Estimation of average concurrent logins (top servers)

[Scorpienne]

Well, looking at Harbinger, it definitely has a lower cap. Much lower. Its registering full when Jedi Cov is skirting Heavy/Very Heavy territory.

 

I'm currently trying to figure out the pop settings for each of the destination servers. Once we have that, torstatus's numbers are useful again, since they collect the status from each server. We'd just have to make sure we accounted for the difference in each server individually. But more work, but once we have the ranges for each server, its easy as pi ;-)

 

That's the spirit! I see a lot of surveys happening. :-)

 

Anecdotally, I was on Sanctum of the Exalted (an RP PvE EST Origin server) to disband our Imp and Pub guilds so we could get guild bank tabs moved over, and there were about 20-30 people per side on the server between midnight and 1:30.

 

So that's a shred of anecdotal info on what's going on in one light server.

 

 

Paige

Edited June 24, 2012 by Scorpienne

[Emeda]

Well you guys and girls have fun my sub runs out again some time tomorrow. Maybe ill be back in a few months again.

 

To the people posting that they are wrong and anything they do doesnt matter. We know it doesnt matter but some people like to figure out things that they are not being given.

 

Looks like you have alot more people helping you too so thanks to them for helping you.

 

One thing I have kind of noticed but nothing to back it up as true. I noticed the last few times I checked the status 3 or4 times in the last week that all the servers were about the same level as in heavy or very heavy or full. I really hope Bioware isnt just playing with all the numbers just so you guys give up since some people have reported numbers that just dont make sense. Maybe their new meathod for server status is to have one person just hit standard on the servers then once one hits heavy hit heavy on them all then keep doing it to full then back down to light after peak.

 

Have fun all

[SnakeCL]

So my data on the JC server ranges is incomplete, but it might not actually be needed to calculate.

 

This is for Jedi Covenant:

Light is 1000 or less

Standard is 1001-2500

heavy is 2501-3500

Very heavy is 3500-4XXX (we don't actually need this since we don't have any pops above 3.0)

 

Using the old method.

Average of 0 and high light (0 + 1000)/2 = 500

Average of the high light and high standard. (1000 + 2500)/2 = 1750

Average of high standard and high heavy. (2500 + 3500)/2 = 3000

 

Most recent swtor status figure was 2.10 for Jedi Cov. So 500+1750+300= 2550.

 

So 2550?

 

The newer method comes to this.

2500(.9)+3500(.1) = 2250 + 350 = 2600

 

2600

 

Somebody check my math. I'm no math whiz.

Edited June 24, 2012 by SnakeCL

[Scorpienne]

 

I'll do some math-fu tomorrow when I'm on my work machine and have real excel, not just google docs. :-)

 

You're a hero for doing so many surveys!

 

Paige

[SnakeCL]

This is what I managed to pull from the survey data (after the transfers). For data points from only one faction, I doubled the number to reach a total. This is only an estimate, but there's already a bit of a trend forming.

 

I'm listing the server, then the population level, and then the range of numbers that was reported for each population level. For "light" i only listed the highest number, since the low-end if 0.

 

Managed to come up with a graph showing what few data points we have now. Looks like a pattern is starting to emerge for the larger servers.

 

http://img703.imageshack.us/img703/3327/graph2t.jpg

 

 

 

JC
light 807
standard 1057
Heavy 2742-3316
Very Heavy 3678

Shadowlands
light 706
standard ?
heavy ?
very heavy ?

Ebon Hawk
light ?
standard 1600-2105
Heavy 2165-2290
Very heavy 3215-3419

Begeron Colony
light ?
Standard 638-708
Heavy 1270

Canderous
light ?
Standard 988-1025
heavy ?
very heavy ?

Edited June 24, 2012 by SnakeCL

[Valkirus]

This is what I managed to pull from the survey data (after the transfers). For data points from only one faction, I doubled the number to reach a total. This is only an estimate, but there's already a bit of a trend forming.

 

I'm listing the server, then the population level, and then the range of numbers that was reported for each population level. For "light" i only listed the highest number, since the low-end if 0.

