What I’d really like to see are transfer counts, including where most of the transfers off a given server went to, correlated with which servers were matched with which per time period, and who won that week.
I.e., I’d like to see data that supports or refutes the idea that the majority of transfers are from losing to winning servers.
The last data I saw had an overwhelming majority of transfers from servers with long queu times (generally from winning servers), to servers with short queu times (generally to losing servers). There are some people going to the winning servers, but the simple truth is the vast majority are not.
That being said, we are working on long term improvements to queus, transfers, and WvW content in general. We’ll talk more about this in the future as features/plans get closed to being finalized.
The last data I saw had an overwhelming majority of transfers from servers with long queu times (generally from winning servers), to servers with short queu times (generally to losing servers). There are some people going to the winning servers, but the simple truth is the vast majority are not.
That being said, we are working on long term improvements to queus, transfers, and WvW content in general. We’ll talk more about this in the future as features/plans get closed to being finalized.
Well i’ve got some news for you “Colin” if that is your real name. I transferred from a losing server to a winning server, and that makes it the majority because this is the internet.
The last data I saw had an overwhelming majority of transfers from servers with long queu times (generally from winning servers), to servers with short queu times (generally to losing servers). There are some people going to the winning servers, but the simple truth is the vast majority are not.
That being said, we are working on long term improvements to queus, transfers, and WvW content in general. We’ll talk more about this in the future as features/plans get closed to being finalized.
On that note, any news on the Guesting system? Because if I understand correctly, it is the whole reason the free unlimited transfers exists in the first place.
Thanks for the response, Colin. I appreciate it. That could help forestall some arguments that pop up regarding server populations and why servers keep losing. The common argument is that the server has too few people remaining to field enough people for WvWvW to hit the cap, let alone build a queue because all their people leave for better servers that win more.
If in truth the majority are moving to lower-queue-time servers and not just jumping to whichever server trounced them that week, then the common argument would appear to be wrong, based on actual data rather than conjecture. At least, the reason typically given for the lack of WvWvW participation is incorrect.
Again, thank you. It’s always a welcome sight when an ArenaNet employee responds with useful information.
The last data I saw had an overwhelming majority of transfers from servers with long queu times (generally from winning servers), to servers with short queu times (generally to losing servers). There are some people going to the winning servers, but the simple truth is the vast majority are not.
That being said, we are working on long term improvements to queus, transfers, and WvW content in general. We’ll talk more about this in the future as features/plans get closed to being finalized.
If you read between the lines he basically just said that the whole WvWvW point system will be eliminated and as soon as your server is getting beat too badly you will then log in as a dragon and YOU will be a DE for the enemy team to try and kill.
AND
You will still be able to do the jumping puzzle but at the end the chest will contain a big bad mother golem that will come out and give you a spanking for spending too much time on yourself and not helping your team. (You will also get one badge of honor)
Queue size data. I can give exact queue data for Vabbi (EU) for the last 7 days. Waiting times = 0.0 seconds. lol
Even at peak time (8pm) on Saturday and Sunday I could enter Eternal Battlegrounds instantly. Granted, the server was being comprehensively beaten due to very good nightcapping at first by Ring of Fire over the last week and by the end of it, RoF held most of the maps constantly. Don’t know how much we pulled back though. But if more people came to servers like ours to get away from the queues, then we could put up more of a fight.
I’ve also been wanting this data-set updated because those older graphs in no way reflect what is going on with queues anymore.
And as much as everyone hates the idea, there are going to be alot of people transferring around in the next week. It’d be better to have these people wanting to move to lower queue times rather than just picking the next winning server due to lack of information.
The last data I saw had an overwhelming majority of transfers from servers with long queu times (generally from winning servers), to servers with short queu times (generally to losing servers). There are some people going to the winning servers, but the simple truth is the vast majority are not.
That being said, we are working on long term improvements to queus, transfers, and WvW content in general. We’ll talk more about this in the future as features/plans get closed to being finalized.
