Balance

Web log of development issues.

Balance

Postby adamsderk » Wed Jul 23, 2008 8:41 pm

Greetings,

It's been a while since I showed anything, but the mind has been working.

I was successful at getting the terrain to be generated before the mission, so I can show it during unit selection. I'm thinking of allowing the player to select their drop point as well (based on possible locations). It did shed light on the nav point selection process, which is where the errors and subsequent freezes come from. I hope to make a new method free from computing errors even if it makes for some strange terrain modifications (making a flat area).

But the real issue that I've been working on is the piloting skill. There are concepts that if you say them fast enough it sounds good and straight forward. But when you go to implement it, all the detail rush in to make it impossible. Or at least until you can sort it all out and come up with a solution.

The piloting skill is a ball on two axis that will move from center when you do something risky. It represents the Mech's center of gravity which for the game I am calling "balance". Now you as a pilot have some control over this point and can move it at a certain rate. Based on your control of this ball, your piloting skill can be determined. Sounds good right? But let's look at the details.
1. How does a point increase in the piloting skill roll equate to this moving ball?
2. When do you check to see if there is a possibility to fall?
3. How much control does the player get?
4. How do you give some "play room" before the fall can take place?
5. How does an increase in piloting roll based on speed relate to an increase in piloting roll based on a damaged gyro and relate to an increase based on destroyed hip actuator?
6. How do you move a 10 second roll to an constant instantaneous roll?

My initial direction was to do something like an aircraft flight envelope. This would show where the pilot was within it and when they crossed over. This didn't work as the balance envelope would change at the same time as the ball and wasn't predictable.

So here is how it is currently going in, which will of course change as I actually do it :)

1. The piloting roll can be between 2 and 12.
2. The average piloting skill is 5.
3. The maximum piloting skill is 0.
4. Balance will move at a rate based on the piloting modifier (PM).
5. The player can move the balance at a rate of 5 PM (to bring it to 0 skill if perfect).
6. There will be an inner circle buffer of 2 (based on minimum die roll and skill of 0).
7. The outer circle of 12 will be a guaranteed fall.
8. Balance between the two circles will generate a fall at the appropriate rates; the farther out, the greater the chance.
9. The balance will be "pushed" based on the current modifiers.
10. The balance can be "pulled" by player.
11. Stopping (or falling) will bring the balance back to center.

So an extreme example. A player who didn't want to chance a fall, but push the limit, could be running at less than 5 PM and move the balance to the opposite of the maneuver they want. They could then pop the maneuver of 4 PM which would slam them to the other side of the inner circle. Actual maneuver would be on the lines of turning fast while running fast or even normal movement with lots of damage. The amount of balance change will depend on the actual speed, but pushing a Mech that fast is costly. Remember, just sprinting (which is full throttle) will cost 2 PM, I'll let you do more by turning, but not without risking falls.

If the player is maneuvering at 5 PM or more will only be able to slow the offset to a guaranteed fall, but not stop it.

I'm pretty excited at what this will bring to the game, but it will be some work to get it in mostly because it used so many places; speed, Mech damage, weapons hitting, terrain, water, jumping, etc.

I would like to thank Sandman for his statistics work on the jamming rates for the RAC and UAC. He got frustrated with some of my choices, but the final result has worked really well, which is why...

Sandman,

Could you provide me with a chart of the odds for rolling 2D6 from 2 to 12 and then make the odds for achieving the same odds in 20 rolls on a 1D100. Ex. Odds of rolling 7 or less is about 50%, what would be the number I would have to roll less if I rolled a 20 times and wanted it to be 50% for all the rolls combined?

Thanks.
Image
adamsderk
Site Admin
 
Posts: 1404
Joined: Tue Apr 19, 2005 11:52 pm
Location: Western Washington

Postby felderup » Wed Jul 23, 2008 10:31 pm

that....reminds me of the, i think, cougar, that was falling over in the mw4 training mission, bad balance could explain that, just an observation, as, falling over when you're just standing there could be something interesting. i'm thinking, a limiter for pilots, good pilots get less limiter, great pilots get no limiter, the green troop would be plodding along at a speed and turn rate that would make it dang difficult to tip.

maybe, for controls, if a hat switch were used to control lean amount, lean forward for a jump to full speed from a standstill, lean back when coming up short, left and right for hard turns at high speed, could be interesting.

mind you, i'm tired, and i say freaky things when i'm tired, you can ignore me if you wish.
felderup
 
Posts: 127
Joined: Sun Jan 28, 2007 1:29 pm
Location: dartmouth, ns

Postby adamsderk » Thu Jul 24, 2008 3:03 am

Felderup,

That is exactly what will happen. I plan on having a "movement control" toggle to keep the beginner players from falling all the time.

