Guild Wars 2

Switching maps is going to royally kitten up my reports. Different objectives with different ids.

But the people want their lag and blob fights.

Samuel Loretan.8601:

• I agree that asking new players to chose a World before doing anything in the game is odd, now that Worlds are a WvW concept. This isn’t something we can change easily, however, but I’ll keep it in mind if an occasion arises.

This seems to be at the heart of the matter. It may not be easy to change, but it may now be necessary. Disconnecting the WvW entities from the general community will allow the implementation of hard population caps and absolutely should be a part of any WvW 2.0 implementation. The metaserver has rendered the general concept of server meaningless.

Samuel Loretan.8601:

• I won’t detail the way the new algorithm works, because I want to minimize the risk of people trying to game it. That being said, I am confident it is robust against such attempts. It’s not looking at a short period of time, and it’s looking at many aspects of the ways players are interacting with WvW to determine a server population.

Security through obscurity never works. If the game mechanic itself cannot stop someone from gaming the system, the assumption should be that someone will be gaming it. Never underestimate the cleverness of your player base.

There is always talk about balancing server population but never any talk about how exactly population is calculated.

I know there is no hard cap for population. But without this Anet can never know the true activity potential of any server and any given time.

For example, a server that seems inactive for long periods of time may keep their population status superficially lower so they can still recruit and add population, and then during their push weeks they would outnumber their opponents. Given that WvW is won by population this could give some servers an advantage during seasons or other critical times.

Does Anet have a system to keep this type of thing from happening?

Offair.2563:

TorquedSoul.8097:

So can the old alpine maps hold as many players as the new maps? Or will the map queues be longer now?

That’s the 1 million dollar question. No one knows the exact numbers of players allowed. Anet never revealed that info.

It seems to me that the more players you have close to you, the greater the lag becomes. So I would expect the carrying capacity of those smaller maps to be less simply because of lag.

So can the old alpine maps hold as many players as the new maps? Or will the map queues be longer now?

When did the new WvW start?

Unless server link + blob = WvW 2.0.

Blackgate has the highest capture volume and highest overall activity level this week. Either they are pushing or they have the highest overall population.

Sea of Sorrows did much better than I expected. For some reason they are fighting more than they did over the past month. I would love some insight to what changed there.

I will post my estimated activity levels vs the actual activity levels for this week when I get the chart together.

EDIT: Capture volume was down globally, -10%. Kills were up globally, +45%.

cafard.8953:

Faux Play.6104:

Guild alliances are the main driver for WvW success. Why not scrap the concept of servers.

Because glory-seeking server-hopping guilds are causing the population balance problem in the first place. Making it easier for them to stack on the same side is the last thing we should be aiming for.

Servers are quasi-alliances already since guilds and pugs come and go.

The problem with servers is that there is no room for the players to create there own “server/alliance” and that servers do not have a hard population cap leading a steady stream of imbalanced competitions as guilds and players move around. This renders the glicko meaningless as we are seeing since the server linking. Most servers were not properly placed by the glicko because of rapid population shifts.

I think the biggest barrier to removing WvW servers is the loss of revenue that Anet would get hit with because alliances should not require transfer fees.

Alliances would actually make the ratings meaningful. Although they should ditch Glicko and just use Elo with a high K value to reflect the inherent volatility of the system.

Ven Zehn.6573:

Wait, so the server that has won 1st place in 2 out of 3 seasons, continues winning when their playerbase feels like it, counts as an oddity to you?

I wonder how the folks on that server feel about a hard population cap?

Teon.5168:

Jayne.9251:

Teon.5168:

I have a feeling this is coming from someone who hasn’t played a great deal of wvw/rvr in mmorpgs.

And then we get mad at Anet when they respond to these ideas, simply because they’re the most vocal.

I doubt that Anet will respond to this absurd idea, because most everyone in the thread thinks it is a stupid idea. Creating one thread about some bizarre idea the OP had while in a drug/alcohol induced haze is far from being ‘the most vocal’……lol.

This isn’t a new idea.
https://toughlovecritic.wordpress.com/2015/09/10/putting-guilds-back-into-guild-wars-wvw-alliances-part-1/

I like any system that would impose a hard population cap. The way that population is currently calculated can be gamed and lead to larger population imbalances.

Phantom.8130:

So….. kick PUGS out of WvW entirely…?

The suggested implementation of Alliances I read allowed for pugs (aka mercenary/militias) to join teams every week. I think the one I read required a certain percentage of pugs so that they wouldn’t be boxed out of the game.

These pugs allow for a recruiting pool for guilds.

Those that don’t like alliances are generally those servers that are gaming Anets rather poorly implemented population metric to overstack their servers.

If anet implemented alliances with a hard population cap, those people would not be able to win the population overstacking game. Right now winning in WvW is just a matter of population so the meta is to overstack.

Naithe.4271:

Thanks Torq! Appreciate the respect, but Rev is right on coverage.

