I don't think your ranking while tied with another guild in progression matters that much. I suppose if you're in the top 10, a couple spots up or down make a difference.

I think your suggestion of only counting loot from the latest tiers is a good one.

Also, I think the stars are meant to lock the scores. So a huge guild can't come in and be #1 all the sudden after they kill illidan.

I think it's most likely related to completion as when you mouse over you can still see their "score". The only way another guild can gain a spot is if Aurora or D&T were to miss a week of "new" loot, which would not increase their own points and allow other guilds to catch up. But technically, if everyone clears everything every week, the standing should remain the same for already completed content guilds (baring the amount DEing).

My memory is a bit hazy, but I believe we got the plate dps belt, which was crap compared to what our dps warrior was wearing. The other item was definitly leather, I want to say it was some healing leather, but it may have been the moonkin belt.

Edit: looking at the loot table, I am pretty sure it was the moonkin belt.

Log out and let the world know that I have an unsocketed and unenchanted T5 breastplate (not my fault!) or log out and cost my guild 0.01 points in this newest progression race!

But on a serious note, the crawler update page is neat, but would it be possible to add a little something somewhere on your guild page telling you when your guild's ranking was last updated?

Rylai, that's already in there. Check the left-hand side of the page when you click your guild's detail listing.

The one thing I'd like to see is user-requested updates, as with Antiarc's Armory Tools. If your guild snapshot is less than 24 hours old, then any update requests would have no effect.

Polling all of the top 1000 guilds is a waste of his server resources. Most of them don't raid every single day.

Antiarc's setup gives him the best of both worlds: low server utilization and up-to-date results. I hope that Aspir considers something similar.

I like the idea of tiering in this manner, and I think I'll implement something like this when I get to updating the scoring system.

The stars are to represent two things:
1) Completion of all current instances
2) Locked ranking
We felt that some recognition should be given to the people who completed all the content first. For example, I dont' think it would feel right if some other guild came up and pushed ahead of a guild like Death & Taxes now just because they had more gear. D&T is at a point where they can't have any significant progress only because they've done everything already. So the idea was that as guilds reach completion we lock in their ranks.

When Zul'Amun comes out this will break completion for all guilds, but to me it still seems like it would be wrong if the top guild loses its position because they didn't clear a mid-tier instance first. So we're tossing around the idea of giving "first kill" bonus points to guilds to help with that.

I'm not set with a particular direction here, so I'm open to any suggestions.

It might be surprising but I actually do get loot updates for most of the top 1000 guilds even doing nightly scans.

Regarding manual updates: it's always been my plan to allow these, but I've just been swamped with my "day job" and haven't been able to add it yet.

---

Some recent updates:

- Initial EU crawl is about 30% complete

- I'm in the process of writing a massive sql query to pick out the guilds who appear to have suspicious progression, but in the meanwhile we've been going through and performing manual adjustments on some guilds where the transfers are completely obvious (everyone else in karazhan while one person is decked out in gear up through BT)

- There were quite a number of guilds that hadn't gotten an update in about 2 weeks due to a timestamp getting messed up somewhere. I've reindexed all the affected guilds now.

On an unrelated note (hopefully this isn't inappropriate to ask here), does anyone have a solid web hosting service they'd recommend? I am having some issues with my current one where the wowjutsu webserver will just go down for ~10 or so minutes throughout the day. I'd need PHP5, Mysql5, shell access, ample disk space, lots of bandwidth, decent CPU/memory limits, and most importantly *stability*. I just wanna find out what options are out there.

I have a new issue that I've just encountered. A guild has been updated on its page to show a loot from last night, yet when I go to the server page they have not been updated with their new score/kill. In addition, they have not been updated on the overall rankings. Is this just a matter of the different lists being populated/revised at different times?

Yeah that issue is sorta mentioned in one of the recent notes (WowJutsu: World of Warcraft Guild Rankings: Development Notes):

I've just gone ahead and refreshed the pages though, so you should see the latest changes now.

