You don't even realize his retort is EXACTLY what we were saying before do you?
He is saying what Mors did originally which is that GW2s peak happened early and around that time they had more like 1 million in sales as opposed to you saying that GW2 had a peak of 460k and has 3.2 million sales as if those peaks are happening now.
That all along is why we've been saying that you can't use the 3.2 million number and yet when this guy says it with a compliment to you at the end you decide to finally accept it.
To be even more clear, I don't have a dog in this fight as I could not possibly care less about FF and if it does good or if it does poorly. I simply say what the other guy was saying and his logic was more sound than yours so I was helping to point out what he was saying.
I got Mors (and your) point a LONG time ago, you never answered my challenges to it.
As I have said again (you must have missed it) if the concurrent users of GW2 during early days having sold 1 million was in the 400k range, a 40% concurrent to total player would only allow for FFXIV to have sold 360K copies, which is completely unbelievable - which is a point that was never addressed by Mors or yourself.
I did the math on Mors 1,000,000 estimate on boxes sold, which worked out to 18% concurrent and is close enough to my numbers to be pretty much the same argument.
Is it possible FFXIV has sold more like a million or more boxes, and has the capacity for only 180K, yet the desire for concurrent players is closer to the 40% range meaning that hundreds of thousands of players want to play at any given time but can't?
Yes, of course. Never said it wasn't.
Which is why I asked - does SE know how many players are stuck? If they do know, would they half ass their expansion to only let in a fraction more of them?
So now that you are caught up...
Here's the part you miss, I'm not trying to figure our copies sold because without them saying it, we can't.
My point all along has been that your logic was massively flawed and if you wanted to try to determine sales in the way that you were then you had to approach it the way Mors did. You didn't want to do that, not because it was illogical, but because you didn't like the number it represented which can only mean 2 things. One option would be that FF isn't doing well, the other option is that your approach doesn't work at all.
Our arguments were on the logic of your approach. Your's was "well that can't be cause I don't like the resulting number".
I will also toss on there that even if they raise it to 200k concurrent can be supported and there was a queue to get in, that wouldn't mean that 500-600k copies sold wasn't right. I wouldn't be the least bit surprised to see a new subscription games have 33-40% of the player base trying to play early on (especially when they can't get in so they are more likely to go idle while doing things that will keep them logged in so they don't have to sit in a queue when they do want to play).
I did the math on Mors 1,000,000 estimate on boxes sold, which worked out to 18% concurrent and is close enough to my numbers to be pretty much the same argument.
Is it possible FFXIV has sold more like a million or more boxes, and has the capacity for only 180K, yet the desire for concurrent players is closer to the 40% range meaning that hundreds of thousands of players want to play at any given time but can't?
Yes, of course. Never said it wasn't.
Which is why I asked - does SE know how many players are stuck? If they do know, would they half ass their expansion to only let in a fraction more of them?
The 13% you came up with was based on flawed numbers, and someone else did the math here based on the numbers provided by the eve website and came up with 8%, so your argument that 3 instances were all at 13% doesn't work... and 8% to 13% to 18% are actually significant variations.
Not only that, but GW2 had players who were going off-line, so that concurrent users mark was only players who were actively playing... whereas ffxiv has no afk-kick, and people are staying in indefinitely so as not to get locked out... so you have to include players who wouldn't normally be online in your concurrent users estimate... So the number of players actively trying to enter would be artificially elevated by the players who are in-game but not actively playing.
As for the rest, do they know? who knows? they have no queue system, so I am not sure how they would tell... maybe keep a log of requests per unique ID? I will tell you that their management decisions so far have been questionable as far as I am concerned, and there have been several times where they effectively came out and said that they were going to START taking metrics long after those metrics should have been necessary.
Originally posted by BadSpock
40% of say 1,000,000 total players trying to play = 400,000 concurrent players.
Maximum capacity of 200k after upgrades = 200k not able to play still.
50 servers, 3600 per server at 180K, to host 400k players server expansion would have to more than double to 8,000 per server (unlikely) or they'd have to add another 50 servers (also unlikely).
So is SE half assing the expansion / upgrades, so that only 1/2 or so of the effected 1017 / queue players can play the game?
If you are making an order for new hardware... why stop at 200k if that is true?
Assumptions:
SE can see the numbers of players able to authenticate, but unable to log into a world.
SE can see the numbers who fail at authentication.
SE can see the numbers who fail to load into an instance after loading into the world, and/or are disconnected when trying.
A couple of things you may be missing... We opened a new datacenter in California last year, and ordered 3 dozen servers... they only had 8 in-channel, the rest of them took 3-4 weeks to get. I am not sure who they are going through, but assuming they have a fixed config that they use, it really depends on how many they have in-channel. So that could be why they are stopping at 200k, they just can't get enough hardware in time.
