The scoring changes are the wrong solution

[Experimentee.7612]

Ever since the game's launch, a significant problem in WvW has been that it is almost impossible to ensure that matchups are fair because the number of players can not be guaranteed to be equal on each team at all times. This can lead to big discrepancies, especially in off-peak times, where one team with extra players can dominate the matchup for a long period of time.

To address this, Arenanet recently launched some major changes to the scoring system that apply weights based on the time of the day and the expected activity (based on historical data). It basically means that the contribution of players during off-peak times is worth significantly less than the same contribution by players at peak times. While it does help mitigate the big discrepancies at off-peak times, the solution is fundamentally unfair and doesn't really fix the core of the issue. Instead it puts a bandaid over one of the symptoms. To elaborate a bit further, I think it is unfair because there are many players who can not choose when they play the game due to real world constraints (work, family, timezone/location, etc...) and penalising those players for this by making their contribution less meaningful, for something they usually can't control, is completely unfair, and perhaps even crossing into the realm of discrimination. Furthermore, the core problem - unbalanced team sizes especially at off-peak times- remains entirely unaddressed by this solution.

I also think there are great risks in using static weights:
1) The data that the weights were based on are not made transparent or reviewed by anyone other than Roy, CC and possibly Trig.
2) The weights assume that all teams follow the same pattern of activity, which is certainly not true.
3) Player activity patterns will change over time - the weights need to be dynamic. Do we have any information on how regularly these static weights will be reviewed or updated? And isn't it just adding more overhead for Arenanet to that manually? Shoudn't that process be automated?
4) I am concerned that players can exploit the static weights by deliberately timing their activity so that it coincides with the peak times, out of a desire for the contribution to have the greatest impact. And I'm referring here to those players that would otherwise perhaps play at others times of the day.

In my opinion, a better solution would have been to weight the scores dynamically based on the number of people actively playing on each side on each map at any given moment. This ensures that even when player populations are not equal, each team can still remain competitive. Dynamic scoring would encourage all players to pull their weight and be mindful that they are contributing efficiently. 

I also think that increasing the score per kill from 2 to 3, together with many other changes in recent WvW patches (including world restructuring), mean it now caters mostly for the fighting guilds. This is concerning for casual players.

 

 

[Sheff.4851]

5 minutes ago, Experimentee.7612 said:

1) The data that the weights were based on are not made transparent or reviewed by anyone other than Roy, CC and possibly Trig.

 

 

 

Do you have a source for this? I'm very curious to see it.

Edited July 24 by Sheff.4851
Fixing quote

[Experimentee.7612]

Another problem of the static weights that I forgot to mention in my initial post are that they do not differentiate between activity on weekdays and weekends.

Edited July 24 by Experimentee.7612

[Experimentee.7612]

2 hours ago, Sheff.4851 said:

Do you have a source for this? I'm very curious to see it.

It was mentioned in the competitive update stream 2 weeks ago. You can watch the VOD from this timestamp onwards for a few minutes:
https://www.twitch.tv/videos/2194238192?t=00h29m18s

 

[Sheff.4851]

7 minutes ago, Experimentee.7612 said:

It was mentioned in the competitive update stream 2 weeks ago. You can watch the VOD from this timestamp onwards for a few minutes:
https://www.twitch.tv/videos/2194238192?t=00h29m18s

 

Cal and Roy are the people who presented the changes, yes. I'm asking about "reviewed by anyone other than Roy, CC and possibly Trig", since that would suggest some knowledge about ArenaNet's internal workings and leadership structure for reviewing and pushing changes like this.

[Experimentee.7612]

31 minutes ago, Sheff.4851 said:

Cal and Roy are the people who presented the changes, yes. I'm asking about "reviewed by anyone other than Roy, CC and possibly Trig", since that would suggest some knowledge about ArenaNet's internal workings and leadership structure for reviewing and pushing changes like this.

Not only did they present the changes, but they also gave a glimpse into the the internal workings of how the spreadsheet was put together.

30:52 - CC says: "If these numbers felt a little all over the place to you guys, we can confirm that Roy has a mad scientist sheet that he put together to figure out the right values."

31:38 - CC says: "No one understands the spreadsheet except for Roy. He bought it to me and says(?)  'can you look at this' and I'm like 'can you explain to me what's happening on this spreadsheet?', but once he explained it, it all made sense."

Some of that was humour, of course, but it suggests to me that besides Roy and CC (and maybe even Trig), nobody else was involved in putting it together or reviewing it.

