Matchmaking, is it working as intended? I think it needs a bigger brain.
First: This is by no means me saying, "in EVERY case, my team wins when we have a larger kill participation %" or, "I know that answer to the universe!!!" (also, I realize that there is no "BEST" way to match people in any game, period, I do get that). -Of note: I don't have 1000 games played, to give a truly accurate representation, please overlook this for the purpose of short-term evaluation. -This game in question is one of several games I've had, (in season 7) where I got players on my team, of equivalent silver rating...but the other team got prior Gold tier players who're currently digging their way out from under this season's reset. I bring this up to poke at my main point: I'd like to see an overhaul/redo of the current evaluation system that is in place.
The game I'm using as a basis for evaluation: http://matchhistory.na.leagueoflegends.com/en/#match-details/NA1/2403229262/39343699?tab=overview My team: http://na.op.gg/summoner/userName=kakkarotlol http://na.op.gg/summoner/userName=jeffworthy http://na.op.gg/summoner/userName=epicak47 http://na.op.gg/summoner/userName=mismagius http://na.op.gg/summoner/userName=pyroteknick The other guys: http://na.op.gg/summoner/userName=chunkiestsalsa http://na.op.gg/summoner/userName=siensavanh http://na.op.gg/summoner/userName=kotsuga http://na.op.gg/summoner/userName=travaran http://na.op.gg/summoner/userName=joshyspiffy
I realize, that there are literally, an unending list of complaints, critiques, etc....but, please hear me out. With some consistency, teams I've been on that have higher (overall) % kill participation, tend towards winning. The attitudes trend towards positivity, and the team functions better because...they're all looking out for each other, and as a bonus, they're more aware of the map.
A breakdown of the above: My team: Other team: kakkarotlol - 36% kill participation (kp) chunkiestsalsa - 43% kp (S6 Gold) jeffworthy - 38% kp siensavanh - 48% kp epicak47 - 39% kp kotsuga - 44% kp (S6 Gold) mismagius - 53% kp (S6 Gold) travaran - 49% kp (S6 Gold) pyroteknick - 49% kp joshyspiffy - 57% kp (unranked)
I will say again, very clearly, this doesn't hold up in EVERY case, and because I have so few games played, I can't give you an avg number of times it holds up; However, I can say that with the current season favoring bruisers so heavily (and with two dmg items + 3 tank items benefiting them in a more impactful way than say...adcs with 6 items), and other changes that have been made, it is vastly more difficult for a single player to overcome a deficit in another lane and in turn, more difficult to climb brackets. This change has me craving a revision/update/re-investigation of the current system that performs player elo analysis.
Looking at the http://na.op.gg/ranking/ladder/ and cherry picking some names I recognized: Doublelift: 56% kp Pobelter: 52% kp C9 Sneaky: 51% kp Cody Sun: 46% Shiphtur: 58% kp C9 Ray: 51% kp Dardoch1: 58% kp C9 Meteos: 51% kp
You'll see that a bunch of trash silvers (read: PyrotekNick) are within 10-20% of the pros...so how does this stat even help at all? Up front...it doesn't. What I am working towards with this over-written wall of text with numbers mixed in, is that I'd like to see Riot not just update season-to-season, but to upgrade their rating evaluation calculations to include variables for a player's preferred position as it relates to: cs/kill participation/kda/wards/objectives and maybe other variables. Why? because I love this game. A LOT.
Implementation would most likely be difficult, and would require a great deal of planning in itself. I believe the best way to implement these kinds of changes, is in stages: -Evaluate players based on the current system (I don't know what statistics are utilized, but I'm pretending it's just Win/Loss ratio) First season of changes: -Evaluate players on an additional statistic (we'll say KDA), create a baseline from previous seasons' data for each tier's brackets, weight that value based on it's usefulness as compared to the current system (I'll theoretically say that KDA would be rated very low compared to win/loss ratio, say 5% value) Second season of changes: -Evaluate players on an additional statistic (we'll say Kill Participation %), create a separate baseline from previous seasons' data for each tier's brackets, weight that value based on it's usefulness as compared to the current system (I'll say that KP% would be rated at 20%) Third season of changes: -Evaluate another statistic (we'll say cs/min), create a separate baseline from previous seasons' data for each tier's brackets, weight that value based on it's usefulness as compared to the current system (I'll say that cs/min would be rated at 25%) -Evaluate another statistic...and so on.
We take the base rating system, assign it's value, we'll say 50%. Each evaluation will need to generate a player rating of some kind. Evaluation output could be: Current system, W/L: 50% KDA evaluation: 5% KP% evaluation: 15% cs/min evaluation: 30%
The sum of the parts, with each outputting a value, multiply it's value by it's usefulness percentage and add the values together (less accurate method, but simpler, I think). Or: Each evaluation is compared to a base-line for players in a given role, in their current bracket and tier (i.e., bronze 1,2,3,4,5, silver 1,2,3,4,5, gold 1,2,3,4,5, diamond 1,2,3,4,5 are separated as 20 individual 'zones'). Each individual evaluation would give a score. Regions of deviation would need to be created, and based on deviation (positive or negative) from the base-line, players would be scored based on what regions they fell into in the different categories. The sum of these individual scores would then become a given players' elo.
I've been playing for several years. I enjoy most aspects of the game. I've had the joy of winning two games (one of them ranked, and this season). where someone on my team afk'd, and we pulled a win from the asscrack of the LoL Gods themselves. I also like the handful of matches where I did get a very tough (but clearly even, skill-wise) matchup, and due to my own (or a teammates') tactical mistake, lost the game. I appreciate the dynamic of personal skill, team skill and tactical awareness that this game brings together.
TLDR: I believe drawbacks that impair a players' ability to experience this game in a reasonable skill match-up tend to ruin the experience. For the 'over-skilled' player, they have no competition, and the game won't appeal to their desire for a challenge. For the 'under-skilled' player, they didn't have a chance in the first place. They were pitted against a foe that was clearly their superior. On both sides of the match-making system, it's crucial that a level playing field is created. The system used will always improve, and I suspect that it will always fall short as the meta changes from season to season. However, I also believe that investing in a system that can evaluate more variables in the pursuit of a fair match-up would be beneficial to the long-term health of the League community.