Or it could be another one of their brilliant administrative decisions, like deciding not to test major fixed for problems affected >20% of the userbase.
Originally posted by BadSpock As I have said again (you must have missed it) if the concurrent users of GW2 during early days having sold 1 million was in the 400k range, a 40% concurrent to total player would only allow for FFXIV to have sold 360K copies, which is completely unbelievable - which is a point that was never addressed by Mors or yourself.
Actually, Mors pretty much hit that one in his/her initial post: GW2 allowed people to log in via overflow servers; ARR does not. That will introduce a significant difference in the numbers, since GW2 only would have disallowed logins once ~all~ the servers were full, whereas ARR will disallow a login whenever any specific server is full.
That throws off the concurrent number, as well as the number who are being turned away. 0.3% sounds extraordinarily low for as much press/flak/attention as the 1017 error is getting (and from my own personal observations at getting 1017'ed).
I don't doubt the GW2 number, but to compare that to ARR concurrent users you have to make some assumption based on the every server/any server problem, and it's definitely not a 1:1 correlation. That's why your numbers "feel" off when you try to account for that.
I'm seeing a lot of contempt for Yoshi and his development crew too over recent events. Let's know who to aim for before we start firing.
I'd bet dollars to doughnuts (or gil to whatever the Final Fantasy equivalent to doughnuts is) that it was a corporate decision to have the number of servers and the capacity that they can handle.
Because let's think about this from a financial standpoint. FFXIV didn't do that hot. I doubt that SE corporate executives wanted to take a big risk so they cut the budget towards the servers prior to launch because they didn't forcast so many people purchasing the game.
Ideally, 1000 servers with 30k people on each is the target...but that's ludicrous expensive and a massive surplus. So you find a reasonable amount of servers, in this case 50. You find a reasonable amount of people per server that keeps things stable, in this case 3800 give or take. And you hope to god that things don't go pearshaped either way.
They made a "good" business decision by keeping their expectations low. If the game didn't sell that well, it was a low risk. But because the game sold far better than they thought, the customers lose and turns out that low risk came with a high cost after all. That cost is customer satisfaction.
I'd bet dollars to doughnuts (or gil to whatever the Final Fantasy equivalent to doughnuts is) that it was a corporate decision to have the number of servers and the capacity that they can handle.
I agree that all of the problems are based on corporate decisions...
At the end of B4 they put in the fix for all the network problems they were having, and then ended the beta. Do you really think that Yoshi couldn't have said "let's run another 24 hours and make sure this is actually fixed?" and that is just one example. As for the devs, how about the fact that they effectively said that they had not taken performance metrics per user beforehand? Why didn't they know how many resources each client connection required(i.e. bandwidth, iops, ram)? No one on the dev team thought to do that, no one on the IT team thought to do that, and no one in management thought to ask about it?
Originally posted by Mors-Subita Just FYI, VGChartz is showing 720k sales... And from my experience their figures are usually about half of total sales.That would make 1.45 million units...
Ok, I went and looked at the VGCHartz numbers - those are almost entirely 1.0 sales. It only lists 62k units globally in sales in 2013.
That may be correct for the 1.0 SKU (people picking it up on clearance to use for ARR, for example), but we can pretty safely assume that doesn't include recent ARR sales.
I seem to recall hearing that 1.0 sales were right around 1M before it went totally off the rails and imploded. All of those people count as an ARR box sale - because they have access to the game. Now subtract out all the people who have both the PC and PS3 versions, because those are duplicate accounts. Correct for the people who mistakenly made duplicate service accounts because of the error in MogStation and PS3 codes. Then add in the actual digital and box sales for ARR (both PC and PS3).
There are a lot of variables in there - unless S/E releases some figures (they may have, I looked, but google is a bit overrun with news of the digital sales suspension as of now), I don't know that we will be able to get a good number for actual box sales - aside from the fact that we know
Comments
Here's the part you miss, I'm not trying to figure our copies sold because without them saying it, we can't.
My point all along has been that your logic was massively flawed and if you wanted to try to determine sales in the way that you were then you had to approach it the way Mors did. You didn't want to do that, not because it was illogical, but because you didn't like the number it represented which can only mean 2 things. One option would be that FF isn't doing well, the other option is that your approach doesn't work at all.
Our arguments were on the logic of your approach. Your's was "well that can't be cause I don't like the resulting number".
I will also toss on there that even if they raise it to 200k concurrent can be supported and there was a queue to get in, that wouldn't mean that 500-600k copies sold wasn't right. I wouldn't be the least bit surprised to see a new subscription games have 33-40% of the player base trying to play early on (especially when they can't get in so they are more likely to go idle while doing things that will keep them logged in so they don't have to sit in a queue when they do want to play).
