Greedy Goblin

Tuesday, January 25, 2011

Motivating progression

Quick business tip: check the price of inferno ink on your server. It is reagent for offhands, relics and darkmoon trinkets. It is often purchased by non-scribes who seek scribes with materials for crafting. If you mill a stack of Cinderbloom, Stormvine, Azshara veil, Heartblossom, you get 5 Blackfallow inks, 1 Inferno ink. If you mill a stack of Twilight Jasmine or Whiptail, you get 6 Blackfallow inks and 2 Inferno inks. Since you can trade 10 Blackfallows for 1 Inferno, you get 1.5 from a stack of lower and 2.6 for a stack of higher herbs. On my server Stormvine goes for 40G/stack, inferno ink goes for 90, that's 95G profit for milling a stack of herbs.

The PuG update:
3 tanks, 3 healers, 4 DD, 8K pot. We are permanently recruiting, just read the rules!

This lockout was great and bad at the same time for the guild. On Wednesday (reset day in Europe), we killed all 4 farmbosses in a single raid. Considering that we firstkilled 2 in the previous week after whole days of progression raiding, it is great.

The bad is that we couldn't raid on Thursday, Friday, Saturday, and only 1.5 hours in Sunday, due to lack of people. No need to say that there was overbooking for the farm raid. When finally people cared to show up on Monday... see above.

This problem is common among all guilds: people prefer the farm raids where you kill lot of bosses easily for lot of loot, and avoid the raids that may kill one boss. In HC guilds the solution is simply the attendance rule. You can't keep your attendance if you skip progression raids. Also, if there is DKP system, DKP is given for time spent wiping. This has obvious drawback: if someone joins the guild later, he'll be 10 billion DKP behind.

Both systems are directly against our rules, but something has to be done. The solution is the all-powerful booster's gold system: If you are killing this boss first time, you are considered newbie and you must give up some gold from your pot share (if your share is less than this gold, you won't go negative, simply no pot for you). That part of the pot is distributed evenly, providing booster's gold for those who killed that boss already. So if you come to progression raid, you'll be rich next farm raid.

Again: this is not a direct guild rule, it's the decision of the raid leader, but of course the "firstkill gold", along with the "leader's gold" must be told before the raid start, preferably on the calendar. I'm thinking about 2000G, because a firstkill takes about 3 hours and 800G worth consumables and repair, so it's still just 400G/hour payment.

The following table describes how much gold do firstkillers and veterans get, depending on their number for a boss. I assumed that the loot is providing 6000G pot, as this is the average that we had on our raids. The second and the third coloumn is for one boss. The fourth and fifth describe the situation when the "newbies" had half of the bosses already, firstkilled the other half (the data is per boss)
Newbie count
Newbie share
Veteran share
1/2 share
2/2 share

Please note how the system automatically adjust to the situation. If just a few newbies are in the raid, they slow the raid down just a bit, and accordingly the pot share increase is small. On the other hand if there are only a few veterans, they will most probably wipe several times while they iron out every little mistakes in the raid, so their high share is well earned. Someone who joins the guild late is not 10 billion DKP behind, he simply has to raid one without pot share and from that point he is "veteran" too. But most importantly: if there is no dead boss, there is no share to distribute, so no one is motivated to be warm body somewhere.

This system motivates coming to progression raids. You have to firstkill a boss once, and the only way to do it while getting gold is coming to a progression raid, where everyone else are newbies too. Also, it's a very strong motivator for veterans coming from more progressed places to join guild-firstkills, as such people could get huge pot. The same applies to guildies who did the firstkill, and next week they see only a handful of their old team online and bunch of newbies. The huge pot motivates them to do another progression raid instead of just skipping this day and wait for more veterans coming next day.

The system also motivates including newbies to your raid, instead of forming cliques: if you bring newbies along, you get more gold. This will allow forming 2 10-men raids when enough people are available instead of one. The two will wipe some, but at the end, everyone will be happy: veterans got lot of gold, newbies seen bosses, earned their veteran status.

Calculation of the share is not straightforward, as the result is lower-limited by 0. The simplest case: 10K pot, 10 players, no newbies: everyone get 10K/10 = 1K gold. If anyone made a 300G worth wipe, I just subtract 300 from him and add it to the pot (these are remembered by an addon). So everyone will get 10.3K/10 = 1030, except him who will get 730G. However I can't use the same straightforward method for 2000G, as it would provide 1200G pot to everyone except him who gets -800. If I just ignore that number and pay 1200 to everyone else, I paid that 800G from my own pocket.

