (2014). Gang Beasts [Video Game]. Double Fine Presents.

Bandai Namco & Sora Ltd. (2014). Super Smash Bros. for Nintendo Wii U [Video Game]. Nintendo.

7 Studios & Shaba Games. (2005). Shrek SuperSlam [Video Game]. Activision.

Eberly, D. (2015). Minimum-Area Rectangle Containing a Set of Points. Geometric Tools, LLC. Retrieved from https://www.geometrictools.com/Documentation/MinimumAreaRectangle.pdf

Bandai Namco Studios. (2015). Tekken 7 [Video Game]. Bandai Namco Entertainment.

[isthisprogress]

Drunken Tavern Brawler - Control Scheme Analysis (Week 10)

One of the biggest criticisms I got was that the control scheme was a bit too complex. To be honest this is what I was going for, I wanted something that was hard to master and was unpredictable. Designing a control scheme has been hard and this was expected. However, with where the project currently is at; it may be a lot smarter to redesign it now.  

I will have a look at games that are similar to the one I am making. 

Shrek SuperSlam 

(7 Studios & Shaba Games, 2005)  

This is what I am trying to emulate the most. 

Players have to know a bare minimum of 7 commands to grasp the game:

Fast Attack

Strong Attack

Grab/Throw

Block

Jump

Evasive Dash

Equip/un-equip weapon

Gang Beasts

(Boneloaf, 2014)

This what I was initially inspired by in the early stages of my project. However, I diverged and wanted to add more originality to my project.

Players have to know a bare minimum of 6 commands to grasp the game:

Jump

Left Punch/Grab

Right Punch/Grab

Lift

Duck/Headbutt

Kick

Tekken 7

(Bandai Namco Studios, 2015)

I’ve played way too much of this throughout the semester.

Note: By default L2 and R2 are not mapped to anything.

Players have to know a bare minimum of 7 commands to grasp the game:

Left Punch

Right Punch

Left Kick

Right Kick

 High/Mid Block (Hold back)

Low Block (Hold Back+Down)

Rage Art

WWE 2K Series

(Visual Concepts & Yuke's, 2016)

I took inspiration from the wrestling games due to them being a good party game, because they involve a large amount of players which can lead to very unpredictable results. However, from the image you can see the controls are not the easiest thing. Funny thing is that they’ve been simplified in the later games ;)

Players have to know a bare minimum of 10 commands to grasp the game:

Strike

Grapple

Irish Whip

Signature/Finisher

Run

Reverse

Taunt

Interact with objects

Limb Target

Target opponent 

Dragon Ball Xenoverse Series

(Dimps, 2015)

This stood out because of its large environments that player fight in.

Players have to know a bare minimum of 11 commands to grasp the game:

Weak Attack

Strong Attack

Ki Blast

Boost Dash

Jump/Ascend

Descend

Lock On

Skills Display (Shows Supers)

Ultimate Attack

Guard

Item List

Super Smash Bros. 4

(Bandai Namco & Sora Ltd., 2014)

Kind of necessary when talking about part fighters.

Players have to know a bare minimum of 5 commands to grasp the game:

Attack

Jump

Special Move

Smash Attack

Shield

Reflection:

DTB currently has 7 commands.

After going through these schemes, I have noticed that I first have to decided what audience do I really want to cater to. Throughout the semester, my focus has changed and I have wanted to make this fighter a lot more technical. As a result, I have lost my way and ended up with a half baked product. Thus, I will take a step back and redesign the combo system into something much simpler. Instead of limb based combat, it will just be strong and light attacks.

Furthermore, I will start removing unnecessary mechanics and abilities, such as:

Double Jump

Moving while blocking: will just trigger the roll.  

If all goes well, the number of commands will drop to 6.

References:

7 Studios & Shaba Games. (2005). Shrek SuperSlam [Video Game]. Activision.

Boneloaf. (2014). Gang Beasts [Video Game]. Double Fine Presents.

Bandai Namco Studios. (2015). Tekken 7 [Video Game]. Bandai Namco Entertainment.

Visual Concepts & Yuke's. (2016). WWE 2K17 [Video Game]. 2K Sports.

Dimps. (2015). Dragon Ball Xenoverse [Video Game].  Bandai Namco Entertainment.

Bandai Namco & Sora Ltd. (2014). Super Smash Bros. for Nintendo Wii U [Video Game]. Nintendo.

[isthisprogress]

Drunken Tavern Brawler - October GDM and Losing Myself (Week 10)

Last night I presented at the game developer meetup and oh boy was it an interesting one. Ben wanted me to present on the projector and gave me a MOA screen. He wanted me to use the projector, because it would allow others to spectate. Little did I know, the projector was connected to 2 other side screens  Hence, I got to present using 4 screens. It was an unexpected honour that just felt a little too good to be true. 

Enough with the gloating, back to the meetup. Throughout the meetup, gameplay bugs became more apparent and I got a better understanding on how they are triggered.

Bugs:

Physics assets spazzing out.

Players being unable to attack.

Players catching thrown objects and having them get destroyed while they are being held.

Unconscious players ricocheting around the environment.

The feedback, I received was also interesting. 

Make one button do multiple actions. E.g. When players strike an object, make that pick it up. This will simplify the complexity that already exists in my project.

Make objects catchable. Something that I will do, if I have time.

Punish players for button mashing. Which already existed, but was removed to compress a bug for the meetup. 

Have a push back, force on the targets struck. 

The general feeling I got from this meetup was that the project had a complex control scheme and system. Something I was actually aiming for, where it can be picked up by people that play fighting games; easily. Nevertheless, I have to find a good balance that has difference (skill) levels of play. Where expert games have an entirely different play from casual games. This has made me want to take a step back and look at what exactly is vital.