[XenesisII.1540]

Pretty baffling they would use weights based on the entire region instead of per match.

I know they think WR is suppose to equally spread players and everything is suppose to be sorta comparable for all worlds now, which we saw for five weeks before scoring change wasn't quite working yet. Reality is they will never be able to fully read how much of a commitment a player will make day by day, let alone groups, time zones, or an entire world. There's guilds that don't play every single night, so doesn't even make sense setting prime time to always produce maximum points, then you have worlds just floating on freebie points they didn't earn themselves.

We could have a t1 server with 4 map queues earning maximum points for winning, while a t6 server that has only 1 queued map but earning the same amount of points for winning in that same period. It's a poorly thought out system that just reverses what we had all those years ago before skirmish mode, where a time zone could produce runaway scores, now it's shifted back to opposite time zones, at least back then you had to actually work for those bonus points, in this system it's just auto given it because you exist in a certain time zone no matter how many players you actually have logged on a given night.

Worse yet there is no real opportunity to play catchup now because those points hit that peak for only 2 hours of the day and then it's lowered the rest of the time, so if a server manages to get people organized the next hour they can't even work to immediately catch up the points from the previous 2 hours. Imagine if they implemented this in a baseball game, hey guys runs in the first inning is worth 5x the runs, good luck catching up in the last 8 innings if you mess up. At the very least the ramp up scoring shouldn't apply to victory points, but to warscore.

This is pretty much the last stamp that says, "this isn't a competitive game mode", not that it was before anyways, but the idea to try and make "winning worthwhile" again by producing bigger runaway scores is pretty ridiculous.

 

48 minutes ago, Sheff.4851 said:

Cal and Roy are the people who presented the changes, yes. I'm asking about "reviewed by anyone other than Roy, CC and possibly Trig", since that would suggest some knowledge about ArenaNet's internal workings and leadership structure for reviewing and pushing changes like this.

You kept arguing Roy is on the wvw team and not the balance team, working on things like scoring changes... and now you want to say he's just a presenter on the scoring changes and not actually working on them too? Pretty sure there's a good reason they have both Roy and Cal presenting changes for balance and wvw, and not just handing over presenter duty to Rubi as usual.... 🤷‍♂️

Edited July 24 by XenesisII.1540

[Sheff.4851]

Just now, XenesisII.1540 said:

Pretty baffling they would use weights based on the entire region instead of per match.

You kept arguing Roy is on the wvw team and not the balance team, working on things like scoring changes... and now you want to say he's just a presenter on the scoring changes and not actually working on them too?

I don't doubt he contributed to them, but to suggest that CMC and Roy are the only two people who reviewed those changes is an interesting take. I'm not going to speculate on who else might have, because that's the kind of stuff that gets you departnered. I'd just be surprised if that was the case, is all.

[Gorani.7205]

After all theses years, where ANet has a hard time balancing people (which they can't, because players transferred and react "unpredictable") and now time slots (by scoring them differently and creating "abusable" time windows by design that have not existed before) I really wonder why the Devs have never tackled the variable they can control better:
Breaking the old wheel of the two strongest teams feeding on the weakest (1st and 2nd fight 3rd) instead on rewarding 2nd and 3rd third going against 1st. 🤷‍♂️

[Sheff.4851]

Quote

and creating "abusable" time windows by design that have not existed before

These time windows always existed, it's why servers would purchase guilds during Seasons to cover off-hours time slots. They were inherently more valuable to the overall matchup because of the lower activity on other teams. The scoring changes just align them with natural periods of activity within the region, moving the higher-value timezones into NA prime, where map queues will prevent any abuse of the high value skirmishes by gating the number of players who are able to participate in them.

[XenesisII.1540]

7 minutes ago, Gorani.7205 said:

Breaking the old wheel of the two strongest teams feeding on the weakest (1st and 2nd fight 3rd) instead on rewarding 2nd and 3rd third going against 1st. 🤷‍♂️

Pretty much the first thing they should have been looking at if they wanted to achieve any kind of competitive environment, but everything they do just always seems geared to reward the bigger mass instead.

[Sheff.4851]

17 minutes ago, XenesisII.1540 said:

Pretty much the first thing they should have been looking at if they wanted to achieve any kind of competitive environment, but everything they do just always seems geared to reward the bigger mass instead.

You could argue that JQ and TC (#2 and #3) going after BG (#1) is the main reason that Seasons never came back. I'm not sure that any of the three servers involved in that season would call it competitive, but you'd have trouble asking them about it, because most of them quit.