The proper calculation method at the end of the raid: I write to a piece of paper every "newbie's" name and how many times 2K they would have to pay if if they would have enough pot share. I add the total number to the pot in a pocket calculator and divide it by 10. I compare this pot share to the "to be paid" numbers on my paper. If the "to be paid" is larger, I cross out the name on the paper and remove the guy from the raid. I repeat the above method until no one in the raid has negative share.

Example: 10K pot, 2 bosses, 3 0/2 newbies, 1 1/2 person, 6 2/2 veterans.
First pass: the newbies add 3*4000, the 1/2 person adds 2000, pot is 24000, share is 24000/10 = 2400. 2400 is smaller than 4000, so I remove the newbies from the raid without payment. As 2400 is bigger than 2000, the 1/2 person stays.
Second pass: the 1/2 person adds 2000, the pot is 12000. Individual share is 12000/7 = 1715G. 1715G is smaller than 2000, the 1/2 person is removed too.
Third pass: The pot is 10000, individual share is 1667, that's what the veterans get.

One more question emerges: is it fair to cut the payment at zero? Wouldn't it be better if the newbies could go negative and must pay 2000G/boss upon firstkill (this way the example would end at first pass: 2400G share for veterans, 400G for the 1/2 guy, -1600 for the newbies. Many said it would scare off people joining, but I never cared about freeloaders not joining. The real reason of the cutoff is that the cost of carrying a newbie diminishes over time as the veterans get more experienced and geared. Boosting a newbie over Halfus is a trivial task unless he is utterly useless (which case we simply kick him from the raid). So charging him the full 2000G for that trivial job is rather stealing than legitimate service. So a 0/5 newbie on an 5/5 farmraid (30K pot from loot) just loses his 3000G pot share, which is a fair price for carrying him over 5 bosses, while 7000G payment plus losing 3000G pot would be robbery for a farm run.

The 2000G kicks in exactly where it should. The 4/5 guy, who was busy farming last week's 4/4, but skipped the progression wipes on the 5th boss gets 2000G less than the veterans. Next week he'll join the progression!

There is one problem: what if someone wants to come, but can't, because we can't bring 12 people? Bad luck. However since 400g/hour is not much, if he spends the time while we are wiping by making gold, he'll be easily able to pay the "newbie"-fee.


Azzur said...

In society, the solution is the government - where they provide the services & regulations that would not have eventuated from a complete free market system.

Thus for WoW, I believe that there are 2 other options:
- A "raid tax" where x% amount is held over to fund your progression raid. If the participants refuse to pay the tax, they can choose not to come to the farm raid.

- The guild bank can be used to fund progression, which is useful since cata has the automatic guildbank contribution.

Anonymous said...

You need to check those numbers on those inks, 4.5 Blackfallow Ink is not 5 and 0.5 Inferno Ink is not 1
Twilight Jasmin does average 6 Blackfallow and 1 (not 2) Inferno Inks per stack. At least that's what I came up with in the nearly two months I've milled Cataclysm herbs.

Jumina said...

Actually newbies are never "10 billion DKP" behind. Older members have to spend DKP on some expensive things while newbie takes them later for minimum bid. That is my experience. Other possibility is EPGP. It's based on priority (points for attendance divided by points for loot). The trick is that once a week a "decay" is applied to the points (for example minus 10%). That prevents players from stockpiling the points.

But all of this does not prevent people from preferring the farm before the progress. I'm looking forward for the results of your experiment.

Taemojitsu said...

<irrelevant preface omitted> If the pot from early bosses is paid out for killing later bosses for which only part of the raid has experience, it rewards leaving early (during break) when additionally it isn't known if the bosses will be killed. Despite the reward of earning 'veteran' status on a boss, the design goal of the situation automatically adjusting so that more veterans means a lower boost reward does not hold when enough bosses are killed so that the full bonus can be taken from the pot for all veterans due to bids from earlier bosses.

Conversely, making boost proportional to veteran status would not reward veterans on bosses for which there are few veterans in the raid, as their higher share would be averaged out with easier bosses. Since in the example chart for 6k pot the total 'boost' amount maxes out at 7 newbies (-600 * 7 = 4200, compared to -400 * 8 = 3200 for 8 newbies) what about giving a "boost number" for each boss based on the number of newbies and multiplying it by a gold amount.

