Premade Group Finder improvements

[HuxNeva]

Sorry for the long post. I will argue where and why the current Premade Group Finder is not optimal, and suggest some improvements. This is not meant to be about wetter there should be any group facilitation tools or not. There are already plenty of threads showing opinions vary on that subject.

The Premade Group Finder (PGF) tool was introduced to help people firm groups for running ingame content. Alternatives exist(ed), oQueue, OpenRaid, but an ingame Blizzard utility made this it far more mainstream. Going by participation, the tool has been very successful.

Blizzard has other group creation facilities in the game, namely Raid Finder (LFR) and Dugeon Finder (DF). These reach an even wider audience. The most significant differences between these latter 'finders' and the PGF is that for the player looking for a group to do a dungeon or LFR raid, they allow the player to 'click a button', and continue playing waiting for an eventual queue to pop, and that the player is 'transported' to the content. Matchmaking is purely algorithmic, a group leader is determined by the machine, but that leader has no a-priory influence over group constitution and composition. In contrast to that, in PGF the self appointed raid leader can set minimal requirements and have control over how (s)he composes the group based on limited information (in practice ilvl, maybe some achievements, and role).

Many people have argued that this control by the raid leader is needed to ensure a reasonable success with more challenging content (although i personally believe machine learning techniques have been shown to succeed at prediction/composition tasks far more complex than this). For people willing to organize groups (a tiny percentage of players), the system works very well. However, for a significant group of players, the PGF tool is a frustrating experience consisting mostly of a long period of active application and rejection, symptoms of which are for example the 'only groups asking for irrational ilvl requirements that far exceed the ilvl dropped by the content' complaints voiced on forums. I guess even the most ardent achiever can understand being submitted to hours of rejection is not a 'fun' experience for what should be a reasonable application (running content that still has a chance to drop an upgrade).

Where does this situation come from. Simply put, it is a self-reinforcing system. People willing to start groups will succeed in running content more than those just looking to join groups. Their ilvl will thus rise faster. Most will organizers try to improve subsequent success and 'smoothness' of the runs by putting their higher ilvl requirements for their group, attracting just the players that can meet the criteria as well. Thus nearly all group organizers join a subset of the player-base that accelerate 'ahead' of a very large group of players that do not organize groups, leaving the latter uncatered for. One could argue these people should 'just start their own group', but there are many psychological/sociological reasons why this is not an option for most.

So what could be improved?

Most certainly the 'active' part of the submission could be changed. Let people the option to just select the content they wish to queue for, and get on with their game play, rather than forcing them to keep submitting applications to specific groups and facing long bouts of rejections. The system could indicate a prediction based on the request and the player profile of how long the queue could last.
For the organizer, give the option to accept a 'suggested group', indicating a machine predicted chance of success for the group, rather than forcing them to select each individual player based on ilvl. While certainly a lot of raid leaders will still whittle out the weaker links in the proposed group, or make room for paid carries, the 'lazy' option should be available to just accept the group with a reasonable predicted success option.
Worst case: for the applicant the 'queue never pops', but at least you can play in the meantime, for the organizer worst case you ignore the system and still handcraft the group.

[JemiS]

We already have this. It's called LFR/LFD.

I'm not sure what you're getting at with this post.

Is it that the LFR/LFD system needs to be expanded to different content? There have been hundreds of posts on why that's a bad idea.

To be perfectly honest, while I care very little about ilvl compared to experience and willingness to listen, the type of player that wants to "just get into a queue and get on with playing" is not someone I can see myself wanting to invite to a raid.

It's usually not someone who chats in groups and makes friends (thus getting battle tags and getting invited back), and it makes me worry that they're not really interested in the content we're doing. And lets be honest, while not LFR even normals take a modicum of organization and people listening and caring to not be a frustrating experience.

Finding PuGs in the group finder really isn't that bad. I was able to PuG my way through HHFC on a relatively under geared character mid-way through the tier without many problems. I whispered group leaders, I looked for raids on OpenRaid, and I applied to a lot of different groups. The difference is that once in a group, I frequently made friends, and used those new friends to find other groups.

[Soeroah]

What they're getting at is that people who want to find a group can set and forget, and if a leader finds them useful, they'll get an invite, rather than having to sit around looking every two minutes to see if a group has come up.

The difference is LFR/LFD at some point automatically puts you in a group with no biases, while this proposed change allows you to basically list yourself for X dungeon and Y raid or whatever and go about your business, confident that someone looking to put a group together can easily find you with a search on who's interested and invite you.

[HuxNeva]

We already have this. It's called LFR/LFD.

I'm not sure what you're getting at with this post.

Is it that the LFR/LFD system needs to be expanded to different content?

Absolutely not. Proposal is to merge the good parts of the queue'ing system for the person looking for a group, and improve the experience further for the person creating and assembling a group.

the type of player that wants to "just get into a queue and get on with playing" is not someone I can see myself wanting to invite to a raid.

