Guild Wars 2

Alright folks, just write your new super idea here and I will take the time to explain why it might or might not work.

My favorite idea is PPT cap; a server can’t tick higher than 255. So, you fight to lower your opponents’ score rather than fighting to increase your own score.

Feel free to try to destroy my idea too, if you can…

N.B. Try to keep it short

1. PPT stays the same for each match and determines the match winner.
2. Winner gets 1 point. The two losers get 0 points.
3. Server rankings are determined by the points from #2.

• This removes the reason to fight for 2nd place and encourages the two smaller servers to doubleteam the bigger one.
• …which helps alleviate the population gap between servers.

For seasons I agree with Kuru.

But for regular play some adjustment to the scoring would be necessary to encourage attacking the dominant server. Easiest would probably be to award more points for capturing the dominant servers assets.

Another thing I think they should try for non-season play is to seperate matches into time slices. And rate/rank each time slice seperately. And each time slice will be matched up against other severs who have similar ratings for that time slice. So in NA prime it might be AvBvC but in Ocx-Sea prime it might be AvDvF and in EU prime it might be AvCvH.

Hello, I will just copy-paste what I proposed in another thread, where it went completely ignored but I think it’s somehow in the right direction:

- Scoring SHOULD be influenced by the total amout of people that each server individually has in WvW at each time (not divided per map, it would be nonsense since border hopping is a reality), in a way that (in a simplified WvW with just 2 servers facing each other) if a server has 60% players in WvW of its total capacity and another has 40% players in WvW of its total capacity then if the first server owns a number of structures that would give him the 60% of the total score (60% of 695), and the other owns 40%, then they both get THE SAME PPT (excluding bloodlust/finishing points) because they are both giving their full potential and they should be equally rewarded for that.
- all fixed Time-based score influencing that has been suggested is a bad idea, there’s no such thing as a more important time slice, as some of you have suggested, which leaves me baffled.
- any static way of influencing the score is bad, the system needs to be dynamic and adjusted as soon as more people log in and/or out of WvW (with some buffering to not pressure servers with too much new real-time work to do)
PROS:
- Fair scoring around the clock even in case of unbalanced matchups !!!
- others to be thought of (I’m lazy and feel I’ve done enough for today with achieving the previous point)
CONS:
- there will be people whining stuff like “oh come on, we have nearly everything in every map and we are just scoring a tad more than our opponents”.. to which a good GG for fighting no opponents is well deserved.. :/ (and well, minor gg to aNet for giving the guys a very unbalanced matchup too)
- probably not any other unlaughable one.

fedecane.6259:

Hello, I will just copy-paste what I proposed in another thread, where it went completely ignored but I think it’s somehow in the right direction:

- Scoring SHOULD be influenced by the total amout of people that each server individually has in WvW at each time (not divided per map, it would be nonsense since border hopping is a reality), in a way that (in a simplified WvW with just 2 servers facing each other) if a server has 60% players in WvW of its total capacity and another has 40% players in WvW of its total capacity then if the first server owns a number of structures that would give him the 60% of the total score (60% of 695), and the other owns 40%, then they both get THE SAME PPT (excluding bloodlust/finishing points) because they are both giving their full potential and they should be equally rewarded for that.
- all fixed Time-based score influencing that has been suggested is a bad idea, there’s no such thing as a more important time slice, as some of you have suggested, which leaves me baffled.
- any static way of influencing the score is bad, the system needs to be dynamic and adjusted as soon as more people log in and/or out of WvW (with some buffering to not pressure servers with too much new real-time work to do)
PROS:
- Fair scoring around the clock even in case of unbalanced matchups !!!
- others to be thought of (I’m lazy and feel I’ve done enough for today with achieving the previous point)
CONS:
- there will be people whining stuff like “oh come on, we have nearly everything in every map and we are just scoring a tad more than our opponents”.. to which a good GG for fighting no opponents is well deserved.. :/ (and well, minor gg to aNet for giving the guys a very unbalanced matchup too)
- probably not any other unlaughable one.

Which is exactly as it should be (bolded above). PvDoor and points earned over time should not be a measure of success in a PvP arena. Those who claim otherwise are those who benefit from the current scoring mechanic by being rewarded for outnumbering their opponents; nothing more.

Nor attempt to claim there’s skill involved in outnumbering an opponent; even indirectly. A stacked server isn’t stacked because that server is uber pro or better at recruitment. The reasons for why some servers have large populations and others don’t are legion and have nothing to do with skill. For example, players may have joined because they liked the server name. The snowball effect then takes over and players flock to higher population servers, not wanting to be on a lower population server with fewer people to play with.

Then, with the current broken scoring system that rewards numbers above all else, such servers win WvW; but not because they’re skilled. I play on a server in the middle of the EU pack and occasionally we get matched against a rank 2 or 3 server. We routinely wipe players from such servers over and over; their PvP skills are just so bad. All they have are numbers; there is no skill involved. Unfortunately, ArenaNet – in their infinite wisdom – settled on a scoring mechanism that encourages the antithesis of skilled PvP play; human wave tactics and PvDoor.

The argument that there needs to be given equal consideration to players who want to “play their way” is a bunch of BS. That may have held true while there was only one realm vs. realm arena. With the advent of EotM, though, the karma training zerglings who want to play PvDoor for mad lootz are now being catered to in two arenas; EotM and WvW.

Perhaps it’s time to finally separate the two playstyles. EotM for those who want to k-train doors and NPCs and WvW for those who want deeper strategy and PvP. But that’s not going to happen until the scoring mechanic is fixed and disincentives are put into place for blobbing in WvW.

Suggested this in feedback threads and forum threads many times.

Adding Tennis/NA Playoff rules

Retain 24/7 match. Break up each 24 hour day into three 8 hour sets. Each 8 hour set is like a mini-match. A score is kept for the 8 hour set and the highest score wins the set. This gives the winning server of the 8 hour set a Win Point. As well, there is a running score for the entire week. The week-long score counts for one Win Point as well. At the end of the week, the server with the most Win Points then wins the match.

