Temporary Matchmaking Test

[Ysaran]

I am strongly against the unified qué system. There are some scenarios that are badly designed, frustrating and just not fun.

Stonetroll crossing
Twisting tower
The graveyard with the undeads NPC's (can't recall the name)
The ironclad
All the 24vs24 SCs.

[Rubius]

Epic stuff!

Also, we can see a Dungeon tab in the queue window.... WIP?

Early idea that one of the designers added in as a concept. No news on this as an actual feature yet (SC Matchmaker has priority), but if we have more news someday, we'll let everyone know.

[TrainInVain]

Awesome step in the right direction for player retention in imo.

Can you extend the event for a few days if the testing impedes progress on it?

[georgehabadasher]

We understand that players sometimes prefer different scenarios or game modes and would prefer to play them less frequently. Therefore we have added a scenario blacklisting system.
Blacklisted scenarios will be selected less frequently when searching.



Instead of separate queues per scenario map, there's now a single random scenario queue, which will improve queue speed and map variety. We're just testing this single queue on PTS, and behind-the-scenes we're discussing how best to handle Weekend Warfronts and other Scenario types, which we'll share more on later.

Please heavily consider reimplementing weekend events as well as individual map queueing.

One of the best parts of good multiplayer pvp games was always a server browser or ability to specific queue maps that players grow to love and want to play often (think how many players choose to play mostly de_dust2 on counter-strike or blood gulch in halo).

Especially in this game, where many of the maps are poorly designed or disliked by the community (twisted tower, bright wizard college, ironclad, etc.). Please don't say "we'll rework the maps down the line to make them more fun/balanced". Because I can guarantee most players don't want to wait another 2-5 years for twisted tower to be reworked in to something less abominable, while spending those years being forced to play it at random.

I think it’s rather important to maintain the ability to choose what maps you’re playing. Disallowing it or removing weekend events that target a specific SC and funnel players in to that SC while giving bonus rewards is a step backwards. Else you are going to see a lot of ppl having quitter debuff when they get twisted tower and decide to just go back to roaming in orvr instead.

The blacklist feature is clearly made to alleviate this concern, right? In theory it seems like that is the case but due to it being listed here that: "Blacklisted scenarios will be selected less frequently", I believe that the current blacklist won't do enough. It should not make them occur LESS frequently, it should block them entirely if you choose to blacklist a map. That's what a blacklist is after all.

They're moving to this system to get more players in the same queue, meaning that the matchmaking system will have more 'material' to work with to create balanced matches. Presumably, they're working under the assumption that people would rather have a more balanced match on a map that they dislike than an imbalanced on one they prefer.

[gersy]

We understand that players sometimes prefer different scenarios or game modes and would prefer to play them less frequently. Therefore we have added a scenario blacklisting system.
Blacklisted scenarios will be selected less frequently when searching.



Instead of separate queues per scenario map, there's now a single random scenario queue, which will improve queue speed and map variety. We're just testing this single queue on PTS, and behind-the-scenes we're discussing how best to handle Weekend Warfronts and other Scenario types, which we'll share more on later.

Please heavily consider reimplementing weekend events as well as individual map queueing.

One of the best parts of good multiplayer pvp games was always a server browser or ability to specific queue maps that players grow to love and want to play often (think how many players choose to play mostly de_dust2 on counter-strike or blood gulch in halo).

Especially in this game, where many of the maps are poorly designed or disliked by the community (twisted tower, bright wizard college, ironclad, etc.). Please don't say "we'll rework the maps down the line to make them more fun/balanced". Because I can guarantee most players don't want to wait another 2-5 years for twisted tower to be reworked in to something less abominable, while spending those years being forced to play it at random.

I think it’s rather important to maintain the ability to choose what maps you’re playing. Disallowing it or removing weekend events that target a specific SC and funnel players in to that SC while giving bonus rewards is a step backwards. Else you are going to see a lot of ppl having quitter debuff when they get twisted tower and decide to just go back to roaming in orvr instead.

The blacklist feature is clearly made to alleviate this concern, right? In theory it seems like that is the case but due to it being listed here that: "Blacklisted scenarios will be selected less frequently", I believe that the current blacklist won't do enough. It should not make them occur LESS frequently, it should block them entirely if you choose to blacklist a map. That's what a blacklist is after all.