I'm not a big fan of PvE rankings, but I really admire the work that Aspir has done and the commitment he's put into making them acceptable among any competitive PvE guild. A lot of people take pride in the progression of their guild and a website like WoWJutsu really reinforces that. Bosskillers was previously the only site that has ranked guilds before, but their ranking system is very ambiguous and untransparent. How can Last Resort, the second guild in the world to complete TBC content, be ranked #9 in the world?

Anyway, PvE rankings are ultimately subjective, but I think rankings should be based solely on how many Tier X bosses they've killed and when did they kill their latest boss.

For example:

#1 Guild A: 3/5 Hyjal 4/9 Black Temple
#2 Guild B: 3/5 Hyjal 2/9 Black Temple
#3 Guild C: 1/5 Hyjal 4/9 Black Temple
#4 Guild D: 2/5 Hyjal 2/9 Black Temple

-Guild A has killed seven bosses, Guild B and C have killed five, and Guild D has killed four.
-Although Guild D has killed more Hyjal bosses than Guild C, they are ranked lower because they've killed fewer Black Temple bosses, resulting in a lower total between Hyjal + Black Temple.
-In the case of Guilds B and C, each of the two guilds have killed five bosses from each of the two respective instances. The 'tie-breaker' should be determined by this question: "Which guild has killed their fifth boss first?" The guild who has reached the 5/14 status first should be leading.
-Total loot obtained should not be a determining factor. If Guild C was stuck on a boss for a few weeks while Guild B progressed rapidly to come from behind, naturally Group C would have more gear from the last few bosses than Guild B, but ultimately Group B had surpassed them at one point.

This is just my opinion. As of right now, Death and Taxes and Aurora have killed Illidan. If another guild kills Illidan, and suppose the two aforementioned guilds have stopped raiding or some of their Illidan-loot holding members have left the guild, that other guild should still not surpass them in rankings, simply because the only thing that matters is who got there first.

Anyway, again, I'm still not a big fan of guild ranking sites due to its subjective nature, but a lot of work has gone into WowJutsu to make it as "accurate" as possible for Hyjal/BT guilds and SSC/TK guilds alike. Aspir's work is pretty hawt. Mad props.

This is a really nice job!

We need them for EU Servers :P

Yeah you shouldn't be ranked up or down simply because a guild kills farm bosses more than you - it should lock on each boss. An example of this would be if Nihilum stopped raiding for a couple weeks and another guild farmed past them in total illidans, they're not now suddenly world #1 - I'm sure your system locks on illidan, but it doesn't before that!

Hum, but when do you count something as "farm"? Say a group gets a lucky/buggy kill on one boss, but can't kill it again for a month. Should they be locked in the rankings ahead of some group that kills it 4 times in the same time period, just because they got the first one a day earlier?

Horses for courses - this tool isn't designed to be an accurate record of first kills, see for example the problems you get if all loot is sharded, the fact that updates are not necessarily daily, etc. Rather, it's a record of which groups are regularly conquering certain content.

I don't think there's any clean way round this. A quick and dirty hack solution would be to make the first kill of each boss (and/or the first few pieces of loot from each boss) count double or triple. That gives a temporary "bounce" when conquering a new boss, however if you aren't able to repeat the effort, you'll be overtaken by the groups behind you.

Love the site, it's about time something like this game along. I had been hoping for a while they would include it in guild armory information.

Excellent job Aspir, keep up the good work, you're providing a great service to the community.

Would it be possible to generate some server-wide stats and thus produce a server ranking along the same lines as the guild ranking?

There's a lot of palaver on these forums about how it's easy (or not) to get into Heroic pugs, or Karazhan pugs, or Karazhan groups (or SSC/TK/etc. groups). My hypothesis is that the cross-server transfer effect has led to a mass "concentration" of raiding, so the servers with well-progressed BT groups tend to have lots of raiding guilds, while other servers suffer a continuous brain drain of talent.

