The scoring changes are the wrong solution

[Mabi black.1824]

10 hours ago, A Hamster.2580 said:

This delta that you calculate is not a good number to use.

This delta is not to be calculated. you just need to detect it. Read the numbers and jump out on your own. I also wrote delta maximum, that's why I only take into account 1st and 3rd place. and maybe it's not a good number for you, because it makes clear evidence of how inappropriate it is.

[Mabi black.1824]

9 hours ago, Chaba.5410 said:

Breakpoint in this context is a gamer's min/maxing term.  Your translation app probably doesn't understand that context.

So let's say there's a 1x multiplier when your team has 80 players on a map.  It changes to 2x multiplier when you have only 40 players on the map.  Then it becomes 3x multiplier when you are down to 20 on a map.  40 and 20 players are the breakpoints at which the score change takes effect.  If you have 22 players on the map, you are close to that 3x multiplier and want 2 people to leave the map.  If you have 38 players on the map, you're much further from that breakpoint.  But you wouldn't want to go down to 10 players on the map.  You'd want as close to 20 as possible without going above that.

You may hear GW2 build theory crafters talk about "wasted stats" or being "overcapped on precision".  That's in the same context.

Maybe I lost something along the way. Does this multiplier work with reference to playing time or the number of players? takes into account all 4 maps or only 1 map? Is it calculated by difference compared to the other 2 teams or does it have pre-established thresholds? I don't really know where you were in this discussion.

Edited July 26 by Mabi black.1824

[Experimentee.7612]

 

11 hours ago, A Hamster.2580 said:

So just to recap, you claim "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."

You essentially treat "players during off-peak" and "players during prime time" as two singular entities completely disregarding how many players may be in each entity. If you treat them as only two singular entities, then sure you are right in that one may earn 33 points in first place in NA while the other may earn only 9 points in NA and that is unfair. But the whole point of the new system is to no longer look at it from your perspective, the perspective where there are only two entities. 

This new system actually attempts to take into account the number of players in each time zone. 

You constructed your whole essay around that one claim. That one assumption that  "players during off-peak" and "players during prime time" are two singular entities.

I gave an example comparing two points in time, one off-peak, another at peak. I do not assume those are the only entities, and I fully understand that there are several different weights.at different times. My 'essay' is not invalidated at all by that.

 

10 hours ago, A Hamster.2580 said:

You know there are tools that allow a player to actively log into multiple accounts at the same time? I have never heard of anyone being banned for using these too. Your model will easily be exploited because it is assuming that no one will try to inflate the number of active accounts on the enemy team by using alt accounts. It is assuming everyone is honorable.

Trust me, this will be omega exploited. Especially when they implement incentives for winning soon.

That is no different to today, where players can exploit their alt accounts and take up limited places in WvW to manipulate matches. This does not invalidate the dynamic scoring I proposed. But I do think it's an issue that Arenanet need to resolve.

 

10 hours ago, Chaba.5410 said:

Breakpoint in this context is a gamer's min/maxing term.  Your translation app probably doesn't understand that context.

So let's say there's a 1x multiplier when your team has 80 players on a map.  It changes to 2x multiplier when you have only 40 players on the map.  Then it becomes 3x multiplier when you are down to 20 on a map.  40 and 20 players are the breakpoints at which the score change takes effect.  If you have 22 players on the map, you are close to that 3x multiplier and want 2 people to leave the map.  If you have 38 players on the map, you're much further from that breakpoint.  But you wouldn't want to go down to 10 players on the map.  You'd want as close to 20 as possible without going above that.

You may hear GW2 build theory crafters talk about "wasted stats" or being "overcapped on precision".  That's in the same context.

I did not think of it as a step-multiplier. The multiplier would be dynamic and change instantly based on the precise number of people on the map.

40 minutes ago, Mabi black.1824 said:

Maybe I lost something along the way. Does this multiplier work with reference to playing time or the number of players? takes into account all 4 maps or only 1 map? Is it calculated by difference compared to the other 2 teams or does it have pre-established thresholds? I don't really know where you were in this discussion.

The 'multiplier' is another name for the weights/ratios I mentioned in my proposal. The idea was for the 'multiplier' to consider the number of players only, and disregard the time of day. I imagined it would be applied per map, but haven't really given enough thought to alternatives such as applying it to the overall WvW population across all four maps. Exactly how the 'multiplier' works is up for debate, but I wouldn't count on it having thresholds.  I imagined it would be dynamic and change instantly based on the precise number of people on the map.

[Swoo.5079]

If you win a skirmish by outnumbering the enemy, why is that worth a lot of points during the primetime and fewer points off peak?

You are still outnumbering, no matter if you have 210v100v100 or if you have 15v5v3.