They're moving to this system to get more players in the same queue, meaning that the matchmaking system will have more 'material' to work with to create balanced matches. Presumably, they're working under the assumption that people would rather have a more balanced match on a map that they dislike than an imbalanced on one they prefer.

While I understand the logic there I think that I speak for many when I say that people would much rather choose which scenario(s) they are going to be placed in rather than have the chance getting at twisting tower, ironclad, stonetroll, 24v24 SC, eternal citadel, tor anroc, doomfist crater, thunder valley, etc. terrible SCs - provided they are given the choice between the 2 options.

A solution could be to temporarily remove maps from the pool which are most obviously disliked by the community. I'm sure the devs have access to plenty of internal data which could be used to suggest which maps have very low participation/play rate when they are available to queue. And I'm betting I could guess which ones those are.

Alternatively, as I suggested, allow the blacklist to fully remove the ability to get a scenario. Bonus points to the devs if the amount of scenarios you can blacklist is unlimited.

[MaxHayman]

Great to see scenarios getting some love! Really enthusiastic about most of these changes!

Some obvious questions arise...
1) What is the purpose of starting a scenario two minutes early and making people wait up to two minutes until it fills? Why not just wait for the requisite number of people to enter the queue and start everyone at the same time?

2) Is the ready check solving a significant problem? Leaders queueing with afk players hasn't been a significant problem that I've encountered personally, or heard others mention as a problem.

3) Regarding weekend warfronts, has the dev team considered doing something similar the RvR weekly? In the RvR weekly, players can play any zone and earn progress towards completing the event. It would be nice if players could play any scenario to progress the weekend warfront event.

1) Because we have the Random Scenario pool, it can create sceanarios which we have different minimum and maximum player caps. So we have a minimum amount of player needed for the matchmaker to create a new instance (say for example 6 per side) in which a random scenario is made. Some scenarios will have a recommended player minimum of 12 per side. So the scenario would launch with 6 people on each realm in the queue then, backfill the players to fill up the other slots now. The matchmaker now is a lot better at backfilling, so we can rely on starting a match early and then filling up. Matches can also start earlier now, if the recommended amount of players the start time jumps to 30 seconds. At the moment (live) it is fixed at 60 seconds.

2) We will review this based on feedback.

3) It has been something that has been considered but we do not plan to adjust things at this time.

Why do you want to test it on Wednesday 31st December ?
Why not choose a Friday/Saturday/Sunday where more people can test/join ?

The real reason is the system is ready to test now and I have the day off. We also don't want to do it on the week end, we would rather this test not be at a peek time in the week. The amount of players active in the week will be sufficient.

So with the new system blacklisted scs are just deprioritised not fully blocked from our queue?

At the moment just deprioritised. We will likely tweak the functionality of how it exactly works, to get it to be effective enough, but not to mean you never get into a match.

Epic stuff!

Also, we can see a Dungeon tab in the queue window.... WIP?

We don't have any specific plans. This was just added in the design incase we wanted to explore having different queues for things on a tabbed layout.

Great news!
- If you click “Give me a minute”, you’ll be automatically teleported in when the timer ends.
Will this happen, even while in combat in RvR ?

Yes this should be how it worked on Mythics servers.

We understand that players sometimes prefer different scenarios or game modes and would prefer to play them less frequently. Therefore we have added a scenario blacklisting system.
Blacklisted scenarios will be selected less frequently when searching.



Instead of separate queues per scenario map, there's now a single random scenario queue, which will improve queue speed and map variety. We're just testing this single queue on PTS, and behind-the-scenes we're discussing how best to handle Weekend Warfronts and other Scenario types, which we'll share more on later.

Please heavily consider reimplementing weekend events as well as individual map queueing.

One of the best parts of good multiplayer pvp games was always a server browser or ability to specific queue maps that players grow to love and want to play often (think how many players choose to play mostly de_dust2 on counter-strike or blood gulch in halo).

Especially in this game, where many of the maps are poorly designed or disliked by the community (twisted tower, bright wizard college, ironclad, etc.). Please don't say "we'll rework the maps down the line to make them more fun/balanced". Because I can guarantee most players don't want to wait another 2-5 years for twisted tower to be reworked in to something less abominable, while spending those years being forced to play it at random.

