{{
`${userDetail.user.displayName}${userDetail.user.cachedBungieGlobalDisplayNameCode ? `#${standardizeBungieGlobalCode(userDetail.user.cachedBungieGlobalDisplayNameCode)}` : ""}`}}
{{getUserClans()}}
4/19/2016 2:00:30 PM
2
W/L is a horrible way to base someone's performance.
When you have someone like me, that almost ALWAYS runs solo in Crucible, it makes that stat extremely skewed. I have no control over my teammates, or if I get thrown into the middle of a match. It doesn't matter how good your are, it's hard to carry a 6 man team, sometimes even a 3 man team. Unless the people you're against are just ass.
IMO, it's better to take ELO, amount of time in the Crucible, and specific game variant k/d's. I.e. I wouldn't think someone with a high k/d in 6v6 variants could excel easily in 3v3 variants because the situations are typically approached differently compared to playing on semi small maps with 6 opponents.
English
-
[quote]W/L is a horrible way to base someone's performance. ... IMO, it's better to take ELO, amount of time in the Crucible, and specific game variant k/d's. [/quote] ELO is just a [i]weighted[/i] win/loss statistic.
-
You have some valid points. However, I consider it far more valuable to be able to adapt and support blueberry players than push individual skill or in some cases cherry-pick teammates kills as a solo in crucible. Is it perfect? No. But no statistical representation of skill is.
-
Re: w/l as skill measurement: [quote]Is it perfect? No. But no statistical representation of skill is.[/quote] Looking at win/loss as a measure of ability in a team game is definitely imperfect. As you said, there are many roles one can play which will contribute to team effectiveness in different ways. A pitcher on an American League Baseball team is often erroneously judged on his win/loss record but he doesn't contribute to the team's offensive score at all. Having looked at alternative skill ratings, I like what Halo 5 is doing with their [b]KDA[/b] stat. It's created with this formula: [b][kills + (assists/3) - deaths]/games played[/b] This creates a number which illustrates how much positive scoring you contribute in an average game. The reason this is better than a straight [i]kill:death ratio[/i], is it differentiates between two players with the following scores: [i]Player 1: 2 kills, 1 death. Player 2: 20 kills, 10 deaths.[/i] Additionally it considers [i]assists[/i], and anyone that understands team slayer/clash understands that assists can be hugely influential in which team wins an encounter let alone who gets the most points on the scoreboard.
-
The thing is W/L would be a great representation of skill but with bungie's SBMM it can't be. I am always getting match against teams of 6 and the teammates I have on my team only get 1 to 5 kills in a whole game. W/L in trials can be a good representation of skill though , because it's your to3 vs the other to3. Right now KD is the best way to judge skill because it shows what the player can do individually.
-
"Combat rating" is probably the best way to measure skill currently.