 

JC

light 807

standard 1057

Heavy 2742-3316

Very Heavy 3678

 

Shadowlands

light 706

standard ?

heavy ?

very heavy ?

 

Ebon Hawk

light ?

standard 1600-2105

Heavy 2165-2290

Very heavy 3215-3419

 

Begeron Colony

light ?

Standard 638-708

Heavy 1270

 

Canderous

light ?

Standard 988-1025

heavy ?

very heavy ?

 

So how does this help us to understand what the current number of players are logged in during prime time? I can take those figures, if based on the amount of servers in NA and conclude ( using a losely formulated estimate for calculating those figures ) less than 100,000 are logged on. But that may be totally inaccurate. Sticking to the topic of the thread..it would mean counting only the top 20 servers there are about 34,000 logged on. Average.

[islander]

Well you guys and girls have fun my sub runs out again some time tomorrow. Maybe ill be back in a few months again.

 

To the people posting that they are wrong and anything they do doesnt matter. We know it doesnt matter but some people like to figure out things that they are not being given.

 

Looks like you have alot more people helping you too so thanks to them for helping you.

 

One thing I have kind of noticed but nothing to back it up as true. I noticed the last few times I checked the status 3 or4 times in the last week that all the servers were about the same level as in heavy or very heavy or full. I really hope Bioware isnt just playing with all the numbers just so you guys give up since some people have reported numbers that just dont make sense. Maybe their new meathod for server status is to have one person just hit standard on the servers then once one hits heavy hit heavy on them all then keep doing it to full then back down to light after peak.

 

Have fun all

 

I am, and I will...thanks

[SnakeCL]

So how does this help us to understand what the current number of players are logged in during prime time? I can take those figures, if based on the amount of servers in NA and conclude ( using a losely formulated estimate for calculating those figures ) less than 100,000 are logged on. But that may be totally inaccurate. Sticking to the topic of the thread..it would mean counting only the top 20 servers there are about 34,000 logged on. Average.

 

Those figures are basically to help us figure out at what population range the servers change their population level. The population numbers we've been using have been off ever since the transfer, since BioWare raised the population caps. ie. Standard on Jedi Covenant now is a lot more than Standard on Jedi Covenant before the transfer.

 

Once we have the corrected figures, we can use torstatus's data for each server, and come up with a better average.

[Thyferra]

Is there any way to get easy access of the Lvl 50 pop on a server? all the other level ranges are quite easy to get but lvl 50 seems to be a pain. I don't really want to check every possible instance of the game to find a few stray lvl 50s. Perhaps i am missing something.

[Scorpienne]

Is there any way to get easy access of the Lvl 50 pop on a server? all the other level ranges are quite easy to get but lvl 50 seems to be a pain. I don't really want to check every possible instance of the game to find a few stray lvl 50s. Perhaps i am missing something.

 

The least screwed up way that I've found (I'm not saying it's good, and I'm not saying it's the only way) is to census the 50s by advanced class.

 

This only works if there are less than 100 level 50s of each advanced class. I've found in some surveys that there are more that 100 level 50s of that class. In which case, you're screwed. Nothing you can do.

 

So put these in your search terms box:

 

On the Pub side

50 Vanguard

50 Commando

50 Guardian

50 Sentinel

50 Sage

50 Shadow

50 Gunslinger

50 Scoundrel

 

Or on the Imp side

50 Juggernaut

50 Marauder

50 Powertech

50 Mercenary

50 Assassin

50 Sorcerer

50 Operative

50 Sniper

[islander]

I do that on a regular basis for my own server. You don't have to do it planet by planet. You can do it by level.

 

LINK

 

 

Paige

 

Yeah, this is how I knew how terrible my old server, Kinrath Spider was. Simple checks by level set (1-10, 11-19, 20-29, 30-39, 40-49, and 50). If 50 gave me 100 results, I then turned around and searched by archtype (jugg, mara, powertech, merc, etc) and the latter never did me wrong.

 

I still log on once a day to make sure I don't have guildmates who were inactive logging in and just making sure they know what the deal is now.

 

Kinrath still has like 8-10 imperials online during primetime. I actually have one guy who doesn't want to transfer to Fatman (because the people there are jerks, in his words). I'm like oookay but if you change your mind, we'll be here for you!

