History repeats itself

[NecrotechMaster]

100,000 (+/- 50,000) is still a very large core group of players, prolly 85% of those players have been here for 5+ years

and thats with 0 marketing from ncsoft to attract more poeple

i was actually talking to someone i work with about coh and they were actually sad they couldnt play it because just the character creator was exactly what they were looking for, yet they had never actually heard about the game until i had told them about it today

[Aggelakis]

A quick guess would be circa 100,000.

That's quite unlikely.

Solid yellow dots for everyone but Vigilance, Zukunft, and Defiant (Defiant blipped yellow during a couple of my multitudes of character swaps). Virtue went red first, Freedom later in the evening. When Virtue crashed, each time there was between 1000-1400 in the queue. Judging by that (1000-1500 = red dots ... color didn't PRECISELY translate into population but it was pretty close), I would guess there was around 5,000, maybe 7,500 ... maybe 10,000-15,000 all told over the course of the evening/night.

Don't take me wrong. That's still bleepin' ridiculously awesome turnout. But going by financials, City had (best guess) 20-40k subscribers by the end -- and 70,000 former subscribers *DID NOT* show up last night to round out your number. All the servers would have grayed out from being full.

[Victoria Victrix]

That's quite unlikely.

Solid yellow dots for everyone but Vigilance, Zukunft, and Defiant (Defiant blipped yellow during a couple of my multitudes of character swaps). Virtue went red first, Freedom later in the evening. When Virtue crashed, each time there was between 1000-1400 in the queue. Judging by that (1000-1500 = red dots ... color didn't PRECISELY translate into population but it was pretty close), I would guess there was around 5,000, maybe 7,500 ... maybe 10,000-15,000 all told over the course of the evening/night.

Don't take me wrong. That's still bleepin' ridiculously awesome turnout. But going by financials, City had (best guess) 20-40k subscribers by the end -- and 70,000 former subscribers *DID NOT* show up last night to round out your number. All the servers would have grayed out from being full.

Number of subscribers pre-announcement 40k.  Number of pure f2p, 40k.  Number of freemium 20k.

(numbers verified)

[Aggelakis]

Regardless, my statement stands. If 100,000 people had showed up, all of the servers would be grayed-out-full.

[Adrenalin]

Virtue went splat.

It's the Unity Rally all over again!!

https://dl.dropbox.com/u/33551029/Capture1.12.2012.-6.15.56.jpg

This should be turned into a meme.

[Colette]

"Number of subscribers pre-announcement 40k.  Number of pure f2p, 40k.  Number of freemium 20k. (Numbers verified)."

Thank you, V.V. That's still a lot of people. In sum, CoH commands about 50,000 "whales." That's a sweet chunk of change there.

[Adrenalin]

"Number of subscribers pre-announcement 40k.  Number of pure f2p, 40k.  Number of freemium 20k. (Numbers verified)."

Thank you, V.V. That's still a lot of people. In sum, CoH commands about 50,000 "whales." That's a sweet chunk of change there.

I might actually place it higher than that, F2Pers sometimes spend a lot more than VIP members all told. I pop in and out of the SWTOR boards and have been watching players buy multiple Cartel Coin packs... we're talking $40.00 a pop and sometimes as many as 5 at a time. I would be very curious to know what kind of funds the CoH store was raking in based on that.

[Timelord Tom]

I might actually place it higher than that, F2Pers sometimes spend a lot more than VIP members all told. I pop in and out of the SWTOR boards and have been watching players buy multiple Cartel Coin packs... we're talking $40.00 a pop and sometimes as many as 5 at a time. I would be very curious to know what kind of funds the CoH store was raking in based on that.

I was on a yearly subscription, so that was at a discount, but probably paid out that same amount over the course of a year on Paragon Points.

There were probably a few that spent even more than I did.

[NecrotechMaster]

i know depending on the stuff released on the market that month, i spent between $0 and $100 on top of my 3 month sub plan

[Aggelakis]

I know a couple people personally who spent about/over $40/mo on pp, on top of their subs. One of them had THREE subs (technically two and one for his son) and still spent that much - though that's less astounding because with three subs, he had to get things three times over if he really liked them.

