DUST Search - Replace MU matchmaking system with something else

Nothing Certain wrote:

ReGnYuM wrote:

CCP Rattati wrote:

The-Errorist wrote:

For better matchmaking, a personal skill level stat should be used to match player of similar skill levels. Win/Loss ratio shouldn't heavily impact that number and needs to incorporate more than just WP/s and KDR.

Here's what came up with a long time ago and the spreadsheet.

Here's a summery the formula:
(Warpoints/Kills/30)^(3/4) + (Kills/Deaths)^(1/2) + (Warpoints/150)^(9/10) (1-(1.5*Deaths*Price/10)^(8/10) + 1.5*sqrt(Wins/Losses) = Personal Effectiveness Stat

I have demonstrated before that Mu is a very good indicator of player skill, by comparing it to KDR, WP/D and a multitude of other more complex stats I experimented with.

WP/D is my favorite simple method btw.

Here's my problem, I usually solo ambush yet I consistently get but up against organized squads. I just don't understand how my solo MU is being matched up and put up against consistent Proto squads?

Because you are a good or reasonably good player. Team A gets the squad, which with the multiplier gives them the 6 highest Mu. Team A then gets populated by the lowest Mu players, you don't make the cut, you have to get worse.p

Pages: 1 [2] :: one page

Author	Thread Statistics \| Show CCP posts - 3 post(s)
ReGnYuM Kameira Lodge Amarr Empire 3558	Posted - 2015.03.07 23:39:00 - [31] - Quote Nothing Certain wrote: ReGnYuM wrote: CCP Rattati wrote: The-Errorist wrote: For better matchmaking, a personal skill level stat should be used to match player of similar skill levels. Win/Loss ratio shouldn't heavily impact that number and needs to incorporate more than just WP/s and KDR. Here's what came up with a long time ago and the spreadsheet. Here's a summery the formula: (Warpoints/Kills/30)^(3/4) + (Kills/Deaths)^(1/2) + (Warpoints/150)^(9/10) (1-(1.5DeathsPrice/10)^(8/10) + 1.5*sqrt(Wins/Losses) = Personal Effectiveness Stat I have demonstrated before that Mu is a very good indicator of player skill, by comparing it to KDR, WP/D and a multitude of other more complex stats I experimented with. WP/D is my favorite simple method btw. Here's my problem, I usually solo ambush yet I consistently get but up against organized squads. I just don't understand how my solo MU is being matched up and put up against consistent Proto squads? Because you are a good or reasonably good player. Team A gets the squad, which with the multiplier gives them the 6 highest Mu. Team A then gets populated by the lowest Mu players, you don't make the cut, you have to get worse.p Here lies the problem, my experience is not that of a balanced system. There honestly needs to be stronger consideration for solo players paired up against squads.
Nothing Certain Bioshock Rejects 1569	Posted - 2015.03.08 00:14:00 - [32] - Quote ReGnYuM wrote: Nothing Certain wrote: ReGnYuM wrote: CCP Rattati wrote: [quote=The-Errorist]For better matchmaking, a personal skill level stat should be used to match I have demonstrated before that Mu is a very good indicator of player skill, by comparing it to KDR, WP/D and a multitude of other more complex stats I experimented with. WP/D is my favorite simple method btw. Here's my problem, I usually solo ambush yet I consistently get but up against organized squads. I just don't understand how my solo MU is being matched up and put up against consistent Proto squads? Because you are a good or reasonably good player. Team A gets the squad, which with the multiplier gives them the 6 highest Mu. Team A then gets populated by the lowest Mu players, you don't make the cut, you have to get worse.p Here lies the problem, my experience is not that of a balanced system. There honestly needs to be stronger consideration for solo players paired up against squads. Furthermore for the record, I am not some random noob (sorry new bros) I am 70million + SP veteran , and I would like to believe I have a firm grasp of the role and gunplay of Dust 514. Right, we can't have both balanced teams AND squads, you just can't make it happen. We can separate based on metalevels or Mu or some other similar factor and this can possibly fix the problem. There won't be the discrepancy in ability just the synergy of squadding to account for. Hmmmmm, ReGnYum, seems familiar....nope never heard of you. You sure you play the game? Because, that's why.
Kallas Hallytyr Skullbreakers 1378	Posted - 2015.03.08 00:49:00 - [33] - Quote CCP Rattati wrote: Pretty much this, also, the teambuilder incorporates a squad multiplier. Is this based on squad size at all? A squad of 6 mid-level mercs will beat the tar out of 2 high-level mercs: 6 high-level mercs are worth more by far than two squads of 4 mid-level mercs. Frankly, I've been put into a bunch of battles recently with one big squad that seem pretty useless vs two big squads of proto players (presumably high Mu) and my squad, usually only two or three strong, is just not a match. Simply, multiple noncommunicative small squads are not a match for one large, fully communicating squad. Alt of Halla Murr. Sentinel.

