View Single Post
Lt. Commander
Join Date: Dec 2007
Posts: 120
# 25
11-30-2011, 06:43 AM
Quote:
Originally Posted by Roach View Post
The point of my post was to put forth the question, " what do we measure to determine player skill and placement in a ranking system?"

In the end if it is just wins that determine a players rank and by defualt analysis those whom run premade teams will rise to the top and the casual player will not even come close to the same rank due to PuG inconsistancies, it may be better to just submit the names of those players whom are known to be at or near the top for thier own Ques as the number of mid level and lower level PvP'ers I think is greater in number.
I think there are many premades out there, but they are not all equally good. A match-making system can help identify PuGs that have enough potential to beat one of the weaker (or even one of the stronger) premades.

Of course, every players wish may be to have a top rating, but not being in a Premade is a good sign you never can, and I don't really see a way to change this.

Quote:
What will happen is that those whom have shown themselves superior will play as often as their group allows while the PuG'ers will still be repleate with cries of "unfun" and "no fair" as those mid level not-quite-so-profficient PvP'ers will rise to the top in PuG and the cries of "seperate PuG and Premade" will continue.
I am not sure I can follow you here. I think the main issue people have is that they fight the same Premade again and again and don't see a way to win. That's why they want the Premades gone. If they don't keep fighting a Premade because the premade is just rated so much higher, their concern is addressed. If they keep fighting it, they are either better than they think (and remember, every time you lose, your opponents rating rises and your own does not or get worse, so if it is requally an unequal match, you should stop seeing them.), or there are just you and the Premade in the queues and seperating the queues would lead to neither of you playing.

Quote:
At the same time you will haev those whom flub thier own records so they can remain the Big fish in the small pond, so to speak.
But can he? How can you be a "Big Fish in the small pond" if you have to intentionally lose matches? I don't think this is really something people can pull off well.

Quote:
I think the system may need to be more than just who wins, number of won matches, etc.
A good system may be more complex than just a single number. But if it's just one single number with no mathematical tricks, it has to be match wins. No other single value is more representative.

A better system with more numbers taking into account should probably consider #matches and the ratio of win/losses.
Say, people that have played 100 matches or less are in their own "league" compared to people that played 1000 matches. It may also require to cut off after a certain amount of time. It makes sense to me to track win/losses for your entire career, but actually accounting for match-making should probably be limited to something like the last 100 matches or the last 30 days.


Relevant Matches for Win/Loss Ratio: Last 100 matches or all matches in the last 30 days, whichever number is higher.
RATIO Rating: 10 Wins to 0 Losses = 10; 9 to 1 = 9; [...] 0 to 10 = 0;

Rating Adjustment based on experience: +1 per 100 matches played.
Team Rating is: Sum of all Adjusted RATIOs across all team members.

Now we need a match-making algorith that actually creates sensible matches without locking people out of play forever because no match can be found, or ignoring ratings because it only looks for wait time. Maybe it's best to classify players into categories based on how long they are queued when considering teams.

*ponder*