This is a specious response… at best. “generally from winning servers”. The real information we want is not how individual people transfering are, but filtering out the noise to find the signal. One only has to look at the Blackgate performance 2 days ago versus yesterday to realize that it is not about where the majority of people are transferring to and whether those servers are “winning”, but the number of wvw battles that are being impacted by blocks of people who play certain times transferring together. Blackgate wasn’t a “winning” server when the mass transfers happened. So yes… I am sure you aren’t lying about the statistic. But they are winning now by a large margin and THAT is the issue. Thank you for pretending not to see it.
I know it is common for developers and their forum mouthpieces to think their users are too stupid to notice logical fallacies in arguments… but this one is pretty glaring. To even attempt to make the argument that free transfers aren’t going to impact the top tier or 3 wvw battles is to completely ignore some pretty obvious evidence to the contrary…. and is sufficiently condescending that I would have stopped paying for this game if you had a monthly fee.
Most of the people I know have already stopped playing.
I think we need a little more explanation on this “transfer population’s” distribution.
Example:
100 people transfer from HoD to Kaineng
100 people transfer from JQ to Crystal Desert
100 people transfer from SBI to Blackgate
200 people transfer from Kaineng/Crystal Desert/Blackgate to HoD.
So in this example, a vast majority of players are transfering off high population servers, yet one of the most populated servers actually increases in population because more people have that as their destination.
I think it’s time to forcibly move players that have transferred to some of the highest population servers after X date. That will take care of the leechers, reduce populations and queue times.
Give these players two weeks notice so they have the opportunity to transfer to a lower population server of their choice. Once transferred off, lock out their ability to transfer back to that server.
Honestly, I don’t see any other option at this point beyond releasing more WvW maps to field a bigger population.
I think we need a little more explanation on this “transfer population’s” distribution.
Example:
100 people transfer from HoD to Kaineng
100 people transfer from JQ to Crystal Desert
100 people transfer from SBI to Blackgate
200 people transfer from Kaineng/Crystal Desert/Blackgate to HoD.
So in this example, a vast majority of players are transfering off high population servers, yet one of the most populated servers actually increases in population because more people have that as their destination.
Interesting point of view. I think this might be the real matter here.
Core 2 Duo E6850 @ 3.5 GHz – 4GB RAM – GTX 470
Windows 8 Professional
This is a specious response… at best. “generally from winning servers”. The real information we want is not how individual people transfering are, but filtering out the noise to find the signal. One only has to look at the Blackgate performance 2 days ago versus yesterday to realize that it is not about where the majority of people are transferring to and whether those servers are “winning”, but the number of wvw battles that are being impacted by blocks of people who play certain times transferring together. Blackgate wasn’t a “winning” server when the mass transfers happened. So yes… I am sure you aren’t lying about the statistic. But they are winning now by a large margin and THAT is the issue. Thank you for pretending not to see it.
I know it is common for developers and their forum mouthpieces to think their users are too stupid to notice logical fallacies in arguments… but this one is pretty glaring. To even attempt to make the argument that free transfers aren’t going to impact the top tier or 3 wvw battles is to completely ignore some pretty obvious evidence to the contrary…. and is sufficiently condescending that I would have stopped paying for this game if you had a monthly fee.
Most of the people I know have already stopped playing.
I’m not familiar with this situation at all. I’d simply like to point out a few things. You blast Colin for logical fallacies while providing nonsensical anecdotal “evidence” for your barely coherent point.
Colin was pointing out that the majority of transfers came off of high que time servers which tend to be winning servers. High que times correlate with wins and/or high tier status. He also pointed out that the destination servers tended to be low que time servers which tend to be losing and/or low tier status servers. Seems pretty straight forward to me.
If you have this data, what’s the harm in being transparent with it. Transparency would go along ways to alleviate the frustration of perceived in-action on these core WvW issues.
Fine, if you want to protect your total server numbers as industry secrets, but publish Deltas ( aka Population transfers in, population transfers out – Delta / Velocity ).