Imagined differences in outcomes:

Blue v Red v Green
Old System:
Blue and Green play evenly during prime time. Red falls behind. During off-peak hours, Green has maybe 10 players. Red has maybe 25. Blue is fielding 60 players during off-peak hours. Blue karma train rolls through the maps at night.

Blue and Green go from a close match, to Blue having a sizable lead. After the weekend, Blue has a 20k point lead. Neither Red or Green can dominate enough during prime-time hours to make up for this. Blues lead grows expotentially. Match becomes a blow out.

New System:

Blue and Green play evenly during prime time. Red falls behind. During off-peak hours, Green has maybe 10 players. Red has maybe 25. Blue is fielding 60 players during off-peak hours. Blue karma train rolls through the maps at night.

Green and Blue split the prime time hour fights over the three days of the weekend. Blue wins the night sets. At the end of the weekend, Blue 6 win points and Green has 3. Green works hard to win the prime time sets the next few days. At the sixth day, Blue has 9 win points and Green has 8 win points. Red manages to win an early day set and get 1 win pont. Last day, Green pushes 24/7 to try to get the final 3 win points. Blue actually has to work to win.

The main difference is it would make comebacks possible even if you don’t have night coverage. Currently, if you have any semi-signifigant coverage gap, you aren’t coming back. This makes many matchups boring. The new system would allow comebacks despite coverage gaps. Because of this, coverage wars would matter much less. Since any set is worth the same win points, no time of day is worth more than the other time of day. The only advantage is total week points, which counts for only one Win Point.

This can be adjusted to allow win points for 2nd place, but more for 1st place within a set. The only thing to worry about is making sure there wouldn’t be ties. The week long score Win Point is there to help break possible ties.

Second Proposal:
The outmanned buff is now almost entirely meaningless with the removal of repair costs. Instead of changing the outmanned buff on players, I would add a buff/debuff to objectives.

Strained Resources Debuff:
Any time a server has X amount of lead, any new Tower or Keep they take is given the Strained Resources Debuff for 20 minutes. Whenever a losing server takes a Tower or Keep with the Strained Resource Debuff, that Tower or Keep is worth 50% more PPT for the next 45 minutes.

Result:

Karma train flipping by a stomp-rolling server can be used by the other servers to gain points. Retaking a Keep or Tower with the debuff and defending it will get the server more PPT than the leading server constantly flipping it. Servers can still karma train for loot and karma if they want, but it could cost them PPT if they aren’t careful.

This change should also help incentivize defending.

I’ll clarify/mod the points from winner more:

Points from #2 should include:
1) Points rewarded for capping structure/sites from 1st place. Different structure weird differently.
2) Bloodlust should be adjusted so only 2nd and 3rd place per current score are able to use. This should be modified to be strictly be only usable against 1st place. 1st place does not need to be able to use bloodlust to even more steamroll the competition.

The above and below will help promote a more exciting and even matchups!

Lord Kuru.3685:

1. PPT stays the same for each match and determines the match winner.
2. Winner gets 1 point. The two losers get 0 points.
3. Server rankings are determined by the points from #2.

• This removes the reason to fight for 2nd place and encourages the two smaller servers to doubleteam the bigger one.
• …which helps alleviate the population gap between servers.

New Scoring System for WvW

The following is an amalgamation of my own ideas as well as the suggestions of others throughout the forum for improving the state of the WvW scoring mechanics. I’m not seeking to take credit for all of them as my own. Rather, I’m attempting to pull all of what I consider the best ideas into a unified whole. Feel free to quote anything I write below and claim credit for it if was truly your idea to begin with (preferably with a link to where you first posted it on the site, of course).

I’ll individually address each aspect of the WvW game mode from which World Score points would be earned (or not earned) as well as the reward structure under this proposed system.

Player Kills

• Defeating a player without stomping them awards 1 World Score point.

• Defeating a player with a stomp awards 2 World Score points.

Defeating a player by either method awards the same amount of loot. This is in recognition of the fact that some classes have an inherent advantage over others when it comes to stomps.

The loot for defeating players will be substantially increased over current levels for two reasons:

1) To compensate for the elimination of several of the current loot sources per my further suggestions (as will be detailed throughout this post).

2) To incentivize PvP in a PvP arena.

Borderlands Bloodlust

Currently, this buff awards:

“An additional World Score point for each defender finished while any bloodlust is active.”

This portion of the buff just quoted will be eliminated. The means by which Borderlands Bloodlust is acquired is just another mechanic that favors numerical superiority. To then award on top of that additional World Score points for finishing moves is grossly unbalanced.

The buffs to power, precision, toughness, vitality, condition damage, and healing power will remain in effect or a different set of buffs could be developed.

Outnumbered

As armor repair costs will be eliminated from the game with the April 15th Feature Pack, having no armor repair costs as a benefit of the Outnumbered buff will become obsolete.

Under my proposed system, “no armor costs” will be replaced by a bonus to World Score points from player kills. While Outnumbered:

• Defeating a player without stomping them awards 2 World Score points.

• Defeating a player with a stomp awards 3 World Score points.

It may also be appropriate to consider increasing the personal loot awarded to players on making a successful kill as a benefit of the Outnumbered buff. This is to recognize that outnumbered players are at a numerical disadvantage. If they score a kill under such conditions, they and their server are given a greater reward for having beaten the odds.

Killing NPCs

No reward which can directly or indirectly be converted to gold will be given for killing NPCs in WvW; not karma, not coins, not badges, not laurels, not materials, not siege blueprints, nothing.

The rewards previously associated with killing a Veteran Supervisor at a supply camp and Champion or Legendary Lord at a tower, keep, or Stonemist Castle will now only be rewarded on successfully capturing the objectives associated with these NPCs and only when such objectives are actively defended by real players (see below).

The only possible exceptions to this are awarding the following on killing an NPC:

1) Personal experience.