Conceptually, "award bonus gold to veterans" seems an easier way to think about distribution than "subtract boost gold from newbies". In the 3x 0/2, 1x 1/2, 6x 2/2 example, using the current ruleset, it becomes simply
1) pay 2k per veteran for last boss (or "for being veteran of every boss")
2) *total: 12k, more than pot, so split pot between 6 people = 1667 as described.
3) if pot had been 20k, then next step: pay for "veteran of all bosses except one, or less". 7 people, so out of 8k left after step 1 would split for 1143 gold for 1/2 and 2/2 complete veterans.

With a "boost number" system, in this same example the 2/2 complete veterans would have 4 + 3 = 7 boost points each, while the 1/2 complete veteran would have 3 boost points. If each is worth 300g, then from a total of ([7 * 6 = 42] + 3) = 45 boost points worth 13500 gold of boosting, the 2/2 complete veterans would receive (10000/45) * 7 = 1556 each, while the 1/2 complete veteran would receive (10000/45) * 3 = 667 gold.

stop ignoring my comments

Ðesolate said...

Question about the newbie-status. If the Boss is killed in a non-Guildgroup is he still a newbie, since he made the boss? (only difficulty is that he possibly performed bad but weren't removed from the raid, thanks to "nice ppl")

A negative pot share would scare the newbs off. I personally would not pay for a Bosskill unless I can go afk or my personal performance is of no matter. Not gaining a share of the pot may be passive payment but at least you can leave the raid +/-0. My pot share would be experience.

In my former Guild progression-runs usually had a x3 DKP-modificator which also applied to the guildfirstkill bonus. Theese runs usually only aimed fot the yet unfinished bosses. Translating this to GDKP would be absolutly stupid. Unless the payment would be made by the 10% Guildearnings (what is farmed by other players so it would be a bit "unfair").

Orenja said...

I'm still not convinced, but it sounds a lot better to me today than it did on the weekend.

One point I'd be interested in your thoughts about: You're mentioning those that can't be on the firstkill because there's no slot for them - what about those that can't be on the firstkill because they are not raid-ready yet, e.g. just became 85? I for example would love to participate in progression raids but I'm simply gettingt my gear together right now as I didn't immediately start leveling upon Cataclysm-release; and even though you're well known for being a proponent of "gear doesn't matter", there still is a certain minimum that should be aimed for (and, given same skill, those with better gear will simply outperform me and therefore likely take my spot).
So, if I spend the time not with getting gold, but gear, your argument cannot be used here.

I hardly have a problem with not being on the firstkill from a money perspective, especially since we established that there is no actual gold to be paid, but rather pot to be minimized, I'd just like to hear your thoughts on it.

Riptor said...

Con only Scribes trade Blackfallow Inks for Inferno Inks or is this possible regardless of the own Professions?

Anti said...

Stormvine @ 2g each would be nice.

gold is worth practiacally nothing on my server. there is not even normal GDKP runs. this means most decent players have stockpiles of gold from months with nothing to spend it on.

is it worth so much more on your server? regular GDKP runs and your guild especially probably has a large effect. players have always had something worthwile to spend their gold on.

does this mean my server price of 5g for cinderbloom and stormvine is artificially high because of the high amount of gold in circulation?

nightgerbil said...

Your making your loot rules more and more complicated as times goes on. I will be honest my maths isnt that strong, I seriously doubt that I will be able to function as a raid leader/lootmaster if your trend continues. This is a bad thing right? You wanted more raid leaders to emerge?

Now I grant you saying that the above addition might be optional, but if veterans want to raid under that system they would join "x"s raid as opposed to mine if I didnt use that particular reward system. To remain competitive with the raids organised by yourself, I have to use similar (or more popular) loot systems to you. Now I grant you could just call me an idiot and tell me to just go and join a raid and not lead them if I cant handle the maths, but I think the more complicated this gets the fewer will understand how to administer them and that will defeat the purpose of the project as I understand it.

Gevlon said...

@Azzur: taxing and government has the problems that somehow the tax never reaches those who work for it. In this system if X boosts Y, Y pays to X.

@Anonymous: my numbers are backed by both Wowhead and my own experience.

@Jumnina: "point for attendance" is a bad idea as it motivates being there, and not killing the boss. So in a "pay for attendance" system, you get lot of raiders watching TV.

@Taemojitsu: if someone always leaves in the break, he never sees higher bosses than Halfus and Magmaw. If he ever wants to see Cho'gall, he must either be there on the progression kill, or pay once for firstkill.