To tag onto the comments above, when a really close match-up is changed dramatically by the perception of transfer ‘funny business’ ( switching sides, massive influx etc ) this is frustrating your customer and will result in the loss of active users and a decline of ARPU.
@ColinJohanson: There is an other issue. One way to avoid Queues is to move to a different regions server that wont have queues during your primetime. However no matter that its no queues at all, in any zone, during EU primetime most US servers are full and thus dont accept transfers? Fix this and you will see a lot of evening out of populations.
General rule of transfer for activev WvW guilds: You want to avoid queues when you transfer but you also want a server with an active community that isnt a graveyard and still has a shot at good organised opposition in WvW (That means not just pugs with randoms).
If you go for a server in the same region as yourself that fills above said kriteria, you will find none as EU servers that has a chance on a good matchup and an active community are quickly attracting people during EU primetime that they get queued to all hells. Low pop servers has no shot at fulfilling this either as they have no chance in climbing the rankings to get a matchup against organised opposition = no good fights.
However moving between regions fixes this. No queues during EU prime on US server and vice versa, organised opposition (dont go for low pop servers as that means PvDoor though) great community and so on and the original inhabitants of the server is happy to have the cover of their out of prime timeslot. So a Win Win situation for all.
So make this “server is full” autobalance mechanic see the differance of timezones and actual serverload so guilds can move between regions to even out the populations on a server more around the clock. Would also limit the “night capping” complaints as more servers would get coverage.
The last data I saw had an overwhelming majority of transfers from servers with long queu times (generally from winning servers), to servers with short queu times (generally to losing servers). There are some people going to the winning servers, but the simple truth is the vast majority are not.
This is consistent with the theory that players in overpopulated servers are switching to servers against whom they are currently matched to spy on and sabotage them.
The last data I saw had an overwhelming majority of transfers from servers with long queu times (generally from winning servers), to servers with short queu times (generally to losing servers). There are some people going to the winning servers, but the simple truth is the vast majority are not.
This is consistent with the theory that players in overpopulated servers are switching to servers against whom they are currently matched to spy on and sabotage them.
The number of accounts transfering to spy would be statistically negligible in this data set. I’m not saying its not happening, it definitely is. This data wouldn’t support that necessarily though.
Well yes, but common sense does. Guilds aren’t sending more than 1 or 2. We can probalby assume that most of the higher tier servers are doing it and for the most part negating the others. Its cheap and unsportsmanlike, but I don’t think we’ll see that traffic stand out in the collected data of all transfers.
This is a specious response… at best. “generally from winning servers”. The real information we want is not how individual people transfering are, but filtering out the noise to find the signal. One only has to look at the Blackgate performance 2 days ago versus yesterday to realize that it is not about where the majority of people are transferring to and whether those servers are “winning”, but the number of wvw battles that are being impacted by blocks of people who play certain times transferring together. Blackgate wasn’t a “winning” server when the mass transfers happened. So yes… I am sure you aren’t lying about the statistic. But they are winning now by a large margin and THAT is the issue. Thank you for pretending not to see it.
I know it is common for developers and their forum mouthpieces to think their users are too stupid to notice logical fallacies in arguments… but this one is pretty glaring. To even attempt to make the argument that free transfers aren’t going to impact the top tier or 3 wvw battles is to completely ignore some pretty obvious evidence to the contrary…. and is sufficiently condescending that I would have stopped paying for this game if you had a monthly fee.
Most of the people I know have already stopped playing.
I’m not familiar with this situation at all. I’d simply like to point out a few things. You blast Colin for logical fallacies while providing nonsensical anecdotal “evidence” for your barely coherent point.
Colin was pointing out that the majority of transfers came off of high que time servers which tend to be winning servers. High que times correlate with wins and/or high tier status. He also pointed out that the destination servers tended to be low que time servers which tend to be losing and/or low tier status servers. Seems pretty straight forward to me.