No doubt. YB’s OCX appears lacking in a major way letting BG tick big points in the middle if the night. Just like YB used to do to you.

Not a big concern for YB as it looks like Anet is going to nerf OCX scoring soon rendering those guilds irreverence.

Svarty.8019:

TorquedSoul.8097:

There are ways they can be fair and fix the unbalanced passive scoring problems. They simply need to ditch the tick.

Then where is the incentive to recap anything when you are massively outnumbered? But of course, you have now moved on and never read this rebuttal, despite people typing it every time you make your suggestion ><

If you get rid of the tick you will still have a portion of the score coming from yaks which act as a contestable passive scoring element.

Did you comment in my scoring proposal thread?
https://forum-en.gw2archive.eu/forum/game/wuv/PPT-less-Scoring-model-Fights-and-Dolyaks/
I also address how fight activity could be scored to encourage engagement in the same thread. By making the fight portion fixed to 40% of territorial control score, it forces teams to fight or give up a large block of points.

There are ways they can be fair and fix the unbalanced passive scoring problems. They simply need to ditch the tick.

Naithe.4271:

We took Torque’s advice and decided to pop redbull and snort some crack.

Torque are we doing this right? Cool.

Absolutely. Thanks for the shout out.

It’s good to see that my words were sufficient motivation to get Blackgate back in the game after your long hiatus. As someone new to the game, it’s awesome that I can have such a strong impact on a veteran group of players. You’re Welcome. Anytime you need a boost, I’m here for you.

And I can’t condone the use of crack on GW2’s official forum, but if it helps you secure the win, I’m not one to judge. Whatever gets your fingers clicking.

I’m quite happy with the result. And while I don’t speak for YB or their commanders, I can say based on the chatter in teamspeak, they are happy to have competition again.

As they say, things are sweeter after the struggle and I am confident that Yak’s Bend has the strength of character to push through this population re-balancing intact and with a positive attitude.

Stay classy, and I will see you on the borderlands.

lil devils x.6071:

TorquedSoul.8097:

lil devils x.6071:

TorquedSoul.8097:

lil devils x.6071:

TorquedSoul.8097:

lil devils x.6071:

What needs to be on the poll is :

• Make scoring based on PPK over objectives rather than rewarding PvE game play in a PvP game mode, and I do not see that option there.

Why can’t it be both. As it stands PPK represent less than 15% of the overall score. I think the weight of fighting should be increased but territorial control should not be completely thrown out.

IMO.

As long as the fighting over the objectives is rewarded more than the objective themselves it will improve the way players are rewarded for their efforts, both offense and defense. PvD Should have never been part of the score in a game mode where you want to encourage players to fight one another instead of NPC’s.

The problem when you allow for " avoiding fights" to be rewarded is it will encourage players to do exactly that in a PvP game mode. What is the point of playing a PvP game mode if the players who do not do so are the ones who win?

Sorry, but I think that asymmetrical combat most certainly should be supported and sometimes that involves hit and run tactics and ducking fights.

The problem right now is that nearly 90% of the scoring can happen passively. Dropping that number to about 40%, and having fixed scoring for objective upgrades and captures, and boosting fights to about 40% will balance things nicely.

I like the idea of guerrilla warfare in WvW … and that revolves around objectives and deceptive play.

Pure PPK scoring would reduce the map to an gladiatorial arena and produce low complexity game play.

The gameplay has already been reduced to " dumbed down mode" due to the impact of PPT and not using the scoring to actually reflect the effort. This is why I think it should not just be able " PPK" but ALSO he objectives. If the highest PPT in the game is given to players who are actually fighting over objectives rather than the objectives themselves, it will STILL matter that they have the objectives, just will creating value for the objectives in a different way. The defenders get slightly higher rare loot drops and PPK than the attackers so you do actually want to own the structure to increases the loot and PPk for the players on your team, which gives a reason to defend them outside of " PPT" which does not offer personal reward.

This would remove PvD value form the gamemode and bring players to where the action is instead of just rewarding PvD or " outnumber or run" mentality as it currently does.

If you don’t want someone to PvD, don’t let them. Get in their way.

I don’t see why Anet needs to accommodate your play style exclusively.

If you blob up and cant track down the havoc groups doing the PvD, then maybe you need to de-blob to create a counter offensive.

I think strategic thinking and countermeasures make the game more interesting. Narrowing the scope to fit only one play style is pretty lame.

You would be trading one problem for another.

PvD happens because no one cares to defend objectives due to this design flaw. They fix that and people will care to defend. You will still have " strategic and countermeasures" due to objectives being more valuable than they currently are to players, just in a different way. They will be more valuable since you cannot receive the highest PPK and loot in WvW without them.

Well hopefully the scoring change will win the poll and we can start breaking it all down.

