You are here

Quixy's and Viigoo's research lab

Before posting make sure you are aquainted with our forum rules.

72 posts / 0 new
Last post
Quixy's picture
Offline
Quixy's and Viigoo's research lab

Hi everybody,

Me and my friend, Viigoo made an Achtung Die Kurve lab:

What is the Achtung Die Kurve lab?

The Achtung Die Kurve does research to things in Achtung Die Kurve. We search for all kind of statisctics in Achtung Die Kurve. For example we can find out how much holes you get in 1 second or how long a power-up lasts. 

About this topic

Every statistic we found in Achtung Die Kurve will be post in this topic. So in this topic will be a lot of helpfull information about Achtung Die Kurve.

Questions

Viigoo and I are looking for things we can research. So do you have a good question for the Achtung Die Kurve lab? You can post it here. 

Greetings Quixy and Viigoo

-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

Results

How long does it lasts to turn 180 degrees? And with power-ups?

It takes 1,25 seconds to turn 180 degrees.

It takes 1,75 seconds to turn 180 degrees with the green turtle power-up.

It takes 0,75 seconds to turn 180 degrees with the green speed power-up.

It takes 0,75 seconds to turn 180 degrees with the red speed power-up.

It takes 2,50 seconds to turn 180 degrees with th red turtle power-up.

 

Frequency of making a gap/hole

We did a test of the frequency of making a gap/hole. The total holes we did make is 915.

We made them in 1952 seconds. So you get an average of 0,46875 holes per second. That's almost one hole in 2 seconds.

We think it's programmed to 0,5 holes per second.

 

Orthogonal lucky line through move

How big is the chance that you go through a line with just making a lucky hole? One thing is sure: You can't go through a line when your head and the line make an angle less than 30 degrees. We tried 317 times to go through a line with an angle of 90 degrees.

We failed 296 times and we had 21 times succes.

That means that you have chance of 6,62 % to go through a line with making a lucky hole in an angle of 90 degrees.

Another research about power-ups

 

Alfy wrote:

One month ago I did another experiment with help from ADKPro. We wanted to find out what determines the winner of a point if both players die at the exact same time. So we made a scaled room and at the start of every round both players went outside the screen before the lines settled. At the beginning we thought it was random who would win the point, but the result was 10-0 for me. We then changed colors and then ADKPro won 10-0. So we came with the hypothesis that the winner is determined by the colors. This means that some colors are better than others. We then tested it with every color and got these results.

1. blue        (beats 6 colors)
1. light green (beats 6 colors)
2. light blue  (beats 5 colors)
2. green       (beats 5 colors)
3. red         (beats 3 colors)
4. orange      (beats 2 colors)
5. yellow      (beats 1 colors)
6. pink        (beats 0 colors)
 

We can confirm this and it also works in ffa

 

Offline
Champion

How many seconds takes it to turn a half circle? And with green PU and red PU?

krass's picture
Offline
Clan: Boss

It takes 1,25 seconds to turn 180 degrees.

It takes 0,75 seconds to turn 180 degrees with the green turtle power-up.

It takes 1,75 seconds to turn 180 degrees with the green speed power-up.

 

Corrected you. Smile But how did you find out? Are you just using a clock or sth? Biggrin

Quixy's picture
Offline

Hmm you corrected me wrong, because you're faster when you have a speed power-up.

How I did find it out? I just made a simple circle and I timed how long it takes to make one. And then I devided it by 2.

With the power-ups I just picked a big hole power-up. Then I picked up the turtle power-up and I timed how long it takes to make a circle.

Btw I did this many times so I think this is secure enough.

Offline
Champion

What's the average amount of premium powerups in a second? like 0.5/s
And what with basis powerups? 
 