[XenesisII.1540]

6 minutes ago, Sheff.4851 said:

You could argue that JQ and TC (#2 and #3) going after BG (#1) is the main reason that Seasons never came back. I'm not sure that any of the three servers involved in that season would call it competitive, but you'd have trouble asking them about it, because most of them quit.

Nope. That happened in season 2, pretty much the only time a sustained double team happened for multiple weeks. We still had another season after that in which BG went on to win, again. Tournaments never came back because of player burnout, apparently, not a double team that worked the way it should work for a three way game mode, and only lasted 9 weeks out of 12 years. 

Here were the points for all three seasons btw.

Season 1 - BG 35, JQ 27, TC 23, SoR 19, SoS 13, Mag 9

Season 2 - JQ 35, TC 33, SoS 29, BG 29, Mag 23, DB 13

Season 3 - BG 20, SoS 14, TC 13, JQ 13, FA 12, Mag 8

Season 2 didn't look competitive?

🤷‍♂️

[Sheff.4851]

Just now, XenesisII.1540 said:

Nope. That happened in season 2, pretty much the only time a sustained double team happened for multiple weeks. We still had another season after that in which BG went on to win, again. Tournaments never came back because of player burnout, apparently, not a double team that worked the way it should work for a three way game mode, and only lasted 9 weeks out of 12 years. 

Here were the points for all three seasons btw.

Season 1 - BG 35, JQ 27, TC 23, SoR 19, SoS 13, Mag 9

Season 2 - JQ 35, TC 33, SoS 29, BG 29, Mag 23, DB 13

Season 3 - BG 20, SoS 14, TC 13, JQ 13, FA 12, Mag 8

Season 2 didn't look competitive?

🤷‍♂️

BG won S3 because TC and JQ fractured after the double-team in S2, TC specifically. Source: conversations with BG, JQ, and TC server leaders from that period of time. Yes, players quit because of burnout during Seasons. Sure. What caused the burnout, specifically? In part, the within-server disagreements involved with coordinating a double team, because nobody likes to be told who they are and are not allowed to fight in a sandbox PvP game mode.

[XenesisII.1540]

1 minute ago, Sheff.4851 said:

BG won S3 because TC and JQ fractured after the double-team in S2, TC specifically. Source: conversations with BG, JQ, and TC server leaders from that period of time. Yes, players quit because of burnout during Seasons. Sure. What caused the burnout, specifically? In part, the within-server disagreements involved with coordinating a double team, because nobody likes to be told who they are and are not allowed to fight in a sandbox PvP game mode.

The difference here is the players had to organize and actively force them themselves to do the double team.

But what if the system naturally promoted it for you through game play?

Also the burnout wasn't just for "that event", it was tournaments as a whole, every single server trying to play to win, playing over time, and also getting the achievements done.

🤔

[XenesisII.1540]

6 minutes ago, Sheff.4851 said:

BG won S3 because TC and JQ fractured after the double-team in S2, TC specifically. Source: conversations with BG, JQ, and TC server leaders from that period of time. Yes, players quit because of burnout during Seasons. Sure. What caused the burnout, specifically? In part, the within-server disagreements involved with coordinating a double team, because nobody likes to be told who they are and are not allowed to fight in a sandbox PvP game mode.

P.S Season 2 looks more competitive does it not? That's what we're striving for right? I mean the whole point of messing with scoring is to achieve this right?

[Sheff.4851]

 

Just now, XenesisII.1540 said:

P.S Season 2 looks more competitive does it not? That's what we're striving for right? I mean the whole point of messing with scoring is to achieve this right?

Nobody who I have spoken to on those servers calls S2 a competitive environment. Were you there yourself? I'd love to interview you for a server history project that I'm working on, if you have some personal perspective from that period of the game. It would be nice to juxtapose your opinion against the two dozen or so roamers, server leaders, commanders, and other players who called it one of the least fun periods of the game.

[XenesisII.1540]

1 minute ago, Sheff.4851 said:

Nobody who I have spoken to on those servers calls S2 a competitive environment. Were you there yourself? I'd love to interview you for a server history project that I'm working on, if you have some personal perspective from that period of the game. It would be nice to juxtapose your opinion against the two dozen or so roamers, server leaders, commanders, and other players who called it one of the least fun periods of the game.

