Greedy Goblin

Wednesday, March 16, 2011

The myth of the elitist jerks

The chart below distributes the lvl 85 members of the 4+ bosskills alliance-side raiding guilds on our server (EU-Agamaggan) into three groups, those who killed at least 1 boss (raiders), those who only killed Argoloth and those who killed nothing:

As you can see, besides the server first guild, in every guild the raiders are a minority. It's surprising on it's own, why do someone avoid raiding if he is in a raiding guild? I guess the socials are ready with an answer: the guild has a raiding core and the rest are friends and family or newbies who don't get a chance to raid, as the elitist jerk raiders only go with other highly geared players to ensure fast runs and minimize loot competition.

To answer this, I displayed the raider percentage of the guild versus the progression of the guild:
It says that with very strong correlation, the more progressed the guild, the more people get the chance to raid. So these "elitist jerks" are much more likely to give a spot to a newbie than those "freindly heplfull ppl" who still need more gear to kill Valiona (because ilvl 350 is definitely not enough). Important note: only the first 2 guilds are around 50 characters, the rest have 100+, so the argument "they don't even invite you in the guild if you are not pro" does not hold.

Also, the elitist jerks gave the biggest free source of knowledge to the community. The participants of that forum help "selflessly" more than all the "freindly heplfull ppl" combined. And what do they get in return? Handhold-demanding questions from M&S who can't/don't want to read the damn first post, or run the sim for himself.

The elite is not at all elitist. They don't want to exclude new or struggling people, they want and do help them. They only don't help those who can't be helped: the morons and slackers. The M&S claim to only want "chance" but they have all the chance in the World. They actually want freebies and boosting.

About alts: since there can be alts in every kind of guild, they don't distort the results, unless the alts are handled differently in the differently progressed guilds. Without further info, it's reasonable to believe that they just inflate the numbers but don't effect percentages.

PS2: our guild The Pug is the red dot on the chart. As you can see we are the furthest above the trendline, so we give the most chance to newbies compared to our progression. One more reason to join!

40 comments:

Unknown said...

Once again you are forgetting about alts. As an example in my guild we have between 15-19 raiders. We have 1 guy leveling and 1 guy on the edge of kicking off for inactivity (haven't log in for more than month). But we have already over 40 lvl 85 characters in the guild. Every member gets a spot if he wants to join the raid. So answer can be alts. In my guild all but 3 members have already at least 2 lvl 85 characters.

Sum said...

"As you can see, besides the server first guild, in every guild the raiders are a minority. It's surprising on it's own, why do someone avoid raiding if he is in a raiding guild?"

Alts surely? I have my raiding main and alt + 3 nonraiding alts in my guild. Most others have something similar, although of course not everyone has alts at all.

Or did you somehow account for alts in your graphs?

two said...

I understand and agree with your point. Yet I have the feeling 0.71 is still a bit low for this chart. Doing the same chart in a high pop realm might be interesting as well.

Frank Fujita said...

I'm not on your server, but in my guild, I'm a part of a 10/12 normal raiding group, and there are lots of people who aren't even interested in raiding. But even if you only counted the alts of the people who raid (who have never killed a boss) the "raiders" would be a minority.

Anonymous said...

It seems your data set is missing some important information: number of accounts and alts per account in the guild. without that information your statistics don't show anything about players, only characters, and is therefore pretty worthless for purposes of your argument.

Anonymous said...

How were you able to take Alts into account? They could account for a high percentage of the actual member numbers.

nightgerbil said...

So these "elitist jerks" are much more likely to give a spot to a newbie than those "freindly heplfull ppl" rubbish. total codswallop. You dont get into those guilds if yr a newbie. They have application procedures that filter out the rif raf.

What your seeing is that successful raid guilds dont have many hangers on, just a few alts and trials. Trials get sacked or demoted to "social" status if they were fun, but ya know if ya got to the stage where you GOT a trial in a full on raid guild, your a raider! you dont want social status, you want and need a raid spot and without it you will go back down the guilds until you get one. So the ones not quite up to making the team leave, the ones not upto making the team in the first place dont get a guild invite.

Your begging the question Gevlon. Your "facts" dont support your conclusion

Anonymous said...

Interesting work. What was the method in obtaining the data, and how did you distinguish between Friends/Fam (F&F), alts, and raider ranks?

I would imagine the correlation is due to good guilds progressing faster (e.g. Paragon clearing 11/12 HM Icc in <2 hrs in wotlk), freeing scheduled raid time to gear alts/F&F, making the guild have a higher % of progressed toons. Guilds mid-progress might have 2-3 initiates per 25m, carrying them one night, then dumping them back to their old guild if they can't compete with raiders, which could skew results.