In the meantime, here is my proposal for scoring change if you want to continue the discussion.

https://forum-en.gw2archive.eu/forum/game/wuv/PPT-less-Scoring-model-Fights-and-Dolyaks

lil devils x.6071:

TorquedSoul.8097:

lil devils x.6071:

TorquedSoul.8097:

lil devils x.6071:

What needs to be on the poll is :

• Make scoring based on PPK over objectives rather than rewarding PvE game play in a PvP game mode, and I do not see that option there.

Why can’t it be both. As it stands PPK represent less than 15% of the overall score. I think the weight of fighting should be increased but territorial control should not be completely thrown out.

IMO.

As long as the fighting over the objectives is rewarded more than the objective themselves it will improve the way players are rewarded for their efforts, both offense and defense. PvD Should have never been part of the score in a game mode where you want to encourage players to fight one another instead of NPC’s.

The problem when you allow for " avoiding fights" to be rewarded is it will encourage players to do exactly that in a PvP game mode. What is the point of playing a PvP game mode if the players who do not do so are the ones who win?

Sorry, but I think that asymmetrical combat most certainly should be supported and sometimes that involves hit and run tactics and ducking fights.

The problem right now is that nearly 90% of the scoring can happen passively. Dropping that number to about 40%, and having fixed scoring for objective upgrades and captures, and boosting fights to about 40% will balance things nicely.

I like the idea of guerrilla warfare in WvW … and that revolves around objectives and deceptive play.

Pure PPK scoring would reduce the map to an gladiatorial arena and produce low complexity game play.

The gameplay has already been reduced to " dumbed down mode" due to the impact of PPT and not using the scoring to actually reflect the effort. This is why I think it should not just be able " PPK" but ALSO he objectives. If the highest PPT in the game is given to players who are actually fighting over objectives rather than the objectives themselves, it will STILL matter that they have the objectives, just will creating value for the objectives in a different way. The defenders get slightly higher rare loot drops and PPK than the attackers so you do actually want to own the structure to increases the loot and PPk for the players on your team, which gives a reason to defend them outside of " PPT" which does not offer personal reward.

This would remove PvD value form the gamemode and bring players to where the action is instead of just rewarding PvD or " outnumber or run" mentality as it currently does.

If you don’t want someone to PvD, don’t let them. Get in their way.

I don’t see why Anet needs to accommodate your play style exclusively.

If you blob up and cant track down the havoc groups doing the PvD, then maybe you need to de-blob to create a counter offensive.

I think strategic thinking and countermeasures make the game more interesting. Narrowing the scope to fit only one play style is pretty lame.

You would be trading one problem for another.

lil devils x.6071:

TorquedSoul.8097:

lil devils x.6071:

What needs to be on the poll is :

• Make scoring based on PPK over objectives rather than rewarding PvE game play in a PvP game mode, and I do not see that option there.

Why can’t it be both. As it stands PPK represent less than 15% of the overall score. I think the weight of fighting should be increased but territorial control should not be completely thrown out.

IMO.

As long as the fighting over the objectives is rewarded more than the objective themselves it will improve the way players are rewarded for their efforts, both offense and defense. PvD Should have never been part of the score in a game mode where you want to encourage players to fight one another instead of NPC’s.

The problem when you allow for " avoiding fights" to be rewarded is it will encourage players to do exactly that in a PvP game mode. What is the point of playing a PvP game mode if the players who do not do so are the ones who win?

Sorry, but I think that asymmetrical combat most certainly should be supported and sometimes that involves hit and run tactics and ducking fights.

The problem right now is that nearly 90% of the scoring can happen passively. Dropping that number to about 40%, and having fixed scoring for objective upgrades and captures, and boosting fights to about 40% will balance things nicely.

I like the idea of guerrilla warfare in WvW … and that revolves around objectives and deceptive play.

Pure PPK scoring would reduce the map to an gladiatorial arena and produce low complexity game play.

ZombieLeach.5862:

TorquedSoul.8097:

ZombieLeach.5862:

I can’t read it because the page is sliding off the left side of my phone screne and won’t load that side. I’m missing 2-3 words per line. :/

Turn your phone sideways.

Tried that as well as zooming out. The left couple inches or so of the page isn’t there at all. It’s like the page loads seceral inches off center.

Well get to a computer! The future of WvW is at stake. Unless you are voting for QoL, then I wouldn’t go through the trouble.

ZombieLeach.5862:

I can’t read it because the page is sliding off the left side of my phone screne and won’t load that side. I’m missing 2-3 words per line. :/

Turn your phone sideways.

KOPPER.1458:

QoL suggestion: Build Templates.

I would love these as well. A one click build change would be awesome in WvW. We would have to be out of combat but it would still be a great improvement.

lil devils x.6071:

What needs to be on the poll is :

• Make scoring based on PPK over objectives rather than rewarding PvE game play in a PvP game mode, and I do not see that option there.

Why can’t it be both. As it stands PPK represent less than 15% of the overall score. I think the weight of fighting should be increased but territorial control should not be completely thrown out.