2) WXP.

3) Maybe account-bound items such as Empyreal Fragments and the like.

Dolyaks

The current World Score points for killing Dolyaks will remain intact.

Under the new system, World Score points will also be rewarded for Dolyaks which successfully deliver supplies to their destination:

• For each destination along a route to which a Dolyak successfully delivers supply, 1 World Score point will be awarded if the supply camp from which it traveled is no longer under the control of the server to which the Dolyak is delivering supplies (i.e. the supply camp flips while the Dolyak is in transit but it still manages to make its delivery).

• For each destination along a route to which a Dolyak successfully delivers supply, 2 World Score points will be awarded if the supply camp from which it traveled is still under the control of the server to which the Dolyak is delivering supplies (i.e. the supply camp and the destination are both under the control of the same server when the Dolyak makes its delivery).

The question of whether or not escorting and/or defending Dolyaks should reward something in addition to World Score points is a thorny one. On the one hand, we have as precedent the escort bots scripted to follow the Dolyaks when the game was first released and such rewards still existed; we don’t want a repeat of that.

(continued)

(continued)

On the other hand, those who engage in defense, support, and escort are a vital part of the war effort but go largely unrewarded; they really need to have their efforts recognized. I’ll leave this up to ArenaNet to come up with a workable solution.

Sentries

The current World Score rewards for killing/capturing Sentries will remain intact.

Upgrading Objectives and Points Per Tick (PPT)

• An objective – supply camp, tower, keep, and Stonemist Castle – with no upgrades completed (not just on order; completed) earns NO PPT for the owning server.

• As upgrades are completed at objectives they will earn increasingly greater amounts of PPT. This doesn’t have to correspond to the current amount of PPT that objectives are worth; perhaps re-evaluating the amount of PPT each objective contributes is in order. Regardless, I’ll use a supply camp in the context of its current PPT value as an example:

1) Server captures the supply camp. It earns no PPT for the server while it has no upgrades completed.

2) Supply camp is upgraded with “Hire Caravan Guard”. When this upgrade completes, the camp will then start earning the owning server 1 PPT.

3) Supply camp is upgraded with “Increase Supply Deliveries”. When this upgrade completes, the camp will then start earning the owning server 2 PPT.

4) Supply camp is upgraded with “Hire Additional Guards”. When this upgrade completes, the camp will then start earning the owning server 3 PPT.

5) Supply camp is upgraded with “Strengthen Guards”. When this final upgrade completes, the camp will then start earning the owning server 5 PPT. The extra point of PPT is a bonus for having completed all upgrades at the supply camp (and conveniently aligning its PPT value with what supply camps are currently worth).

For reasons I’ve detailed at great length in other posts, personally I’d like to see PPT eliminated entirely (TL;DR: it’s not a measure of skill but, rather, a scoring system that rewards grossly outnumbering your opponent, PvDoor, and absentee landlordism).

I begrudgingly accept that the elimination of PPT is unlikely to ever happen. However, with this compromise, it only awards PPT to servers actively engaging in upgrading and defending what they own rather than karma training and leaving objectives empty and un-upgraded.

The other advantage this suggestion offers is – if servers want to earn any PPT points – they will have to more carefully manage their supply and balance it between offense and defense.

No longer will capturing an objective and raiding it of all its supply allow servers to double-dip by having both an objective earning them passive PPT and a fresh source of supply with which to run to the next objective and build siege weapons. They must make mutually exclusive cost : benefit choices:

• Order an upgrade at the recently captured objective in an effort to have it produce PPT for the server?

OR

• Raid the supply depot, abandon the objective (which will not produce any PPT), and run to the next objective to build siege with the supply obtained?

Of course, they could choose to do both; but they will have to put at least some focus into and assign some personnel the tasks of escorting Dolyaks, upgrading, and defending what they capture rather than abandoning it to continue karma training.

In short, if servers want to earn PPT, they’re going to have to put forth an active effort to do so beyond the current method of rapidly capturing territory and earning PPT passively from empty and un-upgraded objectives.

Capturing Empty, Undefended Objectives

Capturing an empty objective (supply camp, tower, keep, and Stonemist Castle) in which:

• no real players are present at any point from the time the objective becomes contested to the time it is captured,

• and with no significant break between “Defend the X” events triggered at the objective (in other words, a more-or-less continuous assault from the time it’s contested to the time it’s captured),

will award ZERO World Score points, ZERO guild influence, and ZERO rewards which can directly or indirectly be converted to gold; not karma, not coins, not badges, not laurels, not materials, not siege blueprints, nothing.

The only possible exceptions to this are:

1) Personal experience.

2) WXP.

3) Maybe account-bound items such as Empyreal Fragments and the like.

(continued)

(continued)

The intent here is to completely disincentivize karma training and objective flipping. I concede personal experience and WXP as rewards to give those who are new to WvW an opportunity to earn something for their efforts without being thrown right into a meat grinder of a PvP clash.

Such players could flip empty objectives while learning the basics of WvW at a slower pace until they’re ready to start facing real opponents. In the process, they’d be able to level their character and their WvW ranks.

I’m loathe to include #3 as there may come a day when account-bound rewards can be converted to gold by one method or another (reincentivizing karma training again). I include it anyway because I realize what a tough sell it will be to convince ArenaNet to so drastically reduce the rewards for this particular activity in the first place. The compromise is to keep them account bound, though.

Capturing Actively Defended Objectives

An actively defended objective is defined here as:

• any supply camp, tower, keep, or Stonemist Castle

• which has a real player(s) inside who is/are member(s) of the server which owns the objective

• and is in a contested state by any enemy action.

The state of being “actively defended” goes into effect as soon as a real player is present within the objective while it is in a contested state (e.g. a previously empty objective would become “actively defended” if a defender enters it any point while it is still contested). This state of being considered “actively defended” persists even if the real player defender(s) abandon the objective at some point during the siege.