 

- made a sticky topic -

Quixy's picture
Offline

Updated with red speed and red turtle

Alfy's picture
Offline
Champion
Clan: buny

I've previously tested the orthogonal lucky line through move the same way as you and i got about 6.5 %, so that confirms your result. I also want to mention that when you go through right after the head of the other player, the chance of success doubles. This can be used in situations when you are squeezed to the wall by your opponent. Here is a screenshot to illustrate: http://i.imgur.com/7pQvh.jpg
Orange is squeezed to the wall, but tries to be lucky by going through the blue line. The probablility of orange making a hole at the right time is as you said, 6.62 %. But there is also a chance that blue makes a hole at the exact time when orange goes through. I estimate that this is also 6.62 %, so the probability of orange going through is actually 13.24 %! In a real game situation you will probably not hit exactly 90 degrees on your opponents line, so the chance is a bit lower, but it's still suprisingly high.

 

One month ago I did another experiment with help from ADKPro. We wanted to find out what determines the winner of a point if both players die at the exact same time. So we made a scaled room and at the start of every round both players went outside the screen before the lines settled. At the beginning we thought it was random who would win the point, but the result was 10-0 for me. We then changed colors and then ADKPro won 10-0. So we came with the hypothesis that the winner is determined by the colors. This means that some colors are better than others. We then tested it with every color and got these results. 

1. blue        (beats 6 colors)
1. light green (beats 6 colors)
2. light blue  (beats 5 colors)
2. green       (beats 5 colors)
3. red         (beats 3 colors)
4. orange      (beats 2 colors)
5. yellow      (beats 1 colors)
6. pink        (beats 0 colors)

Detailed results: http://www.text-upload.com/read,3846693772745

We did the experiment in team mode to not loose rank (this was before team ranking was implemented in the game). Afterwards we found that it was different results in FFA, so our results only applies to team games.

The conclusion was that in team games dark blue and light green are the best colors because they beat most colors. The results also shows that team coool is better than team hot. In fact, all the colors of team coool beats all the colors of team hot when dieing at the exact same time. 

 

 

Viigoo's picture
Offline
Clan: HC

Thank you very much for your support and involvement Alfy!
Unfortunately Quixy is off for few days and I'm unable to make some of the researches mentioned by you.
But as soon as he comes back he'll surely update the topic and we'll get back to work Blum 3

1) You are absoultely right about increasing the chances of survivng in head close situation - "I also want to mention that when you go through right after the head of the other player, the chance of success doubles"- but if we want to be precise, I don't agree with that the chance of succes doubles - (imagine situation when you have 50% chance of surviving in orthogonal move, in this situation it would not be 100% :P, it would be slightly lower)

2) I'm 99.99% sure that certain power ups have different spawning chance. And you are right 10 tries is not a relieable source of information, but it was just opening information and the base for further analysis. I am making statistics in Libre Office (something like Excel, if you've not heard of it before). Why is that? It allows me to update results and add more statistics, parameters, formula etc. without disturbing previous results... moreover, adding results will just increase the accurracy of our researches, and limit the risk of any kind of mistake.

3) I made with Quixy similiar experiment with connected with duration of certain powerups, and we ALMOST agree with you in 100% :P, in our research we've noted that fat power up lasts 7 seconds, not 8. But don't worry we'll make additional tests to ensure that there was no mistake.

4) I'm shocked with this experiment with winner of a point if both players die at exact time! I thought about it before, but I suspected that the one who is closest to the battlefield wins, anyway we'll surely examine this problem soon!

Haha and of course thanks for updating our "TO DO List" :D,

Offline
Clan: sk4f

This is very nice work!!! I'm promoting this research :D Let's think of something...

Quixy's picture
Offline

Im back of my holiday.

Alfy, wow I didnt know that the Dirol team is in the advantage with team.

I don't think the hole frequency change when you're fat. It looks like you get more holes, but that is because the holes are much bigger.

And thank you for the ideas, we definitely will pick some ideas and work them out.

And everybody feel free to ask question and help with the research.

