The Sonic Center
Sonic Central => Competition Central => Topic started by: Luxray on May 29, 2010, 12:13:46 pm
-
Now please note, before I begin, that for now I have only taken the top 50 players on the Sitewide Rankings (http://www.soniccenter.org/rankings/sitewide). If this gains enough appeal, I will expand it to 100 or perhaps 150.
The SSR rating is taken from your stats across TSC, assigning a value (which is the SSR Rating) and then ranking you alongside other TSCers. So using the following formula we are able to determine a rating of your stats which we are calling the SSR:
Sitepoints -1
---------- = SSR = Efco
# of stats
Efco rating accurately measures your stat efficiency.
This is the chart as of 29th May 2010:
Rank | Player | SSR |
|
|
|
1 | maggot | 1.339 |
2 | SprintGod | 0.760 |
3 | stanski | 0.693 |
4 | pjmaster | 0.484 |
5 | DarkspinesSonic | 0.458 |
6 | Miles9999 | 0.434 |
7 | Romulo The Brazilian | 0.420 |
8 | frankoredstrato | 0.405 |
9 | fuzzerd | 0.378 |
10 | yoshifan | 0.363 |
11 | mike89 | 0.312 |
12 | Shadow Jacky | 0.297 |
13 | eredani | 0.289 |
14 | Zeupar | 0.277 |
15 | SpinDashMaster | 0.272 |
16 | bertin | 0.267 |
17 | truesonic53 | 0.266 |
18 | douglas | 0.253 |
19 | SonicTheWerehog | 0.244 |
20 | Jawzun | 0.243 |
21 | Stefan | 0.240 |
22 | OmegaDJ | 0.237 |
23 | Judgement | 0.228 |
24 | Paraxade | 0.215 |
25 | PfoSonic | 0.210 |
26 | SkyLights | 0.206 |
27 | Sonic62 | 0.204 |
28 | Psyknux | 0.191 |
29 | Auriman1 | 0.184 |
30 | RPGnutter | 0.177 |
31 | Thorn | 0.171 |
32 | Ring Rush | 0.168 |
33 | sonicam | 0.160 |
34 | Strong Bad | 0.152 |
35 | PsyBorg | 0.151 |
36 | blueblazer_8 | 0.147 |
37 | Sephiroth | 0.143 |
38 | Quartz | 0.142 |
39 | sonichero | 0.125 |
40 | Cruizer | 0.115 |
41 | sonicandamy | 0.108 |
42 | Flying Fox | 0.107 |
43 | magnum12 | 0.102 |
44 | MiStEr_ShAoLiN | 0.094 |
44 | teamblackarms | 0.094 |
46 | Groudon199 | 0.091 |
47 | GerbilSoft | 0.082 |
48 | knuckles_sonic8 | 0.079 |
49 | eggFL | 0.073 |
50 | Rolken | 0.051 |
| Average | 0.258 |
I have attached a file with a detailed version of this chart. It lists Efco for each player.
Be back in a month with the updated stats.
-
There is also another rating which I call the Efco rating, which accurately measures your stat efficiency. This should help competitors get their stats higher and higher on the charts and give them more sitepoints.
Haha, no. What this does is encourage exactly what we've been trying to avoid: refusing to submit a stat unless it's near the top of the charts. It also discourages playing multiple games in lieu of playing one game over and over, because the leadership and championship points contribute to your chart. You could argue that the Sitewides already award more points for better stats in one game than above average ones in multiple games, but your chart actively penalizes players for submitting in games they haven't become proficient at. Yes, I'm aware that's the point of the chart, but please, please do not use this as a way to sway how people compete, as it defeats the spirit of the Sitewides.
Also, here's that calculus Luxray was talking about:
Sitepoints -1
---------- = SSR = Efco
# stats
I hope that didn't make your brains explode. :P
-
^ DON'T REVEAL MY SECRETS D:
-
WHAT SECRETS
ALL I DID WAS CLICK MAGGOT'S TWO RATINGS AND LOOKED IN THE FORMULA BOX ZOMG
-
But according to TSC I should be #1 in SSR :(
-
Haha, no. What this does is encourage exactly what we've been trying to avoid: refusing to submit a stat unless it's near the top of the charts. It also discourages playing multiple games in lieu of playing one game over and over, because the leadership and championship points contribute to your chart. You could argue that the Sitewides already award more points for better stats in one game than above average ones in multiple games, but your chart actively penalizes players for submitting in games they haven't become proficient at. Yes, I'm aware that's the point of the chart, but please, please do not use this as a way to sway how people compete, as it defeats the spirit of the Sitewides.
The antirecords charts does the same thing, penalizing players for submissions in games they are not proficient at?
-
No, anti-records just points out some of your more sub-par stats. Aside from them being listed on the page you aren't penalized for it. Thorn is referring to penalization on the actual rankings; eg, you submit some sub-par stats, and your rank goes down instead of up. The way TSC works right now, submitting stats will always make your rank go up, on both the individual game rankings and on sitewide, even if they're bad, though better stats are better-rewarded. This encourages people to submit stats rather than not submitting at all to games they aren't really good at.