IMO.

Scoring will have the biggest impact in terms of balancing game play but do like the QoL improvements.

Definitely prioritize scoring changes.

Norbe.7630:

here to help you track the kills and deaths
http://wvwintel.com/

as for the camps, towers, and keeps
you’ve tracked it yourself, you just have to put 1 of each instead of the 5, 10, 20, 30 thing,

or you can use that site too to track the captures you want every 5 mins if you like

to make it simple,
1 capture
1 kill
1 death

That is a great website. But its not helpful for what I am doing. I have no desire to manually collect weekly capture data.

If they have weekly statistics, I would like to see them so that I could compare them to my own to gauge accuracy. The FAQ says they collect data every 10 seconds, but it looks like its the client making the API call so I don’t think they have a data store on their end. Just a guess.

Norbe.7630:

TorquedSoul.8097:

I am open to anyone providing another method for measuring activity. I am using a simplistic measure to avoid putting too much of my own biases into the result. I personally feel that I should weight kills/deaths a bit higher but since it was arbitrary I decided to just use the raw numbers.

edited it for you, see above ^
if you cant see the relationship then i have no more to offer
that’s the simplest form of explanation i can provide

It really seems to me that you stuck on semantics. There are general activities that I am tracking (captures and killing and dying) and specific activities that you mention which I don’t have the data to track.

I track what I can track. I don’t track what I can’t track.

I’m sorry this frustrates you. If you believe the posted chart is meaningless, please disregard.

I appreciate you trying to inform me of the inherent complexities of WvW, and I will track BlobvPlayer when anet provides that information in the API. In the meantime, I make do with what they provide.

Norbe.7630:

hmmm, who is being hostile? lol
is it wrong to ask something you don’t understand, i guess it is… its called being hostile….

TorquedSoul.8097:

One capture volume point is not one capture. A camp is worth 5, towers 10, keep 25, and castle 35 capture volume points. Basically the PPT value.

uhmmm, ok.

TorquedSoul.8097:

I am not tracking PPT, I am tracking individual captures.

uhmmm, what again?

i think you should have said in the first quote that,
A camp is worth 1, towers 1, keep 1, and castle 1 capture volume point.
(this also applies to PvD when no one is defending as it existed in WvW 1.0)

your graph is all about activity right, i am just correcting you

i came to help you on something while you accuse me of being hostile?
get a grip

I really don’t know what you are getting at. I may have worded somethings poorly but I think it is clear at this point what I am doing. When I say PPT value, I mean PPT tick value. Not total PPT score.

And yes your posts seem a tad antagonistic … references to fortune tellers and allusions that the data is inaccurate. If you wish to challenge the accuracy of my data you should be explicit in detailing the error. Playing games with semantics isn’t helpful. It is always possible that my code has a bug. I have found and corrected many already. If you want to help you should state why you believe the data to inaccurate … not just allude that it is. You are not going to hurt my feelings by pointing out a problem. Besides, I love a good argument.

If you feel that capture volume should literally be 1 point for one capture regardless of the asset, I disagree. I’m trying to estimate activity levels and in regards to activity, a camp should not be weighted the same as a Castle. Some could argue that a Castle should be weighted more than 7 camps.

I am open to anyone providing another method for measuring activity. I am using a simplistic measure to avoid putting too much of my own biases into the result. I personally feel that I should weight kills/deaths a bit higher but since it was arbitrary I decided to just use the raw numbers.

Ali.7041:

According to this data, JQ’s ocx/sea is on par with yb’s.. lol.

This uses last weeks data which may be a bit skewed as a result of the 20% overall activity increase from the update.

The pairings may have adjusted the balance a bit also.

Regardless, JQ+ has about a 10% higher activity in OCX/SEA than YB+.

Norbe.7630:

you should have taken snapshot every 15mins not 5mins for the correct ppt model you know…

lets be clear to all readers, this is not the current matchup data
this is WvW 1.0 data, this is not WvW 2.0 data

it simply combined the individual activities together for the purpose of making a chart
this doesnt mean that this is the merged/linked chart

ok do whatever you like best,
i just passed by to ask validity of the chart thats all
fortune telling is not my thing, and i dont believe in those…….
(when i ask fortune tellers how they do it they always get angry, always…. lol)

TorquedSoul.8097:

it gives an good overall activity level estimate since the server linking.

: )

I don’t know where your hostility is coming from but I will engage anyway.

I am not tracking PPT, I am tracking individual captures. This is why I grab a snapshot every 5 minutes rather than 15 minutes. Why 5 minutes? Because the RI last 5 minutes after an objective is captured. It is the smallest fragment of time I can use if I want to see all the captures.

The first post states that this is based on last weeks data and that it is an estimate. I’m not really sure what your point is.

PPT is a terrible indicator of activity which is why I do it this way. If you prefer to use something else that is your business.