IMO, with higher guilds, people will do alt runs with alts from other high end guilds, or mains from middle-ranged guilds (i.e. alt A brings in main B for BoT while main B's guild is on BWD). This could also skew results.

I would also imagine that F&F status in higher guilds denotes former raiders who have gone casual, but are still capable of doing at least 4 bosskills in their occasional raid night. We have plenty of capable F&F who come on our off-nights raids.

Gevlon said...

Two parts added to the post to prevent more of the rather strange comments above.

PS to mathematically challenged: mentioning "there are alts" does not help unless you can meaningfully prove that the number of alts change with the progression. In absence of such proof we must assume that every guild have equal amount of alts, so the charts (which are in %) are not affected.

Important note: only the first 2 guilds are around 50 characters, the rest have 100+, so the argument "they don't even invite you in the guild if you are not pro" does not hold.

Squishalot said...

"As you can see we are the furthest above the trendline, so we give the most chance to newbies compared to our progression."

'Highest chance relative to the number of bosses killed'. Prescorus, far and above, gives the most chance to newbies. Trendline means nothing in this context, absolute means everything.

Jumina said...

Don't forgot guild leveling and guild achievements. On my server many guilds have both PvE and PvP sections for this reason. And PvP members can sometimes join alt runs as well as our guild friends.

Anonymous said...

Your dismissal of the alt complaint is nonsense. you are assuming a result which you have no evidence for simply because it suits your argument. most likely, the high progressed guilds have more alts which are *also raiders* which potentially complete contradicts your argument: 'noobs' aren't getting more chances because the well geared alts of raiders steal those chances (a phenomena which I have seen in an actual server first guild).

Sum said...

"PS to mathematically challenged:"

No reason to call us mathematically challenged. It is my opinion that not knowing the relative amounts of inactives/alts in each guild is such a large source of error that your data can not be used to draw any statistically meaningful conclusions.

"unless you can meaningfully prove that the number of alts change with the progression. In absence of such proof we must assume that every guild have equal amount of alts, so the charts (which are in %) are not affected."

I don't think the burden of proving this lies on us; I would say that since you are the one asserting a claim that based on this data things works so-and-so it falls on you to prove that you can in fact make your assumption without it significantly distorting your statistics. I would be almost certain that large casual raid guilds have more alts and inactives than dedicated raid guilds, but of course I can't prove that. However, that theory is at least equally valid to your "no difference" one in the absence of proof. As a result, I'd say no conclusions can be made based on this data with such a large unknown variable in play.

As someone else said, "good" raid guilds don't accept people without something to show for themselves - my guild doesn't recruit anyone for raiding unless they have previous raiding experience, which they can work for in some other guild or by pugging. On the other hand, everyone accepted for raiding also gets to raid - we don't take people and then not give them a chance.

I'm not saying you might not be right, but the way you are using data with a lot of unknowns in it to prove a point that you obviously set out to prove seems dubious. Calling those who disagree "mathematically challenged" while using at least somewhat flawed logic is pretty lame.

Anonymous said...

How can you compare "normal" guilds which usually have lots of alts with The PUG where no alts are allowed?

Camiel said...

The Pug still has more "nothing" than raiders. I thought the whole purpose of the guild was to give everybody a chance to raid, so why have so many members never killed a boss?

Azuriel said...

Incidentally, I don't think anyone, not even the M&S, would make the argument that the Elitist Jerks guys are elitist jerks. The EJ forums answers a specific kind of question: what is the best spec/talent/gems/enchant for highest DPS/healing/survivability?

It is the lower-case elitist jerk(s) that tells you NOT using EJ-stamped build means you fail.

Gevlon said...

@Last anonymous: that's part of the point. Alts of a raider are usually also raiding, at least baradin hold, so if everyone has 2 alts, the nothing % is not changed.

@Zazkadin: I don't know. I can surely say that no one was ever turned down permanently. It's rather "they don't even try to raid".

Anonymous said...

Gevlon claims:

"It [the data] says that with very strong correlation, the more progressed the guild, the more people get the chance to raid. So these 'elitist jerks' are much more likely to give a spot to a newbie than those 'freindly heplfull ppl' who still need more gear to kill Valiona"

No. The data tells you that (proportionally) more characters have the opportunity to raid, not that more people do. It is quite possible -- and, based on my anecdotal experience, likely -- that the additional raiders found in more progressed guilds are alts. Consider also that more progressed guilds are better able to carry an alt through a raid than a guild that is still struggling on Maloriak. The argument here isn't that "there are alts" but rather that "there are alts... and they raid".