-
There is also another rating which I call the Efco rating, which accurately measures your stat efficiency. This should help competitors get their stats higher and higher on the charts and give them more sitepoints.
Haha, no. What this does is encourage exactly what we've been trying to avoid: refusing to submit a stat unless it's near the top of the charts. It also discourages playing multiple games in lieu of playing one game over and over, because the leadership and championship points contribute to your chart. You could argue that the Sitewides already award more points for better stats in one game than above average ones in multiple games, but your chart actively penalizes players for submitting in games they haven't become proficient at. Yes, I'm aware that's the point of the chart, but please, please do not use this as a way to sway how people compete, as it defeats the spirit of the Sitewides.
Also, here's that calculus Luxray was talking about:
Sitepoints -1
---------- = SSR = Efco
# stats
I hope that didn't make your brains explode. :P
This.
-
I cannot fathom how this ranking thing sounds like a good idea
-
To be fair, this list seems better at putting people where they should be based on demonstrated skill. However, as far as the actual rankings are concerned, I agree with everyone above who said we should use the old system. Even if it is less accurate, it encourages competition.
-
To be fair, this list seems better at putting people where they should be based on demonstrated skill. However, as far as the actual rankings are concerned, I agree with everyone above who said we should use the old system. Even if it is less accurate, it encourages competition.
Well this system is technically still being developed and its aim is to work alongside the existing sitewides as a supplement to encourage more competition.
-
I am left confused as to why we need an alternate version of the Sitewide Rankings and how the ranking divided by an arbitrary number constitutes a new and improved system...
-
The variable in the denominator is not arbitrary. >_>
I am left confused as to how the ranking divided by an arbitrary number constitutes a new and improved system...
It doesn't.
its aim is to work alongside the existing sitewides as a supplement
-
So after some feedback on this, i have changed the formula around to something more game balanced. The SSR Rating shall now be found in each game, and then summed up across all games across TSC. The rating is still the same, in that it is the sitepoints earned in a game divided by the number of submitted stats in said game. Submitted stats does not include totals, so it becomes more rewarding with filled stats. To make it look complex, i have an image of the formula:
(http://img706.imageshack.us/img706/3057/86577639.jpg)
I am still undecided how to apply this into the Efco rating. It'll either be the inverse of the SSR in each game summed up, or the total SSR Rating inversed.
-
Fair enough Zeupar, but that still doesn't answer my first question of why this is better than just using the Sitewide Rankings to compete. When it comes to competitive value, I can see that this shows the the value of the statistics a given person has submitted, but the way this is working out, it factors in all of the things that would not be relevant to statistics in this case, and by that I mean Championships, Leaderships, and Percentage, and ends up forcing those things to become relevant when determining the... whatever SSR stands for. Can someone tell me what SSR stands for in this context? Is it Sitepoints/Stats Ratio?
As for the formula, I'm assuming you intended to substitute the formula on the sitewide FAQ? Besides my obvious criticism that your formula currently cares too much about our friendly neighborhood site mechanics and instead be straight-up dependent on the statistics themselves, it doesn't appear to be quite the same. The c variable is missing the 2x. And unless I'm reading it incorrectly, the n and s variables are essentially the same thing, so your formula should probably account for that.
-
Fair enough Zeupar, but that still doesn't answer my first question of why this is better than just using the Sitewide Rankings to compete. When it comes to competitive value, I can see that this shows the the value of the statistics a given person has submitted, but the way this is working out, it factors in all of the things that would not be relevant to statistics in this case, and by that I mean Championships, Leaderships, and Percentage, and ends up forcing those things to become relevant when determining the... whatever SSR stands for. Can someone tell me what SSR stands for in this context? Is it Sitepoints/Stats Ratio?
As for the formula, I'm assuming you intended to substitute the formula on the sitewide FAQ? Besides my obvious criticism that your formula currently cares too much about our friendly neighborhood site mechanics and instead be straight-up dependent on the statistics themselves, it doesn't appear to be quite the same. The c variable is missing the 2x. And unless I'm reading it incorrectly, the n and s variables are essentially the same thing, so your formula should probably account for that.
Don't remember saying its better, since it does rely on the sitewides. This is only a supplement to make competition more spread across all games and an incentive to get people to submit all their stats. And yes you would be correct in what the acronym stands for.
...actually i should remove the leaderships and championship part of the formula :U
And the n / s variables are only the same if a person fills in all their submissions. In fact, the s variable includes totals, whereas n doesn't (i think?). Eventually, this formula will rely on its own take on the statistics and not the sitewide formula. The formula was just a basis to work off.
-
If one of your aims for this project is to have your formula encourage people to submit statistics to all categories of a game, then your formula should probably account for that. However, once it does, it will probably start serving the same functionality as the sitewide rankings formula, and less the functionality of determining the value of a person's average statistic, which is what I meant when I asked why this would be better than just using the Sitewide Rankings to compete. The closer you make the formula's purpose to being a robust do-everything ultimate trick, the less exclusive the information is.