It is just pandering to bigger guilds.

The server system wasn't great, but transfering community building from server to guild isn't fantastic either. Teams have no long last value since they get dissolved.

 

[Zok.4956]

21 minutes ago, Swoo.5079 said:

If you win a skirmish by outnumbering the enemy, why is that worth a lot of points during the primetime and fewer points off peak?

You are still outnumbering, no matter if you have 210v100v100 or if you have 15v5v3.

I agree, the score/points should be the same.

22 minutes ago, Swoo.5079 said:

It is just pandering to bigger guilds.

Yes. They apparently want to close WvW in the off-hours without closing WvW in the off-hours.

[A Hamster.2580]

2 hours ago, Mabi black.1824 said:

Never seen before, I have always seen points delivered to the team/server. The design of this mode has been designed with reference to the server. The points earned are from the server. This design doesn't give a kitten about the individual player. 

 

The old system didn't care about the individual player. 70 players in prime time or 15 players in an off hour time zone giving their best could only gain a maximum of 5 points. This new system attempts to account for the difference in number of players so it does care about individual players.

[A Hamster.2580]

2 hours ago, Mabi black.1824 said:

This delta is not to be calculated. you just need to detect it. Read the numbers and jump out on your own. I also wrote delta maximum, that's why I only take into account 1st and 3rd place. and maybe it's not a good number for you, because it makes clear evidence of how inappropriate it is.

Once again, you only used two time slots like those two are the only two that exist (ex: two 12 hours time slots that the players only have). Prime time zones and off hour time zones are a lot more than just two time zones. That's why using your definition of "delta" is just inaccurate.

[A Hamster.2580]

1 hour ago, Experimentee.7612 said:

I gave an example comparing two points in time, one off-peak, another at peak. I do not assume those are the only entities, and I fully understand that there are several different weights.at different times. My 'essay' is not invalidated at all by that.

Yes you do assume that since we do not know the exact number of players playing in the prime time zones and the off hour time zones. For all we know, the large number of prime time players could still be undervalued on a per player basis than the small number of off hour players even under this new system.

[Experimentee.7612]

4 hours ago, A Hamster.2580 said:

Yes you do assume that since we do not know the exact number of players playing in the prime time zones and the off hour time zones. For all we know, the large number of prime time players could still be undervalued on a per player basis than the small number of off hour players even under this new system.

No, we do not know the exact number of players playing at different times. But if we choose an off-peak time slot and a peak time slot, and control for the number of players, then the contribution of the players in off-peak times is worth objectively less than the same contribution by the same number of players during peak teams, under the recently implemented scoring system.

[Sheff.4851]

13 minutes ago, Experimentee.7612 said:

No, we do not know the exact number of players playing at different times. But if we choose an off-peak time slot and a peak time slot, and control for the number of players, then the contribution of the players in off-peak times is worth objectively less than the same contribution by the same number of players during peak teams, under the recently implemented scoring system.

Right, but there's not the "same number of players", because the entire point of the scoring update is that activity isn't the same during peak times and off hours. That's why they are called peak times and off hours. GW2Mists has activity graphs for EU and NA that you can use to see this for yourself. Scroll down to "Activity in %" on this page: https://gw2mists.com/matches/eu. See the parts that are completely flat? That means that almost nobody is playing. But all of the structures sitting on those empty maps, with nobody playing them, are still generating score. That's the part that is unfair to people who play in prime, and unfairly weights overall match score towards timezones that are lower activity, or completely dead.

[Swoo.5079]

3 minutes ago, Sheff.4851 said:

Right, but there's not the "same number of players", because the entire point of the scoring update is that activity isn't the same during peak times and off hours. That's why they are called peak times and off hours. GW2Mists has activity graphs for EU and NA that you can use to see this for yourself. Scroll down to "Activity in %" on this page: https://gw2mists.com/matches/eu. See the parts that are completely flat? That means that almost nobody is playing. But all of the structures sitting on those empty maps, with nobody playing them, are still generating score. That's the part that is unfair to people who play in prime, and unfairly weights overall match score towards timezones that are lower activity, or completely dead.

In servers, where you could move guilds to different servers and make sure you had activity at certain times, sure, it was unfair.

But now it is players that create guilds and Anet that spread those guilds.

If your guild only recruits players for prime time why should it be protected? Recruit players that play off prime time.

[Sheff.4851]

5 minutes ago, Swoo.5079 said:

In servers, where you could move guilds to different servers and make sure you had activity at certain times, sure, it was unfair.

But now it is players that create guilds and Anet that spread those guilds.

If your guild only recruits players for prime time why should it be protected? Recruit players that play off prime time.

