MidichIorian Posted June 1, 2012 Share Posted June 1, 2012 (edited) My main server dropped a good 120 or so in a week and it seems to correlate with the population drop I've seen in-game. That in itself is bad but what's worse is that it didnt drop many places (2) on the server list, which indicates that other servers lost just as many. Edited June 1, 2012 by MidichIorian Link to comment Share on other sites More sharing options...
Scorpienne Posted June 1, 2012 Author Share Posted June 1, 2012 Are those numbers coming from the theta calculation numbers or actual ingame surveys? If the former, could it be possible that as we see a greater number of servers with an average theta below 1, we see less accuracy in our population estimates (especially that we estimate high)? If our floor is 255 per server (or whatever it is) because of the estimate we're applying to low theta, then we can't effectively count population as it approaches 255 * number of servers, right? Uhhhh.... I hate to look like an idiot... but what's a "theta calculation?" You mean phi calculation? If so, then yes... the OP has details on how the totals were calculated. They're not from surveys. The OP also has an extended discussion about how a phi value less than or equal to 1 is really uncertain, and could mean 4 people or 400 people, so there's an order of magnitude difference in the average number of concurrent logins there. Paige Link to comment Share on other sites More sharing options...
DarthTHC Posted June 1, 2012 Share Posted June 1, 2012 (edited) Uhhhh.... I hate to look like an idiot... but what's a "theta calculation?" You mean phi calculation? If so, then yes... the OP has details on how the totals were calculated. They're not from surveys. The OP also has an extended discussion about how a phi value less than or equal to 1 is really uncertain, and could mean 4 people or 400 people, so there's an order of magnitude difference in the average number of concurrent logins there. Paige Sorry, phi... I look like the idiot. OK so for ø < 1, we're taking 250, but it could be 0-499, right? Eyeballing it, it looks like about 2/3 of the servers fall into that range so let's call it 80 servers? For those 80 servers, the total population is somewhere between 0 and 39,920 but we're counting it as 20,000. I dunno, the estimate for ø to me seems to be enough to suggest that the "leveling-off" we see over the past few weeks may not be an actual leveling off. It might simply be a factor of how we're estimating server populations for low-ø servers. Edit... I'm not saying your calculations are wrong. I'm just saying they necessarily become less accurate the lower the population of the servers. Because of that forced lack of accuracy, the assumption that the population decline is slowing could be incredibly faulty. I mean, take 10k off the numbers for the final 2-3 weeks, which is easily within the margin of error for the low-ø servers, and see what the percentages tell you. Edited June 1, 2012 by DarthTHC Link to comment Share on other sites More sharing options...
Scorpienne Posted June 2, 2012 Author Share Posted June 2, 2012 I'm not saying your calculations are wrong. I'm just saying they necessarily become less accurate the lower the population of the servers. Because of that forced lack of accuracy, the assumption that the population decline is slowing could be incredibly faulty. I mean, take 10k off the numbers for the final 2-3 weeks, which is easily within the margin of error for the low-ø servers, and see what the percentages tell you. I completely agree. The way I figure the math gets *very* uncertain the closer to a phi value of 1. That's a very good point, and one I missed. Yes, unfortunately, the leveling off is def within the MOE for the low-phi servers. :-( I hadn't realized that at first, but you're completely right. My hope was springing eternal... Paige Link to comment Share on other sites More sharing options...
Scorpienne Posted June 2, 2012 Author Share Posted June 2, 2012 http://www.vg247.com/2012/06/01/swtor-mmo-subscriber-numbers-are-funny-things-with-lots-of-variables-to-consider-says-bioware/ "There have been many analyst reports over the last few months citing drops in subscription numbers for SWTOR. Recently, a SWTOR player took it upon himself to do some server population figuring, and found that the average game server has 344 players online at a given time." At least these guys figured out what the numbers more-or-less mean before mentioning them. :-) Link to comment Share on other sites More sharing options...
erictlewis Posted June 2, 2012 Share Posted June 2, 2012 (edited) http://www.vg247.com/2012/06/01/swtor-mmo-subscriber-numbers-are-funny-things-with-lots-of-variables-to-consider-says-bioware/ "There have been many analyst reports over the last few months citing drops in subscription numbers for SWTOR. Recently, a SWTOR player took it upon himself to do some server population figuring, and found that the average game server has 344 players online at a given time." At least these guys figured out what the numbers more-or-less mean before mentioning them. :-) Damage control, wow boy howdy. 344 is the number of players on an average server??? Don't they mean the 10 servers that are constantly showing as standard I guess those are the average server. There is no way on this earth that the other 200 servers that are showing light has 344 average players on at a single time. Edited June 2, 2012 by erictlewis mispelled words Link to comment Share on other sites More sharing options...