 

Alfy's picture
Offline
Champion
Clan: buny

Viigoo wrote:

1) You are absoultely right about increasing the chances of survivng in head close situation - "I also want to mention that when you go through right after the head of the other player, the chance of success doubles"- but if we want to be precise, I don't agree with that the chance of succes doubles - (imagine situation when you have 50% chance of surviving in orthogonal move, in this situation it would not be 100% :P, it would be slightly lower)

Of course -.-

It will be 1-(1-0.0662)2 = 0.114  = 11.4 %

 

 

Geert's picture
Offline
Developer
Moderator
ffa eu top 46%
team eu top 55%
Clan: MOD

Did someone had a peek in the source code?! Nice thread guys. I just verified some numbers and in general it looks pretty ok.


 

businessman's picture
Offline
Champion
ffa eu top 8.7%

Whate determines the winner of a "kiss" (two or more players kill each other)?

userious?

Offline
Clan: sk4f

This:

Alfy wrote:

One month ago I did another experiment with help from ADKPro. We wanted to find out what determines the winner of a point if both players die at the exact same time. So we made a scaled room and at the start of every round both players went outside the screen before the lines settled. At the beginning we thought it was random who would win the point, but the result was 10-0 for me. We then changed colors and then ADKPro won 10-0. So we came with the hypothesis that the winner is determined by the colors. This means that some colors are better than others. We then tested it with every color and got these results.

1. blue        (beats 6 colors)
1. light green (beats 6 colors)
2. light blue  (beats 5 colors)
2. green       (beats 5 colors)
3. red         (beats 3 colors)
4. orange      (beats 2 colors)
5. yellow      (beats 1 colors)
6. pink        (beats 0 colors)
 

Quixy's picture
Offline

No, the winner of a kiss is always the player that is slightly a bit further than the other player. These results are only for if you die at the same time in the wall

Offline
Clan: sk4f

Oh yeah I understand what you mean. If I'm turning away from the other players line while 'kissing' I win most of the times.

Quixy's picture
Offline

With the new update it will be much easier to test things, so, who does have a question?

Offline
Champion

Duration one finger

Royal's picture
Offline
Champion
ffa eu top 0.3%

1. Is the frequency (distance; maybe in % of the playing field) of holes normally distributed?

2. If you are really motivated and good in statistics xD: If its not a standard normal distribution (what i guess Smile ) please tell me the Expected Value and the Variance Biggrin

 

If I am expecting too much with this question, please only tell me the Expectet Value. This is no joke, im really interested! Smile

(Expected Value = Average Distance (in this case))

Viigoor's picture
Offline
Champion
Clan: HC

salietje wrote:

Duration one finger

One finger - 5 secs

Check Emperor Jay forum for more info about duration

Offline

AND when you are fat . You make a faster turn .

 
 
 
 
Offline
Clan: sk4f

Is this true? I thought it only seemed that way, the radius is the same I thought.

Offline

No . I think it is true . You make a smaller circle. 

Quixy's picture
Offline

Same radius, but if you're fatter, it looks like your radius is smaller.

benniegek's picture
Offline
Clan: FUCK

i tested something:

 

I tested how many fat powerups you have to pick up to get a curve as big as the earth:

Details:

The earth is +/-12742km. (12742000000mm.)

My screen is 31cm

The normal curve lines are 2mm. (0.000002km.)

With one fat they are 4mm.

 

I took the thickness of a normal curve and did it

x2,x2,x2,x2,x2,x2,x2,x2,x2,x2,x2,x2,x2,x2,x2,x2,x2,x2,x2,x2,x2,x2,x2,x2,x2,x2,x2,x2,x2,x2,x2,x2,x2.

the number was: 16675000000 => 16675km

 

So, the number of fat powerups you have to take to get a curve bigger then the EARTH on my screen is: 33!

Blum 3

 

Saturday CurveFever!!!!   (saturday night fever)

 

Fire's picture
Offline
Champion
team eu top 11%
Clan: SW

:P


~.~

benniegek's picture
Offline
Clan: FUCK

do you think this is right, Fire?