The pool of players is much smaller in off-hours than it is for prime, so there are fewer players to go around. Inviting off-hours players to your big primetime guild so that they can work coverage for you just fractures the off-hours communities further than they are already.

[Doghouse.1562]

On 7/24/2024 at 3:43 PM, Experimentee.7612 said:

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.

It also misses the fact that "peak time" is something somewhat arbitrary and arithmetic. For those teams (servers as were) who're dominating at "off-peak" times, those ARE peak times - the times when most of your players are active. It's not your fault if other people aren't around, and there's nothing much that you can do about it. Fix it, if you feel it need fixing, by matching teams that tend to play at the same times, by all means. But don't penalise and dishearten people and trivilaise their efforts simply because they happen to play at what you've chosen to define as the "wrong" times. That's a TERRIBLE idea.

Edited July 26 by Doghouse.1562

[Sheff.4851]

6 minutes ago, Doghouse.1562 said:

It also misses the fact that "peak time" is something somewhat arbitrary and arithmetic. For those teams (servers as were) who're dominating at "off-peak" times, those ARE peak times - the times when most of your players are active. It's not your fault if other people aren't around, and there's nothing much that you can do about it. Fix it, if you feel it need fixing, by matching teams that tend to play at the same times, by all means. But don't penalise and dishearten people and trivilaise their efforts simply because they happen to play at what you've chosen to define as the "wrong" times. That's a TERRIBLE idea.

Exactly, like how any NA prime guild that got dragged into T1 by coverage guilds in off-hours got penalized and disheartened by the lack of anything to do in prime. This is a shift to a more equitable solution, where your VP are calculated relative to activity in your region, rather than being entirely independent from activity.

[Swoo.5079]

14 minutes ago, Sheff.4851 said:

Exactly, like how any NA prime guild that got dragged into T1 by coverage guilds in off-hours got penalized and disheartened by the lack of anything to do in prime. This is a shift to a more equitable solution, where your VP are calculated relative to activity in your region, rather than being entirely independent from activity.

So is it prime time activity or is it prime time activity for your guild? I mean if there is no one else there to fight...

[Sheff.4851]

1 minute ago, Swoo.5079 said:

So is it prime time activity or is it prime time activity for your guild? I mean if there is no one else there to fight...

Correct. We have both identified the issue with the old scoring system, and the way that it gave flat VP while ignoring activity over a 24 hour period.

[Dawdler.8521]

1 minute ago, Sheff.4851 said:

Correct. We have both identified the issue with the old scoring system, and the way that it gave flat VP while ignoring activity over a 24 hour period.

It didnt ignore activity though. That's what PPK added to score..

[Sheff.4851]

1 minute ago, Dawdler.8521 said:

It didnt ignore activity though. That's what PPK added to score..

2 points per kill accounts for about 10% of match score overall, and when PPK was added to the game it didn't realistically change the position of servers overall. The largest contributor of overall score, by far, is the points you get from holding a structure as it ticks away every five minutes, and what players found unfair is that a T3 keep would sit on an empty borderland, generating score every five minutes even though nobody was attacking or defending it. That's just inactive gameplay.

[Experimentee.7612]

20 minutes ago, Sheff.4851 said:

Right, but there's not the "same number of players", because the entire point of the scoring update is that activity isn't the same during peak times and off hours. That's why they are called peak times and off hours. GW2Mists has activity graphs for EU and NA that you can use to see this for yourself. Scroll down to "Activity in %" on this page: https://gw2mists.com/matches/eu. See the parts that are completely flat? That means that almost nobody is playing. But all of the structures sitting on those empty maps, with nobody playing them, are still generating score. That's the part that is unfair to people who play in prime, and unfairly weights overall match score towards timezones that are lower activity, or completely dead.

Don't you see the unfairness? Let's say that due to life circumstances, you were only able to login to the game at midday to play. This change means that your contribution is suddenly worth a lot less than if you had logged in at 7pm.

I think you are missing the real problem here. The problem with score generated in off-peak times is not the level of activity but rather that the population of each team is not balanced. If at 4am, score is generated when all three teams have only five players online, then I believe that that score is just as valid as score generated at 7pm when all three teams only have 70 players online. So, in my view, the solution should not be to weight the scores based on activity levels. The solution should be to balance the teams at all times. Obviously, it's not possible to control when players are active. So an alternative is to artificially balance the teams by scaling up the score generated by the team with fewer players by a factor that accounts for, precisely, the difference in the number of players on each team. 

Also, as a sidenote, please remember that the data on GW2mists is based on only the players that are registered on that site, and may not be a true reflection of reality.

