Greedy Goblin

Monday, June 21, 2010

The 1/3 rule in gold bid

The PuG update: 77 people in the guild, come!

Undergeared update: The soccer championship hit us hard, all three of our resident tanks were away. So "impossible" Festergut was tanked by usually holy paladin and a bear who never tanked in raid and never-ever was kitty in his life. One of the healers was our resident elemental shaman who never healed in raid. We had no warrior, so we had no commanding shout, so the pala healer had less than 40K at Festergut. We didn't have +3% hit debuff either. Nor totem of wrath/moonkin aura, +6% healing, sunder armor and lot of little ones. And Festergut was down to 450K (video). Less than 5%. I think he is doable with this setup if healer shamans chain fire elementals (205K), and all healers do just 500 DPS (that can be done if by rolling SWP/Flame shock with focus macro. Or if the cat's DPS increase. Or if I suck less when vomiting (there is a little time after debuff arrives when I can still cast a barrage). Festergut will surely be down if any of our resident tanks return, if we get one more hunters or 25% buff arrives. So while the naysayers belonging to the cult of Festergut gained one more week, their time is surely running out.

Have you noticed that in PuGs and 5-mans always hunters and death knights pull the lowest DPS? They use to say "lol i just haz no gear". Next time send them the link of this post to show what one shall do in 3300 GS. I especially recommend the "active time" field to be mentioned them:


As I already wrote, I think the second highest bidder should be compensated when the loot is distributed via gold bid (akwardly called "GDKP"). My suggestion was to give him 1/3 of the money the winner paid. The reason is that every raider have "raid optimizing" power, trying to exclude anyone from the raid who is a possible competitor. It's strongest with the RL and his buddies, but everyone has some: "I don't join to raid where I have competitors". Such "optimizations" weaken the raid and delay its start. I believe (and the raid experience in the ganking project proves) that the 1/3 rule successfully lure people with high competition into the raid (due to the high compensation).

However another aspect must be examined: can the system be abused? The simple way of doing it is someone without any need for the item bids just to be second highest bidder. He'll get 1/3 bid and in the case of ending up #1 he sells the item for someone who wants it. I'd like to prove that this is impossible, therefore the 1/3 rule should be applied in gold bid runs. The content below is highly technical, if you are not interested, the conclusion is: "no, the system cannot be abused profitably, the abuser just hurts himself"

For simplicity let's say that the minimal increment is 100G and everyone bids with minimal increment. At first let's discuss the outcome of the bidding without "bid ninja":
  • Different people are bidding, the bid goes up.
  • People quit the bid one by one, only 2 are left, keep bidding.
  • The bid reaches the point where the second guy (SG) no longer wants to bid (second guy breaking point, SGBP) so he places no more bid.
  • The item is sold for SGBP or SGBP+100G depending on who placed the last bid when SG choose to quit. Please note that the bid ended near SGBP and not "first guy breaking point"
  • SG gets SGBP/3 as compensation
Now let's add bid ninja to the system, and let the first guy use the optimal strategy: don't bid against bid ninja.
  • Different people are bidding, the bid goes up.
  • People quit the bid one by one, only 2 + BN are left, keep bidding.
  • The bid reaches SGBP, so SG places no more bid.
  • If FG placed the last bid, BN places 100G more
  • FG does not bid against BN
  • item sold to BN for SGBP+100
  • FG gets SGBP/3 +33
  • BN offers the item for sale. First guy already knows SGBP, as it was obviously announced by SG himself who stopped bidding. So FG offers SGBP and not a cent more. It's unreasonable for SG to bid more now if he did not bid 20 seconds ago, and BN has no options. He must accept any gold to decrease his losses.
  • The item is resold to FG for SGBP
  • Financial summary: FG paid SGBP-(SGBP/3+33) = SGBP*0.6666-33, BN got SGBP-(SGBP+100) = -100
The bid ninja cannot end up second, unless either FG or SG make a mistake. Such mistake can be stopping before SGBP. But the 1/3 system strongly discourage it, since if BN is second, either FG or SG is third. The third guy gets nothing, so he is encouraged to bid to get the item (which is valued higher than current bid, remember we are before SGBP).

There is one exception though: if bid ninja is not completely a ninja but someone who wants the item, but not for that high. He is actually third guy, but when he reaches his own breaking point (TGBP), he keeps bidding, in hope of financial income. First and second guy do not know he is behind his breaking point and assume he is a legitimate bidder who wants to use the item. So FG bids against him after SGBP. TG/BN don't bid if he is second, so the item is sold for SGBP+200 for FG, TG/BN being second, he got his SGBP/3+66 (taken from SG).