On a successful capture of an actively defended objective, the victorious server is awarded World Score points equal to a base value plus the PPT value of the objective as determined by the number of upgrades it possesses.

For example, let’s assume capturing an un-upgraded supply camp awards 5 World Score points (but only if actively defended; an undefended supply camp would award ZERO points). Using the list from above in the “Upgrading Objectives and Points Per Tick (PPT)” section, a supply camp with two fully completed upgrades would award the capturing server 7 World Score points (the base 5 points + 2 points for the upgrades = 7 points).

On successful capture, the victors are also given personal loot equal to the current rewards for having captured an objective and defeating it’s primary defender NPC (Veteran Supervisor or Champion/Legendary Lord). Recall that earlier I had suggested eliminating rewards for killing NPCs; this is the compensation mechanism. Personal experience, WXP, karma, and guild influence are handed out according to the current schedule for such rewards, as well.

In short, all the current rewards for capturing objectives plus World Score points will be awarded to the victor; but only if the objective is actively defended at some point by real players.

Actively Defending Objectives

For each “Defend the X” event triggered at an objective in which real players of the defending server are present and which runs its full 3 minute course, the defending server is awarded World Score points.

In recognition of the fact the defenders have the advantage, these World Score points will be a fraction of what an attacker would earn on successfully capturing that same objective. I suggest a 10-to-1 ratio for starters, to be tweaked as testing and balance dictate.

For example, let’s say there’s a keep which – between its base capture value plus the value of its completed upgrades – would be worth 50 points if captured. The defending server would earn 5 World Score points for each “Defend the X” event that runs its full course while the keep is contested if there are real players of the defending server present at some point during each “Defend the X” event.

• The keep comes under attack.

• There are real player defenders inside.

• A “Defend the Keep” event triggers.

• The defenders are all killed during this first “Defend the Keep” event.

• Three minutes later, the defending server earns 5 World Score points because there were real player defenders present at some point during the event.

• The attackers maintain their siege and a second “Defend the Keep” event triggers.

• The defenders respawned but were unable to reach the keep before the second “Defend the Keep” event expired. The defending server earns no World Score points from this second event.

• The attackers maintain their siege and a third “Defend the Keep” event triggers.

• The defenders manage to slip inside the keep while the third “Defend the Keep” event is active.

• Regardless of the outcome, the defending server will earn another 5 World Score points for “Defend the Keep” event number three on account of having had real players inside the keep during the event.

(continued)

(continued)

I also suggest some small reward be given to players who participate in defense events. Again, it needs to scale to a much lower amount/value in comparison to the rewards given to the attackers. This is to account for the fact that defenders can participate in multiple “Defend the X” events over the course of a single siege while attackers have to contend with an all-or-nothing scenario (they capture the objective and are rewarded, or they fail and get nothing).

To those who would critique the above as opening up a potential for exploiting or gaming the scoring system, I would counter by saying it would not. The reasons are as follows:

The most likely attempt at exploiting the system would be for a server to pay the fee to transfer one of its members to an enemy server they are facing in the current match.

The transferee would then contest their home server’s objectives in which their comrades are camped in an effort to trigger “Defend the X” events for the express purpose of farming free World Score points using the mechanics outlined above.

The reason this will not work is – on transfer – a player is barred from playing WvW in both the current match and the following week’s match. It becomes physically impossible to engage in these kinds of shenanigans on short notice.

Of course, a server could simply pay to send such exploiters to all servers they may potentially face far in advance of any matches. These sleeper agents would then remain camped on the opposing servers for weeks or months on end waiting for a WvW match up with their home server.

Could it happen? Perhaps. Would it happen? Unlikely, since it would mean the sleeper agents are effectively exiled from ever playing for their home server on the rare chance they may get to exploit the scoring system on their behalf.

The Elephant in the Room

No discussion of redesigning the scoring system would be complete without also addressing the issue of transfers, server stacking, and monetization.

As it stands now, I would hazard a guess that most WvW-ers have less interest in cosmetic gem store items than their PvE brethren. As such, ArenaNet must find other methods for monetizing WvW. At present, transfer fees appear to be that method.

Unfortunately, we see the consequences of ArenaNet’s current transfer implementation in conjunction with their WvW scoring and rewards systems. Players flock to high population servers to better their odds at earning the rewards, which then induces more players to join those same servers in a self-fulfilling cycle.

The gross imbalances this brings to WvW in both population and coverage results in the disparities we see today, leading to suggestions of redesigning WvW scoring mechanics in an attempt at redress.

However, until we suggest a method by which ArenaNet can profit from WvW while avoiding the scenario just outlined, the situation will continue unchanged. I don’t yet have such a suggestion. I’m simply throwing it out there in an attempt to stimulate a dialog on the issue. How do we even out the WvW populations while simultaneously monetizing this game mode for ArenaNet?

Thank you for reading.

Lord Kuru.3685:

1. PPT stays the same for each match and determines the match winner.
2. Winner gets 1 point. The two losers get 0 points.
3. Server rankings are determined by the points from #2.

• This removes the reason to fight for 2nd place and encourages the two smaller servers to doubleteam the bigger one.
• …which helps alleviate the population gap between servers.

This idea would be an interesting experiment

I say experiment because it will probably not be a viable system to rank the servers. Here’s why :

1. It’s too dependent on the player reactions to this new ranking system. You make a big _assumption that the players will completely change the way they play if we introduce this scoring/ranking system. Will players really start to double team the biggest server or most just don’t care because they only play for rewards/fun. Fighting the winning server is not always rewarding or fun, especially when they outnumber you very bad. It makes taking any structure from them very hard or simply winning zerg fight nearly impossible. The 2nd is an easier target for 3rd and it’s also much funnier to play against an opponent where you stand a chance.

I think the quality of the gameplay provided by fighting an opponent well matched with you is what influence most players to fight for 2nd rather than the ranks. But, it would be very interesting to see how it turns out.