@Desolate: of course if you killed that boss anywhere, you are a veteran.

Actually the share of the newbie is more than experience: chance to bid on loot, valor points and the "veteran status" that allows him to claim pot next time with us or to join "link achie" trade pugs.

@Orenja: if you become raid-ready later, all you have to do is join one farmraid where you get no pot and earn your veteran status.

@Riptor: anyone can trade them.

@Nightgerbil: there are addons for everything. If this system works, it will surely be included to GDKP addons.

Anonymous said...

The milling results are way off. I milled hundreds of stacks for the faire and far more this expansion for glyphs and I am certain the low level herbs give on average per stack

5 blackfallow ink
.5 inferno ink (1 ember)

And that the high level herbs give on average per stack

6 blackfallow ink
1 inferno ink (2 embers)

This is backed up by comments in wowhead of some large trials eg: "I milled 49 stacks of cinderbloom today and got:
51 Burning Embers
512 Ashen Pigment

"For some Milling data:
520 Twilight Jasmine Milled = 104 mills.
307 Ashen Pigment
57 Burning Embers"

That would be half the number of burning embers you've stated over 300+ mills, the chances of which are infinitesimal. The key is not looking at just the wow drop table because although you *can* mill 1-3, it's not equally likely you will mill 1 or 3. You're far more likely to mill 1.

scribus said...

@Nightgerbil: I agree, I actually have quite a strong math/econ background but feel a bit confused. It seems to be a bit more complicated than necessary, at least to me personally.

I think there might be demand for more than one raid level at this stage of the PuG. I personally have not seen any of the bosses. I want to gear up first, as you say (head enchant, heroic drops, reputation rewards, leveling LW etc.). Not necessarily because I must, but because I want to. I am not in a rush. Once I am satisfied with myself and time permits (living in the US means server time is not always convenient), I might enter a raid in the calendar myself.

But I am sure if someone "steps up" and enters a raid in the calendar to attempt some bosses which are on "farm" there would be demand. Maybe also a bit less pressure than joining (even as a "newbie") into the established raid with Glotan (Do I get kicked if I stand in the fire for 0.5 seconds? What if I mess up a mechanic that I read about but couldn't immediately execute, will I be seen as "M&S" even if not called out?)

Orenja said...

Okay, so maybe this is what irks me: It doesn't really concern me that I have to miss out on the pot once. And it shouldn't, because if it were seriously punished than we'd just be off to were we don't want to be: kind of forcing raid attendance.
To reinforce that: noone should ever feel obligated to be in every raid because otherwise they might have a longlasting disadvantage - I think we agree on that.

But then again maybe this is were I don't see this idea working that well: Where's the incentive to go to progression raids, now? This system only get's rid of disadvantages progression raiders had so far, that is, not getting paid for their time.
What it doesn't introduce is an incentive to actually do progression raids, as one can simply wait for farm status, sit out on the pot for one run and be in the same position as the person having been in the instance for weeks and actually helping with progression.
This whole system is based on the assumption that noone went to progression raids now because they didn't get paid; I think that a lot of people simply don't enjoy wiping over and over again that much and they'd rather use the time doing something "fun". (As I already said, I very much enjoy progression raiding, but I'm pretty sure that's not the majority of people)

I simply don't see that much of an incentive to do progression raiding over waiting it out and losing on the pot once, therefore I'm not sure whether this idea will work out. It hardly harms anyone, so why not try it, but right now I'm not expecting it to change the situation radically.

Also, I see the system failing in the future as soon as some boss exceeds the estimated time of tries and therefore g/hour sinks. To avoid this, the 'booster gold' would have to be a dynamic amount with a fixed g/hour rather than a dynamic g/hour with a fixed total amount. This would be a logistical nightmare to manage, especially in this guild where there is no fixated raiding team and therefore individuals will likely spend a different amount of time 'making progression' on a boss.
I'm not saying this WILL actually happen, I'm just saying that if your assumptions are right it would be the only logical thing to do: Wipe on a boss as long as the g/hour is better than farming/playing the AH/..., and the second it goes underneath it, quit it and do something else. This will only work as long as all estimates hold true all the time.