I will also say that activity doesn’t always translate to wins. Some servers can simply play smarter to overcome small activity deficits. In the current system this amounts to defending better, on-tick captures, and having a better KDR, i.e. better use of manpower. Large activity deficits are not likely to be overcome.

Norbe.7630:

TorquedSoul.8097:

These estimates are based on last weeks data so may still be a bit inflated due to the population changes from the update.

The chart shows three timezone blocks (NA, OCX, EU). SEA was combined with OCX.

The number next to the server names is the average glicko rating of the combined servers.

The activity scores are calculated by adding capture volume, kills, and deaths.

seems legit…
ill take the matches from the top and bottom of the chart (kill,deatch,cap)

x is the capture volume on your chart

Kills + Deaths + Capture Volume = Activity Score

SoS+GoM (33.5k + 42.9k + x = 80.5k) x=8.6k
HoD+EB (31.9k + 38.9k + x = 120.3k) x=49.5k (49,500 capture volume?, thats way bigger than their kills or deaths)

i had to stop here because i think there’s something wrong about the chart, or not…..

You seem to be consistently misinterpreting how I define capture volume. One capture volume point is not one capture. A camp is worth 5, towers 10, keep 25, and castle 35 capture volume points. Basically the PPT value.

These are also estimates based on last week’s data … not this week’s.

SkyShroud.2865:

TorquedSoul.8097:

These estimates are based on last weeks data so may still be a bit inflated due to the population changes from the update.

The chart shows three timezone blocks (NA, OCX, EU). SEA was combined with OCX.

The number next to the server names is the average glicko rating of the combined servers.

The activity scores are calculated by adding capture volume, kills, and deaths.

I can’t tell the full impact of the queues until next week but it gives an good overall activity level estimate since the server linking.

how was the data collected?
did you use a script to capture the data at the beginning of each timezone and end of each timezone to get the differences?

or is juz base on ur assumptions?

No assumptions. This is raw data. I take an WvW API snapshot every 5 minutes and then collect and analyze data points at the end of the week. I use PERL to parse the data and run the reports. Ad hoc reports like this one, I use excel to create the graph.

And use the data normalized to hour of day and then clumped together into a larger time block of NA, OCX and EU.

dubstax.9102:

Escorts are the most boring thing to do in WvW. Furthermore, good luck escorting and not (maybe better chance if on a BL) running into a gigantic blob. The population shifts have made WvW 90% blob fighting in almost everywhere. I’d rather not escort just to end up being zerg melted or better seeing the zerg and running for my life so as not get zerg melted.

Yaks are worth points now and don’t typically have escorts. I doubt that dumping PPT will result in any significant changes in that regard.

In what I propose, the yaks are worth about the same as they were before to the total score (maybe a little less.) So there is actually less incentive to escort a yak.

Right, if you were calculating maximum potential points then PPSC would be higher than PPT by about 8.5%.

But max PPSC is never achieved and PPT points will always be the same regardless. PPSC is usually around 60% of its max.

Oh wait. You are EU. Nevermind. I have no idea what happened to EU. And you probably don’t want to travel to Anvil Rock then either.

shrug.

Chaba.5410:

I still don’t understand why an OCX server (IoJ) was paired with a SEA server (JQ) when NA was needed to provide for better matches.

My understanding is that they simply folded the glicko ratings top to bottom although I’m not sure what week they grabbed the glicko ratings to get the line up that they did.

These data are just last weeks, but I also checked a 4 week average and the results aren’t much better. A little smoother but still imbalanced globally and by time blocks.

Unless someone gets clever and overstacks a lower tier server. But that will never happen … right?

Lets pick one to stack today. How about we start at the bottom.

EVERYONE GO TO ANVIL ROCK!!!!

It looks lo me like they want to direct transfers to the lower tier servers. But since no one knows how permanent the linking is (it’s beta) I doubt any significant amount will move from the upper tiers to the lower tiers. Swapping among lower tiers could happen though it will not fix any population balancing issues.

If you want your guild to grow, you will need to move it to a lower tier.

These estimates are based on last weeks data so may still be a bit inflated due to the population changes from the update.

The chart shows three timezone blocks (NA, OCX, EU). SEA was combined with OCX.

The number next to the server names is the average glicko rating of the combined servers.

The activity scores are calculated by adding capture volume, kills, and deaths.

I can’t tell the full impact of the queues until next week but it gives an good overall activity level estimate since the server linking.

Chaba.5410:

Increase in population across the board since patch lead to an increase in capture and fight activity? No surprise there. -_-

Surprising or not, it happened.

The more interesting question is whether the global activity will drop as a result of server mergers.

Sviel.7493:

TorquedSoul.8097:

I’m pretty sure I am close with my estimate. The 4 minute spawn rate isn’t for each camp but rather for each supply connection each camp has. So a camp that supplies 2 objectives will spawn 2 dolyaks over a four minute period. A camp that supplies 3 objectives will have 3 dolyaks spawn. and so on.