 

 

Saturday CurveFever!!!!   (saturday night fever)

 

Etzomat's picture
Offline
Clan: JI

benniegek: Did you like..measure the thickness of your line with a ruler? :)
Because if I use a smaller screen (lower resolution) my line will be smaller causing me to have to pick up more fat PU to get to the same result. Well, it would only change for 1 or 2 PU I guess..

Quixy's picture
Offline

This thread hasn't been updated for a long time, because Viigoo is in the exam year. 

I'm going to do some other research soon alone, when I have time Smile

Offline

I would like to know how the chances are to jump over a wall when you are fat. As I experienced that often it works, as when you are fat you create bigger holes.

Geert's picture
Offline
Developer
Moderator
ffa eu top 46%
team eu top 55%
Clan: MOD

oops, sorry! By accident I deleted the double post with a very nice reaction attached to it :( Can you please do it again :(


 

Fire's picture
Offline
Champion
team eu top 11%
Clan: SW

Geert wrote:

oops, sorry! By accident I deleted the double post with a very nice reaction attached to it :( Can you please do it again :(

Let the actual mod of the forum do it :)


~.~

Viigoor's picture
Offline
Champion
Clan: HC

Quixy wrote:

This thread hasn't been updated for a long time, because Viigoo is in the exam year. 

This is true that I have some work with exams nowadays, but I actually miss some of our research. I suggest to continue and reopen our lab.

Just let me know Quixy if you have time soon.

WranDon's picture
Offline

What's up? When will the reopening be?

WranDon - the forum slut

Offline
Clan: ES

maybe you test if there is  a rule who will get the better place in  a ffa game if same points, for example: 1. 50points, 2. 45points, 3. 45points

WranDon's picture
Offline

No one, as far as I know.

WranDon - the forum slut

Incred's picture
Offline
ffa eu top 1%
team eu top 35%

I remember Quixy but i think he quit a long time ago. Haha

WranDon's picture
Offline

He was quite good, as far as I know.

WranDon - the forum slut

BeerMaker's picture
Offline

Alfy wrote:

One month ago I did another experiment with help from ADKPro. We wanted to find out what determines the winner of a point if both players die at the exact same time. So we made a scaled room and at the start of every round both players went outside the screen before the lines settled. At the beginning we thought it was random who would win the point, but the result was 10-0 for me. We then changed colors and then ADKPro won 10-0. So we came with the hypothesis that the winner is determined by the colors. This means that some colors are better than others. We then tested it with every color and got these results.

1. blue        (beats 6 colors)
1. light green (beats 6 colors)
2. light blue  (beats 5 colors)
2. green       (beats 5 colors)
3. red         (beats 3 colors)
4. orange      (beats 2 colors)
5. yellow      (beats 1 colors)
6. pink        (beats 0 colors)

 

I know it's been a long while since this thread has been started, but are you guys able to measure this thing again including colors: white, purple and rainbow? I would it very much.

The Maker of Beer has joined the game!

WranDon's picture
Offline

Incred wrote:

I remember Quixy but i think he quit a long time ago. Haha

Looks like he hasn't quit at all ... Mabe I should PM him?

WranDon - the forum slut

Fire's picture
Offline
Champion
team eu top 11%
Clan: SW

WranDon wrote:

Incred wrote:

I remember Quixy but i think he quit a long time ago. Haha

Looks like he hasn't quit at all ... Mabe I should PM him?

He's back


~.~

WranDon's picture
Offline

And quite a noob now ... Wink I'll PM him.

WranDon - the forum slut

Offline
Clan: ALKP

is this still correct? because i once tested it and there dark blue loses always vs yellow, but also lose vs light green, so he would win 5 and not 6?!

TheNewSound's picture
Offline
Clan: Baas

what is the normal speed of a line going straight in pixels/second?