I figured that the hat is where the balance control would be, but the keyboard will have to have four keys.

Thanks.
Image
adamsderk
Site Admin
 
Posts: 1404
Joined: Tue Apr 19, 2005 11:52 pm
Location: Western Washington

Re: Balance

Postby Hana Yuriko » Thu Jul 24, 2008 4:22 am

adamsderk wrote:Sandman,

Could you provide me with a chart of the odds for rolling 2D6 from 2 to 12 and then make the odds for achieving the same odds in 20 rolls on a 1D100. Ex. Odds of rolling 7 or less is about 50%, what would be the number I would have to roll less if I rolled a 20 times and wanted it to be 50% for all the rolls combined?

Thanks.


While I may not be Sandman, a quick search online can show you the bell curve for rolling 2d6.

http://www.thedarkfortress.co.uk/tech_r ... _rolls.htm

http://en.wikipedia.org/wiki/Dice#Probability

Plus, if you wanted a probability calculator, there's this program: http://www.fnordistan.com/smallroller.html
Hana Yuriko
 
Posts: 274
Joined: Sat Apr 23, 2005 4:18 am

Postby adamsderk » Thu Jul 24, 2008 4:00 pm

Hana,

Neat little program :)

However, if I ask Sandman to run the numbers for me, I can ask him for an equation that I can use that will cover all of the odds. :twisted:

Thanks.
Image
adamsderk
Site Admin
 
Posts: 1404
Joined: Tue Apr 19, 2005 11:52 pm
Location: Western Washington

Postby SandMan » Sun Jul 27, 2008 6:54 pm

Derk,

Sorry I haven't gotten back to you, but I've been out camping the last two weeks. I think I understand what you're asking for, so I'll see what I can do. Combinatorics for the 20 rolls combined will of course be the most interesting part, so that might take a little time to make sure I've got it right. I'll email you when I've got something. Expect cool graphs and equations you never knew existed.

//SandMan
For Davion!
ImageImage
SandMan
 
Posts: 172
Joined: Mon Jan 07, 2008 4:48 pm

Postby adamsderk » Sun Jul 27, 2008 10:59 pm

Sandman,

Hope you had a good trip. I look forward to the report.

Thanks.
Image
adamsderk
Site Admin
 
Posts: 1404
Joined: Tue Apr 19, 2005 11:52 pm
Location: Western Washington

Re: Balance

Postby SandMan » Tue Jul 29, 2008 8:15 pm

Derk,

Check your inbox. It wasn't completely clear from your description, but I assumed that you wanted the odds of at least one failure in 20 1D100 trials to equal the odds of one failure in one 2D6 trial.

(I also had to go to 20 1D1000 because a D100 is too coarse, as you'll see on the spreadsheet).

So, for example, the odds of failing one roll on 2D6 with a TN of 7 is 41.67%

The odds of failing at least one roll in 20 1D1000 with a TN of 27 is 40.96%


Also, I chose to always rule in favor of the player, when two values were close. For instance: in the above example, a TN of 28 gives 42.16%, which is about 0.2% closer to the true value. However, it will result in slightly more failures than CBT calls for, and more failures = less fun. Not that a fraction of a percent is going to make much of a difference either way, but there you have it.


So, were we on the same page for all that?

//SandMan
Last edited by SandMan on Wed Jul 30, 2008 4:05 am, edited 1 time in total.
For Davion!
ImageImage
SandMan
 
Posts: 172
Joined: Mon Jan 07, 2008 4:48 pm

Postby Smoking_Mirror » Wed Jul 30, 2008 3:56 am

Arg! Maths! >_<
Smoking_Mirror
 
Posts: 370
Joined: Wed Jul 04, 2007 2:38 pm
Location: South Korea (English)

Postby SandMan » Wed Jul 30, 2008 4:10 am

Smoking_Mirror,

Not sure if I can post the excel file here, for those not-Derk to peruse and ogle. But I could host it for a little while, if you're interested. Or, I could just explain the combinatorics to anyone who isn't clear on them. Unless, of course, that was an actual grunt of pain at the maths :P

//SandMan
For Davion!
ImageImage
SandMan
 
Posts: 172
Joined: Mon Jan 07, 2008 4:48 pm

Postby Hana Yuriko » Wed Jul 30, 2008 7:22 am

I'd hope that this random chance of failure doesn't translate to, "I was able to maintain my balance then suddenly for no reason, I fell over!"
Hana Yuriko
 
Posts: 274
Joined: Sat Apr 23, 2005 4:18 am

Postby Smoking_Mirror » Wed Jul 30, 2008 11:31 am

Maybe it will just mean the indicator moving faster some times rather than others?