Why not? What do you care whether the person is farming, questing, watching YouTube or playing Hearthstone while waiting for a group to be accepted in?

The difference is that once in a group, I frequently made friends, and used those new friends to find other groups.

I tend to agree with you here on the larger social implications of 'convenience', but tried to explicitly keep that out of this thread.

- - - Updated - - -

What they're getting at is that people who want to find a group can set and forget, and if a leader finds them useful, they'll get an invite, rather than having to sit around looking every two minutes to see if a group has come up.

The difference is LFR/LFD at some point automatically puts you in a group with no biases, while this proposed change allows you to basically list yourself for X dungeon and Y raid or whatever and go about your business, confident that someone looking to put a group together can easily find you with a search on who's interested and invite you.

Yep. That and a slight support in group composition suggestions to shift from 'every person in my group has to meet this criterium' to 'the group as a whole needs to meet this criterium', which the group composer is free to ignore of course.

[Deleted]

I'll give you that the option to queue for content rather than particular groups would make sense and be an improvement.

The one thing that doesn't make sense is this- "indicating a machine predicted chance of success for the group". There are too many variables to take into account for it to be doable i.e. gear and how it should affect performance, the ability of that certain player, how that certain player fares with the content that will be done, whether or not he's drunk out of his mind, whether or not he is healing with one hand and using the other to play jenga etc.

[Kokolums]

I struggle to understand where Blizzard stands regarding PGF.

Diablo 2 has PGF and it is very successful. However, when Blizzard announced D3 they said they were eliminating PGF altogether in favor of an automated matchmaker. There were several 25 page threads on the Blizzard forums demanding the return of PGF, but none of it mattered. Eliminating PGF made D3 much more antisocial. Blizzard argued time and time again on the D3 forums and in interviews that PGF sucked because if the community got too large, the group you wanted would fill before you could click the button. So therefore, PGF was never coming back.

Several years ago, I used to argue time and again that Blizzard could, at the very LEAST institute BOTH AMM and PGF in D3. But that fell on deaf ears.

Then out of NOWHERE, PGF appeared in WoW in WoD. And suddenly......WoW had both AMM and PGF. Its like they took my idea and applied it to the wrong game.

Blizzard has never once stated WHY they suddenly reversed their opinion on PGF. I don't know if this is just a test of PGF that they plan on ripping back out of their games, or if it is a fundamental philosophical change. It boggles my mind that it showed up at all. I don't know what to suggest for improvements because after years of being told by Blizzard that PGF sucked, I half expect them to rip it out of WoW again.

[pmkaboo]

Why not? What do you care whether the person is farming, questing, watching YouTube or playing Hearthstone while waiting for a group to be accepted in?

because that person will more than likely leave after first wipe: "cba with this, i should rather go back to queue and do what i was doing than continue wasting my time here".

also i like how youre glorifying lazyness here:

One could argue these people should 'just start their own group', but there are many psychological/sociological reasons why this is not an option for most.

I struggle to understand where Blizzard stands regarding PGF.

Diablo 2 has PGF and it is very successful. However, when Blizzard announced D3 they said they were eliminating PGF altogether in favor of an automated matchmaker. There were several 25 page threads on the Blizzard forums demanding the return of PGF, but none of it mattered. Eliminating PGF made D3 much more antisocial. Blizzard argued time and time again on the D3 forums and in interviews that PGF sucked because if the community got too large, the group you wanted would fill before you could click the button. So therefore, PGF was never coming back.

Several years ago, I used to argue time and again that Blizzard could, at the very LEAST institute BOTH AMM and PGF in D3. But that fell on deaf ears.

Then out of NOWHERE, PGF appeared in WoW in WoD. And suddenly......WoW had both AMM and PGF. Its like they took my idea and applied it to the wrong game.

Blizzard has never once stated WHY they suddenly reversed their opinion on PGF. I don't know if this is just a test of PGF that they plan on ripping back out of their games, or if it is a fundamental philosophical change. It boggles my mind that it showed up at all. I don't know what to suggest for improvements because after years of being told by Blizzard that PGF sucked, I half expect them to rip it out of WoW again.

the difference between diablo and wow is that in diablo everyone is selfsufficient so even if you get matched with 3 brain dead monkies you can still enjoy the game. in wow raids however you rely on most people(for some bosses its all people) understanding whats going on and performing decently at least.
now the difference between AMM and PGF is that AMM can easily and quickly match people together based on select criteria, but there is 0 interaction between the players before they are already groupped so they basically dont give a shit about what happens with the group, because if things go south they can just leave the group and join the queue again and try in a different group. PGF makes people apply to groups knowing that other real person will have to accept this application, and this tiny interaction already makes people care about their performance, at least a bit, because there is always the possibility of being in a group with them again and also you know you might be replaced if you dont do well, and more importantly it also takes some time to find a group with PGF and you have to put some "work" into it, so when things go south you will think twice about leaving.

from all this it should be pretty clear why AMM is perfect for diablo and PGF works well in wow.