If my estimate was wrong my score approximations would be way off rather than just 1-2%.

I could bore you with the details if you like. I think I still have the spreadsheet somewhere.

Here’s an incomplete list of travel times for yaks. I don’t have time to dig up the more polished stuff but it should suffice.

Here are the travel times for Dolyaks from camps to objectives.

NEC to Airkeep – 2:05, 2:12
NEC to Rampart – 3:57

NEC PPT = 5 + 21 + 12 = 38

NWC to Firekeep – 2:10
NWC to Rampart – 4:36

NWC PPT = 5 + 21 + 12 = 38

NC (west) to Rampart – 2:48
NC (east) to Rampart – 2:22
NC to NET – 3:54
NC to NWT – 3:16

NC PPT = 5 + 12 + 12 + 9 + 12 = 50

SEC to Airkeep – 3:14
SEC to SET – 2:08

SEC PPT = 5 + 12 + 21 = 38

As you can see, during any 4 minute period, each camp spawns at least 3 yaks.

As for the ratio, the borderlands have 145 points from objectives and a potential ~220 points from yaks. If we factor in an EBG corner (75 points) for a team that doesn’t own SMC (no yak points) then it becomes 220 points from objectives and 220 points from yaks—that is the absolute limit of objective PPT contribution: 50%.

If you were intending to include EBG w/o SMC, then it was only 10% off, which isn’t too far.

edit: FUN FACT—since the change to the south Airkeep layout, that yak now runs in slow motion to preserve the travel time.

There is much easier way to calculate the PPT contribution just by looking at the total scoring of tier.

I’ve attached a table with Tier 2 NA values to show scoring contributions.

Regarding PPT, this is pretty much a fixed value every week. Total Points per Tick x 4 x hours of day x days of week. So 695 × 4 × 24 x 7 =467,040.

Points per kill is 1 point per kill so using Tier 2 data, 103,306 points.

So the contribution of PPSC = Total score – PPT – PPK. PPSC = 922,836 – 467,040 – 103,306 = 352,490 or 38% of total score.

But its contribution relative to PPT (PPSC/PPT) is 75%. So PPT is currently worth more than Yaks. And prior to the implementation of PPK, PPT was routinely between 58%-62% of the total score. The level to which PPSC contributes to points is dependent on how effectively teams keep the camps connected to their supplied objectives. I think a theoretical maximum for PPSC would be something like this total supply connects x minutes in week * 0.75. Total supply connects is something like 15 per BL plus 22 for EBG which gives a total of 67 connections. Max PPSC = 67 × 0.75 * 10080 = 506,520. So for tier 2 last week their overall supply efficiency was in the neighborhood of 70% which is pretty good.

However you were right about the yaks and their travel times. I reached my conclusion about the yak spawn rate by working backwards through the data.

Using tier 2 as an example again, I estimate that the PPSC is 348,928 which is off the above value of 352,490 by 1%. That is a pretty good estimation. And this estimation is based on the belief that 1 yak worth 3 points spawns every 4 minutes OR 0.75 points per active supply connection between a camp and objective.

The way I calculate my PPSC is by calculating how much time in minutes an objective is being supplied by a camp and then multiplying that total by 0.75.

Its interesting that this actually works given the variety of the travel times of yaks. So I decided to do a little check myself using the BL maps. I watched yaks coming from NEC to NET, Garri and Air Keep. The camp was owned by the owner of all three objectives. So here is what I saw:

1 yak would leave from the camp and travel to the tower and the same yak would leave the tower and then proceed to Garri. Travel time to garri ~ 4 min.

1 yak would leave the camp and go to the air keep with a travel and respawn time of ~2 min.

So every 4 minutes (12 minutes total supply time) this camp delivered 3 yaks for 9 points. 9 points every 12 minutes ~ 0.75 per supply connect minute. I think the actual number I use for my calculation is slightly less than 0.75 (0.7255) which mean slightly slower global spawn rate.

While model didn’t accurately predict the actual movements of the yaks the numbers still work out. Which is interesting. This means that some yak routes may be more valuable than others although I don’t know that I am ambitious enough to time every yak route to sort this out.

SkyShroud.2865:

capture volume means ktrain?

And kill volume means Blob v Blob. And KDR means who consistently holds a numbers advantage, i.e. Blob v Player (BvP)

You can interpret the data any way you like.

Capture activity was up 20% overall last week and kills increased about 30%. This is before server linking. I wonder if the long queues produced by the linking will reduce activity for next week.

While I agree that reseting all glicko scores would be a better long term solution, there is no point in doing it while they are beta testing server linking.

If they reset the scores and undo the linking you will have tier 8 servers getting matched with Tier 1 servers and that will be a mess for about a month.

And without truly fixing the population balance both by server and by timezone it will never matter under the current scoring system.

