Wood Tier & Variance: A Statistical Consideration
Many of you are aware of the classic Bell curve.
http://i.imgur.com/01htpR6.gif
The idea is fairly simple, but the 'average' person sits in the middle, with extremes moving out toward the edges. Now, I'm not suggesting that we have Master and Challenger equivalents (because it's not healthy and nobody--except trolls--really want to be last). Before, I look at the numbers, then, it's clear that we currently need at least one tier below Bronze. This would allow a breakdown as follows:
Wood, Bronze, Silver-[Average Player]-Gold, Plat, Diamond ~2%, ~14%, ~34%-[no one is actually average]-~34%, ~14%, ~2%
Now, people could argue that with millions of players, even 2% is too much for Diamond and too few for newbies, maybe. Maybe you skew these numbers a little to something like
~5%, 15%, 40% [false avg] 30%, 9%, 1%
But what does League really look like? #This:
http://na.op.gg/statistics/tier/ Wood, Bronze, Silver || Gold, Plat, Diamond [0%], 36%, 38% || 17%, 7%, 2%
This means the ACTUAL average player is somewhere in low silver? Really?
Here's the point. The simple reality is that there's no 'safe' way for new ranked players to properly queue together. The learning curve from "true" Bronze V (that is a new or inexperience LoL player) to the average player [should be between silver/gold] is VAST. And there's no stomping ground to learn. I truly believe that much of the extreme variance we see in 'elo hell' is due to the hodgepodge of player types and abilities. We need a system that collates data more intelligently and much more quickly, I've argued that all throughout the boards. But baring that, we can solve some low hanging fruit, and this is one.
Now, one complaint could be that people who don't play often are just collecting dust in Bronze (well, first off, this is a new season, so there hasn't been much time for dust). But, maybe. Having this lower tier still allows a position for the 'dusty' and one that SHOULD be relatively easy to climb out of with a little practice, since it will be filled with only the new or very inexperienced. This provides an opportunity for reward as one learns more about this game and soon CLIMBS to Bronze.
And here's the very important rub.
If Riot cannot create a matchmaking algorithm and system that results in a curve like the one above, then this is clear evidence of an issue with MMR and matchmaking in general. Now, for the more astute masses, I realize that it's likely better that the distribution of players follows something akin to a Paretian-Bell hybrid. And, that's true, which identifies a need for Master and Challenger tiers. However, we can still do better than we are.
Thus, here's a vote for a little Wood... tier. [slayer-jinx-wink]
#EDIT: Here's a strawpoll since I forgot one. :) http://strawpoll.me/7235809