2. You still keep PPT unchanged. Sure you expect that the players will drastically change the way they play and focus the currently winning server fixing the population imbalance. But if that doesn’t happen, simply because 3rd prefer to have good interesting fight with 2nd rather than being wiped non stop by 1st, then nothing is fixed. We still have the broken PPT system.

3. What about the ranking in general? How do you rank the server exactly according to your system. Currently the server are ranked by their glicko rating. If you change taht by point you have to provide a new ranking system.

What about the rank of 2nd and 3rd? They both have 0 points… 2nd might actually be nearly competitive with 1st while 3rd might be better off in the tier below. Or maybe 2nd and 3rd are nearly equal in strength but 1st is a lot stronger and should move up in tier. Or maybe it’s a very balanced match up and we should keep pairing these servers together.

Too much is not explained in this idea to see if it would really work or not and it also makes big _assumptions.

So for now, unless the missing details are explained, I would say it’s not a good idea.

fedecane.6259:

Hello, I will just copy-paste what I proposed in another thread, where it went completely ignored but I think it’s somehow in the right direction:

- Scoring SHOULD be influenced by the total amout of people that each server individually has in WvW at each time (not divided per map, it would be nonsense since border hopping is a reality), in a way that (in a simplified WvW with just 2 servers facing each other) if a server has 60% players in WvW of its total capacity and another has 40% players in WvW of its total capacity then if the first server owns a number of structures that would give him the 60% of the total score (60% of 695), and the other owns 40%, then they both get THE SAME PPT (excluding bloodlust/finishing points) because they are both giving their full potential and they should be equally rewarded for that.

Population based solution! I have been a huge fan of these types of solutions in the past until I realized a few very important problems

Let’s start by looking at the current PPT system to understand the problems. The current scoring system, PPT, is actually quite good to rank the server according to their strength in wvw. When a server is stronger it wins in the score, when a server is weaker it loses in the score. It’s very good to kitten the strength of every servers.

So, where does it fail and why so many people want to change it? Well, maybe it is too good for it’s own good… The problem is that when a server is stronger, it is able to build a big lead in the score really fast and probably even too fast. It’s so fast that after only 1-2 days most match up are already decided and there is no point to keep trying for the rest of the week.

Another big problem is that it gives huge power to certain timezone that vastly outnumber their opponents…. But back to your idea

1. Making the scoring dependent on population might produce unexpected results. I play on a lot of different servers and the population vary a lot. Even in the same tier between 2 opponents that perform equally well. That is to say that the “skills” of the players vary a lot. We often say that the population win the match up but it’s rather the effective population. Some server are able to do less with more and some server do more with less. It’s just how it is and I bet a couple of players would not like to learn that number wise they suck.

Arena Net only gave hint to population numbers 1 time if I recall correct. Call me crazy on that one, but before the league started Devon Carver said that some server might stand a chance fighting against server way higher than theirs : example BP in bottom tier 4 at the time and Dragon Brand middle tier 2 at the time. Having played on both servers, I can tell that strangely the wvw population looked similar on both servers…

Let’s see how HoD perform against FA next week… HoD probably have more numbers than FA currently since they can queue many maps at a time. If HoD beats FA, we can review this counter argument.

2. Arena Net doesn’t want to release population information. I don’t know the exact reason but they don’t. If we tied the score to the population in some way, people will be able to calculate the relative population of each servers and Arena Net doesn’t want that.

3. Your system is probably too fair. This is the biggest issue. As the current one is too good at finding the strongest server, yours help too much the losing servers. We need to kitten the strength of the servers. If a server is losing very badly, it’s a not a very good idea to magically boost it’s score to give him the opportunity to win.

Even if a server has a better score/population ratio, if it’s getting destroyed in the maps, it should also get destroyed in the score. If they don’t they will keep being match with opponents they have nothing to do with.

Take for example Sanctum of Rall. Very recently they were Tier 1 and a good tier 1 server. But then they start losing a lot of guilds and began to get destroyed in all the maps. Now let’s imagine the remaining players had a good score/population ratio. Then Sanctum of Rall would still be competitive with the other tier 1 server. It doesn’t matter if they tick at 50 ppt all week long, they will stay in tier 1.

That scenario would suck for everyone. If you get destroyed for whatever reason then you should move down. If you roflstomp your opponents then you should move up. Servers need to be match against server they are competitive with.

Conclusion

For the reasons mentioned above, this idea is not a good one.

Johje Holan.4607:

Another thing I think they should try for non-season play is to seperate matches into time slices. And rate/rank each time slice seperately. And each time slice will be matched up against other severs who have similar ratings for that time slice. So in NA prime it might be AvBvC but in Ocx-Sea prime it might be AvDvF and in EU prime it might be AvCvH.

I have also been a big fan of time slice match up a long time ago. Even made a thread about it

The system certainly have a very good point :

You won’t be able to find another system that match the server according to their respective strength better than this one.

It’s simply the best to provide the most balanced match up. It also has the huge benefits to give more power to the players since they can have a bigger effect to the match up in which they usually play since it last only 8 hours instead of 24.

But it also has a couple of flaws

1. It’s probably very hard to implement compared to other scoring/ranking/match up system. Devon kinda hint at this point when he mentioned the 21 matches in a week scoring system in last CDI. Looking at how hard it seems (for them) to provide color change to the commander tag, I wouldn’t expect to a change like that before 2020.

2. It brings more downtime to wvw. Every 8 hours the match up are basically resetting. This means that for at least 5 minutes wvw will be down. For most it’s not a problem but if the 5 minutes is right in the middle of the 1 hour you are able to play wvw during week day, it would certainly suck.

3. To continue on point 2, having multiple resets further reduce the effective play time. Will you try to attack a big fortified tier 3 keep that might take 1 hour to take if there is only 45 minutes left in the match up? Probably not. This means that wvw will be “dead” way longer than the 5 minutes down time for reset.

