SpeedingBus, on 08 November 2014 - 08:56 AM, said:
The whole fact PGI can't do anything to matchmaker is due low population its why you see veteran clans playing against fresh noobs.... there's not enough players to really separate players by elo
And yet in the solo queue the matching in general works just fine, and the Elo variances are quite tight (50-100 was IIRC the range of deviation last time Russ said something "hard".
IME I`d say that it`s primarily an issue the group queue, where some groups have Elo scores between 2500 and 0 in teh same group that have to be matched with equivalent players if possible, combined with the fact that CW hasn`t fully kicked off yet (it`s a known fact that there are somewhere between dozens and thousands of players that quit at some point, stating they WILL be back when and if real CW starts, something I personally hope for because then Zerb`s Claws will once again be a complete entity and not a lone placeholder.), further agitated by many of the smaller groups now splitting into solo drops because the matchmaking id subjectively better in solo. Especially those with less skilled players that complain about getting matched up against groups of 12 eery drop, which is of course hyperbole, but a grain of truth exists
Quote
Why were we forced to 12vs12's can't there be 8vs8 as well? Oh wait nope that would split player base making matchmaking even worse.
That is a design decision that may or may not have merit. There is also a significant financial factor that plays into it, the same one which requires Premium time for private matches. A server qith 16/24 players is running at 2/3 capacity, and therefore wasting valuable resources that gould go towards setting up another 12 vs 12 match.
Interpreting that to be entirely because the playerbase is too small is a tenuous connection at best, with no actual data to support it. It may in fact be true, but there is absolutely zero actual proof.
In fact, if the population were in fact "too small", switching back to 8vs8 would almost be the golden bullet to solve all the problems, wouldn`t it? So, if it were in fact the case, there`s no real reason for PGI not to revert back to 8vs8.
Quote
While I may not know the numbers matchmaker clearly indicates something is wrong with this games population size.
Again, solo queue is just fine IME, and teh group queue is thoroughly borked. However, having set up clan wars, LAN parties and all sorts of other events over the years, including random FFAs, I must say one thing: If there is something that has to happen to make something go off without a hitch, the chances of it actually happening are close to zero.
For example, to use MWO matchmaking, a 12 man w 2000 elo average gets matched with a 9 man that also has a 2000 average. The 9 man needs a 3 man /2k avg to be "properly filled" and ahave a "perfect" match.
But all the available groups with 2k averages +/- are 2 man or 4 man +, becasue the players at that level are usually competitive unit members, meaning they`ll generally be dropping with their unit. There ARE numerous 3 man groups currently active, but they`Re all tied up in a match at this time.
How many hours should the game wait for that perfect match? Or should the matchmaker just bite teh bullet and grab teh first 3 man group that hits search so that ANYTHING will happen before it turns into 15 minutes of wasted time because those perfect 3 were nowhere to be found?
Do you think people would rather play a match at all against an iffy opponent, or waste 15 minutes, potentially multiple times an hour?
What I`m trying to say is that you can write the most perfect matchmaking system ever devised to man... but when the people that would fit that perfect criteria just arent there at the exact time when you need them, that perfectiont very quickly falls apart.
But that`s not a "playerbase size" issue, it`s a "goddamn it, why can`t I order a taxi that seats 8 people?" "You can , sir, but they`re currently all booked and the next one wont be free for4 at least 30 minutes" issue..... And if you`ve ever worked as a taxi dispatcher, esp on New Year`s Eve or other major holidays, then you already know that even if ALL of your taxis were 8 seaters, they still wouldn`t be enough, something would still come up and do it`s best to monkey everything up.
I think for the most part the matchmaker has the same issue, it`s not necesarily population size, but rather population constellation. If the average MWO player sucks donkey balls, and there are no top pros left to match with (and there will always be a limited supply of them, because for them to win someone else has to lose), they you will get an average donkey ball sucker.
But that`s not "low player count" that`s low GOOD player count.... judging from teh numbers of packs purchased (the generqally release the numbers after pre-order), that sounds more in line with teh insane number of mechs that cross teh table during preorders..... or there are a whole lot of people with 2-3-4-5-6 accounts and too much money
I see fewer and fewer names repeatedly outside the group queue. Something that leads me to believe that the playerbas is in no way huge, but far from dying and still growing overall, esp. considering I`ve played with and /or against the majority of forum regulars at least once or twice in the last 3 years (many significantly more often).
I think when "real "cw comes out we will probably get a good impression of just how large the playerbase is. We`re almost certainly a far cry from teh millions of players that WoT or WoW cater to, but I do think that a healthy half million or more
active players is not out of the question. We were up to something like 300K IIRC when teh counter was removed post CB, and sometime last year teh 1 million account milestone was passed, so assumming linear to slightly exponential development we should be somewhere around 1,5 million accounts. Applying the "30% of total are active" Rule of Thumb we used in CS for one of the large industry monoliths, that leaves approx 500k active players.
EDIT: Yikes, that ended up a lot longer than expected, but it`s a bit hard to add a proper TL;DR to it other than: you may be right, you may be wrong, the future will tell the truth
Edited by Zerberus, 08 November 2014 - 09:45 AM.
5
