Discussion RE: Removal of LFR Kill Counter

Dungeons, Raids and Scenarios
Overview
I know there are a million and one threads complaining about the fact that the boss kill counter for LFR has been removed, however I just thought I would offer some suggestions to work around the main issue. As I've seen it, the root problem is as follows:

Some LFR groups lose members partway through, either due to wipe attrition or some players not needing the loot from the next boss(es). This causes the encounter to become a much larger time commitment than intended for those players still left in the group.

It has been said in blue posts (I can find them, if absolutely necessary) the removal of the kill counter from LFR has been accompanied by a silent buff that allows people to re-queue and have priority for fresh instances. I’ve also heard that players ought to complete the second instance because it still yields valor, which would be useful for the new gear iLvl buff system.

Problems
1) In all but the most specific cases, staying for bosses you've already downed will not be worth it. This entails that I will most likely quit once I've finished the encounters I need on the second run, as 45 valor for 1 or 2 encounters worth of time plus a second queue timer is not worth it. To put it into perspective, a heroic dungeon run, which takes 20 minutes tops plus 10 minutes for dps queue (30 min), rewards 80 valor initially, and 40 valor on subsequent runs. The LFR run takes 5 minutes per boss fight, 10 minutes for trash in between (this is generously quick), and a second queue timer of 20 min (25-40 min depending), which rewards 45 additional valor. Both sets of time estimates assume NO WIPES. The Yields (measured in valor points per minute) for each will be:

Heroic: 80vp/30min=2.67vp/min
Heroic (2nd run or beyond): 40vp/30min=1.33vp/min
LFR (1/3 w/ Trash): 45vp/50min=0.9vp/min
LFR (1/3): 45vp/40min=1.125vp/min
LFR (2/3 w/ Trash): 45vp/35min=1.286vp/min
LFR (2/3): 45vp/25min= 1.8vp/min

Even if I were to increase the heroic queue to match the LFR estimate (which goes beyond what I’ve experienced), the reward is still 2vp/min compared to the most generous LFR scenario of 1.8vp/min (unless the player is doing the heroic for a second time, then the 2/3 without trash would be the only LFR scenario to be more efficient). An efficient-minded player will take this into account, and make the time/effort maximizing decision. Also, this assumes VP is the main focus of the player. I understand the deserter buff for quitting halfway through a LFR prevents queuing for 30 minutes on anything (heroics including), however in that time I can do dailies, farm, level an alt, or enjoy my real life elsewhere. Also, if a player is focused on gear, not VP, getting a deserter debuff has no effect on the other methods of acquiring epic level loot (10-25 mans, dailies/rep).

2) It perpetuates the issue. If I join a group mid raid, but need loot from the first and/or second bosses, I will have to re-queue. On my second trip, I will likely drop out after the bosses I need (for the reasons mentioned in problem #1). This passes the buck onto the group I just left, and they will have to in turn drag fresh bodies into a partially completed LFR, and restart the cycle of pain. Also, people quitting as soon as they zone in and see the LFR is partially completed will cause the actual downtime to be longer, as the group has to re-enter the queue and compete with other groups ahead of them.
Solutions
1) Combine raids that have re-joined the LFR queue. If players avoiding partially completed dungeons are an issue for such a significant part of the population, it stands to reason the former group can be redistributed to make full LFR groups without dragging in those looking for a fresh instance. If you have five LFR groups with 20 players each, and they’re all on the same encounter, split one of them up and place its members into the other four groups. If a scenario occurs where either the distribution would result in a poor raid composition, or the demand for more players in a partially completed LFR is less than the supply of players, make THEM wait until a similar LFR group has room or the necessary composition can be generated by the queue system.

2) Allow players to choose to queue for a partially completed LFR group. Let’s not forget there are those looking to complete these LFRs as quickly as possible (those actually doing it for valor points, gear off the last boss, etc). They can help even out the revamped queue system discussed in #1. This would also allow players stuck between the cracks of my proposed system to break group and go quest (or farm, etc), but still remain queued for the encounters they need.

3) If you do not take the above two suggestions (There may be minority cases where it does not work), then move those players queuing for the second time to the very front of the queue, to minimize their invested time. Mathematically speaking, if the queue timer for LFR were reduced to 5 minutes, it would alter the VP/min estimates drastically:

LFR (1/3 w/ Trash): 45vp/35min=1.286VP/min
LFR (1/3): 45vp/25min=1.8VP/min
LFR (2/3 w/ Trash): 45vp/20min=2.25VP/min
LFR (2/3): 45vp/10min=4.5VP/min

This would mean the only scenario to be less worth it than a second heroic would be the worst case scenario, which statistically speaking ought to be the minority of cases. This, however, assumes a new instance will be created in 5 minutes from the sample player queuing. It may, take longer or shorter depending on demand for LFR at the time (I do not have access to those statistics).

