Maintenance for the week of December 16:
• PC/Mac: No maintenance – December 16
• NA megaservers for patch maintenance – December 17, 4:00AM EST (9:00 UTC) - 12:00PM EST (17:00 UTC)
• EU megaservers for patch maintenance – December 17, 9:00 UTC (4:00AM EST) - 17:00 UTC (12:00PM EST)
• PC/Mac: No maintenance – December 16
• NA megaservers for patch maintenance – December 17, 4:00AM EST (9:00 UTC) - 12:00PM EST (17:00 UTC)
• EU megaservers for patch maintenance – December 17, 9:00 UTC (4:00AM EST) - 17:00 UTC (12:00PM EST)
Analysis: the Cyrodiil test is not technical, it is behavioral
furiouslog
✭✭✭✭✭
✭✭
✭✭
Something has really been bugging me about the design of the current Cyrodiil AOE experiment, and it derives from trying to figure out how ZOS is going to use these data as a means to understanding if these combat changes will improve Cyrodiil performance.
Some assumptions:
1. ZOS already knows the practical effect of the cooldown changes on server load. They have access to server calculation data, and they know how their code operates, and they already have access to vast amounts of actual combat data provided it were logged. It would be possible to take two hours of a high pop instance, parse the actions sent to the server, and simulate the outcome if they changed the cooldowns by replacing or reorganizing certain castable skills under a logical substitution regime. I suspect that they already used data like this to design their experiment in the first place. There is no reason to actually run a test at all because they can simulate the outcome on server performance.
2. ZOS is aware that some players will not participate in parts of the test (e.g. templars) due to the effects of the individual test conditions, and that this fact would influence the server performance benchmarks, therefore comprising them.
So, taken these as a given, what is this test actually attempting to measure? Theoretical Cyrodiil performance will obviously be better under the global 3 second cooldown because server load will be reduced the most in that scenario. It will also possibly be reduced because, hypothetically, the largest number of people will either choose not to play or will use different characters that are playable under those conditions.
Because they already have an idea of the practical effects of the changes on server load due to the ability to simulate those conditions, and since they know will not be able to get a true benchmark for the testing scenarios due to the opt outs which render the data meaningless as a technical measure, what they are probably actually testing is the consumer population behavioral reaction to the changes. From the test they will be able to learn:
1. Who stops playing in Cyrodiil entirely under the 4 testing scenarios even when provided an incentive (AP increase) to play more.
2. What changes are there in time spent in Cyrodiil by each player compared to their baseline time spent.
3. Who makes alterations, and which alterations, to their "main" PvP build in order to be viable given the scenario.
4. Who brings an alt into Cyrodiil to play, which alts, and under which scenarios.
5. Behavioral changes of grouped players under different scenarios.
From this, they can derive which of the scenarios is most viable from a consumer acceptance standpoint and compare that with the performance benefit of those changes to provide estimates for scenario-specific player participation.
Typically, in order to perform behavioral research on consumers, there are strict legal considerations about the collection of data that are associated with PII, the definition of which and conditions of use were updated today in conjunction with the introduction of the test.
So what does this all mean?
1. ZOS is actually performing a psychological test that is publicly framed as a technical test, which is legal given the T&Cs, but possibly unethical without explicit and transparent consent (a point that is arguable, and in full transparency I did not do a thorough review of the EULA and PII but I'm assuming their lawyers did).
2. If you are unhappy with the changes under a particular test condition, make sure to opt out of participation realistically during the next four weeks to ensure ZOS gets an appropriate measure of the effect on player participation under the 4 scenarios.
Some might interpret this as a "conspiracy theory". It's not. It's an analysis of experimental design intended to determine the purpose of the experiment - anyone is certainly free to contradict my assumptions or attack my reasoning and conclusions.
Is it evil? I don't think so, but it is sketchy. Consumer research happens all of the time, and it's a necessary feedback loop that provides data necessary to support the health of companies and their products. I also understand why a test like this would be necessary, because players, if asked directly which scenario they would want, would respond with "no changes". This design lets ZOS figure out who will actually change their behavior with respect to participation in Cyrodiil and determine if any of the scenarios do the least damage to participation while improving performance, which increases broader player participation and possible recapture of players who became frustrated and left the game due to Cyrodiil performance. So if I am right, I guess my recommendation would be to act accordingly to make sure that ZOS has the data that leads them to the proper conclusion.
Some assumptions:
1. ZOS already knows the practical effect of the cooldown changes on server load. They have access to server calculation data, and they know how their code operates, and they already have access to vast amounts of actual combat data provided it were logged. It would be possible to take two hours of a high pop instance, parse the actions sent to the server, and simulate the outcome if they changed the cooldowns by replacing or reorganizing certain castable skills under a logical substitution regime. I suspect that they already used data like this to design their experiment in the first place. There is no reason to actually run a test at all because they can simulate the outcome on server performance.
2. ZOS is aware that some players will not participate in parts of the test (e.g. templars) due to the effects of the individual test conditions, and that this fact would influence the server performance benchmarks, therefore comprising them.