Edited June 25, 2012 by islander

[Thyferra]

I know its nots much but it is a start.

 

Just took my first data from Nightmare Lands.

Its labled as standard and i counted 1596 players online. This took me about 10 - 15 mins to get this data so its not 100 % accurate. Once the server jumps to heavy i will do it again.

[Scorpienne]

I know its nots much but it is a start.

 

Just took my first data from Nightmare Lands.

Its labled as standard and i counted 1596 players online. This took me about 10 - 15 mins to get this data so its not 100 % accurate. Once the server jumps to heavy i will do it again.

 

Oooh! Awesome! Did you put it on the survey?

 

Survey link here

 

Paige

[Scorpienne]

OP updated with today's numbers.

 

Total logins keeps skyrocketing. We've added another 4k logins and the estimated concurrent logins is 74,402. This represents a population of 27,064 on destination servers (36% of the total estimated logins) and 47,3383 on origin servers (64% of the total estimated logins).

 

If you're looking at servers with a phi greater than 1 then we're looking at 28,334 concurrent logins.

 

A *LOT* of people are contributing data. I'll run some math to see if we've got some idea of what the new server caps are and/or if there are new server caps.

 

 

Paige

Edited June 25, 2012 by Scorpienne

[Thyferra]

Oooh! Awesome! Did you put it on the survey?

 

Survey link here

 

Paige

 

Just added the data to the survey. I did not have numbers for the fleet so i put in 100 as a ph. will try to get that data next time as well.

[Scorpienne]

Oh, and... I got distracted by some real actual billing-the-client work, but, I meant to point this out... look at the list of top servers...

 

Do you see who isn't in the top 3?

 

1. Tomb of Freedon Nadd (PvP EUR English) Avg Logins ≈ 1889.375 Phi = 2.33 Change = 0.76
   
2. Drooga's Pleasure Barge (PvE US West) Avg Logins ≈ 1773.125 Phi = 2.27 Change = 0.6
   
3. The Harbinger (PvE US West) Avg Logins ≈ 1676.25 Phi = 2.22 Change = 0.75
   
4. The Fatman (PvP US East) Avg Logins ≈ 1521.25 Phi = 2.14 Change = -0.47
   
5. The Bastion (PvP US West) Avg Logins ≈ 1443.75 Phi = 2.1 Change = 1.12
   
6. Jedi Covenant (PvE US East) Avg Logins ≈ 1405 Phi = 2.08 Change = -0.14
   
7. Prophecy of the Five (PvP US East) Avg Logins ≈ 1327.5 Phi = 2.04 Change = 1.06
   
8. The Ebon Hawk (RP PvE US East) Avg Logins ≈ 1308.125 Phi = 2.03 Change = 0.75
   
9. The Red Eclipse (PvE EUR English) Avg Logins ≈ 1180 Phi = 1.93 Change = 0.48
   
10. Canderous Ordo (PvE US East) Avg Logins ≈ 1170 Phi = 1.92 Change = 0.6
    

 

 

The Fatman continues to demonstrate a precipitous population decline. Nearly all the other destination servers have large growth.

 

 

Paige

Edited June 25, 2012 by Scorpienne

[Thyferra]

That is something that was to be expected. The Fatman was the go to server wheny our server was dead. So many people created toon there because their server was empty. Now that they transferred all their main toon they will play those again.

 

Edit: New numbers from Nightmare Lands. The server is now labled heavy and we got over 2208 players logged in. We got more than 100 level 50 sorcerers. I doubt it will be much more but i cant give a 100% accurate number because of this.

Again this has been taken over 15 mins.

 

Edit: I doubt i can do much more. As soon as we hit very heavy quite a few classes will be over 100. Sorcerers already are.

Edited June 27, 2012 by Thyferra

[Scorpienne]

Okay, here is a NEW AND DIFFERENT spreadsheet.

 

New Spreadsheet.

 

Obviously, we need to know that status X equals population Y. If we know that, we can turn the statuses from Torstatus into reasonable estimates of the number of people logged in on various servers.