4. Is it worth to play in a match up that is already decided? If you log in at 6 hours in the match up of 8 hours. Chance are that you already know if you will lose or win this particular match up. So, if your play time is near the end of a match up, you might begin to think that your contribution doesn’t matter. But you will always be there for the final karma train

5. If your play time is in the middle of 2 match up, will you want to actively play in both? For example, one time zone might be very strong while the other is very weak. Do you want to play in both or would you only care about a particular one. I think this is a good question to ask ourselves if we decide to implement any system that split the day in a couple of matches.

Conclusion

It would provide the most competitive server matching and give more power to the players but it has a couple of flaws. So this is an interesting scoring/ranking system, but maybe a bit complicated to implement.

A thing that I would absolutely keep if we ever do it, would be to keep the upgrade between 2 consecutive matches : you restart with the upgrades from the previous day when the next day start.

etiolate.9185:

Retain 24/7 match. Break up each 24 hour day into three 8 hour sets. Each 8 hour set is like a mini-match. A score is kept for the 8 hour set and the highest score wins the set. This gives the winning server of the 8 hour set a Win Point. As well, there is a running score for the entire week. The week-long score counts for one Win Point as well. At the end of the week, the server with the most Win Points then wins the match.

It’s a good idea and it would produce a result nearly identical to PPT cap. But it also has some flaws that PPT cap doesn’t suffer from :

1. The players that have their play time right in the middle of 2 time slot might feel like their contribution has less impact. For example, they might perform very well for the last 2 hours of a match up than very well for the first 2 hours of the next match up but still lose both match up. It might feel to them like all their work is loss because of this 8 hours match up system.

2. It might be too fair. It’s not as drastic as the population based solution but this one might also produce unwanted rankings.

For example, we have server A, B and C. A wins match up 1 with average PPT of 250, B wins match 2 with average PPT of 250 and C wins match up 3 with average PPT of 600.

Obviously, with the current ranking system, C would most likely win the match up. But with your system, it would have exactly the same number of points as A and B until the end when the final point is awarded. This doesn’t look like a very convincing win from C.

If a server is suppose to win, it should win. Same thing if a server is suppose to lose, it should lose. PPT cap does right that but it also make it more fair for the currently losing servers, hence it improves the current situation while not introducing new problems.

3. You need a new ranking/matching system. You basically get rid of the glicko ratings so what will you use now? How do you differentiate between the performance of A and B in the previous example, they have the same number of points… This question needs an answer before we can take an advised decision on this scoring system.

etiolate.9185:

Second Proposal:
The outmanned buff is now almost entirely meaningless with the removal of repair costs. Instead of changing the outmanned buff on players, I would add a buff/debuff to objectives.

Strained Resources Debuff:
Any time a server has X amount of lead, any new Tower or Keep they take is given the Strained Resources Debuff for 20 minutes. Whenever a losing server takes a Tower or Keep with the Strained Resource Debuff, that Tower or Keep is worth 50% more PPT for the next 45 minutes.

I’m all in for giving more utility to the outnumbered buff, but the question is what to do? I think you came up with a good idea here.

This system encourage a server to actively retake their territory as fast as possible after the big bad karma train winning server has come and flipped everything. I like it because it encourages active play, it’s simple and well defined.

Also, it uses what is already in place and bring an improvement, so we don’t have to reinvent the wheel here.

It might also encourage server to team up against the winning server since they will gain more points by taking the structure they just took. This, as you said, might also mitigate the power of the karma train and force them to defend for at least 20 minutes. I think the 20 minutes is a good idea too because after 20 minutes the winning server will likely stop defending as actively giving a little chance to the losing server to retake their territory.

Conclusion

Probably the best idea mentioned so far in my opinion.

A bonus: have you considered merging it with PPT cap? The 2 can work well together.

Gudradain.3892:

1. Making the scoring dependent on population might produce unexpected results. I play on a lot of different servers and the population vary a lot. Even in the same tier between 2 opponents that perform equally well. That is to say that the “skills” of the players vary a lot. We often say that the population win the match up but it’s rather the effective population. Some server are able to do less with more and some server do more with less. It’s just how it is and I bet a couple of players would not like to learn that number wise they suck.

Arena Net only gave hint to population numbers 1 time if I recall correct. Call me crazy on that one, but before the league started Devon Carver said that some server might stand a chance fighting against server way higher than theirs : example BP in bottom tier 4 at the time and Dragon Brand middle tier 2 at the time. Having played on both servers, I can tell that strangely the wvw population looked similar on both servers…

Let’s see how HoD perform against FA next week… HoD probably have more numbers than FA currently since they can queue many maps at a time. If HoD beats FA, we can review this counter argument.

2. Arena Net doesn’t want to release population information. I don’t know the exact reason but they don’t. If we tied the score to the population in some way, people will be able to calculate the relative population of each servers and Arena Net doesn’t want that.

I’m talking about the instantaneous number of people that each server has in WvW. The number of people that are in a WvW map fighting for your server (not counting EoTM, it’s a PVE map, you know). Server population has little to do with this.

Gudradain.3892:

3. Your system is probably too fair. This is the biggest issue.

Wasn’t fairness the target of this whole discussion? I mean, if we wanted a system in which stacking servers is encouraged and there’s no chance for a low active WvW population server to do anything in WvW beside roaming / small scale fights I think we don’t really need to change anything!

Gudradain.3892:

As the current one is too good at finding the strongest server, yours help too much the losing servers. We need to kitten the strength of the servers. If a server is losing very badly, it’s a not a very good idea to magically boost it’s score to give him the opportunity to win.

The current system is too good at finding the server with the bigger active WvW population around the clock, and that’s it. It has little to do with strength, skillful gameplay, tactics and such, that SHOULD be the metrics to define a server stronger than another. Up to now a server is stronger than another if it has better coverage.

And I never spoke of magically boosting the score. What I meant is that if 2 servers have a big gap in active wvw population at one time, the one that outnumbers should naturally be able to get a major percentage of the map and it should not be rewarded for doing so, because it’s just normal.
And if in this situation the outnumbered server manages to capture a tower, this will grant this server more PPT than the same tower would grant to the outnumbering opponent.