Gevlon said...

@Last anonymous: If the more progressed guilds are more able to carry an alt, they are also more able to carry a newbie.

Maldwiz said...

PS to mathematically challenged: mentioning "there are alts" does not help unless you can meaningfully prove that the number of alts change with the progression. In absence of such proof we must assume that every guild have equal amount of alts, so the charts (which are in %) are not affected.

These people should take a look at this first.

Riptor said...

"It says that with very strong correlation, the more progressed the guild, the more people get the chance to raid"
I disagree. In "elitist" Guilds the Raiders usually stick together. They have, aside from the Main Raid, usually their own 10 Man Groups or in some Cases even an Alt-25 Man run. But due to the Fact of them being elitists, they despise to play with People they do not know or have to assume that they do not hold up to their Standards.
These Players are used to raid in a very professional Environment where they expect, and also demand, the other 24 Members (or 9 nowadays) to play at a certain level. Also these Guilds do this anywhere between 4 and 7 Nights a week so they are used to a level of mutual understanding others can hardly imagine. This is why they in many Cases try to only raid their alts with Guildmates or People from other Guilds of which they know that their Level of Progression is roughly the same.
It is true that if you are a newbie and you already know someone in a hc Guild you will learn a lot of things about WoW in a very short Period of time, but I, from my own experience, know that if a high ranked Guild has 60% Raiders in your Char it is roughly the same 25 Players behind all Toons.

Leeho said...

Raid don't need to carry an alt, just compensate a little to his gear, and even that isn't usually needed. You are the one who tries to prove that gear means nothing, so why alt would be significantly worse than main? Player already knows fights, he usually is not way worse on playing alt than main. He lacks some certain experience, but again if he's brought in he's more likely needed than carried. Sometimes it's easier for raid to take a slightly undergeared holy pala over a geared resto shammy because of certain abilities and fight mechanics. Sometimes you even need to bring in 10 green-geared kitties to kill a big dragon with strange abilities. So usually it's a player who geared and prepared an alt is helping to the guild progress, and not that guild is carrying him helping to get some achievements.
Newbie is the other thing, though. He isn't used to mechanics, he learns and makes mistakes, slowing down the whole raid.
And there's no way a newbie can join a good raiding guild and get into the raids. Raiding experience is required. You can join as friend, be carried with alt runs and such, but noone will take you to progress runs to learn how to raid. It's pointless and a total waste of other raiders time.
I'd say that there's a different exlanation to your math. People that successfully raid on mains usually raid a bit on alts too. If you can kill 6 bosses on hc on main, you can go and clear normal difficulty on less-geared alt in your spare time. On the other hand, if 5 bosses on normal is the max you can kill on main, you can do only BH on alt. So alts in those guilds are more progressed. Next, most of social spots are taken by people who raided but then stopped to, more often due to unability to raid at raid times or burnout. They usually pug something or join alt runs, and as they are still good at game their progress is high even with pugs.

nightgerbil said...

Gevlon they dont carry the "newb" cos they dont let them in! an alt on the other hand only needs gearing to fill a gap in the guild line up, you dont have to worry about skill, experience or it being a loot whore. Its a guild asset. Like wise my holadin is an asset to my guild as is my enhance shaman. A guild asset. It wasnt my holadins achievments that got my entry to that guild though. You might say prove they are alts, that begs the tomatooes/tomatoas reply of prove they arent alts.

different note; why are so many "pugs" not raiding? why cant I get a group going for anything in a morning run? why do I always see people spamming lfm hc for 15mins+ until the group finder fills their need? Is it the time I play? or is the PUG, in fact full of alts itself and most people leveled and then pop in now and then to see how its going?

Anonymous said...

Conversely, I don't see how you can assume the % of alts is constant. ( We really don't know. )

1) My believe is that the % of alts is obviously smaller at the very upper end. If you are in an world first level guild you may be essentially a paid employee. Regardless of whether you might like to try out the goblin starting zone, you have a job to do.

2) A couple of years ago I was told that the reason why the #1 guild on my realm had an "alt guild" was that one of the web sites that ranked guilds took into account the % in the guild who had killed X. So if you were insecure enough to care about your ranking, you would help your ranking by making sure the alts were not in the raiding guild.