Reflection:

At the beginning of the semester, I declared that I was changing my approach entirely to a test driven approach. Where, I would slow down and slowly add little bits to the game; then do a play test. This approach worked wonders, I managed to have a testable demo by week 3. As a result, the progression was very smooth and productivity was high. 

Then the September Game Developer meetup happened...... After receiving a lot of feedback. I got so absorbed and wanted to overhaul the project and build it around ALL the feedback, I found useful. As a consequence, I reverted to old habits and kept developing without testing. Progress was made, but at the meetup more problems were  present than any other part of the project.

Therefore, I have decided to take a step back and re-evaluate the project. Where, I may go back and re-design elements of the project.

[isthisprogress]

Drunken Tavern Brawler - Game Update (Week 10)

Blocking:

There's now an animation and shield material for blocking. The material begins to fade as the shield decays. The decay is speed up by damage taken. The number is for debugging. 

Roll/Dash: 

Players can roll by tapping R1/RB while moving. It has a 3 second cool down. 

Dash Cancel: 

Pressing block while dashing stops movement.

Flat Cursors:

Harry has changed the cursor to something more flatter.

Before:

After:

Health Pack:

 I have added health packs that recover either 25% or 50% health. As well as a power up manager that procedurally spawns them.

[isthisprogress]

Drunken Tavern Brawler - For The Sequel (Week 9)

The slow weeks have started to hit me, I am working a lot less and my production rate has dropped. This was kind of expected, due to how much work I had gotten done prior and a cross critique session that made me feel as though I would have gotten more from not coming in. The inevitable slope of slack and demotivation was bond to come. However, with this being my 6th semester; I have a better idea on how to deal with this. Hence, I have looked back at the Trello board and had a thought about what really is essential to the final product.

Upon reflection I have spoken to the others and decided to remove:

Weapon system

Object health and smash

Catching thrown objects

Tackling

Air combos

https://trello.com/b/SmoWX2ni/drunken-tavern-brawler

Furthermore, late last week I decided to count up the hours and give myself another 3 weeks to work on development; because I will need that week and a half to work on bug fixing and optimisation. I have assigned a maximum of 30 hours a week, something that is manageable and won’t drive me to the brink of insanity as all the bugs slowly creep up and make me want to cry myself to sleep knowing that it’ll never be anywhere close to what I dreamed and won’t put too much stress on me.

Another Round of Playtests

Playtest 1:

I ran another three rounds of play tests with a total of different 11 people. During these tests, I noticed that combos were not utilised and players were still not sure if they got hit. One piece of feedback that 2 testers recommended were to have characters flash red or a distinct colour when hit.

I made a quick mock up to test this feature and it did its job, but needs a visual improvement to blend in smoother.

Hit React Colour Feedback:

Aside from design improvements these play tests also made more of the bugs obvious.

Players unable to attack.

Players falling out of the map entirely.

The upper body glitching out:

Reflection

Looking back on all this information and what I have learnt. I know that I will have to keep cutting tasks and focusing on proactive and reactive game balancing. Where fine tuning values are going to be a big task.

Also, I tried working on rolling and got this wonder:

[isthisprogress]

Drunken Tavern Brawler (Week 8)

Dynamic Camera WIP

Drunken Physics Animations Part 1: This Isn’t Gonna Be Too Hard

Drunken Physics Animations Part 2: Pfffffff, I Knew That

[isthisprogress]

Drunken Tavern Brawler - Drunken Physics Animations Part 2: Pfffffff, I Knew That (Week 8)

So, it’s Sunday and I have finished it to a somewhat responsive state. The graph can get stuck and bug out the animation. However, I’ll have to fix that would a couple more checks. It took 3 days, but here it is:

Furthermore, I have spent sometime helping Ryan rig for his other project. During that short time, I was reminded about bone parenting. This got me thinking about, whether I could send a simulate physic command down a selected bone chain. I did this before the last blog post. Previously, I was swapping out the character’s ragdoll and this would cause the arms to reset into place. However, after doing this the movement became smoother. With this method I cold avoid doing all the blend bone by layer and fake the effect of specific animated limbs.

Nevertheless, when I had finished doing all of this. I discovered I could apply this method and get it working with animation montages. *Shrug. Which would have eliminated one of those days. Now, I am thinking of redoing it all with animation montages. It’ll be more work, but will save me when debugging in the long run

On the bright side, I have fixed the jerky dynamic camera.

[isthisprogress]

Drunken Tavern Brawler - Drunken Physics Animations Part 1: This Isn’t Gonna Be Too Hard (Week 8)

One of the big problems with my combat system was that I would have to undo the ragdoll to use any of the attack animations. Throughout the semester, I have been searching for a way to play an animation on a specific limb. However, when going through the animation graph I came across layer blend per bone that allows me to play an animation on a specific set of bones.

image

Cool right!? Well, kind of. Because of this I can’t use animation montages (Something that would have made my life easier). Hence, I have to do this of the animation graph. Basically, the attacks are another action like run and jump; instead of an interrupt. So, I can’t call the animations unless a bunch of conditions are met and they can’t take priority. 

As a result, I had to spend a couple days towards designing combo stream and making then making the animation graph in a somewhat readable manner for debugging. Then figuring out what methods and calls to make in the animation’s notification tab.     

This is what it currently looks like: 

[isthisprogress]

Drunken Tavern Brawler -  Dynamic Camera WIP (Week 8)

Today I started working on my dynamic camera. So far far I still have to pan out a couple values, work on the speed, and fix the bug that teleports the camera away, if player 1 get eliminated.