The 13% you came up with was based on flawed numbers, and someone else did the math here based on the numbers provided by the eve website and came up with 8%, so your argument that 3 instances were all at 13% doesn't work... and 8% to 13% to 18% are actually significant variations.
Not only that, but GW2 had players who were going off-line, so that concurrent users mark was only players who were actively playing... whereas ffxiv has no afk-kick, and people are staying in indefinitely so as not to get locked out... so you have to include players who wouldn't normally be online in your concurrent users estimate... So the number of players actively trying to enter would be artificially elevated by the players who are in-game but not actively playing.
As for the rest, do they know? who knows? they have no queue system, so I am not sure how they would tell... maybe keep a log of requests per unique ID? I will tell you that their management decisions so far have been questionable as far as I am concerned, and there have been several times where they effectively came out and said that they were going to START taking metrics long after those metrics should have been necessary.
A couple of things you may be missing... We opened a new datacenter in California last year, and ordered 3 dozen servers... they only had 8 in-channel, the rest of them took 3-4 weeks to get. I am not sure who they are going through, but assuming they have a fixed config that they use, it really depends on how many they have in-channel. So that could be why they are stopping at 200k, they just can't get enough hardware in time.
Or it could be another one of their brilliant administrative decisions, like deciding not to test major fixed for problems affected >20% of the userbase.
Actually, Mors pretty much hit that one in his/her initial post:
GW2 allowed people to log in via overflow servers; ARR does not. That will introduce a significant difference in the numbers, since GW2 only would have disallowed logins once ~all~ the servers were full, whereas ARR will disallow a login whenever any specific server is full.
That throws off the concurrent number, as well as the number who are being turned away. 0.3% sounds extraordinarily low for as much press/flak/attention as the 1017 error is getting (and from my own personal observations at getting 1017'ed).
I don't doubt the GW2 number, but to compare that to ARR concurrent users you have to make some assumption based on the every server/any server problem, and it's definitely not a 1:1 correlation. That's why your numbers "feel" off when you try to account for that.
I'm seeing a lot of contempt for Yoshi and his development crew too over recent events. Let's know who to aim for before we start firing.
I'd bet dollars to doughnuts (or gil to whatever the Final Fantasy equivalent to doughnuts is) that it was a corporate decision to have the number of servers and the capacity that they can handle.
Because let's think about this from a financial standpoint. FFXIV didn't do that hot. I doubt that SE corporate executives wanted to take a big risk so they cut the budget towards the servers prior to launch because they didn't forcast so many people purchasing the game.
Ideally, 1000 servers with 30k people on each is the target...but that's ludicrous expensive and a massive surplus. So you find a reasonable amount of servers, in this case 50. You find a reasonable amount of people per server that keeps things stable, in this case 3800 give or take. And you hope to god that things don't go pearshaped either way.
They made a "good" business decision by keeping their expectations low. If the game didn't sell that well, it was a low risk. But because the game sold far better than they thought, the customers lose and turns out that low risk came with a high cost after all. That cost is customer satisfaction.
I agree that all of the problems are based on corporate decisions...
At the end of B4 they put in the fix for all the network problems they were having, and then ended the beta. Do you really think that Yoshi couldn't have said "let's run another 24 hours and make sure this is actually fixed?" and that is just one example. As for the devs, how about the fact that they effectively said that they had not taken performance metrics per user beforehand? Why didn't they know how many resources each client connection required(i.e. bandwidth, iops, ram)? No one on the dev team thought to do that, no one on the IT team thought to do that, and no one in management thought to ask about it?
Just FYI, VGChartz is showing 720k sales... And from my experience their figures are usually about half of total sales.
That would make 1.45 million units...
Ok, I went and looked at the VGCHartz numbers - those are almost entirely 1.0 sales. It only lists 62k units globally in sales in 2013.
http://www.vgchartz.com/game/35106/final-fantasy-xiv-a-realm-reborn/Global/
That may be correct for the 1.0 SKU (people picking it up on clearance to use for ARR, for example), but we can pretty safely assume that doesn't include recent ARR sales.
I seem to recall hearing that 1.0 sales were right around 1M before it went totally off the rails and imploded. All of those people count as an ARR box sale - because they have access to the game. Now subtract out all the people who have both the PC and PS3 versions, because those are duplicate accounts. Correct for the people who mistakenly made duplicate service accounts because of the error in MogStation and PS3 codes. Then add in the actual digital and box sales for ARR (both PC and PS3).
There are a lot of variables in there - unless S/E releases some figures (they may have, I looked, but google is a bit overrun with news of the digital sales suspension as of now), I don't know that we will be able to get a good number for actual box sales - aside from the fact that we know