I think it’s rather important to maintain the ability to choose what maps you’re playing. Disallowing it or removing weekend events that target a specific SC and funnel players in to that SC while giving bonus rewards is a step backwards. Else you are going to see a lot of ppl having quitter debuff when they get twisted tower and decide to just go back to roaming in orvr instead.

The blacklist feature is clearly made to alleviate this concern, right? In theory it seems like that is the case but due to it being listed here that: "Blacklisted scenarios will be selected less frequently", I believe that the current blacklist won't do enough. It should not make them occur LESS frequently, it should block them entirely if you choose to blacklist a map. That's what a blacklist is after all.

I am strongly against the unified qué system. There are some scenarios that are badly designed, frustrating and just not fun.

Stonetroll crossing
Twisting tower
The graveyard with the undeads NPC's (can't recall the name)
The ironclad
All the 24vs24 SCs.

The scenarios which are in the pool the Random Scenario can pick from does not contain every scenario in the game right now. We have a shortlist we have picked internally, and we can weight them so some appear more frequently than others. When the rotation happens weekly it will pick from this list. The blacklist system will be tweaked over time. We need to improve the algorithm.

The matching based on a point system is what I also had suggested about 2 years back. Sounds awesome. Really interested how your calculation formula looks like. What aspects are considered.

Looking forward to seeing the the performance improvements. Does this utilize a bucket approach that kicks off SCs every x seconds?

It's not totally finalised right now and we will tweak it over time. The matchmaking system is based on Open Match.

Awesome step in the right direction for player retention in imo.

Can you extend the event for a few days if the testing impedes progress on it?

Depending on if or what goes wrong this could be a measure we take yes.

[FischiPiSti]

I didn't think such interface changes would be possible, but it's great to see. But if it is possible, I think something like the ranked window, where you can see how many players of what roles/classes are queued up, would be a great addition, so players can get more info on "why is it not popping?" I know earlier attempts at matchmaking failed, but I think a large reason why that was the case was that players had zero information on why it wasn't popping, so with the impression that nobody was queued up, they just left. The incentives are great, but it's needlessly vague. Let us see that there are 150 DPS queued up with no tanks, so we can start shaming Order.

People say it won't work because queue times will be too long, but that's BS. There are normally 200+ players online per side. Clearly, there are enough players, but most are literally forced into RvR, either into lag fests or carrying crates because of the sorry state of scenarios. I know people who specifically refuse to play sc because of the cancer. And when the system doesn't work, the correct way is to move forward, try new things, not revert to the old system that demonstrably did not work for a decade. It festered long enough, so I'm very thankful that you guys are trying.

Even the little things matter immensely, not even the matchmaking, but just preventing the sc from starting while half the players are still on their 2. 1 min timer, because even if the composition itself is fine, it doesn't matter if they miss the first group fight, it's all over after that. I would go as far as outright not letting the sc start at all, even if just 1 guy is still on the loading screen. Waiting half a minute more is still better than waiting 4 minutes to surrender. Have a ready check or something, and start immediately if everyone is ready.

[gersy]

The scenarios which are in the pool the Random Scenario can pick from does not contain every scenario in the game right now. We have a shortlist we have picked internally, and we can weight them so some appear more frequently than others. When the rotation happens weekly it will pick from this list. The blacklist system will be tweaked over time. We need to improve the algorithm.

that's great, thanks for the reply. i look forward to the new system as scenarios are my 2nd favorite content (next to city siege )

[ShadowWar]

...I think that I speak for many...

I disagree.

[gersy]

...I think that I speak for many...

I disagree.

I don't. Because I've already spoken to many on it, as SC change is very long awaited, and there are many SC enthusiasts voicing their opinions on it in private channels. My post was largely echoing their some of their concerns/questions regarding the initial list of changes. Not just my own and not subjectively. More choice is almost always a good thing, removing choice is almost always a bad thing. However I got a good reply from the person in charge of the rework, affirming that the aforementioned concerns have already been considered and will likely be ironed out.

Then again you are not an actual SC or smallscale player and could never see reason in the past so... no point to argue with you.