__________________________________________________________

"Then you have squared circles" - 6stijn6

Offline
Clan: FC

what about white, purple , rasta and rainbow colors? who beats who in a kissing battle? Thanks for all the statistics btw, keep up the good work! :D^^

 

<a href="https://signatures.cf/profile/1122268"><img src="https://signatures.cf/img/1122268/default/+1v1_asia+1v1_eu+ffa_asia+ffa_... alt="Rival Proximity's profile on curvefever" style="height: 95px; width: 300px;" / ></a>

PartyBlaze's picture
Offline

Nice work!

 

Indeed, what about white, purple, rasta and rainbow? I've always thought that if two or more people die, that it'll be alphabetically chosen. Like in an FFA battle, when two or more people have the same amount of points.

 

By the way, how about lime and blue? They both beat six colours, but will blue or lime win?

CurveFeverTv's picture
Offline

My friend, Yubi Nashi & I decided to improve the Research Lab where possible.

We saw the post about the difference in colors; that some colors would be better than others. So we decided to redo the whole test and include the new colors like white, purple, rasta & rainbow.

Alfy wrote:

We wanted to find out what determines the winner of a point if both players die at the exact same time. So we made a scaled room and at the start of every round both players went outside the screen before the lines settled.

First we tested red against all other colors and red always ended up losing. After that we did the same for orange and got the same results, orange always ended up losing. We thought something was wrong and decided to make a new room. This time my friend, Yubi Nashi was host (previously I was). Again we tested both colors against all other colors and this time red and orange won every time.
Red (host) vs Yellow = Yellow wins
Red vs Yellow (host) = Red wins
Red (host) vs Orange = Orange wins
Red vs Orange (host) = Red wins
Conclusion: The host gets the least amount of points when multiple players die outside the playing field.

After that, we checked if there would be a difference between the team the color is in. For example, rainbow can be in hot as well as in cooI.
Dark blue (host) vs. Rainbow in hot = Dark blue wins
Dark blue (host) vs Rainbow in cooI = Rainbow wins
Conclusion: Team cooI gets the highest amount of points when multiple players die outside the playing field.

Later we realised guests can also play unranked team games when they get added by a player. We tried it out as well and got interesting results.
Red (guest) vs orange (host) vs pink vs dark blue = dark blue > red > pink  > orange
Conclusion: Guests get the highest amount of points when multiple players die outside the playing field.

We also noticed there is no difference between non-premium players & premium players.
Conclusion: There is no difference between non-premium players & premium players when multiple players die outside the playing field.

We did not get any logical or re-occuring results between colours when we mixed up guests, host & team.
Conclustion: We did not get any logical or re-occuring results between colours when we mixed up guests, host & team. We think the color itself doesn't matter, but we cannot prove it.

 

We then put all those conclusions together to see which variables have the greatest impact.
Conclusion: We know the flowchart of the source code looks something like this.

x = Number of players in hot team
y = Number of players in cooI team

Summary:
We think that when multiple players die at the same time outside the playing field, it doesn't matter which color you are, but we cannot prove it.
The team the player is on has the greatest impact on the amount of points he will get. Every player of team cooI gets more points than every player of team hot.
The host will get the least amount of points of his team. The guest(s) will get the most amount of points of his team.

Example 1:
Players: Yellow (host) vs Pink (guest) vs Red vs Dark blue vs Dark green
Result: Dark blue: 3 or 4 (random)
              Dark green: 3 or 4 (random)
              Pink: 2
              Red: 1
              Yellow: 0

 

This is only for team. We did not test FFA.

Offline
Champion

Gj guys Smile

Offline
Clan: XAXA

Why is yellow always last -.-. For me lgreen and yellow are best basic colors....

Offline

GJ CurveFever TV and Yubi Nashi *GOOD* 

United Arab Emirates -  الامارات العربية المتحدة

 

InfinityMonkeys's picture
Offline
ffa eu top 29%
Clan: IMOG

Maybe people can go further wit this

Pages