 

The first tab of the spreadsheet is called "data from surveys". This is just the data from the surveys (link to the survey, link to the data) with a couple of extra columns at the far right and some nonessential columns removed.

 

One interesting thing is that the bulk data shows that the general population is 47% Reps and 53% Imps. I'm going to call the total server population 2x the single faction population. That might be an under or overestimate of 3%, but heck, that's close enough for what we're doing here.

 

The first set of values is for destination servers only (NOT origin servers) and ONLY for data post-start-of-transfers. Status = server status and Dest Server Pop = 2x the single faction population from the survey. The table to the right is the summary of server status.

 

Status Avg Pop Value (Min - Median - Max)

Light 664 (174-740-1,056) 11 observations

Standard 1,367 (660-1,094-2,216) 13 observations

Heavy 2,401 (1,232-2,338-3,604) 11 observations

Very Heavy 3,408 (2,844-3,316-3,994) 7 observations

Full 3,527 (3,166-3,650-3,766) 3 observations

 

This is a huge difference from what we'd been using previously.

Prev = all analysis up to this point

Avg = average value of destination servers at that status (above)

Median = median value of destination servers at that status (above)

 

STATUS | PREV | AVG | MEDIAN

Light | 250 | 665 | 740

Standard | 1,000 | 1,367 | 1,094

Heavy | 1,938 | 2,401 | 2,338

Very Heavy | 2,688 | 3,408 | 3,316

Full | 3,000 | 3,527 | 3,650

 

INCREASE FROM PREVIOUS # (based on avg and median values)

Light 2.7 - 3.0 increase

Standard 1.4 - 1.1 increase

Heavy 1.2 - 1.2 increase

Very Heavy 1.3 - 1.2 increase

Full 1.2 - 1.2 increase

 

 

So for *destination* servers

Light should probably triple what we're currently using.

Standard should increase 10%-40%.

Heavy should increase about 20%.

Very Heavy should increase 20%-30%

Full should increase 20%.

 

This means that all of my data since the start-of-transfers, or maybe even earlier, should increase maybe 30% overall for destination servers. That's cool. Still this is "only" 45 data observations and it's "only" from seven servers (Begeren Colony, Canderous Ordo, Jedi Covenant, The Ebon Hawk, The Fatman, The Harbinger, and The Shadowlands).

 

I'd love to get some data from Tomb of Freedon Nadd, Drooga's Pleasure Barge, The Bastion, Prophecy of the Five, and The Red Eclipse since those are big servers for which we have no data.

 

For the origin servers, I have 3 observations that are all less than 50 people. The few times I've checked back in on my origin server it's been under 50 people. I think this means I can no longer look at the population of origin servers because it's too uncertain.

 

As we keep getting surveys, I'll fine-tune these numbers. So keep 'em coming!

 

 

Paige

Edited June 25, 2012 by Scorpienne

[Scorpienne]

Edit: I doubt i can do much more. As soon as we hit very heavy quite a few classes will be over 100. Sorcerers already are.

 

Yep, I've even seen that on the Ebon Hawk, particularly with sorcerers. Once one or two or more 50 adv classes go to 100, then there's not much to be done.

 

Be sure and hit the enter key (magnifying glass) a few times. Sometimes search is a little weird.

 

 

Paige

[SnakeCL]

Scorpienne's great work!

 

I'll be doing some more surveying, but watch out for Begeron Colony. Even though its a destination server, its been polling significantly lower than the other destinations, Not enough data yet, but its entirely possible its on a "reduced" cap compared to servers like Fatman and JC.

 

It was running "heavy" with 1200-1300 users, and standard with "700-800" users. The outlier is the "standard" at 1400-1500.

[Scorpienne]

I'll be doing some more surveying, but watch out for Begeron Colony. Even though its a destination server, its been polling significantly lower than the other destinations, Not enough data yet, but its entirely possible its on a "reduced" cap compared to servers like Fatman and JC.

 

It was running "heavy" with 1200-1300 users, and standard with "700-800" users. The outlier is the "standard" at 1400-1500.

 

FWIW, here are the AVERAGE logins at each server status for the servers that we have info. ND = no data. n = is the number of observations. The data is all too sparse and variable as it stands to say too much about them. I'm not sure that we can say that about BC just yet.