Conclusion
The solutions I presented attempt to maximize each player’s freedom to play as they wish, without being penalized for making the time or effort maximizing decision of bailing on a partially completed LFR. If I have missed something in my proposal, please, let me know and I will make adjustments to my arguments, and if needed adjust my position accordingly.
2) It perpetuates the issue. If I join a group mid raid, but need loot from the first and/or second bosses, I will have to re-queue. On my second trip, I will likely drop out after the bosses I need (for the reasons mentioned in problem #1). This passes the buck onto the group I just left, and they will have to in turn drag fresh bodies into a partially completed LFR, and restart the cycle of pain. Also, people quitting as soon as they zone in and see the LFR is partially completed will cause the actual downtime to be longer, as the group has to re-enter the queue and compete with other groups ahead of them.


Yes but as long as people keep joining runs in progress that means nothing.

2) Allow players to choose to queue for a partially completed LFR group. Let’s not forget there are those looking to complete these LFRs as quickly as possible (those actually doing it for valor points, gear off the last boss, etc). They can help even out the revamped queue system discussed in #1. This would also allow players stuck between the cracks of my proposed system to break group and go quest (or farm, etc), but still remain queued for the encounters they need.


No objection to this, one of my regrets about this change was I could no longer fish for a run in progress. But having joined that run in progress I do not want to then wait 25 minutes for a single tank ie the last week prior to 5.1.

Solutions
1) Combine raids that have re-joined the LFR queue. If players avoiding partially completed dungeons are an issue for such a significant part of the population, it stands to reason the former group can be redistributed to make full LFR groups without dragging in those looking for a fresh instance. If you have five LFR groups with 20 players each, and they’re all on the same encounter, split one of them up and place its members into the other four groups. If a scenario occurs where either the distribution would result in a poor raid composition, or the demand for more players in a partially completed LFR is less than the supply of players, make THEM wait until a similar LFR group has room or the necessary composition can be generated by the queue system.


Just not practical, i think the biggest replacement I have seen this past 2 weeks was 1 tank, 2 healers, half a dozen dps. The algorithm would have to send a long time to find a run in progress on the exact same boss to make a combined run a full group. Even something as simple as 2 runs with 1 tank, 3 healers and 9 dps would not work because guess what that's 26 people.

Overall these last 2 weeks of lfr have been the quickest and smoothest ever. 1 run in progress, 18 fresh runs. Very few wipes. 1 Wipe from this last weeks 9 lfr runs. I also think your time estimates are off. I have done both parts of MV in 50 minutes including dps queue time.
12/11/2012 03:42 PMPosted by Cybella
It has been said in blue posts (I can find them, if absolutely necessary) the removal of the kill counter from LFR has been accompanied by a silent buff that allows people to re-queue and have priority for fresh instances.
This has been true forever, it was implemented about a week after LFR was born.

I think you're seeing a bigger problem than there really is. Randomness of drops will help distribute players across bosses, not everyone will be skipping out on the same one repeatedly. Only blizzard knows if the queue times get outrageous, they seem very reasonable whenever I join up as dps.

The only change that really seems worth the effort would be making all LFR completions worth 90vp. It's still too time-consuming to farm for valor that way, as it should be, but provides enough incentive that a few more players would stick around to clean up the number of partial runs in progress.
Solutions
1) Combine raids that have re-joined the LFR queue. If players avoiding partially completed dungeons are an issue for such a significant part of the population, it stands to reason the former group can be redistributed to make full LFR groups without dragging in those looking for a fresh instance. If you have five LFR groups with 20 players each, and they’re all on the same encounter, split one of them up and place its members into the other four groups.


This is so unbelievably impractical. I honestly don't even know where to start to explain how difficult this would be for the system to handle. It SOUNDS really good, but it is not an option. (Just to give you a couple issues (1) Groups of players that have queued togerher (2) almost every group needs healers - a 20 person raid with 2 healers isnt going to fill 3 other 20 person raids with 2 healers in each)

2) Allow players to choose to queue for a partially completed LFR group. Let’s not forget there are those looking to complete these LFRs as quickly as possible (those actually doing it for valor points, gear off the last boss, etc).


I personally would love this, but it doesn't work and it's purely a solution that benefits the INDIVIDUAL rather than every party involved.

