graphs of WR development (request)

Look for replays and levels and ask people's times.

Moderator: Moporators

Post Reply
User avatar
ribot
Not banned
Posts: 2416
Joined: 19 May 2002, 16:20
Location: Miranda: the true state
Contact:

graphs of WR development (request)

Post by ribot »

I have seen this subject touched on other threads but there is no thread about discussing and making graphs of WR developments. (I can't download viper's program (http://www.moposite.com/mopolauta/viewtopic.php?t=2069) so i dont know what it is about.) As far as I know there has been no graphical predictions of the total time limit of the internallevels. I could simply do it with my calculator, but i know there are math students that visit this forum that could do it better on computer.

I haven't studied enough statistics to know what kind of graphs you can make, with natural developments of whatever. But I do know that you can easily generate a curve of development with a calculator and the data of years of WR improvements.

So after my babbling, this is my idea, so someone plz suggest software or do it:
Generate graphs of each internal level to see their theoretical mathematical limit.
Also generate a total time development graph, to find a theoretical limit.
Then you can also make a total time limit estimation by adding the theoretical internal time limits together.

And I still think WR will go under 35 minutes.
-
"leader status in the Elma against-the-system underground" - Abula
-
IncrElastoMania - Elma Simulation - Browser Game 2020
Elma Imager - Command Line Tool 2020
psteve
Kuski
Posts: 481
Joined: 19 Aug 2004, 13:41
Location: Canberra, Australia

Post by psteve »

If anyone is going to make a graph, remake a graph, or post an existing graph, make sure you do it in UNITS OF TIME not units of wr tables
I am a Flying Spaghetti Monsterian. My God has noodly appendage. The reduction in pirates is the cause of global warming.
User avatar
Abula
Moposite admin
Posts: 4471
Joined: 16 May 2002, 23:00
Team: FM
Location: Helsinki
Contact:

Post by Abula »

40:02,71 (151.) | WCup4: 8. | 3x WR | 3x GAA | 11x FEM | KOM | The History of Elasto Mania (1995-2018)
User avatar
ribot
Not banned
Posts: 2416
Joined: 19 May 2002, 16:20
Location: Miranda: the true state
Contact:

Post by ribot »

yeah i saw those graphs and it's not what i'm requesting. for some reason those graphs were made with estimations of the lowest total time, and they are not well plotted at all.
-
"leader status in the Elma against-the-system underground" - Abula
-
IncrElastoMania - Elma Simulation - Browser Game 2020
Elma Imager - Command Line Tool 2020
User avatar
ribot
Not banned
Posts: 2416
Joined: 19 May 2002, 16:20
Location: Miranda: the true state
Contact:

Post by ribot »

ok i have done this now, using the data from phillip's excel table. it is not quite correct though, since it rounds for example table #2 to 43.75, when the value is 43.757667. Also i have not bothered to calculate the time, but only using tables as time unit. you can of course do it more accurately if you want to, and then plot the graph with advanced grapher like i did (free trial, just google). this program also calculates different kinds of curves for you, that are best matching the data of given values (total times or level times or your own times if yu want to).

the only useful curve i found is this:
Y(x)=(9.5795519*10^(-5))*x^2-0.0513973*x+43.2077702

which results in a parabola. since the curve fits so nice to the total time we can pretend it means that the theoretical limit, or in other words, the bottom of the parabola, is 36.31 minutes, or 36:18,82, in table 268.

parabola and total times overview: http://www.mxb.dk/upload2/overview.gif
Image
closeup of the limit: http://www.mxb.dk/upload2/closeup.gif
Image
-
"leader status in the Elma against-the-system underground" - Abula
-
IncrElastoMania - Elma Simulation - Browser Game 2020
Elma Imager - Command Line Tool 2020
psteve
Kuski
Posts: 481
Joined: 19 Aug 2004, 13:41
Location: Canberra, Australia

Post by psteve »

While the parabola might fit nice any mathematician will tell you that you shouldn't use it, saying the limit is at the vertex is pretty clueless. Sorry if this sounds like an angry post, but the parabola is definetely wrong.
I am a Flying Spaghetti Monsterian. My God has noodly appendage. The reduction in pirates is the cause of global warming.
psteve
Kuski
Posts: 481
Joined: 19 Aug 2004, 13:41
Location: Canberra, Australia

Post by psteve »

I personally think the graph is linear and will flatten out very sharply like \_ This graph is not common in nature, but one example that this type of graph fits is the rate in which we can make machines smaller. Since we are trying to make wrtt smaller...
I am a Flying Spaghetti Monsterian. My God has noodly appendage. The reduction in pirates is the cause of global warming.
User avatar
ribot
Not banned
Posts: 2416
Joined: 19 May 2002, 16:20
Location: Miranda: the true state
Contact:

Post by ribot »

well there is no reason the curve should follow any mathematical one at all. at least not one that can be predictable at all. however i thort it was funny that i got this result.

if you look at the plotted graph, you can see that it is going downwards like stairs, and is only getting slightly less steep. so graphically it doesn't seem like we are getting close to the limit yet. but it also looks like it is starting to turn to change less dramatically. however these are just my own illusory patterns.