Really in real life there's not a lot of random chance. When driving down the road there's a slight chance that you will lose control for no reason and crash in to a tree, and in fact, it does happen. However it's not the sort of thing you need to factor in to a driving simulator.

I guess it's more like sometimes when you are running carelessly on a fairly even surface, suddenly you accidentally trip on a big rock and fall over. In normal circumstances there's zero chance of falling, but because of the random rock; oops, I fell over.
Smoking_Mirror
 
Posts: 370
Joined: Wed Jul 04, 2007 2:38 pm
Location: South Korea (English)

Postby adamsderk » Wed Jul 30, 2008 5:55 pm

Sandman,

Always a pleasure to see your reports. You did it perfectly.

So next step, given an input value from 2-12, can you give me an equation that will output the cutoff number? I'll then do a random number generation from 1-1000 and check it against the cutoff number.

Smoking_Mirror,

While I don't love math, I don't shy away from it either. Games are completely constructed from math, so if you don't like it, you won't be successful at it.

Hana,

There are two concentric circles, one at 2 PM (piloting modifier) and one at 12 PM. If you keep the balance within the small circle, you won't fall. But if the balance moves outside the small circle (which it will), your chance of falling is based on how far from the small circle you are. If you are able to stand up while your balance goes all the way to the outer circle, you will instantly fall.

I expect most people will play outside the small circle. How far outside of it is based on their level of risk. But eventually you will fall.

The player will know they are pushing it though, it will be a surprise "when" but not "if." A good pilot will be able to chose methods to minimize their exposure, for instance, if you turn left instead of right, you might push the balance in a favorable direction.

Thanks.
Image
adamsderk
Site Admin
 
Posts: 1404
Joined: Tue Apr 19, 2005 11:52 pm
Location: Western Washington

Postby adamsderk » Tue Aug 19, 2008 4:59 am

Greetings,

It's been a while and I've been struggling with the engine again. I got side tracked from piloting to put in the sprinting speed levels. That lead to the code that "remaps" the speed to make it smoother. I was using 0 to .87 and needed to change it to 0 to 1.00. In doing so, I ended up moving some code around that allowed a divide by zero that took three days to track down (I had to go to a previous version and step through the changes to find it).

So where am I now. I have sprinting in, the display bar on the hud has changed to show a green, yellow and red bar for walk, run, and sprint speeds respectively.

Turning is still limited to run speed. So that means if you are at full run or higher (sprinting), you can't turn. If you are less than run speed, you can use up to run with turning. Standing still will turn you at run MP levels.

Why am I being particular about this... well it matters when you get to the balance marker.

The beginnings of the balance marker is in. That means that your speed will move the balance based on the speed used (including turning rate). If you move forward, the balance will move forward at the rates listed in CBT. If you turn (assuming you are under a run), the MP you used to turn will move the balance to the side (depending on the direction of turn).

The player will use the "awsd" keys to move the balance back at a rate of 5 points per 10 seconds. Considering that the you would have to be using 18 MP to generate more than 5 points per ten seconds, most movements will be easily managed by the pilot. The issues will come in with joint damage along with one time balance modifiers.

Thanks.
Image
adamsderk
Site Admin
 
Posts: 1404
Joined: Tue Apr 19, 2005 11:52 pm
Location: Western Washington

Postby SandMan » Tue Aug 19, 2008 5:14 am

So when do we get to play with this?

:D

Also, did you see the new Hull Down rules in TacOps? Sounds like just the thing for this balancing act. Basically, if you start to fall, you can try to 'take a knee' to avoid taking damage.

//SandMan
For Davion!
ImageImage
SandMan
 
Posts: 172
Joined: Mon Jan 07, 2008 4:48 pm

Next

Return to Development Blog

Who is online

Users browsing this forum: No registered users and 2 guests

cron