CCP Rattati C C P C C P Alliance 18216	Posted - 2015.03.08 12:42:00 - [34] - Quote There are at least two blogs/threads about the exact way the teambuilder works. A match with a single 6 man veteran squad, will be as balanced as possible by having the best players place consecutively on team B until the the sum of MU on Team B exceeds the sum of MU of the squad on Team A "times" the multiplier. Usually that's about 8-10 of the best players pitched against the 6 man team. Given the hand that is dealt the teambuilder, that is literally the best it can do. It is the unequivocal best way. And please stop claiming two squads are pitted against individual randoms, it just doesn't happen (or at least has not been proven to do so by countless examples from players). "As well as stupid, Rattati is incredibly slow and accident-prone, and cannot even swim"

Kallas Hallytyr Skullbreakers 1380	Posted - 2015.03.08 12:56:00 - [35] - Quote Well, I wasn't claiming that it put multiple squads vs randoms, but it seemed, anecdotally, that it was putting multiple small squads against large, high Mu squads which is not really sufficient to counteract the high skill and high communication. The Mu numbers might be lined up as closely as possible, but I was really asking if the squad size was more than just a single modifier, because a large squad is disproportionately more effective than a small squad. So solo is x1 Small squad is x2 Large squad is x2 Or does it account for the larger squad sizes? Solos is x1 Small squad (2 players) is x1.2 Medium squad (3-4) is x1.6 Large squad (5-6) is x2.5 Alt of Halla Murr. Sentinel.
Himiko Kuronaga Fatal Absolution Negative-Feedback 5216	Posted - 2015.03.08 18:44:00 - [36] - Quote CCP Rattati wrote: There are at least two blogs/threads about the exact way the teambuilder works. A match with a single 6 man veteran squad, will be as balanced as possible by having the best players place consecutively on team B until the the sum of MU on Team B exceeds the sum of MU of the squad on Team A "times" the multiplier. Usually that's about 8-10 of the best players pitched against the 6 man team. Given the hand that is dealt the teambuilder, that is literally the best it can do. It is the unequivocal best way. And please stop claiming two squads are pitted against individual randoms, it just doesn't happen (or at least has not been proven to do so by countless examples from players). I still think you need to simply deploy fights at every 5 minute mark. 1:00, 1:05, 1:10, 1:15, etc. It seems like the team builder is doing its job just fine but the matchmaker doesn't honestly have enough quality players to really make the fights interesting. There are enough quality players in the system, but most are already in other fights. Spreading the fights out gives them a little more time in line and lets the game put the good players with the good players and the bad ones with the bad ones. Far fewer compromises. If you visually represent the time to deployment as a countdown, you also won't get the stigma of players feeling like it takes forever to get into a fight. Usually banned for being too awesome.
Aeon Amadi Chimera Core 8937	Posted - 2015.03.09 00:46:00 - [37] - Quote CCP Rattati wrote: There are at least two blogs/threads about the exact way the teambuilder works. A match with a single 6 man veteran squad, will be as balanced as possible by having the best players place consecutively on team B until the the sum of MU on Team B exceeds the sum of MU of the squad on Team A "times" the multiplier. Usually that's about 8-10 of the best players pitched against the 6 man team. Given the hand that is dealt the teambuilder, that is literally the best it can do. It is the unequivocal best way. And please stop claiming two squads are pitted against individual randoms, it just doesn't happen (or at least has not been proven to do so by countless examples from players). What's the Win/Loss Ratio of a single 6 man squad versus a team of individual players? Have a suggestion for the Planetary Services Department? *Founder of AIV*


Pages: 1 [2] :: one page
First page \| Previous page \| Next page \| Last page