5 minutes ago, Doghouse.1562 said:

It also misses the fact that "peak time" is something somewhat arbitrary and arithmetic. For those teams (servers as were) who're dominating at "off-peak" times, those ARE peak times - the times when most of your players are active. It's not your fault if other people aren't around, and there's nothing much that you can do about it. Fix it, if you feel it need fixing, by matching teams that tend to play at the same times, by all means. But don't penalise and dishearten people and trivilaise their efforts simply because they happen to play at what you've chosen to define as the "wrong" times. That's a TERRIBLE idea.

Agreed!

[Dawdler.8521]

21 minutes ago, Sheff.4851 said:

2 points per kill accounts for about 10% of match score overall, and when PPK was added to the game it didn't realistically change the position of servers overall. The largest contributor of overall score, by far, is the points you get from holding a structure as it ticks away every five minutes, and what players found unfair is that a T3 keep would sit on an empty borderland, generating score every five minutes even though nobody was attacking or defending it. That's just inactive gameplay.

Then the obvious answer should have been to reduce the exponential gain of upgraded objectives. 

We know it work fine at reset - PPK alone pretty much decides the first skirmish.

[Sheff.4851]

1 hour ago, Experimentee.7612 said:

Don't you see the unfairness? Let's say that due to life circumstances, you were only able to login to the game at midday to play. This change means that your contribution is suddenly worth a lot less than if you had logged in at 7pm.

The current system is more equitable for playtime, but when a group has felt special for a very long time, and then changes are made that reduce that group's effectiveness, it definitely feels unfair to them specifically, even when it's a more equitable system for all players.

Quote

I think you are missing the real problem here. The problem with score generated in off-peak times is not the level of activity but rather that the population of each team is not balanced. If at 4am, score is generated when all three teams have only five players online, then I believe that that score is just as valid as score generated at 7pm when all three teams only have 70 players online. So, in my view, the solution should not be to weight the scores based on activity levels. The solution should be to balance the teams at all times. Obviously, it's not possible to control when players are active. So an alternative is to artificially balance the teams by scaling up the score generated by the team with fewer players by a factor that accounts for, precisely, the difference in the number of players on each team. 

That would be a good long-term solution! For the short-term, adjusting VP gain by average activity in the region is an improvement over the existing system, though.

[Chaba.5410]

1 hour ago, Experimentee.7612 said:

The problem with score generated in off-peak times is not the level of activity but rather that the population of each team is not balanced.

This is a problem regardless of off-hours vs NA/EU prime skirmish.  That's why historically there's been servers with "weak NA" that existed in higher tiers.  The score generated by off-peak times was causing teams to not be balanced during NA, for example.  So unbalanced teams during off-hours would also cause unbalanced match ups elsewhere.  How do you fix that?

Edited July 26 by Chaba.5410

[Swoo.5079]

13 hours ago, Chaba.5410 said:

This is a problem regardless of off-hours vs NA/EU prime skirmish.  That's why historically there's been servers with "weak NA" that existed in higher tiers.  The score generated by off-peak times was causing teams to not be balanced during NA, for example.  So unbalanced teams during off-hours would also cause unbalanced match ups elsewhere.  How do you fix that?

All of this boils down to too many teams 

And then too many players at prime time - many of those pve players playing only for GoB.

WR has not worked because there are still way too many teams, plus if a team is clearly outnumbered some players will go play another game mode.

When a team/server is strong all these non hardcore wvw players take space but if the team isn't dominant they won't be there.

[A Hamster.2580]

20 hours ago, Experimentee.7612 said:

No, we do not know the exact number of players playing at different times. But if we choose an off-peak time slot and a peak time slot, and control for the number of players, then the contribution of the players in off-peak times is worth objectively less than the same contribution by the same number of players during peak teams, under the recently implemented scoring system.

"control for the number of players" "same number of players during peak"

That's where your claim is wrong. You just assume number of off hour players = number of prime time players to conveniently make your claim correct.

[A Hamster.2580]

19 hours ago, Experimentee.7612 said:

Don't you see the unfairness? Let's say that due to life circumstances, you were only able to login to the game at midday to play. This change means that your contribution is suddenly worth a lot less than if you had logged in at 7pm.

You don't know that your contribution will be worth less unless you know the exact number of players that played during prime time time slots and exact number of players that played during off hour time slot to do the calculations. If your team happened to have unusually more prime time players than what this new system tried to account for, your contribution might actually be worth more as an individual when you play during off hour than when you play during prime time. (Ex: 70+ players playing during prime time time slots and queing all maps while only 5-10 players playing during off hour time slots)

I know it can be counterintuitive, but you gotta calculate by individual players per the new system, not by one single conglomerate.

Edited July 27 by A Hamster.2580