In what way was it not competitive? I was playing on SoS in those years, I remember days I would log on and BG would just blanket maps, and people would just take the week off, we didn't even have skirmish rewards or reward tracks to work on. It took two servers to handle them in that tournament, and those two servers came in 1st and 2nd, were they not proud of that fact alone? It's like one of the biggest accomplishments those two servers did in wvw history. 🤷‍♂️

Yeah there was burnout, in all the tournaments, there was 21 other servers in NA that had experienced burnout too. Not just the three on top everyone knew was gonna win anyways, because coverage, that fact alone is why tournaments would never work in the first place. 🤷‍♂️

[Sheff.4851]

14 minutes ago, XenesisII.1540 said:

In what way was it not competitive? I was playing on SoS in those years, I remember days I would log on and BG would just blanket maps, and people would just take the week off, we didn't even have skirmish rewards or reward tracks to work on. It took two servers to handle them in that tournament, and those two servers came in 1st and 2nd, were they not proud of that fact alone? It's like one of the biggest accomplishments those two servers did in wvw history. 🤷‍♂️

Yeah there was burnout, in all the tournaments, there was 21 other servers in NA that had experienced burnout too. Not just the three on top everyone knew was gonna win anyways, because coverage, that fact alone is why tournaments would never work in the first place. 🤷‍♂️

Some people on those servers were proud of that fact, but other people were deeply frustrated by being told that they weren't allowed to attack any other server but BG, because all BG did was fight in spawn and then log off. Then, once the matchup was mathematically decided (around week six of eight), JQ and TC stopped playing, because they'd accomplished their goal of making sure BG lost, and there was no point to the last two weeks, which also frustrated people that just wanted to play the game, because now BG ran over the remaining players on both sides. That rift, between groups that encouraged the double team and groups that just wanted to play the game and fight either side, was one of the main reasons that TC's leadership community collapsed after Seasons, and never really recovered -- and that much is evident in the Season 3 scores that you linked. If the double team was so successful in S2, why wasn't it repeated in S3? Do you see how JQ went from second, to first, to fourth? It's because the communities on those two servers had absolutely no interest in double teaming again, which is part of why SoS was able to break into T1 in the first place, after being nowhere close to it in S1.

As far as burnout goes, yeah, everybody burned out to some degree, but the problem was much, much worse in Gold league. Servers down in the bottom of the tier system (Ehmry Bay, Kaineng, other T6/T7/T8 servers didn't really have burnout issues, but they also knew, fundamentally, they didn't have the activity to compete in Seasons in the first place, so there was less pressure to tag for coverage and double team enemy servers overall.

I think you'll find these videos very interesting when I'm done with them. It's a lot of oral history that nobody's ever really bothered to write down or talk about, and it's taken me several months to collect and interview the people involved in that period of the game.

Edit: Source for BG, specifically, is that my current guild (Knights of the Temple) is comprised of old BG players from Season 1, including Lythiea, one of the BG War Councils leaders during Seasons. We talk about that period of the game frequently during our raids. Source for TC and JQ is interviews with Jedbacca and GuildMM, JQ server leaders, and interviews with Ryvalia, Atros, and a few of the other TC server leaders from that period as well.

Edited July 24 by Sheff.4851

[XenesisII.1540]

19 minutes ago, Sheff.4851 said:

Some people on those servers were proud of that fact, but other people were deeply frustrated by being told that they weren't allowed to attack any other server but BG, because all BG did was fight in spawn and then log off. Then, once the matchup was mathematically decided (around week six of eight), JQ and TC stopped playing, because they'd accomplished their goal of making sure BG lost, and there was no point to the last two weeks, which also frustrated people that just wanted to play the game, because now BG ran over the remaining players on both sides.

That rift, between groups that encouraged the double team and groups that just wanted to play the game and fight either side, was one of the main reasons that TC's leadership community collapsed after Seasons, and never really recovered -- and that much is evident in the Season 3 scores that you linked. If the double team was so successful in S2, why wasn't it repeated in S3? Do you see how JQ went from second, to first, to fourth? It's because the communities on those two servers had absolutely no interest in double teaming again, which is part of why SoS was able to break into T1 in the first place, after being nowhere close to it in S1.

As far as burnout goes, yeah, everybody burned out to some degree, but the problem was much, much worse in Gold league. Servers down in the bottom of the tier system (Ehmry Bay, Kaineng, other T6/T7/T8 servers didn't really have burnout issues, but they also knew, fundamentally, they didn't have the activity to compete in Seasons in the first place, so there was less pressure to tag for coverage and double team enemy servers overall.