So, I see a number of theoretical problems cropping up here (e.g.: what motivates people, how is being responsible for progression assessed, ...).
This idea might work out, especially in the short run. But I don't think it is worked out well enough to put much hope into it in the long run - it MIGHT still work excellent a year from now, but I wouldn't put my money on it. (and I'm not talking about current raid instances in a year from now, but similar situations regarding progression raiding)
As I said, there's hardly any harm in trying it, and I'm also not expecting ideas like this to be worked on for hours and hours, putting way too much time into effectively just bureaucratizing the guild unlike it is supposed to be. And this works, because you're not afraid to admit defeat and reverse unsuccessful changes to the rules as you've shown in the past.

In short: I see two possible strategies behind your new rule: You either wanna make progression raiding a more attractive option, equal to farm raiding; or you want to play the 'social ppl" into thinking it is (that's what most your PvP-strategies for random groups are about)
If it's the former, I'm not convinced that progression raiding is indeed a better option than it was so far - but I guess I'll do some number crunching on that one to find it out.
If it's the latter (it doesn't feel like it, for various reasons) I'm not sure it will work, e.g. because people's motivations might have been misjudged.

Jumina said...

@So in a "pay for attendance" system, you get lot of raiders watching TV.

My inaccuracy. In the EPGP system EP points can be awarded for time spend on the raid or killing the boss or both. As a raid leader I can decide whether to award EP or not. My point was that the newbie in these systems is not so much "behind" as you perhaps think.

But as I said this is not something which would actually stop players skipping the progress. Only officers watching their players can do.

The DKP/EPGP are actually just systems for distributing the loot. They can prevent long discussions about loot but I am not sure if any system can change this behavior. I am curious about your future experience.

Gevlon said...

@Orenja: currently there is no compensation for coming to progression raid over farm raid. I agree that this system (exactly because it's not overly punitive on newbies) is not a perfect compensation. But more than nothing. We had progression, so many people came to progression raids for free. If this system just increase that number by 10-20%, we won't have to cancel raids anymore with 8-9 people.

@Anonymous: Even if we assume 6 blackfallow ink, 1 inferno ink for whiptail it's still profitable, despite less. (120G/stack whiptail, 1.6*100G ink)

The lower herbs with 5 black, 0.5 inferno for 40G/stack are also profitable.

I'll make a huge, comprehensing milling in the week to get hard data.

Dangphat said...

Gevlon what is your opinion on EPGP?

It attempts to solve the progression raid/players joining the guild mid tier issue.

Eaten by a Grue said...


I think you are probably wrong about gold being worth nothing on your server. I do not know your server, but unless it is so rare that anyone sells anything of value, I really doubt gold is worthless.

There are many ilvl 359 and 346 and pvp BoEs that are still very useful to many players, and gold naturally becomes the medium of exchange for these. Gold is not so easy to come by as to be meaningless. It is not hard to earn gold, but it does take effort, and comes harder or easier to certain people. There is a scarcity of it, despite the seeming inflation.

Perhaps if your side is of such low population that the economy is simply broken.

IrOn TuRtLe said...

"No need to say that there was overbooking for the farm raid."

Um, why not say? Specifically, why not say how overbooked you were, if you're going to bring it up? If it was twelve people, that's just random variation. If it was twenty, why didn't somebody start a second raid?

Breevok said...

And when you do your milling test you will discover that you are wrong

Cinderbloom/Stormvine/Azshara's Veil/Heartblossom - one stack will give 5 blackfallow and 0.5 inferno

Twilight Jasmine/Whiptail - one stack will give 6 blackfallow and 1 inferno ink.

This is the same as we saw in LK deadnettle v icethorn.

I am surprised you post such erroneous data, but you are right it is still profitable to do - if the inferno's sell. Your market is not scribes who all know its cheaper to mill than buy so your market is limited to M&S.

Anti said...


i dont disagree that gold is worth something now. even the entry level pvp blues are currently worth spending gold on for people.

however without GDKP runs for the last 2 years the majority of players have had little to spend gold on since the beginning of wrath.

my faction is the higher population of one of the two fullest oceanic servers (which were all "locked" at release of Cata)

the price people are willing to pay for Darkmoon Faire Trinkets(i sold a tsunami for 49.5k) i would say is caused by the large stockpiles of gold they all have. and this causes the herb prices to have such a high price.

Taemojitsu said...

> If he ever wants to see Cho'gall, he must either be there on the progression kill, or pay once for firstkill.

If he or she is in a raid that kills one new boss each night (or one boss that they haven't seen each night, even if separated by two weeks apart each), they pay four times, one for each boss. If they are in a raid that kills every boss including Cho'gall in one night, they pay only once as a single round of boosting will probably use up the entire pool of shared gold and the other three bosses are given for 'free'.