Current Population In This Game
#1
Posted 31 May 2015 - 03:52 PM
#2
Posted 31 May 2015 - 03:53 PM
#3
Posted 31 May 2015 - 04:02 PM
Seems extremely low.... 50-100 playing at one moment ?
#4
Posted 31 May 2015 - 04:05 PM
We were at the same elo for like, a month or something. Guess he ranked up, cause I don't see him anymore
#5
Posted 31 May 2015 - 04:07 PM
Mudhutwarrior, on 31 May 2015 - 03:53 PM, said:
In addition, when you're just doing pug drops one after another, you should expect to drop against the same people over and over. After all, you're clearly at the same Elo, your queuing up for matches at the same time, etc.
We know there are at least 5 digits worth of active players, but better numbers are tough to come by. CW has hurt things, though, as it takes hundreds of players out of the public queues at any given (non ceasefire) time. Its pretty easy to work out how many people are actively queued up in CW, after all, with a couple accounts in various factions... And the number you see doesn't include groups/solo players who are between matches and not currently queued (in the mechlab, etc). And CW is a very small portion of the player base.
#6
Posted 31 May 2015 - 04:14 PM
Titannium, on 31 May 2015 - 04:02 PM, said:
Seems extremely low.... 50-100 playing at one moment ?
Keep in mind, too, everyone in CW isn't in your queue. Everyone in a group isn't in your queue. Those people CAN'T be put into a match with you.
There need to be orders of magnitude more players playing at any given time.
Then consider: some players select limited game modes. I often only play conquest, for example.
Want to see totally new groups of people? Wait 5 minutes and drop in a mech from a different weight class. This will give the players you're normally matching with a chance to get into a game, and poof, whole new set of players.
But even with a very large currently online population, its expected behavior to get repeated matches with the same players.
#7
Posted 31 May 2015 - 04:41 PM
When major events for CW can only draw 2-4000 during prime time 2 hours before cease fire, and that is a small fraction of the population.... while regular nights barely crack 1000... this is not good.
These numbers are rough guesses based on random looks at CW queues, the closest we have for hard numbers, as well as other anecdotal evidence based on public unit data.
So that is my guess as how many jellybeans are in the jar.
#8
Posted 31 May 2015 - 04:44 PM
I'd expect the reg queue to be that or slightly higher on a regular weekend given the general dislike for cw around the forums.
Edited by Ralgas, 31 May 2015 - 04:49 PM.
#9
Posted 31 May 2015 - 04:46 PM
Titannium, on 31 May 2015 - 04:02 PM, said:
Seems extremely low.... 50-100 playing at one moment ?
More along the lines of 30,000.
You see the same people because Elo is matching you with people in your bucket. On top of it, 90% of the time, the match you are all in ends. You all get put back in the menu screen at the same time, and most of you will click the "play now" button at the same time. Causing you to drop again in the same pool, at the same time. Which is why you end up seeing the same person in multiple drops.
Funny thing is, you noticed the one or two people that kept popping up, but that's 1 out of 24 players. The other 23 are almost always not identical, you just don't notice, because that's not the strange thing.
#10
Posted 31 May 2015 - 05:13 PM
Wintersdark, on 31 May 2015 - 04:14 PM, said:
^ this. It used to be that MWO didn't really elo balance well (at all). And it didn't put into weight classes. Literally. You could have a whole team of assaults vs. a whole team of scouts in random. Now, if you keep pressing start battle in your hunchback, stay until the end of the game, and start again in your hunchback, there's a decent chance you're going to be with some of the same people.
If you jumped into a jenner or waited 5 mins, as stated, you'd likely not be with the same folks.
#11
Posted 31 May 2015 - 05:23 PM
Kjudoon, on 31 May 2015 - 04:41 PM, said:
When major events for CW can only draw 2-4000 during prime time 2 hours before cease fire, and that is a small fraction of the population.... while regular nights barely crack 1000... this is not good.
These numbers are rough guesses based on random looks at CW queues, the closest we have for hard numbers, as well as other anecdotal evidence based on public unit data.
So that is my guess as how many jellybeans are in the jar.
Bad at math?
Lets consider:
Lets assume there's 10,000 people playing at prime time. Of those, 1000 people are in CW.
9,000 people remain. Of those 9,000, lets assume that one third are grouped, and off in the group queue. The actual percentage doesn't really matter, we'll just take a third for napkin math's sake.
Now you've got 2 queues: 3000 and 6000.
6000 people in the public queue. Assuming an average match of 8 minutes. A solo queue searching time of 1 minute. One more minute of Random Stuff Time - though for most people, they're going to be logged in and not searching for/fighting in a match a hell of a lot more than that. That's saying that of those 6000 people, 8/10ths are in a match right now, 1/10th is actually looking for a match. So that's 600 people actually looking for a match. Now, those 600 people are in 1 of 4 weight classes, so not all are even able to be matched with you when you're searching, or against you. Assume 10/25/40/25 percentages for queues - we can see these numbers, at least. That means there are, at any given time, 60 lights looking for a match over the entire 2800ish Elo range.
We know that 90% of matches have Elo's within a 150 pt range inside a team (smaller team vs. team average) and because we're using prime time, we'll assume this holds. 150pts is roughly 5% of the range, but this is a little sketchy because the Elo distribution is very much a bell curve, not an even spread. But lets assume you're a very average player, and 10% of those light pilots are also fitting into that 5% wedge (this is clearly extremely generous). That's 6 light pilots. You need 6 light pilots for ONE match.
Now try to do this with the 3000 strong Group queue.
And that's from TEN THOUSAND people queueing up and playing concurrently, dropping in match after match with 1 minute average "between match" time.
#12
Posted 31 May 2015 - 05:27 PM
Divide that over 5 houses and 4 clans that is just an average of 25 to 38 players per house on average at any one time.
This isn't good news. We also know that most times were piled up in "prime time" for North America, meaning the pacific and asia and Europe were worse.
But this could all be GIGO.
#13
Posted 31 May 2015 - 05:34 PM
With 10,000 concurrent players on at primetime on a given day, I'd really strongly expect easily ten times that number in people who play somewhat regularly. All those people who can't spend hours playing every single day, but combined get you to that 10,000 concurrent players.
So, given all the above, I suspect we have a couple times that number in actual concurrent players in prime time.
#14
Posted 31 May 2015 - 05:40 PM
Next time try without the arrogant condescension.
#15
Posted 31 May 2015 - 05:43 PM
Burktross, on 31 May 2015 - 04:05 PM, said:
We were at the same elo for like, a month or something. Guess he ranked up, cause I don't see him anymore
He's a pretty cool guy in my experiences with him, so at least there's that
#16
Posted 31 May 2015 - 05:45 PM
Kjudoon, on 31 May 2015 - 05:40 PM, said:
Next time try without the arrogant condescension.
Or, you could try without alarmist doomsaying.
And, what is "Healthy"?
Healthy is any number large enough to support PGI, keep the game development going, and keep the lights on in the server room.
Clearly, the game is still being developed(if anything, it's moving ahead faster than ever), the lights are on, and PGI seems to be doing fine.
Edited by Wintersdark, 31 May 2015 - 05:46 PM.
#20
Posted 31 May 2015 - 07:04 PM
Edited by Monkey Lover, 31 May 2015 - 07:04 PM.
3 user(s) are reading this topic
0 members, 3 guests, 0 anonymous users