City Winner History?

[Ramlaen]

I just hope the folks analyzing the times of day when Order loses cities by a larger margin see the correlation between population allowing Order to run more competetive warbands and Order having less flexibility in their competetive warband setups.

[Valarion]

Yea it seemes to be 3-4 hours before EU primetime, or roughly 3-4 hours after EU primetime is where we're getting these hugely biased cities for destro. Not sure what we can infer from that, but it's noteworthy I think.

Late night NA Destro is simply terrible, all days of the week, Terrible isnt even the word for it, its simply craptastic really. This is the best time to level my Order alts btw.

[Foofmonger]

Yea it seemes to be 3-4 hours before EU primetime, or roughly 3-4 hours after EU primetime is where we're getting these hugely biased cities for destro. Not sure what we can infer from that, but it's noteworthy I think.

Late night NA Destro is simply terrible, all days of the week, Terrible isnt even the word for it, its simply craptastic really. This is the best time to level my Order alts btw.

That's a good reason why we aren't seeing these unbalanced cities at late night NA time then.

So what about mid-afternoon EU and primetime NA are leading towards this issue I wonder? If I had to take a guess:

1. Primetime NA Destro is more organized than primetime NA Order (a guess of course).
2. When the awful late night NA destro log off, and the pop is low (NA really isn't on), but it's pre-EU primetime, we also see greater destro organization.

This leads me to believe that what is causing the Order win rate to be less wonky, is EU Primetime Order. The one time that we can discern that we don't see this bias towards destro in the city win rates, seems to be during EU-primetime. That would indicate to me that the organization between Order//Destro is at it's lowest in EU Primtime, and at its highest level of disparity during "NA primetime", and "NA offtimes". This could also be indicative in the inverse, of an issue with NA order in general.

[wargrimnir]

I just hope the folks analyzing the times of day when Order loses cities by a larger margin see the correlation between population allowing Order to run more competetive warbands and Order having less flexibility in their competetive warband setups.

The analysis over the time of day is inconclusive as city sieges across all times of day (apparently) always result in a 60/40 order loss if the numbers are to be believed. The aversion to proper warband composition, be it from a lack of balanced class distribution or lack of desire/leadership to organize a proper comp, affects the entire realm as the responsibility to win trickles down to pug warbands. If you leave more unbalanced pug warbands to cover half of your city instances, they're going to lose more. When those pug warbands are skewed significantly to RDPS classes in City instances that are slightly more favorable to MDPS, you're bringing the wrong tools to a fight who have already been thrown out of the toolbox in the first place.

Throwing any buffs on that dumpster fire of player behavior would need to be targeted at pug-level RDPS classes that overpopulate Order warbands to push any sort of result, and still probably wouldn't matter a whole lot without being clearly broken (see highlighted week of SW's literally murdering people with a single overtuned DOT). From a development perspective, it's a not a class balance issue. Or at least, it's not a problem that could be reasonably be solved by class balance in a balanced manner.

One might imagine the lack of buffs to classes that don't fit the end-game losing narrative of poor warband composition might deter those players from playing undesirable classes in such heavy concentrations, but apparently that's just a beehive we keep kicking that makes noise louder. At any rate, I don't see any particular balance changes coming out of this hot thread that we don't already have in mind. If we did make changes based on this thread, they would be at best a token change that could in theory motivate the realm into some Order pRiDe, ultimately resulting in more pug-level RDPS play in the endgame in an attempt to balance out the win-rates at the expense of proper composition and tactics being used.

Tough choice?

[velenne]

The analysis over the time of day is inconclusive as city sieges across all times of day (apparently) always result in a 60/40 order loss if the numbers are to be believed.

If you're implying that I'm falsifying numbers, it's not appreciated. I didn't gather these numbers, but they pass muster on my brief check of screenshots from the last week or so, and so I have no reason to disbelieve them. As far as the analysis of the numbers goes, anyone here is welcome to contact me and I can provide you the raw data I was given. I welcome you to verify the numbers yourself. edit: You can get the numbers yourself HERE.

The aversion to proper warband composition, be it from a lack of balanced class distribution or lack of desire/leadership to organize a proper comp, affects the entire realm as the responsibility to win trickles down to pug warbands. If you leave more unbalanced pug warbands to cover half of your city instances, they're going to lose more. When those pug warbands are skewed significantly to RDPS classes in City instances that are slightly more favorable to MDPS, you're bringing the wrong tools to a fight who have already been thrown out of the toolbox in the first place.