So, taken these as a given, what is this test actually attempting to measure? Theoretical Cyrodiil performance will obviously be better under the global 3 second cooldown because server load will be reduced the most in that scenario. It will also possibly be reduced because, hypothetically, the largest number of people will either choose not to play or will use different characters that are playable under those conditions.
Because they already have an idea of the practical effects of the changes on server load due to the ability to simulate those conditions, and since they know will not be able to get a true benchmark for the testing scenarios due to the opt outs which render the data meaningless as a technical measure, what they are probably actually testing is the consumer population behavioral reaction to the changes. From the test they will be able to learn:
1. Who stops playing in Cyrodiil entirely under the 4 testing scenarios even when provided an incentive (AP increase) to play more.
2. What changes are there in time spent in Cyrodiil by each player compared to their baseline time spent.
3. Who makes alterations, and which alterations, to their "main" PvP build in order to be viable given the scenario.
4. Who brings an alt into Cyrodiil to play, which alts, and under which scenarios.
5. Behavioral changes of grouped players under different scenarios.
From this, they can derive which of the scenarios is most viable from a consumer acceptance standpoint and compare that with the performance benefit of those changes to provide estimates for scenario-specific player participation.
Typically, in order to perform behavioral research on consumers, there are strict legal considerations about the collection of data that are associated with PII, the definition of which and conditions of use were updated today in conjunction with the introduction of the test.
So what does this all mean?
1. ZOS is actually performing a psychological test that is publicly framed as a technical test, which is legal given the T&Cs, but possibly unethical without explicit and transparent consent (a point that is arguable, and in full transparency I did not do a thorough review of the EULA and PII but I'm assuming their lawyers did).
2. If you are unhappy with the changes under a particular test condition, make sure to opt out of participation realistically during the next four weeks to ensure ZOS gets an appropriate measure of the effect on player participation under the 4 scenarios.
Some might interpret this as a "conspiracy theory". It's not. It's an analysis of experimental design intended to determine the purpose of the experiment - anyone is certainly free to contradict my assumptions or attack my reasoning and conclusions.
Is it evil? I don't think so, but it is sketchy. Consumer research happens all of the time, and it's a necessary feedback loop that provides data necessary to support the health of companies and their products. I also understand why a test like this would be necessary, because players, if asked directly which scenario they would want, would respond with "no changes". This design lets ZOS figure out who will actually change their behavior with respect to participation in Cyrodiil and determine if any of the scenarios do the least damage to participation while improving performance, which increases broader player participation and possible recapture of players who became frustrated and left the game due to Cyrodiil performance. So if I am right, I guess my recommendation would be to act accordingly to make sure that ZOS has the data that leads them to the proper conclusion.
11
-
Edaphon✭✭✭✭✭
✭✭✭furiouslog wrote: »They have access to server calculation data, and they know how their code operates, [...]
That's where I stopped reading.
13 -
caperon✭✭✭✭"They have access to server calculation data, and they know how their code operates"
You are extremely optimistic about ZOS ability to do anything other than put recolored horses on the crown store.7 -
Magdalina✭✭✭✭✭
✭✭✭If my years here have taught me anything, it's that ZOS most certainly do NOT know how their code operates, so I wouldn't go by that assumption if I were you.
My favourite is still when they completely broke (it didn't fire, at all) Overload while fixing infinite dodge rolling
I'm actually really curious about the testing outcome, I mean, at least they're really doing SOMETHING for Cyro performance finally. That's not such a terrible thing imo.1 -
Pyvos✭✭✭✭"Theoretical Cyrodiil performance will obviously be better under the global 3 second cooldown because server load will be reduced the most in that scenario."
Except here's the thing, I've spent the better part of several hours in Standard No-CP Ravenwatch on PC EU, and it isn't any better. If anything, the poor performance is even further emphasized, because now not only is the server not able to handle my abilities within a reasonable timeframe, but now the game is literally preventing me from reacting.Edited by Pyvos on September 8, 2020 5:05PM3 -
furiouslog✭✭✭✭✭
✭✭furiouslog wrote: »They have access to server calculation data, and they know how their code operates, [...]
That's where I stopped reading.
Cynical comments about ZOS's collective understanding of their own software design aren't really helpful, as amusing as they might be. I certainly understand the cynicism given that Cyrodiil performance issues have persisted for the past 6 years.
That cynicism stems from a clear desire to have a better gaming experience and ZOS's failure to meet that desire. You are clearly a seasoned player, given your forum rank. The enjoyment you derive from the game still apparently outweighs its limitations to the point where you make a sizable contribution of time to its community by posting on this forum, from which I infer you are an active player. So whatever problems you think the game has, none of those problems were bad enough for you to quit the game and stop spending time on this forum, which means that if you truly believe that they don't understand how their own code works and performs, you are not providing them with any incentive to learn it, because they are still running a viable business without requiring the level of understanding you appear to expect. As such, your opinion on their competence regarding their code is ultimately a moot point because it has done nothing to change your behavior, which is pretty much the point of this thread.4 -
ZOS_ConnorGmodGreetings all,
After review we have closed this thread as it violates our Community Rules for Conspiracy Theories. We do no allow the spreading of conspiracy theories of misinformation on the forums.
You are welcome to review the Community Rules here.4
This discussion has been closed.