[Amerissis]

The Premade Group Finder (PGF)
Raid Finder (LFR)
Dungeon Finder (DF).

Please use the proper terms for these things, it makes reading so much easier. They are called LFG (looking for group), LFR (looking for raid) and LFD (looking for dungeon).

[Relapses]

The LFG is fine as it stands.

However, I think if Blizzard were to improve the LF guild interface it would be infinitely more beneficial for the player base.

[HuxNeva]

because that person will more than likely leave after first wipe: "cba with this, i should rather go back to queue and do what i was doing than continue wasting my time here".

I agree with you that a certain 'pre-investment' in the process will increase the commitment execution. Much like pre-LFD requiring you to travel to the instance will make you more likely to stick it out for a few wipes since you already sunk in all the time riding there. This said, is the 'being submitted to two hours of rejects' too little,too much or just enough 'determination' to be allowed in? How do you know? For all you know this might be the dudes first application 20 seconds in, or his relentless 3rd hour of applying to any group that pops up for the objective. Are othercommitment 'proves' better (e.g. reputation based? LoL style longer CD's for leaving groups?, ... ?


- - - Updated - - -

also i like how youre glorifying lazyness here:

I don't believe it is laziness in general. Some people are psychologically more keen on leadership positions or risk taking than others. Doesn't mean the others don't put in equal or more effort in other parts of the process.

- - - Updated - - -

Please use the proper terms for these things, it makes reading so much easier. They are called LFG (looking for group), LFR (looking for raid) and LFD (looking for dungeon).

Point taken. I tried going with the 'official' lingo, but it makes it less clear.

- - - Updated - - -

The LFG is fine as it stands.

But can't it be improved upon?

However, I think if Blizzard were to improve the LF guild interface it would be infinitely more beneficial for the player base.

Very true. Ingame guild finder is abysmal. I almost can't imagine people using that.

[Daedius]

Would love it if there were extra tools in the PGF to deter people from abusing it for realm hopping... such as having people ONLY phase to the leader's realm while WITHIN the same zone as him.

[HuxNeva]

The one thing that doesn't make sense is this- "indicating a machine predicted chance of success for the group". There are too many variables to take into account for it to be doable i.e. gear and how it should affect performance, the ability of that certain player, how that certain player fares with the content that will be done, whether or not he's drunk out of his mind, whether or not he is healing with one hand and using the other to play jenga etc.

I would love it if Blizzard would put up full historical (anonymized) group making data and outcomes, so I could have a go at writing the predictor myself. Make it a contest like the Netflix challenge was.

[Relapses]

But can't it be improved upon?

I'm usually of the opinion that if something isn't broken, Blizzard shouldn't try to fix it. There are a few minor things which may improve the overall experience but as it stands, it does exactly what it was implemented to do. I'd like to see perks added to the system, such as a rating system for people who frequently succeed at PuGing (additionally helping identify players who are clearly just putting together a boosting run).

[Deleted]

I preferred the Oqueue style of listing yourself and leaders finding you in the pool. Was a lot faster and less hassle to get into a group. Blizz always seem to do things ass-backwards and decide to make it so groups are listed and players must submit to join.

If I'm running a group and healer leaves I would much rather just have the Oqueue system where I can see Spec/Ilvl and choose what new healer I want, instead of wading through 3 derps who just randomly click every group in the list.

Same as a potential member. If I can just submit myself to what raids I am willing to join and what spec I wish to play and then go about my business instead of having to submit, wait, refused with no reason, submit, wait, join, get message from first group saying "can you join at 3rd boss?" say "no joined another group" etc.

[HuxNeva]

I preferred the Oqueue style of listing yourself and leaders finding you in the pool. Was a lot faster and less hassle to get into a group. Blizz always seem to do things ass-backwards and decide to make it so groups are listed and players must submit to join.

If I'm running a group and healer leaves I would much rather just have the Oqueue system where I can see Spec/Ilvl and choose what new healer I want, instead of wading through 3 derps who just randomly click every group in the list.

Same as a potential member. If I can just submit myself to what raids I am willing to join and what spec I wish to play and then go about my business instead of having to submit, wait, refused with no reason, submit, wait, join, get message from first group saying "can you join at 3rd boss?" say "no joined another group" etc.

It is an information game: A candidate expresses his desire to run a certain objective. even if he wants to, he has far too little info on the potential groups out there in the premade group finder to make an informed decision on which group to prefer over the other. Why does Blizzard expect people to keep selecting and applying to what are in essence random groups for the objective. As I explained because it is a self-reinforcing system, this can take hours for people that have fallen 'behind the curve'. A group creator also has far to little info to make an informed decision on the applicant and so tends to just go by class/ilvl per applicant. In practice she will fill for the needed roles with desired classes at the highest possible ilvl, or make room for a few paid carries. Can't blame her. The 'group suggestion' is there to lure creators into a 'lazy' option, where a balanced group is proposed that is 'good enough', not necessarily the 'max ilvl roles applying at that time'.