[Iron-Emerald]

Solid yellow dots for everyone but Vigilance, Zukunft, and Defiant (Defiant blipped yellow during a couple of my multitudes of character swaps). Virtue went red first, Freedom later in the evening. When Virtue crashed, each time there was between 1000-1400 in the queue. Judging by that (1000-1500 = red dots ... color didn't PRECISELY translate into population but it was pretty close), I would guess there was around 5,000, maybe 7,500 ... maybe 10,000-15,000 all told over the course of the evening/night.

I don't claim to know how these things work, but if there are 1400 queued on Virtue surely that implies a much bigger number than what you're suggesting? After the crash not everyone would have tried to get back into the server at the same time, I know a number who just moved to another server and imagine others might have waited a moment or left for the night. I'd suggest the highest number seen in a queue would be a minimum for the amount in the server at max, with probably quite a few more actually there.

Either way though, I was really impressed with how busy it seemed to be. Every server I visited had enough people that it seemed to be buzzing, whatever the actual number of people was.

[RandomUnicorn]

That's quite unlikely.

Solid yellow dots for everyone but Vigilance, Zukunft, and Defiant (Defiant blipped yellow during a couple of my multitudes of character swaps). Virtue went red first, Freedom later in the evening. When Virtue crashed, each time there was between 1000-1400 in the queue. Judging by that (1000-1500 = red dots ... color didn't PRECISELY translate into population but it was pretty close), I would guess there was around 5,000, maybe 7,500 ... maybe 10,000-15,000 all told over the course of the evening/night.

Don't take me wrong. That's still bleepin' ridiculously awesome turnout. But going by financials, City had (best guess) 20-40k subscribers by the end -- and 70,000 former subscribers *DID NOT* show up last night to round out your number. All the servers would have grayed out from being full.

As others have pointed out, your numbers and assumptions are way off. As you can see below between 12,000 and 50,000 or more players were on at any one point in the final hours and it is quite possible that 150,000 or more players logged in on the final day.

According to NC Soft's own financial reports for June. 2004, CoH had "11 servers, highest concurrent users: 31,085, monthly access, 169,925"  Even if we assume that every server in 2004 was at 100% capacity that means each server could hold almost 3,000 players at any one time. In reality the number is probably far more . A far cry from your "1000 to 1500 people" estimate. Since that time CoH has added 5 servers for players to use (4 European and a Beta server) and a few years back I recall that a Dev mentioned the servers had been upgraded. That means that 50,000 or more people could have been logged on concurrently without graying out the servers, possibly much more.

Even if you ignore NCSofts own financial reports you can verify the minimum server capacity is far in excess of your estimates by several methods.

A single instance of AP holds 70 to 100 people. But don't take my word for it, you can check that at least 3 different ways.

1) Goto the thread "Call to Action: Unity Rally" and look at post 135 by Paula and you will see pictures of some of the Atlas Parks on the unity rally day and in several you can easily count 60+ people in the picture, and that does not include others who were doing missions, marketing, dancing under Atlas or doing Kinky Catgirl RPing in the basement of City Hall.

2) Check out this statement on the making of a youtube video about the protest "This is a crowd of 3700 heroes that gathered on various servers in protest of the closure of City of Heroes. This is a composite shot: we flew through over 50 instances of Atlas Park on various servers recording the heroes..." That works out to around 70 or so heroes on each instance of AP who were at the protest and once again does not count  those elsewher in the zone.

http://www.youtube.com/watch?v=XRsj2DtLCQs


3) Check posts made prior too September 1st for the number of people a single instance can hold and the estimates you will most commonly hear are 70-100 people. I can confirm this and have used the /search feature to find the number of people in a zone and my  personal experience is a new zone spawns at around 80-85 people. Keep in mind that my method will fail to include anyone who is hidden, so the number is probably larger.


Step 2
Now we know that Virtue spawned 33 full instances of AP so a bit of quick math reveals that on that September day the Atlas Parks in AP had between 33x70 = 2310 players and 33x 100 =3300 players. Considering there are several dozen zones besides AP most of which had people playing on that day as well as any people inside instanced missions and it is quite reasonable to suspect that several hundred maybe even a thousand or more people were on Virtue that day when it locked people out. As an example on a Saturday afternoon 700 player could easily be spread across the remainder of virtue with and average of 15 players per zone besides AP, 2 Itrials,  a half dozen regular trials and 100 players doing solo misions or 13 teams of 8 doing missions or any combo between. Perfectly normal fair for a Saturday on Virtue.

Given this information we can see that the low end number is 2310 players and only if we assume that EVERY player on Virtue was at the rally. and in reality the number was probably closer to 4000 or more players at the time it crashed.