3) Whether because of #2 or just tradition, I think all of the top 5 guilds on my server have an alt guild. Whereas I do not know of a single social, casual or semi-serious guild on this realm that has an alt guild or has a main-only policy.

It may not invalidate your premise, but I do know that on my realm the % of alts is not constant; serious 25 man raiding guilds have alt guilds; the rest do not.

mystic said...

I have to agree with the majority of the commentators above, the number of alts most definately does have a bearing on the figures, as do the recruiment policies of the guilds in questions - do they allow alts? do they allow casuals or social members?

I'd also argue that the R squared value you give doesn't indicate a strong correlation at all, R squared is a very subjective measure depending on what you're modelling - the smaple size you've used is simply too small to be statistically significant, or even enough to draw the conclusions you have done with any degree of certainty. Extending this over several 10s of servers or the whole game would reduce the need to bother about alts, there are just over 5M characters on EU and US servers according to wow census (over lvl 20), I'm struggling to find the total number of subs for this area, but with that figure you could calculate the average alts per account, you'd need figures on guild membership too (i.e. not all alts are in one guild, or even the same server).

All that said, I'm not entirely sure what point you're trying to prove, and whether you have a point of view and are using favourable stats to back it up (which is often how stats are [mis]used) or whether you've looked at the stats and drawn a conclusion from that.

chewy said...

Mystic makes a very interesting observation, what are you trying to prove ? I read the post again and it all boils down to the M&S wanting a chance and you've tried to prove that they have every chance.

I sometimes wonder what the M&S did to you that makes you so obsessed with them. They irritate me in the odd pug and occasionally when I meet them on the highway, but generally I just don't care about them, they're irrelevant.

Michael Young said...

It's a very peculiar chart, that simply does not match my experiences and understanding. I certainly could be wrong, but I think it more likely you're not quite interpreting the data correctly.

Alts don't add a flat percentage, and more progressed guilds certainly do handle alts differently. A guild that is clearing content every week will be far more likely to organize alt runs or for individual raiders to seek trade pugs on their alts, inflating the raider percentage.

Just anecdotally, my guild has 40ish accounts, perhaps 160 players. I know there's one non-raider and I bet there are probably 2-3 more that I don't know about because they don't come online often. I personally have one character in the guild who raids, one who does the pvp boss, and 3 alts who don't raid at all. But we've cleared all the normals and are working through the hard modes. We raid three times a week, but I can easily see us clearing in two nights once we've finished current content. That'll leave room to get a second run going with our alts.

I think the situation in my guild is closer to the norm than the situation in The Pug.

Anonymous said...

"PS to mathematically challenged: mentioning "there are alts" does not help unless you can meaningfully prove that the number of alts change with the progression."


The PUG does not allow alts. On your list there are guilds who are more/less progressed that allow alts; therefore the percent of alts changes with progression in an unpredictable way.

One thing to be very careful with is the data point from the server first guild. On the linear fit, this point is highly influential simply because it is far from the the rest. In the absence of this data point things do not look as clear.

What I find very interesting is that for The PUG only 40% of members raid when given every opportunity to do so. Why do you think such a large portion of the wow population makes a choice to not raid?

Anonymous said...

I've found anecdotally that the M&S refuse to play tanks or healers. Good raiders generally have either dual specs or alts and can fill either dps or tank/heals.

I guarantee that if you could break down the specs you would find a HUGE proportion of the 'non-raiders' have no tank or heal spec.

Anonymous said...

Gevlon,

This is anecdote, not data, but might help pondering the alt issue. At least in my guild, the raid members have a higher number of fully leveled alts than the social friends and family do. Our "friends and family" members are far more likely to be level 47 or 62 than max level, our raiding members might have seven level 85s, of which only one or two see the inside of raids.

At least in my raiding guild, the dedicated raiding team mostly brought up several possible characters, and decided which one to use "for real" after seeing what the final mix would be. Also, we tend to lean on each other for flasks, crafting and enchants instead of getting that from the "social" members.

Perhaps cutting it down a bit by looking only at level 85 characters who have their professions maxed would help clarify the data?

At any rate, your post is an interesting one.

Sthenno said...

Once again it's the "my baseless assertion is true unless you can prove it wrong argument." The default position for whether alts have a significant effect on these numbers should be, "we don't know." It's a good enough question that it deserves to be discussed. I think that a lot of people here have correctly argued that bringing alts to raids is nothing like bringing new people.

But this is not just speculation, I have zero interest in bringing new raiders to my raids but I would raid with alts of my raid team any time. You are the one that beat Ulduar in blues to show that good play was better than good gear.