Anet either needs to fix the scoring system to minimize the timezone imbalances or keep the current scoring system and find a way to truly fix population imbalances. Unfortunately short of alliances I don’t see them achieving population balance so they should direct their attention to scoring changes.

Either way, glicko is dependent on the outcome of matches and that is currently imbalanced by the scoring model and/or population. So fixing the glicko doesn’t fix the underlying problem and we will eventually end up back where we started.

Somebody needs to decide what type of game WvW is going to be and suck it up and make the changes.

Norbe.7630:

whats the difference between a door and a dolyak?

a door takes time to kill
a dolyak takes seconds to kill

a door can call for help
a dolyak is dead before anyone can help

if there is no door, then there is no dolyak…
thats the story of a door and a dolyak

The doors will still exist and they can still call for help. There is nothing in this scoring model that precludes the use of objective doors. I would even lobby to put doors on the dolyak pens so that you had to beat them down to kill the dolyaks inside.

I wouldn’t be against giving the sentries doors also. They can just stand there an hold them while you hit them.

There will always be a place for PvD in WvW.

Norbe.7630:

now there wont be anymore PvDoor
there will be PvDolyak lmfao

There will always be PvSomethingNotaPlayer in WvW. I think its unavoidable.

Sviel.7493:

That said, I can’t agree with the way you figure PPK into it. The best killscore always goes to the server with the most people…skill isn’t a real factor. Making it such a huge chunk of the score will only serve to exacerbate the current population issues.

That is not necessarily true. Sure servers that like to blob will do slightly better in fights due to numbers advantage but smaller groups will be less likely to engage in the first place. Also blobbing would present a disadvantage in territorial control.

The goal was to find a middle ground where both strategies could be viable. However more balanced servers would probably perform better overall.

Sviel.7493:

Dolyaks spawn much more frequently than 4m intervals. There’s only kitten delay on their respawn (5m if killed) so they tend to jump out every ~2m or so, depending on the camp. As such, your estimate of PPT to PPYak ratio is way off.

I’m pretty sure I am close with my estimate. The 4 minute spawn rate isn’t for each camp but rather for each supply connection each camp has. So a camp that supplies 2 objectives will spawn 2 dolyaks over a four minute period. A camp that supplies 3 objectives will have 3 dolyaks spawn. and so on.

If my estimate was wrong my score approximations would be way off rather than just 1-2%.

I could bore you with the details if you like. I think I still have the spreadsheet somewhere.

LoneWolfSpectre.6412:

can you post NA as well. would be nice to know who we’re working with

It looks like the top two tiers may not be effected.

I’m basing that on the EU pairings … but there is more complicated and I could be wrong.

DeWolfe.2174:

I… I just….. do you…. are we playing the same game???? Players couldn’t care less about Doylaks. It’s an “Oh look kill it!” as they run past at best. Escorting and getting players to escort is like pulling teeth. I literally had to start paying gold to get players to do jobs like this and bloodlust.

I was offering it as an alternative to PPT. Dolyak scoring is already a thing. Right now dolyaks are worth about 3 points killed or delivered and they spawn at approximately 4 minute intervals. Dolyaks are now responsible for upgrades also and are still dependent on camps.

Its not that I think Dolyaks should be the most important thing, notice the point was to balance the territorial control with fights. I am only suggesting that basing territorial control on dolyaks as opposed to PPT is a better strategy.

The way this scoring system is set up increases the value of fights while making the territorial points more contestable.

What would you do in place of dolyaks? or would you ditch territorial scoring all togther?

TLDR: A scoring system that tries to find balance between Fight oriented play and Territorial control play. PPT is removed and 40% of the score is now the result of fight skill. PPT is replaced with a Dolyak based scoring system (PPC,PPU,PPSC) that gives roamers and havoc groups more impact on score. And by increasing the scoring options, more creative strategies can be developed leading to more interesting gameplay. You still must play to win, but now the outcome is based on Fighting skill, strategy and community organization.

How do you think this type of scoring mechanism would alter WvW strategies and Behavior?

FIGHTS AND DOLYAKS

There are many complaints about PPT regarding nightcapping. While this is a problem it isn’t the biggest problem with PPT. IMO the biggest problem with PPT is the lack of scoring options. Prior to PPK, the score was 60% PPT and 40% PPSC (Points per Supply Caravan). The addition of PPK didn’t alter this balance much. The basic strategy in this scoring mechanism is to hold as much as possible while killing yaks or protecting deliveries. All other events while they could aid in the scoring factors were at best tangential. Creating more options in the scoring mechanism can lead to more creative strategic gameplay.

It addition to the existing systems lack of options, many players feel that it completely disregards the value of fighting oriented playstyle. Due to population imbalances, the score no longer reflected who could defend the best of capture the best because off hour captures offered little resistance. In addition the off hour scoring greatly imbalances the final result.

What I suggest is a somewhat balanced scoring system that takes into account Fights, territorial control and logistics: Fights and Dolyaks.

