Kshahdoo, on 01 January 2017 - 03:48 AM, said:
Of course I mean PUGs, because PUGs play for a personal statistics and units play for wins, but 80% of games (probably more) are played in PUGs, so they do about about 4 times better presentation of the game, than organized matches.
If you play with PUGs and think people are here for the same reason as you, i.e. win the game, it's a big mistake. Wining means nothing, if you have all the mechs you need and don't need XP and CBs anymore. And thinking that w/l is that meaningful, when you play mostly with PUGs, is about as right as thinking people become politicians because they love their countries. But of course the better player the better chance he will help his teams to win more matches, so his w/l can be pretty good as well.
I already explained why WLR is better indication of performance than KDR, in solo-q in another thread. Pugs will be pugs, and do silly things, but in the end people prefer to win, than lose.
El Bandito, on 01 January 2017 - 06:01 AM, said:
On the contrary, high KDR can be easily gotten through devious means, such as intentionally kill stealing low HP targets by holding back your alpha until the right time, or intentionally hiding in a losing match until the enemy kills the VIP/caps the base/captures 750 points, or intentionally playing it safe until everyone is close to death then take the kills.
Only WLR is far more reliable indication of skill in the Solo-Q, because it requires real participation from you to have positive WLR on average, over hundreds of games.
El Bandito, on 01 January 2017 - 06:19 AM, said:
In small samples, yes, WLR can be skewed. Consistency is key. Over hundreds, and thousands of matches, the player's WLR in solo-q can be pretty accurate indication of his competence.
Appogee, on 01 January 2017 - 04:03 AM, said:
We really need to see KMD stats. Between those and Kill stats, you'd get a more accurate picture of a players' performance.
Kill/Death isn't enough by itself, because some pathetic souls deliberately inflate their stats by kill-stealing.
Win/Loss isn't enough, because it's too much at the mercy of PUGlotto, and group carrying.
I think Match Score is probably the best proxy we have at the moment, even though it too can be artificially inflated by LRM usage (much superfluous damage), consumables usage (such as UAV spotting for the LRMers, strikes) and other weird oddities which may not necessarily indicate strong personal performance (eg. Lance in Formation, Hit and Run bonuses etc).
KMDD is not enough, since Assault and Heavy pilots will be too biased compared to dedicated Medium and Light pilots.
Edited by El Bandito, 01 January 2017 - 07:08 AM.