172 replies to this topic

[ElLocoMarko]

Sifright, on 02 December 2012 - 01:16 PM, said:

I'd just like to point out if you drop in a match with only one premade group of four players, you have double the chance of being against the premade than with them

4 slots open on premade side vs 8 on the other.

I was gonna say this! And make graphs... But not now. My guts says this aspect of random matchmaking peaks around 25% players in premade 4's.

I'm not even sure what it is about the premade 4's that equals a win. The few times that I have teamed up we don't even always stick together because of speed differences. But we still win 4 out of 5. My theory - I'm good and one or more of my teammates are better. It's about compounded pilot skill. And the next generation of matchmaker should attempt to balance out that kind of skills gap.

[Malora Sidewinder]

not going to lie, these poll results have pretty much shocked me.

[Bishop Steiner]

Malora Sidewinder, on 29 January 2013 - 04:08 PM, said:

not going to lie, these poll results have pretty much shocked me.

I have found them rather interesting myself, hence the question.. Thank you for taking them for what they are, rather than childishly assuming you know the motivation behind it as Thirdstar likes to do.

[MischiefSC]

So, assumption would be that a poll with self-selected results on the forum would skew disproportionately to people who, of course, post on forums. Assumption would be that these people would skew towards using teamspeak, grouping and the like being more time/energy invested in the game. All that however is assumption.

Even with that though, statistically the gap between pug and premade time it implies is surprising.

If you break it out on a 1-1, .5/.5, .25/.75 basis it comes out to about 61% of the time pugging and 39% premade - even on a forum poll.

Above metric derived from -

Premade only/Pug only each contribute 1 vote towards their respective 'camp'.

50/50 splits their votes between each 'camp'.

Both, but more premade/more pug contributes 0.75 towards whatever their 'more' metric is and 0.25 to their opposite.

Newbies, people who have no interest in the forums just playing the game and people who don't invest in forum activity are totally unaccounted for but I don't feel it's unreasonable to assume they would skew sharply towards 'pug more'/'pug only'.

I still say however that the most significant factor is skill of the people on your team, not having a premade team to drop with. Elo FTW.

[Bishop Steiner]

MischiefSC, on 29 January 2013 - 04:29 PM, said:

So, assumption would be that a poll with self-selected results on the forum would skew disproportionately to people who, of course, post on forums. Assumption would be that these people would skew towards using teamspeak, grouping and the like being more time/energy invested in the game. All that however is assumption.

Even with that though, statistically the gap between pug and premade time it implies is surprising.

If you break it out on a 1-1, .5/.5, .25/.75 basis it comes out to about 61% of the time pugging and 39% premade - even on a forum poll.

Above metric derived from -

Premade only/Pug only each contribute 1 vote towards their respective 'camp'.

50/50 splits their votes between each 'camp'.

Both, but more premade/more pug contributes 0.75 towards whatever their 'more' metric is and 0.25 to their opposite.

Newbies, people who have no interest in the forums just playing the game and people who don't invest in forum activity are totally unaccounted for but I don't feel it's unreasonable to assume they would skew sharply towards 'pug more'/'pug only'.

I still say however that the most significant factor is skill of the people on your team, not having a premade team to drop with. Elo FTW.

show me a poll someone doesn't scream "skew or bias" in and I will die a shocked man. Since I covered all the reasonable angles I could think of at the time (editing a poll to more option after would actually bias it more, after all), I would like to know what is so leading or skewed in the options. Premade only? Pug Only? Both but one more than the other(each way)? Yes. Very leading and biased.

[MischiefSC]

The poll is actually excellent in terms of questions, the problem is a self-selecting pool of participants with no control. No way to know the makeup of the people who participate or get a truly 'random' sample.

The questions are fine though, there's no bias - other than the inherent limitations of a poll on a forum based on how much the participants use a communication tool. The people least likely to use a communication tool are less likely to be on the forums and participating in polls.

[Taizan]

MischiefSC, on 29 January 2013 - 05:16 PM, said:

The people least likely to use a communication tool are less likely to be on the forums and participating in polls.

This. Usually a forum community represents around 6-10% of the whole user base.

[Bishop Steiner]

Taizan, on 29 January 2013 - 05:53 PM, said:

This. Usually a forum community represents around 6-10% of the whole user base.

This is likely true, but I can only pool data from those willing to participate. Dare I say (and I probably sound like an elitist) if a person doesn't care enough to participate in the forums, then they really have no right to complain? (Whereas some folks on here seem to think that Forums exist purely for complaining? oy vey!)

Sadly, I can only compile what I can compile. But better an incomplete view, I feel, than none at all.

[p4r4g0n]

It was a pretty good poll and the ensuing discussion was pretty decent too. Wasn't too surprised at the results although I found the inference that premades comprise approximately 25% of playing population at any one time interesting.

ElLocoMarko, on 29 January 2013 - 04:02 PM, said:

I'm not even sure what it is about the premade 4's that equals a win. The few times that I have teamed up we don't even always stick together because of speed differences. But we still win 4 out of 5.

Yeah, noticed this commonality when casual grouping with no VOIP even. Personally, just being in a group makes me communicate more via text as I know that someone in the team will actually care. Hell, even in all solo teams, more chat usually equals win or at least a better battle.

[MischiefSC]

The issue isn't voice chat or grouped fire. It's stacking your odds - with a 4 man drop you ensure that at least 3 other people on your team know how to play and are willing to play together. That's 50% better odds than the other team gets. It reduces the number of AFKs, disconnects, newbies and suicidal TKing twits by 50% as well.

That's pretty significant.

[Taizan]

Bishop Steiner, on 29 January 2013 - 05:56 PM, said:

Sadly, I can only compile what I can compile. But better an incomplete view, I feel, than none at all.

Sure, that's what the poll function is for. Just saying, it's very vague (unfortunately) - personally I'd like to get much more data / statistics on the game - even if it's just in percentages.

[Cargo Bane]

I only logged on just to vote in this poll. 1st time in a while.

I only pug, and the reason it sucks is because of the lack of features that have been around since CB. I'm sitting on all my PremTime, MC, and Credits until the day it's better. I don't even play anymore expect a few games when a patch comes out.

Edited by Cargo Bane, 29 January 2013 - 07:21 PM.

[Mechwarrior Buddah]

Taizan, on 29 January 2013 - 06:35 PM, said:

Sure, that's what the poll function is for. Just saying, it's very vague (unfortunately) - personally I'd like to get much more data / statistics on the game - even if it's just in percentages.

Theres a group of ppl that could give that to us - PGI.