Given those numbers we can quickly see that there were at least 15,000 and 45,000 people on concurrently on near the end of the final night.

Bare in mind that on the East Coast it was 3AM and players had gone to bed hours earlier. Others had logged off due to work, social engagements, family obligations, mom sending them to bed, there computer no longer able to handle the load, or as I saw in several cases the fact that players found it too emotionally difficult to stay on to the end.

I don't think a 3 to 1 turnover rate throughout the day is unreasonable leading to a total last day attendance of 50,000 to 150,000 players.

[OzonePrime]

 I know Guardian was jammed! People were complaining they couldn't get into AP. Was there for 10hrs, spent 2hrs, early afternoon, on Virtue. Taking screenshots! Have a shot taken seconds before "connection to mapserver lost", on Guardian. Still sorting 12hrs of screenshots.
Was good to see a lot of players i hadn't seen in a while. Several popped on just to say "Goodbye". They said it had been too depressing for them to try to play the last months, knowing it was going dark. Btw, thanks Hitstreak for the kracken(sp?) Great team play!
What is probably the saddest thing, is the amount of people I've talked to who had never heard about the game and would love to play it! No statistics on this to show Disney, unfortunately.(I'm having these people write letters, also.) I want our game back!

/em holdtorch

[Kheprera]

Keep in mind...

The number of people in the queue is the number of people trying to log in[/], it is not the number of people on a server.

[Rienuaa]

Oh indeed, but when a server suddenly dumps the vast majority of people who were logged in...

They only dump a few related zones at a time. Lag in AP wouldn't DC people in Steel, for example.

[Minotaur]

Number of subscribers pre-announcement 40k.  Number of pure f2p, 40k.  Number of freemium 20k.

(numbers verified)

And this was at a particularly low point, I had 5 friends who were unsubbed and not playing, but were about to resub for I24 and had told me so before the announcement and I'm sure this is not unusual.

[Aggelakis]

They only dump a few related zones at a time. Lag in AP wouldn't DC people in Steel, for example.

The Atlas Parks of Virtue crapped themselves once (I wasn't in AP at the time and didn't get kicked). The entirety of Virtue crapped itself twice. It never went 100% down [[EDIT: Until a couple minutes before the end]], so maybe a few people were able to stay on, but there weren't enough APs for 1500 people in the queue to be only in AP every time.

It is basically guaranteed that there was not 100,000 people there. It's very unlikely that 50,000 people were there. It's possible 25,000+ people were there. It is very likely that 15,000+ people were there. It's basically guaranteed that 10,000+ people were there.

[Profit]

Honestly.....don't you think the only reason the servers were this busy was due to game closure.  I can't remember the last time we had that many yellow dots, let alone a red one.

Too bad the only time the community rallied was to attend a funeral.

You may not have been around long enough to know this Tommy, and I don't if you had or not, but at one point the servers were always yellow and red. All of them. It caused a massive lag fest for a long time. Then, the devs implemented a load balancing technique, I forget exactly when but it may be on the wiki or maybe even someone here remembers it happening (I remember because I'm a networking guy and found it interesting but I'm awful with dates) and after that was implemented, the servers almost never got to red dots ever. So the point is, that just because they changed something and it took more people to get to yellow or red, it doesn't mean less people were playing the game or even that very many actually showed to, in your words, attend a funeral.

[Minotaur]

Honestly.....don't you think the only reason the servers were this busy was due to game closure.  I can't remember the last time we had that many yellow dots, let alone a red one.

Too bad the only time the community rallied was to attend a funeral.

Additional to what profit said, server capacity was IIRC upped just before freedom, so it was less likely servers would go red since then.

Also the 100K figure might be accurate for the night, just not at the end. Evidence from stuff I saw in chat channels would suggest that at least as many people logged on during the final few hours, but had logged before the end, as stayed for the switch off. Several people on GMT or GMT+1 in my SG were on in the Euro evening, but didn't unlike me either play all night or get up early on a Saturday morning.

[Triplash]

the devs implemented a load balancing technique, I forget exactly when

I believe that was a few months before Freedom went live. They ran a bunch of stress test events to help gather feedback to make those changes. There was even a badge for attending one of those events, "Defender of Primal Earth" iirc.

Edit: A quick teleport to the wiki tells me that, no, Defender of Primal Earth was not the badge associated with this. If there even was one, now I'm not sure.