Valkirus Posted June 2, 2012 Share Posted June 2, 2012 Damage control, wow boy howdy. 344 is the number of players on an average server??? Don't they mean the 10 servers that are constantly showing as standard I guess those are the average server. There is no way on this earth that the other 200 servers that are showing light has 344 average players on at a single time. Yep. Does not matter what the "average" is if a player on a certain servers knows the pop is extremely low on his server. That is what counts. I could not careless what it is on other servers. Link to comment Share on other sites More sharing options...
Sanxxx Posted June 3, 2012 Share Posted June 3, 2012 Looks like everyone is flocking to the top 3 US servers as they are showing a population increase whilst the rest of the servers continue to decline. Link to comment Share on other sites More sharing options...
snave Posted June 3, 2012 Share Posted June 3, 2012 Thank you for putting in the time and effort to perform this study, I'm sure Bioware already know the state of the population across servers but it's great to see it shared with the community. I'm on the Hidden Beks server EU and at peak time we have about 30 players on the fleet, we have 1 raiding guild and it takes roughly a hour for a warzone to pop. I can't help but feel if there are other servers like this Bioware are losing players hand over fist and need to get these transfers sorted ASAP. Link to comment Share on other sites More sharing options...
Scorpienne Posted June 3, 2012 Author Share Posted June 3, 2012 Looks like everyone is flocking to the top 3 US servers as they are showing a population increase whilst the rest of the servers continue to decline. Very true. Folks, if you're wondering where he's seeing this, take a look here: http://www.torstatus.net/shards/us/stats You can see the growth in the last week for Fatman, Jedi Covenant, and Drooga's Pleasure Barge. When you look at the individual server graphs... http://www.torstatus.net//the-fatman/trends/60d http://www.torstatus.net/jedi-covenant/trends/60d http://www.torstatus.net/droogas-pleasure-barge/trends/60d It's a little harder to determine. I think maybe the growth on these three servers has only started in the last couple of weeks so it's harder to see on the 60 day graphs. Paige PS, out of town tomorrow for a conference. Will update late tomorrow night, or maybe Tues morning. Paige Link to comment Share on other sites More sharing options...
Jewa_taken Posted June 4, 2012 Share Posted June 4, 2012 I completely agree. The way I figure the math gets *very* uncertain the closer to a phi value of 1. That's a very good point, and one I missed. Yes, unfortunately, the leveling off is def within the MOE for the low-phi servers. :-( I hadn't realized that at first, but you're completely right. My hope was springing eternal... Paige This is kinda what I was thinking as well, if the pop never change status from 1 there is no way to guess where in the interval the pop is, at least if it goes from light to standard once in a while, you'll know that it's been in the high areas. Otherwise all you know is that it's online. I saw this when checking against my own server (posted it earlier) where pop at peak seemed to be around 250, not avg pop. That's just one weekend on one server though. With actual headcounts on the low pop servers during off-peak and peak, you'd get a better average to work with, but that's alot of work... With the ones that do change status maybe it would be possible to guesstimate some kind of curve, but that kinda math is beyond me. Still, it seems to have gotten better in the last week on our server, and the ones that are on are active, so even with 20ish ppl on fleet we do have some wzs and there are some people raiding. This might be good info to have depending on how they choose servers for transfer, so thanks again for that. Link to comment Share on other sites More sharing options...