This also encourages to defend stuff from both sides and gives value to tactics like sending few people to capture camps around the map. And completely favours splitting zergs around the map to cover a major part of it.

And most importantly this system would promote skillful gameplay, strategy and “healthy” metrics to compare different servers while vastly reducing the benefit of coverage and stacking servers.
In fact I would say that this would promote destacking and splitting the population equally among servers (because seeing how many slackers there are in WvW these days in T1/T2, people looking for good gameplay will look forward to move to a server without said slackers, and that’s where T8/T7/T6 servers shine! but now you can’t even consider it because with the current PPT system you will just get destroyed every week).

The server who gets the most Karma during the week wins.

A lot of good ideas here and I agree that the scoring system is the biggest flaw in WvW.

Priority No. 1 should be a flexible “tick” rate based on the current player activity.

Assuming each WvW map can host 500 players (165 per server) it would mean 2000 players will up all WvW maps.

When this is the case, the score should tick every 5 minutes instead of every 15 minutes.

When the map empties, the effect should be reversed so when only 100 players are active at 4 AM the score only ticks every 30 minutes or so. Much less frequently than during peak hours.

Additionally 1 kill, regardless of finisher, should award 1 warscore.

Optionally the Bloodlust buff could award 1-2 bonus points for finishers only. But the buff should only open up every 30 minutes so there’s a window of contention rather than constant capping and avoiding the enemy.

@Kraag Deadsoul

I will answer your massive post later today

fedecane.6259:

I’m talking about the instantaneous number of people that each server has in WvW. The number of people that are in a WvW map fighting for your server (not counting EoTM, it’s a PVE map, you know). Server population has little to do with this.

Gudradain.3892:

3. Your system is probably too fair. This is the biggest issue.

Wasn’t fairness the target of this whole discussion? I mean, if we wanted a system in which stacking servers is encouraged and there’s no chance for a low active WvW population server to do anything in WvW beside roaming / small scale fights I think we don’t really need to change anything!

Gudradain.3892:

As the current one is too good at finding the strongest server, yours help too much the losing servers. We need to kitten the strength of the servers. If a server is losing very badly, it’s a not a very good idea to magically boost it’s score to give him the opportunity to win.

The current system is too good at finding the server with the bigger active WvW population around the clock, and that’s it. It has little to do with strength, skillful gameplay, tactics and such, that SHOULD be the metrics to define a server stronger than another. Up to now a server is stronger than another if it has better coverage.

And I never spoke of magically boosting the score. What I meant is that if 2 servers have a big gap in active wvw population at one time, the one that outnumbers should naturally be able to get a major percentage of the map and it should not be rewarded for doing so, because it’s just normal.
And if in this situation the outnumbered server manages to capture a tower, this will grant this server more PPT than the same tower would grant to the outnumbering opponent.

This also encourages to defend stuff from both sides and gives value to tactics like sending few people to capture camps around the map. And completely favours splitting zergs around the map to cover a major part of it.

And most importantly this system would promote skillful gameplay, strategy and “healthy” metrics to compare different servers while vastly reducing the benefit of coverage and stacking servers.
In fact I would say that this would promote destacking and splitting the population equally among servers (because seeing how many slackers there are in WvW these days in T1/T2, people looking for good gameplay will look forward to move to a server without said slackers, and that’s where T8/T7/T6 servers shine! but now you can’t even consider it because with the current PPT system you will just get destroyed every week).

Sorry for not understanding it was instantaneous population instead of average population but the same things applied

1. Not all server are able to do the same things with the same numbers
2. Anet doesn’t want to release information on population number

But these are minor issues

The big one is that your system is too fair. I was a BIG BIG fan of population based scoring/ranking before. Until I realized one big issue : You don’t want to play with a server that completely destroy you or play with a server that you completely destroy.

With your system, Eredon Terrace, the last server, and Blackgate, the first server, could be competitive together score wise. That’s a huge NO. This should never happen. Eredon Terrace is not competitive with Blackgate and should not play against Blackgate. Any scoring system that make server that should NOT be competitive together, “magically” competitive are not good system.

When you lose population because guild transfer off your servers, you should go down in ranking. When you gain population you should go up in ranking. That we like it or not, there will always be population difference in WvW so we need a system that match server according to their strength.

Gudradain.3892:

etiolate.9185:

Retain 24/7 match. Break up each 24 hour day into three 8 hour sets. Each 8 hour set is like a mini-match. A score is kept for the 8 hour set and the highest score wins the set. This gives the winning server of the 8 hour set a Win Point. As well, there is a running score for the entire week. The week-long score counts for one Win Point as well. At the end of the week, the server with the most Win Points then wins the match.

It’s a good idea and it would produce a result nearly identical to PPT cap. But it also has some flaws that PPT cap doesn’t suffer from :

1. The players that have their play time right in the middle of 2 time slot might feel like their contribution has less impact. For example, they might perform very well for the last 2 hours of a match up than very well for the first 2 hours of the next match up but still lose both match up. It might feel to them like all their work is loss because of this 8 hours match up system.

2. It might be too fair. It’s not as drastic as the population based solution but this one might also produce unwanted rankings.

For example, we have server A, B and C. A wins match up 1 with average PPT of 250, B wins match 2 with average PPT of 250 and C wins match up 3 with average PPT of 600.

Obviously, with the current ranking system, C would most likely win the match up. But with your system, it would have exactly the same number of points as A and B until the end when the final point is awarded. This doesn’t look like a very convincing win from C.

If a server is suppose to win, it should win. Same thing if a server is suppose to lose, it should lose. PPT cap does right that but it also make it more fair for the currently losing servers, hence it improves the current situation while not introducing new problems.

3. You need a new ranking/matching system. You basically get rid of the glicko ratings so what will you use now? How do you differentiate between the performance of A and B in the previous example, they have the same number of points… This question needs an answer before we can take an advised decision on this scoring system.