However the story is not over as the same strategy is open to SG too! He can also keep bidding, until FGBP, where first guy quits. The item sells for FGBP+100 or +200 making either FG or the other ninja #2. Since the item worth less to the winning ninja than the price, he has to sell it for FGBP to the first guy, losing 100 or 200G. If he equips the item to conceal he is a ninja, he loses FGBP+100 (or 200) - TGBP. While it's true that either FG or one of the ninjas will be very happy, FGBP is unknown to them, no ninja can position himself to second place, one only ends up there by luck (he bids, other ninja bids, FG don't). So if there are more than one ninjas, one will win the 1/3, but the other will lose more, so on average being ninja does not pay.

As extreme situation it can happen that there are three ninjas bidding against each other up to the sky (until one runs out of gold) making everyone in the raid very happy except the ninja who ends up first who loses sell price-FGBP gold.

So while a third-guy-ninja strategy can work, it depends on being the only bid ninja in the raid which is a bold assumption among people who join to a gold bid raid.


Rohan said...

Your analysis is flat-out terrible.

You assume that the identity of the bid ninja is known. This is a horrific assumption. This is like saying "We can prevent being scammed by not doing business with the scammer." The entire point is that you don't know who the scammer is if they are at all good.

Consider the case with three bidders Anna, Betty, and Nancy. Anna's secret is that she is willing to pay up to 1000g for the item. Betty's secret is that she is willing to pay up to 2000g for the item. Nancy's secret is that she is aiming for second place.

The three start bidding. Eventually Anna bids 1000g. Nancy bids 1100g. Anna drops out. Betty has a decision to make. She's still under her limit, and she does not know that Nancy is a bid ninja. So she bids 1200g. Nancy drops out, and gets 400g for her trouble.

Of course, if Nancy is stupid enough to resell items, or bid for items she obviously can't use, then it's easy to identify her as a bid ninja. But if she's at all smart, she won't do any of these things. She'll take the loss when she has too, and make it up with steady gains.

Gevlon said...

The bid ninja is obviously identifiable. If he bids on an item he can't use or obviously won't (like a bear to a resto item despite he never-ever healed in his life), it's obvious that he is a ninja.

In your example Nancy don't know Betty's price, so she can end up first place. In this case she HAS TO sell the item or she gets a big loss. By selling it, she identifies herself.

If she does not sell the item, she loses even more (up to the full price), making her adventure even more harmful to herself.

The last case is already mentioned: if the bid ninja can use the item therefore can conceal the fact that he is a ninja, he is at the mercy of others being honest. If he is the only ninja AND lucky, he'll get money. If he is not (and one shall assume so in a gold bid raid full of goblins), he can end up being first on a completely inflated bid.

sw0rdfish said...

My worries in a system like this would be people working together to increase their Breaking Point by 33%.

I tell my friend that I want the item, and my breaking point Is 3000G. He bids right along with me until we're the last two, then he stops and I beat him by the minimum bid. Our agreement is that I get the 1/3 back. So really my breaking point is 2000ish, but I know that if I go to 3000, its only goona cost me 2000 when I get my money back.

His benefit is that I do the same for him when an item he wants comes up.

Rohan said...

Gevlon, Nancy just keeps bidding until either Betty or Anna drops out. As soon as one of them drops out, Nancy drops out as well.

Nancy's target is the differential between Betty's max and Anna's max. Nancy knows that it exists, because logically it is extremely probable that it exists. The only case where Nancy will lose is if the differential is very small, and Nancy unwittingly exceeds it in one or two bids.

Also, the point of skilled scammer is that she is willing to take a loss now and then, if they will make it up with multiple gains.

To me, it seems obvious that an intelligent bid ninja's greatest asset is the secret that she is a bid ninja, and so she should ensure that secret is kept.

Gevlon said...

@Sw0rdfish: this scheme has the traditional ninja catch, as you team up to do something immoral AND the scheme demands the team to be moral. I mean you want someone to ninja from everyone else, but not you. Your secret agreement that he'll pay you back the 1/3 is not enforcable, you can't stop him walking away. The "I'll do it for you too" does not work, because you are never in the same position. One of you need more items or having higher competition or ready to pay higher. So one of you pays back much more than the other. This guy's interest tell him to defect and run away with the money (this is most likely will happen with some insane trinket when the 1/3 is a huge sum). You simply can't trust another goblin.