Sure, we could carry a new player if we wanted through most content, but why would we want to?

I do think that I and other people in my guild are helpful and open for the most part. If people want advice many of us would be happy to provide it. What we would not do is give people an invite to our guild or a spot in our raid. There is simply no way.

Anonymous said...

Look, there are 214 members in my guild. We raid 25 mans and a roster of about 35 people. There are a maximum of 15 people that do not raid with us, because of a variety of reasons -- ultimately, they once did, but they no longer do.

214 - (35 + 15) = 164. Do you honestly think we have 164 people floating around in our guild that we just invited from trade? Those are ALL alts. That's an average of 3.28 alts per main.

Absolutely no one gets in our guild without going through a lengthy application process, and the bads are gkicked when they fail their trial. I just don't know where you're getting this idea that progression guilds invite random people for fun.

Oh, and btw, you're usually pretty good with statistics. So I don't know why you think "R^2=0.71" is a strong correlation, because it isn't.

zenga said...

If it can be of any info: we are a guild currently progressing through 25 man hard modes and yet we have to see the first alt raid. And I see the same among other guilds on my server. People tend to prefer spending that 1 alt night on trying to progress further with the mains. Alt's are here mostly used as farmers and pvp chars.

Anonymous said...

Alts in more progressed guilds can handle themselves and don't need to be carried unless the guild really needs that particular alt quickly geared for some reason (OP in some encounter).

In our guild we have official runs in which only main PGs are invited. Outside of the official raids there are "alt runs" made completely of alts. These runs cannot be boosted by mains because mains are either already saved in the official raid or need to stay available. Still the normal encounters are usually steamrolled.

In fact nowadays most fights are very well known by the more progressed players and nerfs already started to pour into live, so composition becomes less demanding and gear is getting less and less relevant.

Boxington said...

"In absence of such proof we must assume that every guild have equal amount of alts."

In a related study of algebra proficiency among students, the researchers did not have data on student's year in school, and so had to assume everyone was in 6th grade. They argued this didn't affect their conclusion that older children had more innate algebra ability.

Seriously though, if an omitted variable is highly correlated with the dependent variable, there is (severe) bias in the results. Moreover, in any serious statistical paper, the burden of proof lies on the researcher to show that omitted variable bias is properly rectified. Sans this, the conclusion is simply not supported.

csdx said...

Actually even aside from alts, I think you're entirely off base. Put it this way, who has a stronger entry requirement? A Hardmode raiding guild, or the guy spamming trade chat for his 'friendly, helpful' guild. This right there invalidates the idea that 'noobs' can get spots easier. Sure, once you're in a high end guild it might be easier to get a spot, but how easy is it to get that invite in the first place?

Also if we assume that every other guild allows alts (who raid in lower proprotion to mains) but the PUG doesn't. With that adjustment, it might fall below the average trendline, and becomes one of guilds with the least chance for new people.

Leeho said...

By the way, at this moment you can see character guild rank in game, even if he's from other guild. I can't find this info on armory, there still are just numbers, but for your server you could check some characters in Stormwind and so identify "rank n" on armory. That way you could calculate the real percentage of alts, if you wanted to.

Anonymous said...

A R2 coeff of 0.70 is not a stong correlation.

Anonymous said...

Gevlon says:

"@Last anonymous: If the more progressed guilds are more able to carry an alt, they are also more able to carry a newbie."

"Able" is not the same as "does". Again, the data tells you that (proportionally) more characters have the opportunity to raid in top guilds, not that more people actually do.

Also, looking back, it's unclear what you mean by "newbie". Is this somebody totally new to WoW? Somebody new to their class, new to the server, new to the guild, or what? I doubt many server-best guilds include many raiders that have played for less than a year.

Anonymous said...

"It says that with very strong correlation, the more progressed the guild, the more people get the chance to raid."

No.

When viewed as percentages, it says only that the more progressed a guild is, the bigger the percentage of active raiders. That's it. Cut. Period.

As the the explanation for this trend, I think it is a much more plausible to assume that hardcore raiding guilds prefer to operate with a tight roster and not to have dead weight hanging around. If you are not on the regular roster (or not meeting your attendance demand), there is no reason to have you around. I for one fell for this banhammer back in the days of Ulduar.

The exception of course being honoury long-time members that does not raid anymore (ie. have become socials) and the girlfriend of the Guildmaster.

It would be much more illuminating to compare the total number of raiders compared to the total number of members in the guild compared to progression (yes, two variables)