The Dolyaks

The dolyaks become the primary focus of territorial control scoring. Unlike PPT which benefits from holding large amounts of assets, dolyaks can be easily contested by small groups. And since the new mechanic for upgrades is dependent on Dolyaks small groups and stifle an enemy with guerrilla tactics by disrupting supply caravans. In addition Dolyaks are only useful if supply connection exists between the camp and objective it supplies. So camp control becomes instrumental and a potential area of high conflict.

Terminology and score calculation for territorial control:

PPC: Points per Capture. Whenever a non-camp objective is captured points are given on a one time basis based on the tier of the objective. Tier simply represents a multiplier (tier + 1) to the objectives old PPT value. So a Tier 0 Tower is worth 10 points (10 X 1) and a tier 3 tower is worth 40 (10 × 4). Camps are excluded because their primary function is now the production of dolyaks.

PPU: Points per Upgrade. Whenever any objective upgrades points are awarded based on the old PPT value. 5 for camps, 10 for towers, 25 for keeps and 35 for castle. PPU rewards defense and creates an additional incentive to capture objectives. These points are dependent upon camps and dolyaks.

PPSC: Points per Supply Caravan. Every time a Supply Dolyak makes it to its objective, 1 point is awarded to the owner of the objective. If a dolyak is killed, a point is awarded to the killing team. Dolyaks become the most important component on the territorial control side as it still represents a form of passive scoring, but is easily contestable by small groups. And unlike PPT, capturing camps can completely shut down all passive scoring and stop upgrades. Roamers and Havoc groups can make a more meaningful contribution to the score by shutting down enemy supply.

Weights and multipliers:
These are used for experimental purposes to balance the score. Scoring occurs as above but then modified by a multiplier to adjust the relative value of each component. My current model applies the following multipliers: PPC x 2, PPU x 3, PPSC x 2.

The Fights

There are two problems with PPK. First, 1 point for one kill says nothing of the skill of the fighters. Simply engaging in more fights will ultimately result in more kills. Excluding deaths from the equation makes PPK a measure of volume of activity. Secondly, the relative value of PPK is dependent on the amount of fight activity going on. If fewer fights are occurring, then the score is quickly skewed towards territorial control. This could be the result of a low activity tier or players intentionally avoiding fights.

To fix these elements I add KDR adjusted killscore and make the fight portion of the score locked at 40% of total score (or 0.667 of the total territorial score.)

Calculating the Fight points:

First calculate KDR by dividing kills by deaths (Max KDR is 3, Min KDR is 0.334). Then multiply the KDR by the number of kills to get a killscore.

The next step is to weight the killscore points to 40% total points. It goes like this:
(PPC + PPU + PPSC) * 0.667 * (server Killscore/sum of all Killscores)

This calculation results in the fight score being equal to 40% of the total score. However the portion of that 40% that goes to each server is dependent on the their fighting skill and fighting volume.

EXAMPLES:

Tier 1 NA (charts attached)

This example assumes that all dolyaks successfully arrive to their destinations. It bases the Tier of the Objective on the old time model. The reason for these two assumptions is that I do not a this time have access to dolyak success or objective tier data in the API. So I am guessing but it is still useful to give a sense of how the score would look and how much each scoring component contributes to the overall score.

The first chart shows the raw data for PPT, Killscore, PPC, PPU, PPSC and PPT before adjusting the weights.

The second chart shows the weighted values and the final total score. PPT and PPK are weighted zero so have no score contribution. overall killscore contributes 40%, PPC ~15%, PPU ~15% and PPSC ~ 30%. While the Killscore is fixed at 40% the other three contributions can vary depending on activity level. Low activity can result in a much larger contribution from PPSC.

Tier 8 NA (charts attached)

This shows how the score contribution changes in a low activity environment.
Killscore is still 40% but PPC, PPU and PPSC are 6.5%, 11%, and 42.5% respectively. This shows a high amount of passive scoring however it would be easier for an out-manned server to compensate by securing camps and hunting yaks. These examples cannot account for how strategies would change since they use historical data based on another system.

Dolyak Tweaks

Here are couple tweaks that could be done to dolyaks but are not represented in the data.

Asymmetrical Dolyak value: Killing a dolyak is always worth the base one point however successful dolyak deliveries are now worth the tier + 1 points. This means that enemies will be less likely to let you coast with a tier 3 objective as it is now producing more points.

Modulating Yak spawn rate: If off hours play is still considered a problem the Yak spawn rate can be slowed down during off peak hours. This reduces not only the PPSC points but slows down upgrades which reduces that scoring as well.

Liston.9708:

That was my take – 2 servers on the same 4 maps with each other. Pro : crickets tiers will see many more people. Con: they could also see a que for the first time since launch.

I’m pretty sure this is what they meant by linking servers. It makes no sense to simply combine scores if you are trying to fix population imbalances.