Overview
This is a new-gen ranking list created with the help of professional analytics. It applies some predictive models that are used, e.g., for estimating performance of a sports team, business project, etc.
How Did This Start?
In short, i have an IRL project that involves an analytical team creating different predictive models, partly intended to analyze risks during exploitation of high-end technology and predict risk behaviour. The model is in its early stages, and I decided – for both business purposes and fun – to test it based on Sins replays to develop something that is basically an algorithmic ranking.
About 100 hours of analytical work were required to create the adaptation of such a complex model to a mathematically predictable game like Sins, but the team made some progress that is also useful for professional area of the project, so I call that a win-win situation.
Analysts looked through the game concept and mechanics and created “focal points” that determine the efficiency of a player in one of 8 select areas. But more on that later.
How Does The System Work?
The model applies 8 different criteria “blocks” with variable values to determine the ranking of a player.
A ranking of 1 is a “basic” level of an average player in competitive Sins vanilla multiplayer.
The criteria “blocks” are as follows:
Microcontrol Skill
Tactical Proficiency
Tactical Versatility
Strategy and Investment
Gaming Influence
Flair and Anticipation
Rush Efficiency
Game Knowledge
Conceptually, the closest in-game ranking system that I know of is applied in Football Manager series to evaluate a player. However, all the elements used in this Sins ranking are original and are adapted for 5v5 games. Efficiency of playing in all three spots: agro, eco and suicide is considered, with the latter given lower value due to higher importance of having a bigger fleet, especially early on.
The model considers purely gaming skills “on average”. It doesn’t take into account things like reliability, e.g. likelihood of rage-quitting, disconnecting/minidumping, etc. Racial variety is not included either: it doesn’t matter if you play with the same efficiency with all six factions or just one faction, like X-Wing or Krusher do.
What Do the Ratings Mean?
As I stated above, the ranking of 1 is a hypothetical average Sins multiplayer participant. In general, the average rating of most skilled lobbies on Saturday night CET – the busiest time for 5v5 games – is about 1.72.
Basically, the rankings mean how many those average hypothetical players is one player worth. It doesn’t mean that a player with 2.1 will quickly destroy a player with 1.5 point ranking in an average agro spot; however, the higher-ranked player is much more likely to have in-game advantage. Playstyles and different set of skills also play a part.
Author’s view on the rankings:
Anything higher than 3 highlights an exceptional player. Such players are comparable to sports superstarts like Messi, Ronaldo, Fourcade, Federer, Phelps, etc.
Rating from 2.5 to 3 mean the highest-quality player who are good enough to dominate the game and are vastly superior to most skilled players.
Rating from 2 to 2.5 mean what most people call a “pro”.
Rating from 1.5 to 2 is very good and is common among third-fifth pick.
Ratings from 1 to 1.5 may be competitive in MP games, but encounters with anyone higher than 2.5 are extremely difficult for them and players must rely on feed or superior spot to be competitive.
The Current Ranking
As of 24th Feb 2020, 38 players have been fully analyzed.
If you want to be added to the list, please inform the guide author. It might take some time to process the request as analysts still have to do their part, even though the algorithm is ready.
Auqia 3,5126
zabik2 3,1078
Matheus 3,0022
Trolik 2,7729
Radioactive 2,3257
Hummer 2,3172
Matzestinkt 2,2756
Krusher 2,1869
BrutalSlayer 2,1426
Cattle Bruiser 2,1410
BigW 2,1004
sbNoobnoob 2,0988
X-Wing 2,0973
HIGGY 2,0557
Coriolanus 2,0058
suzhijiang 1,9384
Hhhhuuuie 1,9289
Rabadapp 1,8651
whirly 1,7316
sb=>ferdie 1,7307
JohnMatrix 1,6186
TheSaint 1,5123
Fantomex 1,4785
MJBriley 1,3909
UNUM 1,3461
Ino 1,3457
WhipperSnapper 1,3139
owll 1,2521
mathematico 1,2198
RECO 1,2137
ShoNuff 1,1624
Sleeper 0,9961
Snowpocalypse 0,7546
Spockimpossible 0,7115
THC Commander 0,5188
KaiKai 0,5178
44447 0,2256
F.A.Q.
Q: May I analyze the algorithm in more detail? See the “inside-out” of it?
A: As of this moment, sadly, no. The project is in its alpha stage and most elements are a commercial secret and know-how. This will likely change in late 2021, when we plan to release the original beta-system to select customers in the market. But I can’t guarantee anything yet.
Q: Is the algorithm reliable?
A: It is simply a mathematical understanding of how good a select player might be in general and how well he might perform in an average game. To me, as a 5v5 regular, these rankings generally seem lucid.
Q: May I help improve the ranking?
A: Uploading your replays to a cloud and sending me the link will greatly help, especially if you are not on the list.
Q: May I understand what are my strong and weak sides according to the algoritm?
A: Possibly. PM me and I will see what I can get from the data.
Progress Log
12 Jan 2020: Proejct initated
16 Feb 2020: First version 1.0.0. launched
18 Feb 2020: 10 more players added
24 Feb 2020: small updates and 3 more players added