66 replies to this topic

[Roadkill]

Mirumoto Izanami, on 09 March 2015 - 09:41 AM, said:

Assumptions.

There are likely to be as many 1 and done losers as there are to be one and done winners.

Not at all true. People who play the game and win are far more likely to keep playing and become actual players.

People who play the game a few times, lose 0-12, and get berated for being a n00b in every game are far more likely to simply quit and find some other game to play.

[Duke Nedo]

There is must be some degree of feeding, but only at low Elo. Difficult to say if that has any major impact at average Elo and above... ? Some effect probably, but I doubt its very much.

I wish pgi would publish our Elos already so we don't need to speculate... And Elo change after each match

[Jman5]

Murzao, on 09 March 2015 - 09:34 AM, said:

When I join group queue I end up getting paired with/against the who's who of comp teams...in which a 1wl ratio or anything close is actually impressive most people would get smashed in 30 seconds repeatedly (especially with a small group vs a larger premade group you have to work doubletime to carry). Then I jump in the top end pub queue and do 2 or 3wl to 'average' it at 1.45.

Why don't you start recording your wins and losses while in the solo queue like OP did and present it as a counter example. I find it difficult to believe that someone could maintain a 2-3 w/l ratio in the solo queue over hundreds of matches. Everyone I know with high w/l ratio are players who mostly stick to group queue.

[Tarogato]

Here's mine for 142 matches so far.
91 matches in solo queue, 51 matches in group queue.




Now here's what my Win/Loss ratio has been doing over time... kinda bleak.
It hovers at 2.0 for a while, then stagnated at 1.6... and now it's starting to fall again.



I also have other data, like my match score ranking, 1-to-12, as well as what mechs and variants I played in each match. Once I figure out how to get OpenOffice to display that data in a pleasing manner, I'll combine it all into one chart.

Here's the first chart again with the weight classes seperated out:

Edited by Tarogato, 13 March 2015 - 08:02 AM.

[Water Bear]

Duke Nedo, on 28 January 2015 - 02:11 PM, said:

Edit 2: Thought I'd update the graph with another 250 matches. A bit weird, suddenly I was deadlocked at 1.0 w/l ratio. Would be interesting to know if anything was changed in the code in the end of January, the perceived effect is rather clear. I don't believe that this has anything to do with my Elo, I am seeing the same guys and before I started recordning this graph there is a baseline of some 4000 games or so, so I should already be at my Elo whatever that is.

I'll just leave it here:

If you're at your Elo, you should expect 50/50 win loss, no?

[Duke Nedo]

Water Bear, on 13 March 2015 - 08:13 AM, said:

If you're at your Elo, you should expect 50/50 win loss, no?

I would have expected to be at my Elo when I started recording this I guess, had some 5k games before. Anyways, if this was a 1v1 game I would expect a graph to look like that when one reaches ones Elo, but not in a 12v12 game. To me Elo is a number largely made up by noise. Probably good enough to separate rookies from pro's after a while, but within the average Joe range, Elo is probably a rather random number.

Will update the figure in the OP, it did turn up again a bit so the deadlock is over.

[HammerForge]

Water Bear, on 13 March 2015 - 08:13 AM, said:

If you're at your Elo, you should expect 50/50 win loss, no?

Well, yes and no. Your ELO is a number that gives you a rating of how good you are versus others. The MM tries to match teams based on ELO so that you are on relatively equal teams. But you could end up on the team that has less of a chance to win more often based on pure chance of when you go to drop. It is not like the match maker is saying, this person is on a losing streak, put him on the better team so he gets a few wins. ELO was developed for one on one games such as chess, where you aren't affected by teammates and how they play, that way regardless of style, people of like ELO should be about the same level of competence and put forth a good game. In something like this, where your team can have teammates with quite different play styles, once the game begins you may not fit into the team mechanic as well.

For example, equal rated teams where one team is made up of very aggressive players versus a team of very defensive players can still be a good match. However, a team of all defensive players or all aggressive players versus a team made up of half defensive/half aggressive players, even if the ELO of the team that is half and half is better, the team of all one style is more likely to win since they are more apt to work as a team. That is why no matter what PGI does with ELO or MM, there are always going to be one-sided matches. Cause 4 ultra-aggressive people got killed early on a mostly defensive team, or because 5 mechs were being really defensive and got left behind and picked off on a aggressive team, and then the route is on.

I can tell you that roughly half the time you will be on a team favored to win in ELO, the other half where you are expected to lose, with some ties in there.

Edited by HammerForge, 13 March 2015 - 11:25 AM.