Secondly: if you find a trustable partner, your scheme soon become obvious to everyone and can counter it. The second guy also get himself a partner increasing his breaking point by 50% (and not 33). This case the sell price will be (SGBP+100)*1.5, the winner's buddy gets (SGBP+100)*1.5/3, paying it back, so the winner paid SGBP+100. Oh wait! That would be the price if no one would play this trick.

Gevlon said...

@Rohan: again. Nancy's trick can work only, if she is the only ninja. Enter Jason the other ninja, playing exactly the same.

After 2000 Anna will not bid and it's unknown to Nancy and Jason. Let's say Nancy bid 1900, Anna 2000. Jason will bid 2100, not knowing that Anna is out and Nancy is a ninja too. At this point Nancy have 2 options.

1: She waits for Anna to bid again what will not happen, Jason "wins", and Anna gets 1/3 (Nancy gets nothing)

2: She bids, expecting to be outbid by Anna or Jason, she "wins" by 2200, Jason gets 1/3.

So in the 2 ninjas + 1 honest situation, on average the honest wins 1/6, the ninjas lose 2/3. I'd like to stress that there is no way for Nancy or Jason to beat each other, the winner depends on luck (who is #2 when the honest quits).

I'd also want to remind you the 3-ninja situation when one ninja gets 1/3, one nothing, one loses the whole of an extremely inflated sum.

Sean said...

Nice! You actually found the way to defeat the system. Your technique is a very nice method.

Btw, finding a trustable partner is not as difficult as everyone imagines. In a once-off game, the incentive is there to cheat because there are no repercussions. However, in a long-run game (in this situation, there will be other GDKPs along the way), the incentive is there to honour agreements. This has been covered extensively by game-theory.

Gevlon comes up with a counter-argument about people using their own "trusted partner" but I believe the scheme isn't that easy to detect. For example, if 2 casters band together and it will be extremely difficult to detect it.

Gevlon said...

@Azzur: if finding trustable partner is possible, then EVERYONE will find one, making the scheme self-defeated and just annoying.

However I believe it's pretty hard to find one exactly because some high prizes break the agreement. I mean you can either run away with the gold or the ITEM! Don't forget that the scheme demands you to tell him your break point, making it easy for him to beat you (with a partner to get 1/3 back!). "Long term" does not work here, as the item is rare, may never drop again. If DFO HM drops, you can't trust your own mother if she's a fellow spellcaster.

TyphoonAndrew said...

What I find odd is that in the examples we're assuming that the ninja is also a pretty wise and thoughtful person; which does not take into account a small proportion of the rich folks who attend GDKP runs. Those people who will up-bid to be pricks, or do not realise until later that they are losing cash. Yes, yes, its rare - but I've seen it.

I've seen a ninja try to sell shortly after the awarded win at much higher price, and refuse to sell for the reasonable balance value. eg. they were willing to bet the value that after the run that somebody would want it, when they might not have spent as much at the time due to saving for the next drop.

I'd also argue that reaching an agreement within the raid is very common. eg. I declare in advance I have low interest in some drops to a potential competitor, but would like preference on others; then we hash out in advance. The items drop, and the bids are kept low. Technically there is no protection if somebody breaks the deal, but it does offer opportunities where the rest of the raid does not get the "optimal" price for the item.

D said...

About the trusted partner : it's easy to have, for social reasons (i know you like them ^^). Family, boy/girlfriend, longtime game partner... all theses things have more value than yellow pixels (maybe not for goblins). And so, group of peoples is more powerfull than single individual, there's nothing you can do about that.
But i don't see that as a major concern. Because if two peoples have twice more gold, they also have twice more costs. So, they won't win every bid. Their power is just about synergy : the "gave me that little more gold i need right now" thing. No big deal.

About the ninjas : we all are ninjas, in a gdkp. When one really want an item, he have to compete against every other team members, who want maximum gold in the pot (and maybe that 1/3 thing, if he feel risky). So, everybody will inflate prices.
But, as opposite to what i thought before, it may be a good thing, to avoid that "lolz, got shadowmourne for 100G, coz i'm the only one who can equip it".
People will have to guess real price of things, and how much gold everyone have in his hand... looks like poker. gdkp made ninjas part of the game, and that's sound brilliant... on the paper.
Can't wait to have returns of that on the long run.