Goretzu Posted June 4, 2012 Share Posted June 4, 2012 (edited) Sorry, phi... I look like the idiot. OK so for ø < 1, we're taking 250, but it could be 0-499, right? Eyeballing it, it looks like about 2/3 of the servers fall into that range so let's call it 80 servers? For those 80 servers, the total population is somewhere between 0 and 39,920 but we're counting it as 20,000. I dunno, the estimate for ø to me seems to be enough to suggest that the "leveling-off" we see over the past few weeks may not be an actual leveling off. It might simply be a factor of how we're estimating server populations for low-ø servers. Edit... I'm not saying your calculations are wrong. I'm just saying they necessarily become less accurate the lower the population of the servers. Because of that forced lack of accuracy, the assumption that the population decline is slowing could be incredibly faulty. I mean, take 10k off the numbers for the final 2-3 weeks, which is easily within the margin of error for the low-ø servers, and see what the percentages tell you. If you are looking at it as a mathmatical challange then it's a sizeable problem which isn't easy to solve (without some sort of sustained count). But really from a server health/game play perspective it is a bit moot, any "Light" server (that never goes higher) is effectively an unviable server as far as SWTOR is concerned (although I'm not saying the odd person won't like to play on one that low), as to sustain healthy grouping, AH, and PvP levels a server really needs to be high standard and I'd even argue that Heavy or even V Heavy because of the intrinsic way SWTOR is instanced (even a full SWTOR server never really feels "full"). By which I mean that by the time a server hits constant "Light" it may as well be effectively classified as "Dead" or "Zombie" in any context, but total game population count. The scary thing is how much of the remaining population are trapped on these "zombie" servers until server transfers come. Edited June 4, 2012 by Goretzu Link to comment Share on other sites More sharing options...
Scorpienne Posted June 5, 2012 Author Share Posted June 5, 2012 Updated with today's numbers. See the original post for details. http://www.swtor.com/community/showpost.php?p=4305330&postcount=1 LOGINS BY TYPE PvE 31,566 PvP 23,407 RP PvE 6,070 RP PvP 2,200 Total 63,243 LOGINS BY AREA APAC 1,950 EUR English 12,568 EUR French 4,643 EUR German 7,235 US East 24,763 US West 12,085 Total 63,243 AVG LOGINS BY TYPE PvE 287 PvP 285 RP PvE 337 RP PvP 367 Average 293 AVG LOGINS BY AREA APAC 650 EUR English 279 EUR French 290 EUR German 249 US East 306 US West 288 Average 293 Paige Link to comment Share on other sites More sharing options...
Blackardin Posted June 5, 2012 Share Posted June 5, 2012 (edited) wow just wow First off I have no stance on the game one way or the other, but man you people must be really bored It is so funny how people see something like this, run with it and post as if it is some kind of a good assumption of anything. If ø is greater than 1, start with 250 plus the % ø is greater than 1 x 1000 (where 1000 is the average of the top end of light = 500, and the top end of standard = 1500). The above example: Does it not bother anyone at all that this is completely and totally made up? Also to take into account that all server pop caps are not the same doing any calaculation based off of a wrong assumption and made up formaulas is just completely without purpose You know anyone can make assumption and support a formula when you already know the end answer (ie population total that was reported) And I can tell you in absolute fact when I did my server it is completely wrong I really wish people would get lives and actually just play the game instead of spending so much time with absolute garbage like this Nothing will be correct, or even close to accurate unless it is released officially, which it never will be. I'm sure in another month some other "genius" will come out with something similar just like the past "geniuses" who tried to figure it out in the past months I've never met so many analysts, accountants, economists, and game designers in my life ;p Honestly, I've seen every type of armchair analysis ever contemplated based on misinformation, conjecture, and just plain ole fabrication. The fact is that this information "could" be accurate, or it could be totally off the wall. It could be produced by someone that is actually being paid to provide negative information. It could be provided by some guy with a sock puppet on each hand...but people are accepting it as rock solid truth. I'm not contesting the data or agreeing with it, just not accepting it as fact. We know that it is totally wrong for your server, as you pointed out. That, along with the many assumptions brings it into question. Edited June 5, 2012 by Blackardin Link to comment Share on other sites More sharing options...
Sanxxx Posted June 5, 2012 Share Posted June 5, 2012 I've never met so many analysts, accountants, economists, and game designers in my life ;p Honestly, I've seen every type of armchair analysis ever contemplated based on misinformation, conjecture, and just plain ole fabrication. The fact is that this information "could" be accurate, or it could be totally off the wall. It could be produced by someone that is actually being paid to provide negative information. It could be provided by some guy with a sock puppet on each hand...but people are accepting it as rock solid truth. I'm not contesting the data or agreeing with it, just not accepting it as fact. We know that it is totally wrong for your server, as you pointed out. That, along with the many assumptions brings it into question. Someone with an interest in statistics has chosen to produce some data based on a defined method, others have chosen to interpret it, that's about it. The actual population estimate may well be wrong , as the author readily admits, but however you look at it the fact remains that populations have declined and continue to do so. Link to comment Share on other sites More sharing options...
ZudetGambeous Posted June 5, 2012 Share Posted June 5, 2012 Is there a way to see the population numbers from previously? If I remember the past numbers correctly fatman seems to be growing, but I don't remember the other 19. Link to comment Share on other sites More sharing options...