I have to ask out of curiosity… if you don’t have the slightest clue what the situation I am referring to is, and that obviously causes you to not even remotely understand the point, why are you responding? I mean the obvious answer is “fanboi” but I figured I’d ask. I guarantee you Colin gets the point.
Here it is in short form for you:
allowing free server transfers, midweek, mid wvw “scoreboard” counting allows for people to freely flee to servers that are not necessarily winning and push them well ahead eventually causing anyone who wants to be competitive in scoreboard watching to eventually localize to a few well-covered servers. this will particularly be the case for concerted night time/early morning players transferring to otherwise competent realms who are already competitive in their current bracket (see: gate, black).
for your reference, blackgate went from being 20k behind after the first few days of the week to being 40k ahead in 1 day following a string of coordinated transfers from late nighters.
@Beld Thanks for the clarification. No, I was not responding out of fanboi impulse. I responded to that not because I disagreed with the point you made. I repsonded due to the “logical fallacies” comment. I subscribe to the “skeptical” world view and when people start throwing phrases like “logical fallacies” I usually expect there to be clear evidence brought forth by that person. In the context of JUST that one post, I had nearly no idea what you were trying to say. Also, the evidence brought forth was anecdotal. Your situation and Colin’s data may very well both be true and your experience is in the minority set of data. To straight up call him a liar and jump off the deep end into conspiracy theories is not justified.
Asking for the data is one thing, and I would love to see it.
Edit: As a side note, does the bottom tool bar of each post kinda work on and off for others too? Sometimes I can quote, other times I can’t.
Asking for the data is one thing, and I would love to see it.
I would also like to see that. However the skeptic in me expects that it does not show anything to the scale that Colin is alluding to. If there were really an exodus from the high pop servers to low pop ones this week’s match in the bottom six would have gone differently.
Furthermore what is classified as a winning server? A server which is currently number one in score or a server which won its match last week? You could possibly have a situation if it is the latter where a server won in a lower tier and met a server which lost in a higher one, the resulting migration between lower ranked ‘winning’ server and higher ranked ‘losing’ server would match his claim but be constant with the general theory of constant upward migration.
Finally, is it just me or is it surprising and disconcerting that Colin misspelled queue twice in his post? Or is that just some strange American spelling?
Asking for the data is one thing, and I would love to see it.
I would also like to see that. However the skeptic in me expects that it does not show anything to the scale that Colin is alluding to. If there were really an exodus from the high pop servers to low pop ones this week’s match in the bottom six would have gone differently.
Furthermore what is classified as a winning server? A server which is currently number one in score or a server which won its match last week? You could possibly have a situation if it is the latter where a server won in a lower tier and met a server which lost in a higher one, the resulting migration between lower ranked ‘winning’ server and higher ranked ‘losing’ server would match his claim but be constant with the general theory of constant upward migration.
Finally, is it just me or is it surprising and disconcerting that Colin misspelled queue twice in his post? Or is that just some strange American spelling?
My interpretation of “winning” is “successful.” ET might come in 3rd place every other week, but its a very successful server. His direct data is in relation to queue times which he associated with “winning” and “losing.”
Edit: After some additional thought, high que times would also correlate with short term success as people like to participate in a winning effort. I guess further clarification could be used here.
Also, Americans hardly ever use the word “queue” outside of the MMO world. It is a british and possibly canadian word that has slowly started to bleed over into the US via the internet. I probably would have mispelled it as well.
My interpretation of “winning” is “successful.” ET might come in 3rd place every other week, but its a very successful server. His direct data is in relation to queue times which he associated with “winning” and “losing.”
There was really not enough information in the single sentence he wrote to properly infer what he meant. I was simply suggesting a situation which was more or less true to what he said, but actually showed a very different situation to what he was claiming.
We can interpret and imply until the sun rises, but as it stands now with the liberal use of “generally” and other non-specific nebulous terms being thrown around, this discussion has probably reached its limit. At least until specific data is released. Which skeptic me would be somewhat surprised to see.