 

 

Server 1 Light

Begeren Colony 199 n=2

Canderous Ordo 789 n=2

Jedi Covenant 721 n=3

The Ebon Hawk ND

The Fatman 880 n=2

The Harbinger ND

The Shadowlands 706 n=2

 

Server 2 Standard

Begeren Colony 1,090 n=4

Canderous Ordo 988 n=1

Jedi Covenant 1,359 n=3

The Ebon Hawk 2,099 n=3

The Fatman ND

The Harbinger ND

The Shadowlands 1,025 n=2

 

Server 3 Heavy

Begeren Colony 1,270 n=2

Canderous Ordo ND

Jedi Covenant 3,200 n=4

The Ebon Hawk 2,216 n=5

The Fatman ND

The Harbinger ND

The Shadowlands ND

 

Server 4 Very Heavy

Begeren Colony ND

Canderous Ordo ND

Jedi Covenant 3,531 n=3

The Ebon Hawk 3,316 n=4

The Fatman ND

The Harbinger ND

The Shadowlands ND

 

 

Server 5 Full

Begeren Colony ND

Canderous Ordo ND

Jedi Covenant ND

The Ebon Hawk 3,166 n=1

The Fatman ND

The Harbinger 3,708 n=2

The Shadowlands ND

 

 

Paige

[sabreene]

I'll be doing some more surveying, but watch out for Begeron Colony. Even though its a destination server, its been polling significantly lower than the other destinations, Not enough data yet, but its entirely possible its on a "reduced" cap compared to servers like Fatman and JC.

 

It was running "heavy" with 1200-1300 users, and standard with "700-800" users. The outlier is the "standard" at 1400-1500.

 

I believe this is probably true. We're at "standard" right now (just ran the numbers) and we have 411 reps and 437 imps, for a total of 848 at standard. From watching torstatus, we should be going into "heavy" soon, and I'll try and catch the break point.

 

Edit: These seem to be the same as they were previously, at about 650 we'd bump up into "standard". So it's possible (not enough data yet!) that while some servers were increased, BC stayed the same as it was before the transfers.

 

I did a few checks on non-destination servers, but at very off-peak times.... on Zaalbar there were only 9 people on at 1am, including both sides. Just for some anecdotal data.

Edited June 25, 2012 by sabreene

[Teefal]

Oh for the love of all that's holy, could someone with SQL access at Bioware at least say, "Getting warmer", "getting colder" We know you're reading this, chuckling to yourself.

 

In fact, they might be changing the caps on the spot as each new guess comes in.

 

Or even .... a range of cap maxes, that randomize each day. (I probably would do this as a goof, knowing that internally there's no need for these arbitrary level labels in their stats)

Edited June 25, 2012 by Teefal

[Scorpienne]

Oh for the love of all that's holy, could someone with SQL access at Bioware at least say, "Getting warmer", "getting colder" We know you're reading this, chuckling to yourself.

 

In fact, they might be changing the caps on the spot as each new guess comes in.

 

Or even .... a range of cap maxes, that randomize each day. (I probably would do this as a goof, knowing that internally there's no need for these arbitrary level labels in their stats)

 

Heh, I know, right? They *have* this data. I could probably write the process flow to show a running display myself. Heck, I'd *learn* enough SQL to do it if they'd let me at the data.

 

I can totally see them at a staff meeting on Mondays, Tuesdays, and Thursdays bringing up this thread and chuckling into their coffee and rolling dice to see what they're going to do to server caps that day just to keep us on our toes! :-D

 

 

Paige

[YanksfanJP]

It amazes me that you people are more concerned with this than you are about playing the game itself. You doomsayers are going to end up being a self-fulfilling prophecy.

 

Does it really matter what the total numbers of players really is? If you're playing and having a good time, no, it doesn't matter. If you're playing and not having a good time, stop playing and unsubscribe. Whether there's 500 or 520 or 380 people on during prime time on a holiday at low tide when Mars is aligned with Jupiter in the seventh position shouldn't have any sort of impact that you or anyone else playing the game should care about.