(Emphasis mine)

I'm not sure if you're implying the onus of responsibility for this rests with the players or the developers? All a guild can do is organize whoever's online in their own space and field the best teams they can. Puggers gonna pug after that, and no one can tell them otherwise. We recruit, we get people in discord, we hold weekly training sessions specifically for city strategies. I don't know what more we can do as players. Last time that I tried to encourage a new player not to play WH, I was muted for 24 hours.

Developers have levers they can manipulate in the game design to incentivize or disincentivize behaviors. (Encouraging more guild/alliance play via content is one way, for example.) I can read your feelings on that matter below:

Throwing any buffs on that dumpster fire of player behavior would need to be targeted at pug-level RDPS classes that overpopulate Order warbands to push any sort of result, and still probably wouldn't matter a whole lot without being clearly broken (see highlighted week of SW's literally murdering people with a single overtuned DOT). From a development perspective, it's a not a class balance issue. Or at least, it's not a problem that could be reasonably be solved by class balance in a balanced manner.

It's impossible to disentangle the effect of the Morale Nerf and the SW buff on city win rates, and it was almost certainly interactive in nature, but one can't argue that the effect following that patch was drastic. Perhaps instead of mega-tuning one class, a scattering of smaller buffs to order's RDPS classes' ability to evade the death ball would help.

One might imagine the lack of buffs to classes that don't fit the end-game losing narrative of poor warband composition might deter those players from playing undesirable classes in such heavy concentrations, but apparently that's just a beehive we keep kicking that makes noise louder. At any rate, I don't see any particular balance changes coming out of this hot thread that we don't already have in mind. If we did make changes based on this thread, they would be at best a token change that could in theory motivate the realm into some Order pRiDe, ultimately resulting in more pug-level RDPS play in the endgame in an attempt to balance out the win-rates at the expense of proper composition and tactics being used.

Tough choice?

(Emphasis mine)

The problem is that one doesn't know if one's class is desirable, or just how undesirable it is, until one hits CR40 and can try to participate in that content. By then the Sunk Cost Fallacy is in full effect and players are compelled to keep banging their heads against the wall. "I have fun in the other areas of the game, why can't I have fun in this one too??"

I think we'd all love to know what you guys have in mind for City. All of this is just reinforcing to me how poorly representative the City Siege Instance is of the game as a whole. I was already in favor of heavily retooling it or scrapping it altogether and this thread has just fueled that belief. I guess you were already there.

[Alfa1986]

Yea, I'm not seeing much. What days of the week were they? To translate to my own EST time:

8 AM EST (NA Morning, Euro Afternoon)
8PM EST(NA Evening, Euro Late Night)
10:30AM EST (NA Morning, Euro Afternoon)
7 PM EST (NA Evening, Euro Late Night)
530 AM EST (NA pre-morning, Euro late-morning).

The only rough pattern I can see here is that the outlier cities are happening at 2 different times more or less.

Time 1: NA mornings/EU afternoons (meaning low NA pop, probably medium EU pop).
My Guess as to whats happening: Generally in my experience these are cities that NA pushes the evening before, but doesn't end up getting to. NA logs off, EU logs on, and finishes the push in their morning/afternoon.

Time 2: NA prime times/EU late nights (meaning high NA pop, probably medium/low EU pop depending on weekday or not).
My Guess as to whats happening: Generally in my experience these are cities that EU pushes the evening before, then they start to log for sleep. As the population dwindles, the NA prime time players push the city.

One thing to consider with city times is also timezone distribution. The reality is that the vast majority of players on this server live in-between PST (Pacific Standard Time, west coast of US), to about what, UTC+3 for RU? With 24 hours in a day, that means that our playerbase heavily covers times from -8 (PST) to +3 (RU), but we are going to have a lag in activity/city pushes in the offtimes (so basically Asia time, the rest of the time between +3 to -8).

it turns out either morning or evening for Eastern Europe. but in general, this is the standard time for the city, 3-4 hours before eu primetime. the population of the server at this time is quite low, especially in order to fill 30 instances))

Yea it seemes to be 3-4 hours before EU primetime, or roughly 3-4 hours after EU primetime is where we're getting these hugely biased cities for destro. Not sure what we can infer from that, but it's noteworthy I think.

yeah, strange coincidences. the main thing is that we can conclude that these deviations are most likely not the result of imbalance.

[Foofmonger]

I just hope the folks analyzing the times of day when Order loses cities by a larger margin see the correlation between population allowing Order to run more competetive warbands and Order having less flexibility in their competetive warband setups.