1. This is possible. One thing I hadn’t decided upon is whether I think the maps should reset every 8 hours. So far, i felt this should not happen, so that whatever work someone was doing at the end of a reset would still continue on to the next. This is also a reward for working hard the previous 8 hour set.

2. To me, “who should win” is decided by the rules in which you play. In any sport or league, a handful of rule changes can drastically alter who should win. In the NFL, new rules to protect players and open up offenses has changed the idea of who should win in a match. In the NBA, hand-checking rules and new alterations to the illegal defense rules changes how the league played, and this changed the idea of who should win. Passing become more important in the NFL and post play became less important in the NBA.

The same concept applies here. Currently, dominating against outmanned servers during a significant time period would lead a server to thinking they should win. My rule changes are making the statement that taking empty keeps all night shouldn’t be the deciding factor. The statement of the rule change is that tightly contested times of day should be on par with uneven times of day. It’s the statement that all times of day should have a roughly equal contribution to who wins. So who should win will win according to the new scoring rules.

That’s sort of the point of this whole discussion.

3. I have felt that ANet should use the weeks between seasons to test out new ideas and new scoring methods. At first, I would hide glicko rating while testing this new method out. It should be a trial run to see how it alters players approach to the game and how well its received. If it’s well liked, then a glicko rating + win/loss performance could be used for ranking. If these are seasons, or if this tournament is a “playoff”, then the rest of the time could be trial periods and pre-seasons. Wins and losses and placings matter.

Hell, if having 24/7 coverage matters less, then you may see a shift in the mentality of player movement. Again, what we value and what rules we make dictate who should win and which servers are best.

Gudradain.3892:

Sorry for not understanding it was instantaneous population instead of average population but the same things applied

1. Not all server are able to do the same things with the same numbers
2. Anet doesn’t want to release information on population number

#1: …that’s kind of how you should distinguish a bad server from a good one.
There’s no reason why with the same numbers one server can’t do what another one instead pulls off.

#2: I think you still don’t get the part about the population… it’s not about server population but just how many people your server has in WvW maps at any given moment. And I never said this number needs to go public.

As for the rest, I think the system should be as fair as possible (also from a customer POV it should be like this. Imagine a new player choosing server randomly and ending up in one of the last ones, with the current system he would be stuck in WvW depression and will be forced to leave for a more stacked server. With what I proposed the server population eventually will balance itself out.) and like I proposed it, it is as fair as it can get.

Anyway it’s clear that there would be a lot of benefits in the long run.. but people want to zerg and to complain so I probably should even stop trying to suggest this sort of stuff.

You guys like…writing a book?

fedecane.6259:

#2: I think you still don’t get the part about the population… it’s not about server population but just how many people your server has in WvW maps at any given moment. And I never said this number needs to go public.

Yes I got it. If you make the score dependent on the population, people will be able to calculate the relative population of each servers. It doesn’t matter if it’s the number at any given moment or the average population over a week.

Anet released an API, that you can easily query to get the current PPT score (in fact you need to look at the list of all wvw objectives, verify who owns each objectives and then calculate the PPT for each server by knowing the PPT value of each different objectives). So, if you modify the PPT score based on population, people will be able to calculate it.

New_PPT = Old_PPT * population_modificator.

There is 3 variables in this equation. We know the New_PPT and Old_PPT, hence you can easily get population_modificator that will give you the relative population between server at any given time. If you store all this information in a database, you could even generates statistics about the relative population of server over any time period.

etiolate.9185:

1. This is possible. One thing I hadn’t decided upon is whether I think the maps should reset every 8 hours. So far, i felt this should not happen, so that whatever work someone was doing at the end of a reset would still continue on to the next. This is also a reward for working hard the previous 8 hour set.

2. To me, “who should win” is decided by the rules in which you play. In any sport or league, a handful of rule changes can drastically alter who should win. In the NFL, new rules to protect players and open up offenses has changed the idea of who should win in a match. In the NBA, hand-checking rules and new alterations to the illegal defense rules changes how the league played, and this changed the idea of who should win. Passing become more important in the NFL and post play became less important in the NBA.

The same concept applies here. Currently, dominating against outmanned servers during a significant time period would lead a server to thinking they should win. My rule changes are making the statement that taking empty keeps all night shouldn’t be the deciding factor. The statement of the rule change is that tightly contested times of day should be on par with uneven times of day. It’s the statement that all times of day should have a roughly equal contribution to who wins. So who should win will win according to the new scoring rules.

That’s sort of the point of this whole discussion.

3. I have felt that ANet should use the weeks between seasons to test out new ideas and new scoring methods. At first, I would hide glicko rating while testing this new method out. It should be a trial run to see how it alters players approach to the game and how well its received. If it’s well liked, then a glicko rating + win/loss performance could be used for ranking. If these are seasons, or if this tournament is a “playoff”, then the rest of the time could be trial periods and pre-seasons. Wins and losses and placings matter.

Hell, if having 24/7 coverage matters less, then you may see a shift in the mentality of player movement. Again, what we value and what rules we make dictate who should win and which servers are best.

1. True
2. Can’t disagree with that. The rules are what makes the sport.
3. You really need to define a good ranking system. The current system, PPT score + glicko2, is quite good to find the strength of each servers in WvW. (Considering that WvW is a competitive game mode where you try to control the most territory most of the time.) I just don’t like the RNG to decide the next match up… but that’s a topic for another discussion.

Your idea is good and Devon Carver even proposed it in the last CDI so it’s on their radar. I think it would be a nice change for WvW if we can find a good ranking system.

Good discussions and suggestions in here which are remaining respectful and adhering to the forum’s Rules of Conduct. Any chance we could get a dev to respond; even if it’s just, “Yes, we’ve read/are reading the thread.”, or even, “We have no intention of changing the WvW scoring system now or in the foreseeable future.” This way we’ll know whether there’s value in continuing to post or if we should drop the subject. Thank you.