i was thinking about ways to measure this more accurately. or i mean to find out more reasonable limits. one idea was to take each level separately, and somehow find out a limit. if you take warmup, which is a very simple level, and look at its wr development, the bigger changes will occur on new styles, and the smaller ones on improvements of old styles. however these slight changes in styles, as in warm up or uphill battle, should be possible in basically every level.

just look at all short levels and maybe all of them have had generations of improvement. steppes, free fall, hooked, flat track, whatnot. the longer levels in turn should have more space for such subtle styles that still can save a second. i mean just look at good old flat track, with zyntis impossible old WRs. and then suddenly a new revolutionary style.

so now when i have been babbling off subject, or off my point anyway, i will tell my idea. then i will probably check out some levels myself, and analyze. it's not likely that freefall will have another huge style change, and therefore it's not really accurate to find its possible time by looking at the old 3X runs. however, if you look at a bigger perspective, as in the total time of elma internals, you cant really know when new mad styles will come. just look at new curva style, who'd expected that?

my idea anyway, after all this babbling that no one will read, is that you can check only the relevant last wr times, with a similar style as the current wr style, to somehow predict devel.ah fuck it.

Is there anywhere I can find the times for the WR lists? I know one way and that is to click on every single WR list, but that takes so much time with my slow connection.
-
"leader status in the Elma against-the-system underground" - Abula
-
IncrElastoMania - Elma Simulation - Browser Game 2020
Elma Imager - Command Line Tool 2020
User avatar
Juski
Kuski
Posts: 2200
Joined: 26 Dec 2003, 20:53
Location: irc://irc.ircnet.org/ranks

Post by Juski »

On statistiks page u can find that im sure^^


And just look at long haul zweq just improved almost a second there recently, how the fuck did he do that?? You guys say that how should we get under 36 when there are no new styles, were does the improvment fit?? The new styles isnt there cause u havent found em yet.
No regrets Image
Are you LOST?
px
Moporator
Posts: 86
Joined: 17 May 2002, 11:28
Contact:

Post by px »

Here kitty kitty, have some dates

1=06.03.00
2=12.03.00
3=18.03.00
4=24.03.00
5=30.03.00
6=01.04.00
7=06.04.00
8=09.04.00
9=15.04.00
11=22.04.00
12=25.04.00
13=28.04.00
14=01.05.00
15=07.05.00
19=27.05.00
21=11.06.00
22=17.06.00
23=25.06.00
24=01.07.00
26=11.07.00
27=17.07.00
28=23.07.00
30=04.08.00
32=13.08.00
33=20.08.00
34=25.08.00
35=03.09.00
37=18.09.00
39=02.10.00
40=08.10.00
43=29.10.00
44=02.11.00
60=08.03.01
66=05.05.01
67=14.05.01
74=23.09.01
75=21.10.01
77=02.12.01
78=10.01.02
79=17.01.02
80=04.02.02
90=18.08.02
91=25.08.02
92=01.09.02
93=08.09.02
94=15.09.02
95=22.09.02
96=06.10.02
97=13.10.02
98=20.10.02
99=27.10.02
100=03.11.02
101=10.11.02
102=17.11.02
103=24.11.02
104=01.12.02
105=08.12.02
106=28.12.02
107=05.01.03
108=12.01.03
109=19.01.03
110=26.01.03
111=02.02.03
112=09.02.03
113=16.02.03
114=23.02.03
115=02.03.03
116=09.03.03
117=14.03.03
118=16.03.03
119=24.03.03
120=30.03.03
121=06.04.03
122=13.04.03
123=20.04.03
124=27.04.03
125=04.05.03
126=11.05.03
127=18.05.03
128=25.05.03
129=01.06.03
130=09.06.03
131=15.06.03
132=22.06.03
133=29.06.03
134=06.07.03
135=14.07.03
136=24.07.03
137=27.07.03
138=05.08.03
139=10.08.03
140=17.08.03
141=24.08.03
142=31.08.03
143=07.09.03
144=14.09.03
145=21.09.03
146=28.09.03
147=05.10.03
148=11.10.03
149=19.10.03
150=26.10.03
151=02.11.03
152=09.11.03
153=23.11.03
154=01.12.03
155=12.12.03
156=21.12.03
157=05.01.04
158=12.01.04
159=18.01.04
160=25.01.04
161=03.02.04
162=09.02.04
163=15.02.04
164=29.02.04
165=08.03.04
166=14.03.04
167=21.03.04
168=04.04.04
169=11.04.04
170=18.04.04
171=26.04.04
172=02.05.04
173=09.05.04
174=16.05.04
175=30.05.04
176=16.06.04
177=28.06.04
178=25.07.04
179=15.08.04
180=30.08.04
181=06.09.04
182=12.09.04
183=21.09.04
184=26.09.04
185=03.10.04
186=10.10.04
187=19.10.04
188=24.10.04
189=31.10.04
190=07.11.04
191=14.11.04
192=21.11.04
193=28.11.04
194=05.12.04
| TT astonishing 47:08,75 | moposite.com | pxx.fi |
Post Reply