The analysis over the time of day is inconclusive as city sieges across all times of day (apparently) always result in a 60/40 order loss if the numbers are to be believed. The aversion to proper warband composition, be it from a lack of balanced class distribution or lack of desire/leadership to organize a proper comp, affects the entire realm as the responsibility to win trickles down to pug warbands. If you leave more unbalanced pug warbands to cover half of your city instances, they're going to lose more. When those pug warbands are skewed significantly to RDPS classes in City instances that are slightly more favorable to MDPS, you're bringing the wrong tools to a fight who have already been thrown out of the toolbox in the first place.

Throwing any buffs on that dumpster fire of player behavior would need to be targeted at pug-level RDPS classes that overpopulate Order warbands to push any sort of result, and still probably wouldn't matter a whole lot without being clearly broken (see highlighted week of SW's literally murdering people with a single overtuned DOT). From a development perspective, it's a not a class balance issue. Or at least, it's not a problem that could be reasonably be solved by class balance in a balanced manner.

One might imagine the lack of buffs to classes that don't fit the end-game losing narrative of poor warband composition might deter those players from playing undesirable classes in such heavy concentrations, but apparently that's just a beehive we keep kicking that makes noise louder. At any rate, I don't see any particular balance changes coming out of this hot thread that we don't already have in mind. If we did make changes based on this thread, they would be at best a token change that could in theory motivate the realm into some Order pRiDe, ultimately resulting in more pug-level RDPS play in the endgame in an attempt to balance out the win-rates at the expense of proper composition and tactics being used.

Tough choice?

Thanks for the updates! Appreciate the viewpoints from the dev community on this subject, now I can stop theorizing as much. XD

Also Velenne, I don't believe war was implying you are falsifying numbers. The way I read that is "times of day are likely statistically irrelevant", from the devs perspective, which could have always been the case. Also I brought it up I think anyway so I don't think you would be the person to blame for bad numbers.

I agree though, I think the correct approach on balancing is giving every class a "niche" in different forms of content, but not necessarily trying to solve win-rates through direct reactionary balancing actions. Just overbuffing the "undesirables" without giving them a true "purpose" without that purpose being "overtuned values", is not the correct approach towards healthy balance of a game. For instance, the issue with the SW class in cities right now isn't "they don't do the damage of the BW class". If they did, what would the point of the BW class be now? Balance is never so simple.

[velenne]

it turns out either morning or evening for Eastern Europe. but in general, this is the standard time for the city, 3-4 hours before eu primetime. the population of the server at this time is quite low, especially in order to fill 30 instances))

Yea it seemes to be 3-4 hours before EU primetime, or roughly 3-4 hours after EU primetime is where we're getting these hugely biased cities for destro. Not sure what we can infer from that, but it's noteworthy I think.

yeah, strange coincidences. the main thing is that we can conclude that these deviations are most likely not the result of imbalance.

What deviations? Sorry, I've lost track of what we're talking about and want to make sure.

It's not the deviations of City Siege wins, because those are all destro. lol.
It's not the deviations in mean instance victories, because those are stable at 60/40.

Do you mean the deviations in the mean win rates? Where destro wins by >70%? It must be this. I think Foof had it right above when he said these were the results of failed pushes that a fresh population gets to come in and wipe up later.

Also, I've copied all the numbers I have into a page a HERE if anyone wants to analyze it themselves. It's got the latest two cities from today and yesterday added to it now.

[Ramlaen]

Throwing any buffs on that dumpster fire of player behavior would need to be targeted at pug-level RDPS classes that overpopulate Order warbands to push any sort of result, and still probably wouldn't matter a whole lot without being clearly broken (see highlighted week of SW's literally murdering people with a single overtuned DOT). From a development perspective, it's a not a class balance issue.

Or at least, it's not a problem that could be reasonably be solved by class balance in a balanced manner.

Since you sidestepped the entire topic of my post, time of day population flux effecting the available player pool and the effect it has on building competetive warbands, let me instead ask a question you might be able to answer given your access to internal discussions.

Were the SW nerfs driven by feedback of experienced SW players interested in class balance or by certain people being 'murdered' by an overtuned DOT?

[Detangler]

Why are you guys so dead set on investigating a 60% win rate v a 70%? With 28 instances, its like 16 v 19 wins total. A couple more wins isn't a huge sway, hence why its an outlier. Flip a coin a couple times and you might get heads 4 times in a row. Could just be that on a certain day Destro got the better of a couple extra instances. A couple wins the other way and you have an Order victory.

But seriously, Order just needs to shelve engineers and play more tanks, and cities will be way more competitive on average. The vast majority of city fights I've had Order only brought 3ish tanks and 5ish healing healers (not counting DPS AMs). How do you expect to win with crappy comps like that?