Jobu said...

I think there are multiple ways to Ninja some money, as the above posters showed.

The main problem with Gevlon responses, IMO, is that each time you tried to answer is if everyone knows everything and are able to do the same as the others. For example, with the "Friend Ninja Scheme" you said, but everyone can have friends.

Well, the issue is that while something is possible for one or two persons, usually it wont be for everyone. Mostly because most of the people are honest and lazy. They won't use this kind of schemes.

Just think on the AH. While everyone could have played it to the best, and could do multiple issues to control the flow and money, most don't. It doesn't interest them.

Same here.

There are multiple ways to ninja money if most people are honest/lazy. Rohan type of scheme was the one the immediately jumped to me too. The scheme is very easy, just raise your bid till you are 2nd, and leave it alone.

Yes, it can fail. Yes, there may be multiple Ninjas, and in that case I may have fallen to my own schemes. But in most cases, I will win. Much like in AH.

I think that the last comments give a much better solution. This GDKP should be handled like Poker. Where everyone understand the issues and ready to fight over the loot much like in poker.

Or, you can say that you use MS/OS rules for decision WHO can have the items for inner big between those guys only.

While this is a bit timid, and the GDKP values may be rather low, it will avoid most of the issues of ninjas. The problem is that tanks bids will be low most of the times, while caster/melee bids may rocket sky high.

Another option, as this is a guild, is to demand that people will use the item (this can be confirmed through the armoury). In that case, the bids will truly reflect the bidders wish. No ninjas in this case.

The last suggestion will lessen the ninjas issues - though the usage of a friend to effectivly lower the price will still work. For that I still don't have a good solution.


Eaten by a Grue said...

Hey they almost killed festergut, no one talks about that?

Gevlon said...

@Jobu: so your point is "honest people in the presence of goblins will suck"? Because it's pretty obvious.

@Eaten by a Grue: since we did not kill him, there is no point to congratulate. Since there can be no doubt we'll kill him, the "cult of Festergut" punks are now busy either crying or trying to find new deity. They don't want to write "Arthas will kill you" because the idea that we get to him is blasphemy against their true God (GS), but they are not sure what else could they write.

Anonymous said...

This is my comments

In summary, a bidder can cancel out the effect of 1/3 bidding rules and take advantage of that. (Pay less for the loot)

Good thought though

Brodster said...

It's not that difficult to be the bid ninja, I do it often.

You are also not factoring in random reasons to bid. Here's an example from one of my earliest GDKP runs I was in:

I am on my Elemental Shaman. My friend is in the raid as well on his Disc Priest. Nobody knows we are friends. Solace of the Fallen drops, he bids up to his breaking point, and is outbid.

Now, I am much deeper in the pockets than he is, and I know he wants that trinket. I already have it, but I am in my elemental set this raid, so nobody knows. Nobody knows we are friends.

I start bidding. People make comments. I say "I want it in case I go resto". Which is plausible, and fair, because in a GDKP run you don't need reasons (they are just there to convince the social you are "legit", whatever that means). I bid the item up to about 15k with another player.

That's as high as I was willing to go with this item, so I drop out. The winner gets it for 15k, the rest of the raid gets a lot more money. I showed everyone my own Solace after that for good measure: everyone but the winner (who paid 13k more than he would have) was happy and laughing.

Point is: you can't know everything all the time. There is no way you would have known that I was not buying it as a "legit" bid.

Gevlon said...

@Brodster: you took a huge risk here. What if the other guy runs out of gold and you end up with another Solace?

Anonymous said...

I know im going a bit off "comments" but i cant help but notice that hunters dps is really something, he must know what he is doing and sequezing every small ounce of dps he can. I consider myself an able player with every class i play and have achived some nice numbers like those dk's with similar gear ( i Dont have the proff tought) but the hunter just blows wow's gs maniacs to pieces. Awsome group you got there. And let me say bye bye to festergut :)

Anonymous said...

Hunters ans DKs are lowest dps? Never seen that myself. Hunters tend to dominate DPS tables (sometimes by up to 15% over non rogues) and while DKs arn't a powerhouse DPS class they can certainly put up good numbers and tend to dominate add intensive fights like LK or Sindragosa (esp as unholy).

If you see hunters doing bad dps you're seeing a very bad player.

Brodster said...


I had no risk, because I was going to give to my friend, and he would eventually pay me back over time. Sorry I did not make that very clear.