3) If you do not take the above two suggestions (There may be minority cases where it does not work), then move those players queuing for the second time to the very front of the queue, to minimize their invested time. Mathematically speaking, if the queue timer for LFR were reduced to 5 minutes, it would alter the VP/min estimates drastically:


So you're suggesting that those who have NOT completed LFR for the week should be placed BEHIND those that have finished it? I mean, I guess this isn't a big deal but what's the point? If it reduces queue time for one person, it increases it for somebody else.
I'd be fine with the ability for people to force a fresh queue, but you only get paired with other people forcing a fresh queue, which means if someone leaves that spot is never filled.

If you aren't willing to help finish runs that had people leave, then the game shouldn't provide people to help finish your runs.

It really isn't a problem that people leave mid run. That will happen for any number of reasons, even if the group is going well and even if said person hasn't killed the later bosses. The problem that they were attempting to solve was it taking an inordinate amount of time to fill those vacancies. With the change, those vacancies have always filled very quickly for me. There seems to be no issue.

Your VP argument is also flawed. If you requeue after a partial run to get loot from the first bosses, the queue time is already a sunk cost. If you got a 2/3 the first time, for example, after you kill the first two bosses the second time the only time cost to finish the run is trash and the final boss, which will usually be more efficient than a heroic unless its your first one for the day. In any case, its not even a problem if you decide its not worth it and leave, the change has made vacancies fill quickly and the group will be full again before trash is over most of the time.
Duboomchikin, I see where you're coming from with regard to my first suggestion, as there is a potential for significant role imbalances (i.e. not enough healers in the pool). I think this hearkens to what Mate was also talking about. I also think this has been an issue since Dungeon Finder first came out back in Lich King, and is a product of population composition, not my proposed system. If you are suggesting the key roles (healer and tank) quitting is what grinds these runs to a halt, I acknowledge the notion as a possible trouble spot for my proposal. I had originally intended for the 1st and 2nd solutions to be used together, so those stuck in a limbo LFR run could break themselves off from the group and stay in the queue for their progression, and be prompted when they are matched up with a group. This would granulate the pool of players, and allow the system to match them up more easily. I'm not sure how often these limbo runs occur, but I would assume it's often enough to warrant the 5.1 changes. I will have to think about this further and get back to you.

With my third suggestion, I was in fact suggesting those people who have already completed a LFR run and want to go on a second run to get the boss(es) they didn't fight, simply to minimize the re-queue times. Since they already went through one long queue to get the partial run, no sense in making them wait through another long queue. Though, upon further reflection, this may not do anything to reduce the average queue times each person would take, as bumping these players to the front of the line will force an additional wait on those behind them, and if those players also get into a partially completed run and get bumped to the front of the fresh run queue, this will force the next round of players to also have to wait longer and so on. Based on that, I shall have to think about this further.

It also seems like most here agree they would like to be able to specify if they want to queue for a fresh run or if they would like to queue for one already in progress. And Asthas' suggestion of a 'Fair is Fair' version of the LFR run, where if members drop out of the group, you cannot requeue is an interesting thought. I think there is a lot of uncertainty as to whether the ability to specify fresh or partially completed would be beneficial to the whole LFR community, or just a few individuals.

(To Asthas) I do not know if queue times ought to be considered a sunk or a variable cost. A sunk cost is one that is incurred one time, when an actor decides to enter a market. A variable cost is one incurred based on production levels (in this case, one has to queue once to receive a number of VP; Valor Points is the product, and the cost of 90vp initially, and 45vp subsequently is the time in the queue). It would be a sunk cost if you had to spend an initial amount of time in order to keep running the raids indefinitely. Efficiency decisions are made on the margins, and if you have to constantly spend something in order to produce, it is not a sunk cost.

My argument is geared toward the additional queue time and the time of re-fighting bosses you've already downed that week as being opportunity costs of the current system relative to other activities for the sole purpose of gaining Valor. The mathematics (which I will admit is fuzzy), was only to illustrate the point that the choice of sticking with the second run is not necessarily efficient. Whether or not it's worth the time and energy spent overhauling the system on the basis of the inefficiency is another question, which I did not address. Further data gathering would determine how much extra time the average WoW raider spends with this system, which in turn would determine if additional changes would be necessary.

Based on a general consensus in this thread so far, the 5.1 system works great for most people. I suppose it is like Breathkeeper said and I am making a bigger issue out of it than it is. I just wanted to spitball some ideas to see if there might be a more elegant solution to the LFR wait times when people quit mid run. I've talked to several people who have still gotten stuck in this sort of limbo despite the changes made in 5.1, and it still seems to be a topic discussed on the forums. Granted this is all anecdotal evidence two weeks into the patch (then again, so are the accounts in this thread in the opposite direction).

As a side note, I like Breathkeeper's idea of making the second run also count for 90VP. Perhaps this could be part of the passive buff that prioritizes the fresh run in the second queuing.