I think you'll find these videos very interesting when I'm done with them. It's a lot of oral history that nobody's ever really bothered to write down or talk about, and it's taken me several months to collect and interview the people involved in that period of the game.

Edit: Source for BG, specifically, is that my current guild (Knights of the Temple) is comprised of old BG players from Season 1, including Lythiea, one of the BG War Councils leaders during Seasons. We talk about that period of the game frequently during our raids. Source for TC and JQ is interviews with Jedbacca and GuildMM, JQ server leaders, and interviews with Ryvalia, Atros, and a few of the other TC server leaders from that period as well.

Servers have broken up a number of times over the years not in tournaments, not to mention tournaments were temporary events, those servers broke up over a temporary event? Might be saying there was a little more going on there. I'm sure the additional pressure of trying to maintain a double team put a lot of pressure on the organizers, but I highly doubt that situation was the reason tournaments were cancelled. Pretty sure this was a volatile time in wvw as well with guilds moving more often as time went on.

In any case wvw has never been competitive, we were all under an illusion that it was for many years. Those tournaments cemented the fact that you needed coverage in order to win, Titan alliance proved that in the first two weeks of the game, and we all refused to acknowledge it then, but then many players started to realize there wasn't any point to playing over time, earning points, winning matches for.... nothing. So why would you even bother with prolonged double teams, everyone just goes for whatever is easy these days, points don't matter, position don't matter, well it kinda does but you don't actually have to do any work to tank a tier.

Trying to manipulate scores to set how much value players time are worth in a time zone now, isn't really going to help, and in the end we still have to ask.. why and for what purpose make these type of changes? To bring back tournaments that everyone burned out on? to emphasize winners and losers strengths more profoundly in matchups while they only jump one tier at a time? Having bigger numbers isn't a motivation to win, they're using the wrong carrot here.

2v1 on the weakest has been a severe issue in wvw since the beginning, it's one of the biggest things they could fix to help player morale and push the game into a more positive competitive direction in a game mode where it barely exist as it is. 50 vs 30/25 could be interesting, 50 + 30 vs 25 is not. The system just needs to be tweaked to discourage you from farming weaker numbers for rewards and gain more for facing more your size or greater. Bigger risk bigger rewards, not less risk bigger rewards. 🤷‍♂️

[Dawdler.8521]

53 minutes ago, XenesisII.1540 said:

Trying to manipulate scores to set how much value players time are worth in a time zone now, isn't really going to help, and in the end we still have to ask.. why and for what purpose make these type of changes?

I mean come on. We know why. Just like alliances, Anet is probably backtracking on trying to weight the overall skills/coverage of guilds and thus the WR distribution algorithm. The easy solution? Forget about coverage and make prime matter the most where queues limit the amount of players anyway.

Btw here is an example of the EU points from wednesday two weeks ago compared to today in respective tier and score deviance highest to lowest:
224/213/199 -> 996/554/467 (13% vs 113%)
241/207/200 -> 811/636/570 (21% vs 42%)
258/196/186 -> 890/711/416 (39% vs 114%)
224/221/203 -> 816/622/579 (10% vs 41%)
237/211/200 -> 776/765/476 (19% vs 63%)
223/215/210 -> 774/651/592 (6% vs 31%)
 

[Sheff.4851]

Did ArenaNet fix the EU/NA victory point issue over the weekend? Or is EU still using NA's timezone weighting and vice versa?

[Dawdler.8521]

3 minutes ago, Sheff.4851 said:

Did ArenaNet fix the EU/NA victory point issue over the weekend? Or is EU still using NA's timezone weighting and vice versa?

Should we expect them too? 😛

Not as far as I know since the points incidently just updated and Lyssa which I will assume won, got 9 points which is NA scaling. But regardless of that, both scales still do +100%/+200% so little in the relations between points would change.

[Sheff.4851]

It had been mentioned that there may be a fix for it over the weekend, but I would have been surprised. It's the last week before team rebuild, so I'd imagine it's pretty low priority. But if EU is still getting NA's scoring, I wouldn't really read into overall performance of the scoring algorithm until it's fixed.

[Dawdler.8521]

6 minutes ago, Sheff.4851 said:

But if EU is still getting NA's scoring, I wouldn't really read into overall performance of the scoring algorithm until it's fixed.

It would be the same overall. Presumably only the scaling are incorrect (ie it happen to be 9/6/3 instead of 15/10/5) and not the actual timezones when it happens. In that case, the score would be different true but the percentages would be identical (ie in the case of T1, Lyssa would still have 113% more points than Khormir).