Now that I think about it, I suppose it's just like giving a loan. Bah!

I do like your idea, it's just fun to play devil's advocate. ;D

Brodster said...

Many hunters and death knights are just terrible, and there are good reasons why there would be worse players in these classes.

Ranjurm said...

On fesetergut, depending on your view of this, you might want to disallow the use of the stackable fire festival food buffs. While this is available to anyone, they do add up and are temporary. Perhaps use only for farm fights (if you don't use extended lockouts) or not at all to prevent ammunition from falling into the hands of nay sayers.

Anonymous said...

As much as I appreciate your project, I sincerely doubt the healers will have time to throw out some DPS during Festergut. Festergut is an intense healing fight. It's one of those fights where the tank can die in a matter of 2-3 seconds.

That being said I have no doubts about you downing him, I just wouldn't rely on the healers for it.

Unknown said...

"BN offers the item for sale. First guy already knows SGBP, as it was obviously announced by SG himself who stopped bidding. So FG offers SGBP and not a cent more. It's unreasonable for SG to bid more now if he did not bid 20 seconds ago, and BN has no options. He must accept any gold to decrease his losses."
Scenario 3:
BN offers the item for sale. First guy already knows SGBP, as it was obviously announced by SG himself who stopped bidding. So FG offers SGBP and not a cent more. It's unreasonable for SG to bid more now if he did not bid 20 seconds ago, and BN has no options. He must accept any gold to decrease his losses.

SG stops at 10k. BN bids at 10.1k. FG drops out, ad leaves BN to foot the bill. BN pays 10.1k.tells FG he’ll sell the item for 10.1k+50% of the 3333g(.33*10.1k)=11766.5
BNmoney=-10.1k+11766.5=1666.5g profit
FBmoney=+3333g-11766.5=-8433 loss
Raidmoney=10.1k*.67=6767g gain

What you’re basically doing is giving the BN a 30% discount if he can find a way to partner with the “#2 slot”. BN being a goblin, he has the opportunity to sell a 15% discount, which comes directly from the profits going towards the entire raid. You’d be better served by just dropping the 1/3, and letting it play out. Also note the 1666.5g profit is enough room if he gets burned occasionally(no reason to if done right, as you’re essentially getting the item for cheaper than the top bid would’ve for the buyer) on miscalculated loot.
-also note that i used big numbers (10k>> 100gbid), which added alot of granularity. These loops holes might break down if items start stelling for 300g on average, because you can't hit that 30% sweetspot, but the 10k price is pretty standard for items like DFO in pugs, from what i've read.

Unknown said...

"The bid ninja is obviously identifiable. If he bids on an item he can't use or obviously won't (like a bear to a resto item despite he never-ever healed in his life), it's obvious that he is a ninja."

Not in a GDKP. I could see hybrids rolling on DFO/DBW no matter the spec, if they have the gold to throw at it. I normally run feral, and i'd happily drop 15k on a DFO for my offspec. The whole point of GDKP is if you have the gold, you get what you want. gold> everything. Even if not, just rolling on role dependant gear leaves you plenty of room to make gold.

"However I believe it's pretty hard to find one exactly because some high prizes break the agreement."
Maybe my guild is out of the norm, but i could find a guildie in less than 10minutes, and it'd be worth it to trust him (he won't ninja a DFO and get booted from the guild.) I've had guildies on alts roll for gear that is questionable, just for another shot at the item.

Hell, i could probably get any person with no morals and half a brain to do it. "hey, if X drops, i want you to bid after me, just after the last guy drops out. After you do, then i'll bid again. I get that extra 200g back, and we'll split the rest of the 1/3, k?"

Scenario 1:
There are two people bidding against item a (let’s say it’s a DFO). There are 3 people involved, first player, 2nd player, and bid ninja. All 3 duke it out, and the price of the item reaches 10k. 2nd player just made the last bid at 10k. Bid ninja ups it to 10.1k immediately, knowing there is a good chance player 1 will follow. Player 1 bids 10.2k. Ninja bid immediately falls out.(if you’re really sure on the worth of an item, say DFO/DBW, or something else really rare, you can risk going higher, but if not, just stop here.) Bam, ninja bidder just made 33% of 10.2k, or 3366g.