Scorpienne Posted June 5, 2012 Author Share Posted June 5, 2012 I've never met so many analysts, accountants, economists, and game designers in my life Technically, I'm an environmental consultant with a PhD in biological oceanography. I analyze biological information nearly every working day. Just sayin' this isn't too far off what I do to pay the mortgage. http://www.linkedin.com/in/paleitman I'm not contesting the data or agreeing with it, just not accepting it as fact. We know that it is totally wrong for your server, as you pointed out. That, along with the many assumptions brings it into question. I respectfully and politely disagree with you. The data isn't "wrong" for my server. I pointed out that the 2 week average of slack time and peak time (about 300ish) is understandably and naturally lower than the peak-hours average (about 1000ish) by a factor of roughly 2 to 3. That doesn't mean the 24 hour 2 week average is wrong. It doesn't mean the peak-hours average is wrong. They are two *completely different measurements*. Heck, if they were the same number, then that would be super weird! I was trying to show an example of *exactly* what the numbers mean by comparing my server's average and peak population - and if you think this shows that the population estimate is "totally wrong" for my server, then I've utterly failed in trying to explain what I mean. The numbers are what they are. If you use them how they're meant to be used, and understand what they truly mean, then cool. Paige Link to comment Share on other sites More sharing options...
Scorpienne Posted June 5, 2012 Author Share Posted June 5, 2012 Is there a way to see the population numbers from previously? If I remember the past numbers correctly fatman seems to be growing, but I don't remember the other 19. Sure. The quick and easy way to do it is to just look at torstatus for Fatman. They'll show you a graph of the past 60 days. http://www.torstatus.net/the-fatman/trends/60d If you're feeling spunky, you can look in the spreadsheet to get the data for every time I've done the analysis. https://docs.google.com/spreadsheet/ccc?key=0Aregkvys5QFodFJ2OWN5U0hwaVFBYWdqUUh1WmdZUFE#gid=24 Look at the other tabs down at the bottom and you'll see the past data. You can copy/paste it into excel if you want to do your own analysis. Paige Link to comment Share on other sites More sharing options...
Scorpienne Posted June 6, 2012 Author Share Posted June 6, 2012 OP updated with todays statistics... as a summary... Top 20 US Servers (6/6/2012) The Fatman (PvP US East) Population ≈ 2451 Jedi Covenant (PvE US East) Population ≈ 1483 Drooga's Pleasure Barge (PvE US West) Population ≈ 890 The Harbinger (PvE US West) Population ≈ 770 Canderous Ordo (PvE US East) Population ≈ 600 The Ebon Hawk (RP PvE US East) Population ≈ 560 Lord Adraas (RP PvE US East) Population ≈ 540 Jung Ma (RP PvP US East) Population ≈ 500 The Swiftsure (PvP US West) Population ≈ 480 Ajunta Pall (RP PvP US West) Population ≈ 470 The Shadowlands (PvE US East) Population ≈ 330 Sanctum of the Exalted (RP PvE US East) Population ≈ 300 Begeren Colony (RP PvE US West) Population ≈ 290 Krayt Dragon (PvE US East) Population ≈ 280 Lord Ieldis (RP PvE US West) Population ≈ 280 Mind Trick (PvE US East) Population ≈ 280 Shadow Hand (PvE US East) Population ≈ 270 Kaas City (PvE US East) Population ≈ 260 Juyo (PvE US East) Population ≈ 250 Sith Wyrm (PvE US East) Population ≈ 250 Top 20 World Servers (6/6/2012) The Fatman (PvP US East) Population ≈ 2451 Jedi Covenant (PvE US East) Population ≈ 1483 Drooga's Pleasure Barge (PvE US West) Population ≈ 890 Tomb of Freedon Nadd (PvP EUR English) Population ≈ 830 The Harbinger (PvE US West) Population ≈ 770 The Red Eclipse (PvE EUR English) Population ≈ 690 Dalborra (PvE APAC) Population ≈ 670 Gav Daragon (RP PvE APAC) Population ≈ 660 Canderous Ordo (PvE US East) Population ≈ 600 Master Dar'Nala (PvP APAC) Population ≈ 600 The Ebon Hawk (RP PvE US East) Population ≈ 560 Lord Adraas (RP PvE US East) Population ≈ 540 Dune Bantha (PvE EUR English) Population ≈ 510 Jung Ma (RP PvP US East) Population ≈ 500 Huntmaster (RP PvP EUR French) Population ≈ 480 The Swiftsure (PvP US West) Population ≈ 480 Ajunta Pall (RP PvP US West) Population ≈ 470 Vanjervalis Chain (RP PvE EUR German) Population ≈ 410 Mantle of the Force (PvE EUR French) Population ≈ 400 Bacca's Blade (PvE EUR English) Population ≈ 370 SERVERS BY AREA APAC 3 EUR English 45 EUR French 16 EUR German 28 US East 81 US West 42 Total 215 SERVERS BY TYPE PvE 109 PvP 82 RP PvE 18 RP PvP 6 Total 215 SERVERS BY AREA APAC 3 EUR English 45 EUR French 16 EUR German 28 US East 81 US West 42 Total 215 SERVERS BY TYPE PvE 109 PvP 82 RP PvE 18 RP PvP 6 Total 215 LOGINS BY TYPE PvE 31,300 PvP 23,396 RP PvE 6,015 RP PvP 2,180 Total 62,891 AVG LOGINS BY TYPE PvE 287 PvP 285 RP PvE 334 RP PvP 363 Average 293 LOGINS BY AREA APAC 1,930 EUR English 12,603 EUR French 4,580 EUR German 6,990 US East 24,776 US West 12,013 Total 62,891 AVG LOGINS BY AREA APAC 643 EUR English 280 EUR French 286 EUR German 250 US East 306 US West 286 Average 293 I know I just updated yesterday. Was out of town Monday and am trying to get back on the MWF update schedule. Paige Link to comment Share on other sites More sharing options...
Allbrass Posted June 7, 2012 Share Posted June 7, 2012 (edited) deleted Edited June 7, 2012 by Allbrass Link to comment Share on other sites More sharing options...
Ashlian Posted June 7, 2012 Share Posted June 7, 2012 568K (US) 432K (EU/AP) Total, 1,000K = 1 million So ~300K people are paying but not playing. Which means you can project another 300K drop in the next few months. (this is w/o counting the impact of the 38 new countries added to the SWTOR) Source: https://docs.google.com/spreadsheet/ccc?key=0Aregkvys5QFodFJ2OWN5U0hwaVFBYWdqUUh1WmdZUFE#gid=1 So you're assuming they aren't going to get anything right in any update including 1.3? Of course, we'll have to wait and see, but since they're doing several things on the backend related to the aforementioned problems with the Hero engine and low server population limits, herding people onto more populated servers with presumably higher adjusted limits, and working on fixes to gameplay mechanics (as highlighted in the Darthater interview - nice to know they're aware of where they went both right and wrong), then I'm willing to bet the remaining pop will stay stable. I sure wouldn't bet on growth, but I know a lot of people currently doing other things who nonetheless will be happy to pay the cost of a cheap dinner entree to check SWTOR back out if some of their still playing friends tell them there's something worth coming back for. I'm betting a good portion of the people not playing, despite all the rabid "I WILL NEVER TRUST BW AGAIN!" posts, will take another look at the next big content update (ie, not 1.3). Most people aren't as fanatic for or against as forum posters. Link to comment Share on other sites More sharing options...
Kendu Posted June 7, 2012 Share Posted June 7, 2012 Must be nice getting in wz, kathol rift is dead and starting to rot. PvP ques take at least a hour. Really don't know what to do, i'm use to raiding in wow a lot and quit wow for swtor, now my guild has quit swtor to go play gw2. Hate my life right now just got gw2, beta weekend yipee:) Link to comment Share on other sites More sharing options...
erictlewis Posted June 7, 2012 Share Posted June 7, 2012 I check torstatus.net every day and watch the numbers continue to fall. 1.3 offers nothing to fix this, other than the transfers that were waiting for an faq on. Link to comment Share on other sites More sharing options...
Scorpienne Posted June 8, 2012 Author Share Posted June 8, 2012 Updated with today's numbers... and we now have a month worth of data. Paige Link to comment Share on other sites More sharing options...
Karl-Just-Karl Posted June 8, 2012 Share Posted June 8, 2012 Updated with today's numbers... and we now have a month worth of data. Paige Just to be clear on this weeks numbers... 3 US servers saw a "gain" 211 US servers saw a "loss" 0 Euro servers saw a "gain" 45 Euro servers saw a "loss" Questions: Is this number very different from previous weeks? Does the availability of PTS have anything to do with that? Link to comment Share on other sites More sharing options...
Recommended Posts