I want to thank everyone for their discussion!
systems been running perfect since change, don't think they need to do anything else, besides ban griefers now.
*snip*


Holy cow, if we had responses like this all the time these forums would be the most productive thing since assembly lines!


(To Asthas) I do not know if queue times ought to be considered a sunk or a variable cost. A sunk cost is one that is incurred one time, when an actor decides to enter a market. A variable cost is one incurred based on production levels (in this case, one has to queue once to receive a number of VP; Valor Points is the product, and the cost of 90vp initially, and 45vp subsequently is the time in the queue). It would be a sunk cost if you had to spend an initial amount of time in order to keep running the raids indefinitely. Efficiency decisions are made on the margins, and if you have to constantly spend something in order to produce, it is not a sunk cost.



If you requeue to get bosses you didn't get the first time, you've already paid the queue time. A sunk cost refers to costs you've already paid and can't recover. Every queue in the past is a sunk cost. You can never get that time back. Every queue in the future is a prospective cost, you can choose to not pay it if you'd like.

Lets run through the decision making progress for someone that wants loot from every boss. (or at least enough bosses that require a fresh run to get them all in one go)

You queue once, get a 2/3. You kill the last boss.
You want gear from the first two bosses, so you queue again and get a fresh instance.
You kill the first two bosses.

Now you have a choice: You either stick around for the final boss for 45 VP, or you leave. You already paid the queue time to try to get loot from the first two bosses. You probably would have queued even if the run offered 0 VP at the end, because the loot is valuable to you. You don't have to pay it again at this point to get 45 VP from this run. You would have to pay it if you wanted to queue a third time for just VP (or for spending lots of coins on the same boss). If you leave the group after the bosses you needed, you don't get refunded the queue time. That is why its a sunk cost. It shouldn't factor into any future decisions. Additional queue times in the future, sure. But not queues you already waited in.
Hey guys, I got a great idea. How about people queueing for roles that are undermanned get extra valor and a bag with more gold, or maybe even some tokens, or maybe even a guaranteed drop bonus of 10% or so on a boss of their choosing.

Holy !@#$, I just fixed lfr queue times without the need for all that convoluted stuff you recommended. And how about that, the system is already in place for lfd, it's almost as if Blizzard is just lazy.

Or maybe they just don't care about lfr queue times because lfr was just a bone they threw to the casuals to shut them up. I'm gonna go with that.
Or they could just make healers FOTM like they did with tanks

Was laughing so hard when I saw reroll numbers because "OMG TANKS DO DAMAGE?!"

Or maybe they just don't care about lfr queue times because lfr was just a bone they threw to the casuals to shut them up. I'm gonna go with that.


And, as was predicted in all the 5000 LFR threads before he came to life,

They've gotten even louder crying over loot.
Or maybe they just don't care about lfr queue times because lfr was just a bone they threw to the casuals to shut them up. I'm gonna go with that.


I thought it was about them being able to see the end game content with no commitment ...did I miss anything?
I do worry about making LFR too valuable in VP also. Allowing it to kill off LFD/scenarios would be a mess. The high-end players getting lots of VP from real raidbosses won't be rerunning LFR as much as those trying to collect all of their weekly valor, so performance metrics will go down and LFR groups will be even *worse*.
2) Allow players to choose to queue for a partially completed LFR group. Let’s not forget there are those looking to complete these LFRs as quickly as possible (those actually doing it for valor points, gear off the last boss, etc). They can help even out the revamped queue system discussed in #1. This would also allow players stuck between the cracks of my proposed system to break group and go quest (or farm, etc), but still remain queued for the encounters they need.


I actually really like this idea. I really don't see any downside other than to the individual person who uses it, and that becomes their choice anyway.
12/12/2012 06:00 AMPosted by Garitou
Or maybe they just don't care about lfr queue times because lfr was just a bone they threw to the casuals to shut them up. I'm gonna go with that.


I thought it was about them being able to see the end game content with no commitment ...did I miss anything?


Yeah, you did. It sounds like you missed pretty much all if TBC, The Birth of Casuals. BC was when Blizzard first realized WoW was a huge hit, and that they were attracting a lot of casual players, and the forums were filled with literally hundreds of threads daily of casuals crying that they pay 15$ too so they should get the same gear as the hardcore raider who spends 4 hours a day farming for mats so he can raid. The whole "getting a currency and buying high level epics with it from a vendor", which is such a staple nowadays, was actually called wellfare epics back then and was looked down upon. LFR was the natural and logical culmination of the big change Blizzard started in BC. And Trollonforums is right, in the end it only resulted in even more crying that they can't get "what their 15$ entitles them to".

Join the Conversation

Return to Forum