Scenario 2:
Me + a guildie in your gdkp. (first, I may or may not drop my guild tag, if I’m worried about you catching me. I can get a reinvite in 2 hours, no problem. Or use alts… etc).
3+ people(first player, me, guildie), dfo is at 10k gold. As soon as first player drops out, me and guildie stop. Say we upped it to 12k total. I pay 12k, guildie gets back 4k. I get 2k for my initial investment, and we split the extra 2k 50/50. (this will always work as long as the extra bids are < 33% return, as swordfish said.)

The problem with all of your conclusions is "let the first guy use the optimal strategy: don't bid against bid ninja". You CAN'T assume the optimal condition. You always always always assume large averages/worst possible situation That's like playing chess and saying "assume player 1 will use every optimal counter-move", which unless you're Deep Blue won’t happen.

cont'd-post cut off by character limit-

Unknown said...

"Hey they almost killed festergut, no one talks about that?"

pre-kill grats? Normally i wouldn't ever say that, and especially not for a boss that Gevlon's been boasting about downing (for what, 3 weeks now, more?), and somehow never got around to it. I'll save my grats (and it really is meant, even if i'll (try to) poke holes in how hard the challenge is in your festergut post), for when it does die.

Eaten by a Grue said...

Hey guys, I think this is one of those ideas to just try out rather than debate. Whatever flaws or benefits the 1/3 system has will be pretty apparent quickly, so just give it a try.

Anonymous said...

Gevlon's response to the comments here don't make sense. He claims that the offered strategies for beating the 1/3 system will not work against optimal strategies. While this may be true, when was the last time anybody was in a raid full of rational actors carrying out optimal strategies? Gevlon's entire blog seems dedicated to documenting ways in which other players are *not* rational actors! Clearly, strategies that capitalize on the less-than-optimal strategies of others have a lot of opportunities for success.

Wilson said...

I cannot help but notice that Utalesko, who was rocking the dps meters to the tune of 20%-50% higher than everyone else, has epic Iceblade arrows eqipped (according to the armory). I don't see how this is permitted. Not my area of expertise, so I don't know how much of an advantage this is, but I expect it is non-trivial.

Gevlon said...

@Wilson: because 2 crystallyzed shadows are "epic".

@last anonymous: people following bad strategies SHALL suffer

Unknown said...

Small trick for Festergut to squeeze out some more DPS. Have the bear tank first. The debuff he gets increases his damage by a large amount. Then when the 2nd tank taunts off him have him switch to kitty and dps. Just make sure he watches his aggro. A hand of Salvation is not a bad idea either. It wont matter if his aggro is low when it comes time for him to tank again because taunt will take care of it.

Wilson said...

@Gevlon -

So what? Lots of banned non-profession-specific epics are made from non-epic materials. Either you're "all blue" or you aren't.

ardoRic said...

"Hey they almost killed festergut, no one talks about that?"

well then, allow me.

First off, there's a little imprecision: it wasn't my very first time healing in a raid as I was healing on our Blood Prince Council kill. It was, however, a very different sort of healing. On Blood Princes I was one of four healers, so there was a lot more slacking going on. This was my first time healing in a healing intensive fight.

I should say that Misako does just fine with just about as much mp5 as I have on my throughput gear, but I had to go get an mp5 set so I could keep up. Last week we had someone link a post in the US forums where someone said something like "ICC in blues is easy for the best players in the world". Koushirou commented on that during raid time, and I concur: we are hardly the best players in the world, we still have a lot of L2P issues... which we get around to solving. That's what makes our group strong, being motivated to improve enough to pull through.

@ Wilson. your figures of 20-50% more are very exhagerated. He does good dps because he puts in the effort to make it. If you check the video, you may notice he moves in stutters to avoid losing auto-shots when going for spores. That's not trivial, and just that makes him gain a lot more dps than switching saronite razorheads for iceblade arrows.

@ Cyan, we were doing that (the Bear tank going first), but it was probably not working as well as it could since the bear's spec isn't as DPS oriented as it could be, but mostly due to Triev never having played kitty before.

Unknown said...

Why 1/3? Why not a more even number for 100g and 250g increments? 3/10 or 2/5 makes more sense to me. Unless there's some non-arbitrary logic to 33% specifically.

Saithir said...


And that's exactly what we did.

There's this issue that apparently we're all hardcore raiders, but for some reason our hardcore tree tauren can't win the dps chart while being in cat form for the first time.

Breevok said...

"Have you noticed that in PuGs and 5-mans always hunters and death knights pull the lowest DPS?"