To take my own server as an example, there are only two groups that have killed Vashj, and none that have killed Kael'Thas. Fewer than 10 groups have killed anything outside Gruul's Lair (edit: yes, including Magtheridon. The numbers for SSC/TK are 5/8 respectively, GG Void Reaver).

I'd be interested to know if this is atypically slow progression, or (as I suspect) around average. Going from the stats on the front page, there are 1622 guilds in SSC, across 222 servers: an average of a little over 7 per server. Median will undoubtedly be lower than that.

Could you produce a page that simply lists the number of guilds per server that are in Karazhan/SSC/TK/Hyjal/BT ? I think it would be an eye-opener. Another metric could be the total points per server, or the average point score per rated guild, or some such. It's less clear whether those would be a meaningful comparison though.

To put the variance in perspective, on my server (The Underbog US CST PvP), we have one guild that has 2 bosses in SSC and Void Reaver down. No one else has killed Magtheridon. There are a total of about 7 guilds that have killed Gruul.

In the new crawl that's going to look for 'suspicious' progression, how is that going to be handled? For example, on The Underbog, I can tell you that most of the top-ranked guilds are there illegitimately (just based on the timing of when WoWJutsu came online). What formula is being used? What conditions determine 'suspicious'?

Maybe this was answered in a previous post, but I've noticed that BT bosses are worth a bit more than Hyjal bosses.

I'll just use my guild for example. We've killed three bosses in Hyjal and BT. Yet our ranking is lower than every guild that has killed 2 bosses in Hyjal and 4 in BT. And you'll notice guilds with more hyjal bosses killed are below the guilds with more BT bosses killed (but equivalent number of T6 bosses killed)

Any particular reason for this decision? The trash in hyjal takes more effort than the first few bosses in BT. It's not a big deal I just think that same tier bosses should be equivalent, with perhaps the final boss of the zone being worth a tad more.

If you disregard loot from players seen in a guild for a month or less, I would note the following effects:

* Guilds newly slurped up by the crawler would have 0.0 rating for a month.
* Such a guild with, say, Kara on farm when the month clock starts might show no new loot on members after that month (their gear is "saturated"), and indeed never show anything new if they have largely moved on, and so rank lower than they normally would.
* A player who wins loot from a new boss within a month of joining, would cause that guild to rank lower than normally. (This would be rare in a guild such as mine, where such players are on "trial" status and less likely to win.)

Given the above, the suggestion sounds like a good one, provided it is waived for guilds just being picked up by the crawler, and perhaps allowing "new player" loot to be counted IF that loot suddenly appeared in, say, week 3 or 4.

Edit, WoW Forumish rant not appropriate for a quality site such as this.

Of course some clown had to try to steal credit for himself: WoW Forums -> Guild Progression

I too am curious if this is happening. In another thread we've seen some work done (by Songster, as a matter of fact) that seem to give fairly simple baselines for 'average' guilds as far as raid progression, but it'd be interesting to see how the servers themselves stacked against one another.

Like is suggested here, I suspect that servers with a lot of skilled raid guilds will attract further likeminded players. Furthermore, the average raid experience and gear level of the server populace will be higher, and as such, mean that people who leave a guild to join or start a new one might have an easier time getting back to higher end content, since they might find fewer people who'd only seen the encounters they're attempting on Youtube, or whatever other source they use for information.

As part of a server that I suspect is at least slighly behind the curve (1 Hyjal kill for the entire server thus far), I've pondered lately where we actually sit, and if the gear/experience level of the populace in general will continue to hinder progress (since apps/new members are likely to have to be geared and trained from nearly scratch).

I think most would agree that T6 difficulty doesn't begin until Archimonde and Gorefiend (and that's debatable). The only difficulty in the first few Hyjal bosses is whether or not your guild can put up with un-rewarding trash. In other words, Hyjal as a whole is an easier instance than Black Temple, thus it should be worth less.

Would it be feasible to add Trash-drop epics to the guild pages, particularly for Black Temple? Obviously they shouldn't count towards the overall score, I just think Wowjutsu has an interesting side benefit of seeing how "lucky" guilds are with particular loot and where that loot goes.