Nope cant say I have. Hunter's especially (mine included) take great pleasure at ensuring we are at the top of the charts - of course there will always be Huntards and DiKs but please don't tar us all with the same brush.

The Gnome of Zurich said...

Gevlon -- you should make the arrow exception clear, I don't remember that. I agree it's reasonable as they are plentiful and cheap, but the recipe is only available to those already raiding ICC (even if just trash), so it kinda goes against the grain. The same is true of many enchants using abyss crystals and other enhancements.

Gevlon said...

@Gnome: you are damn right. ICC ammunition banned.

Anonymous said...

@The Gnome of Zurich

You are right and wrong. recipe is costly because it requires 1 primordial saronite. On the other hand it is unwise to buy recipe. Even now when price of primordial saronite dropped a lot and it sells for 500g price of arrows is at about 3-4g/stack (my server) so for 1 primordial saronite you get 125 stacks of arrows. This amount is more than enough for raiding until the end of expansion.

ardoRic said...

Triev could try using this spec for his feral build.

It goes away with some talents he won't need for single target tanking to make room for a lot more dps talents. This spec should be viable unless the other tank is a DK since it doesn't bring attack speed debuff.

ardoRic said...

@ Anonymous about the ICC ammo,

The problem is not it costing Primordial Saronite to buy, since that can be obtained from running random dungeons. The real issue is you need Ashen Veredict reputation to buy the recipe, which you can't get without raiding.

I always noticed the blue glow of our hunters' arrows, but didn't make much of it because I figured they'd be cleared by Gevlon.

ardoRic said...

Damn, I posted the wrong spec on my previous comment.

This is the spec I was suggesting for Triev to use.

It doesn't cover glyphs, but my suggestion is to have Glyph of Growl as the only tanking glyph, and use the other two slots for dps glyphs, unless we see Triev die.

D said...

I just saw 2 others problems with you current version of gdkp.
The first one is so big than i can't believe i'm the only one to see it (and why only now ? Maybe i'm just fooling myself... if so, blame my poor english skill and lack of sleep).

First problem : winer of the bid (FG) don't pay his price, but price of secong guy (SG). If bids are strictly incremental (aka : minimal bid is also maximum bid), there's no problem. It will take forever for greats items, but it can't be abused. If not, the Ninja Goblin (NG) will wait for the first bid, and right after said a Stupid Price (SP). SP just have to be too big to be reached by anyone, like gold cap, for example. So NG win, and pay the first bid, probably something so low than he can obviously sold the item he win right after with a great benefit.
Fix : strictly incremental bids (bad solution, bids will take forever) or make FG pay his price, not SG's. I don't have a deep knowledge of Vickrey's theory (thanks to Zekta links), but as far as i understand, to pay SG's price is only good whih secrets bids (and that can't be used for obvious reasons).

Second problem : your will to fix the fear of competition is good, and the 1/3 rule seem to adress that (with the problem of cooperative abuse, stated by Zekta, but i don't think it really matter, because of gold limit, as i previously explain).
But, in fact, i think it's counter productive. As a priest, will i join that 6 priests pug ? In a rolling pug, the answer is no (too much competition). In a standard gdkp, the answer is yes : with the competition, the pot will be great, so even if i don't have loot, i will have lot of gold. There's 2 downsides. First : if there's no priest loot, there won't be any competition, so no gold. But it's a luck problem, not a gdkp one. Second : it's still better to be the only one of his class... but not that better (as opposite to a rolling pug, where there isn't any good reason to have competition).
So i think the competition problem is very low in gdkp pugs : the system take care of that by himself. If people still fear competition, it's just because they don't understand that.
In my opinion, you don't need the 1/3 rule.

D said...

A last precision : about cooperative abuse, as stated by Zekta, i said it doesn't really matter, but i need to emphasis "because of gold limit".
It still matter a lot about abusing the pot : everybody lost gold.
So it's another good reason to drop 1/3 rule, in my opinion.

Wilson said...


If you look at the log for the fight that Gevlon linked to, Ul's dps of 5834 is 21% higher than Chopsui's 4812, and 62% higher than Baitra's 3589. So rather than exaggerating, my eyeball estimate actually was low. How much of this was do to the epic arrows, and how much to class imbalance or superior skill or group tactics, I have no idea. I just know that ammunition dps is non-trivial, which is what I originally stated.

Strum said...


The winner pays the amount bid. Bidding more than the required increment is suboptimal.