this guy will prolly never post a graph again to avoid the flames
this guy will prolly never post a graph again to avoid the flames
A lot of players keep saying his data is bad and quoting their own miniscule data sets.
The average WoW player in naxx 25 was terrible, not an elite from an elite guild capable of doing elite DPS.
Notice how his graph says AVERAGE in bolded red text.
The curve through WOTLK is pretty accurate, the conclusion based on the extrapolation is not. Assuming a continuing exponential increase is wrong if you've read about Cataclysm. In all likelihood the curve will go back to a linear version like it was from Vanilla -> BC.
Its entirely not accurate. 2500DPS in naxx was extremely sub-par, it was definitely at least 4k with the exceptional people doing closer to 6k (this is obviously class dependent, since a shadow priest was not going to be doing 6k in those days). I also highly doubt that the AVERAGE dps in ICC is 10k, I see tons of people doing significantly less than 10k in ICC when in a pug.Originally Posted by Shiggity
You are right about his conclusion though, and its mentioned tons in any statistics class... you cant find a trend off of such a limited amount of data, especially when factors that significantly impact the data can change (as blizz stated, gear is better than they had planned to release for this expansion due to hard modes, they will likely fix this in cata which would change the max DPS output).
I read every page and saw so many people posting "LOL WELL I DID" "MY RAID"
I've been in a lot of varying guilds. 2-3k Nax 25 with 1-2 people at 4-6k is what I remember seeing in my raids. AVERAGE people. AVERAGE.
Problems I see for tier gear in Cata are much more nerfs to previous set bonuses. There were only a few during WotLK, but if they plan on lowering dps outputs in Cata then I think we will start using old tier gear again. I don't fully understand how they are going to scale bosses with gear, but yeah.
The first time our guild killed Gruul, there was only Karazhan released (as far as progression went). SSC, Eye, and Mag were there but inaccessible. In order for us to kill Gruul in 17 growths (which was pretty much the enrage for us), we figured that the minimum we could have was 400 dps. The best were around 800 dps at the time (heroic blues and some Kara purples). 2600 is a joke on many levels.
I'm not arguing that it's truly linear; it isn't. But most of that is due to Blizzard not wanting to give people the gear that's optimal for them early on. The numbers I'm using are spreadsheet numbers. If you used averaged throughout you wind up with really odd numbers. What do you consider average? Are you including people who have no clue how to play, or is it an average among top-100, 500, 1000 players? It's too easy to find bias. Spreadsheet data has gotten much better in Wrath than it has been in the past and it's a good thing to look at. Saying there's an exponential curve between Naxx 25, which was basically joke content and ICC 25-H which at this point is reserved for very good players (for now anyway) is silly. You'll be comparing many, many poor players that bring down 'averages' to very good players.Originally Posted by seigfried13
A good example is your comment on the ToGC and Uld10 numbers. You think they're too high, but the Ulduar 10 number in particular is probably too low and should be slightly higher. The ToGC number is hard to figure in because we didn't have Festergut yet. There's a lot of revisionism going on as well. What people can do now in gear is not the same as a few patches ago. DKs were breaking 8k DPS in Naxx in Naxx gear. Ret was coming close to that as well. Why? They got buffed too much. They got nerfed later. The other part is people remembering very low numbers, often because they never had the relevant gear at the time that was progression content. So they remember when Naxx was around they did 2.5k DPS. The part they tend not to remember is that when Naxx was relevant they were in ilevel 200 gear and actually at the Heroic level.
Average players are becoming more scarce the further up you move in LK tiers, though.Originally Posted by Shiggity
In other words, the further up you go, the less bad players you have dragging down the average damage, thereby increasing it even further.
So this only skews his data even more.
You can't use this average over all tiers and expect to get any meaningfull results.
The OP started LK off with extremely average players and then used the absolute high end for ICC HM and included the Buff. That doesn't work.
Since the top end is the only group we can expect to be there for every tier, it's the only one we can compare and get usefull data.
Hmmm, I remember I did 2.1k dps as Arcanemage on Brutallus (ok, with SP+resto shaman in grp and usually 1 innervate). I know there were mages who did better (fire mages and arcane mages who got more innervates, I saw some posts ). And I was not in a "world class" raiding guild.Originally Posted by ElAmigo
I think to remember the fight was like that the average DPS would need 1.8k dps to be able to beat the enrage.
According to internet the top Mage DPS record at Brutallus was 4k dps (100% Arcane Blast cast except when Bloodlust was up, 3 innervates, 2 manatides, 3 bloodlusts). The dps of some people largely depended on the amount of "trickery" done (especially with mages). But definitely DPS could be much higher than 1.7k.
So an average of 1.7k for DPS on Sunwell sounds low to me. I still remember when seeing people in WotLK in Naxx PUG raids I sometimes used to say "I did more dps at lvl 70 than this guy at lvl 80" ^^
MagicSN
Somebody got a Dark Intent and a Focus Magic for Tarecgosa and me? ;-)
I can't speak for his figures before Naxx 25, but in Naxx 25, I was doing like 2-3k. By Ulduar, I was still doing 2-3k and topping the charts. by ToC, I was around 4k and still topping the charts. Borean Tundra is not a very good server. Now I've improved some, transferred servers, and my class got a slight buff. I'm up around 9-10k most non-gimmick fights, and not topping the charts anymore, though I am generally in the top 5.
If you took the Borean Tundra server DPS figures (my previous server) as "average" then his average for most raids would be too high. If you took Mal'Ganis DPS figures (my current server) as average then his average for most raids would be too low.
The problem people are having is, the chart could be averages, but the actual data pool is huge. So, what ends up happening is, any section of the data pool you draw from won't be a representation of the whole. Compound that with no sources on where the data is drawn from, and you have a graph that looks like magically generated numbers. You can rebut any claims of falsification with "it's the average."
There's a reason the previous thread got locked, and it's not because people aren't understanding averages - it's because the graph:
- Was created without showing the data sources.
- Is not clear.
- Ignores the fact that Blizzard won't be following the current trend anyway.
A proper graph would show the DPS requirements to beat the enrage timer of each boss. Why? Because that leaves the human element out, and shows what DPS figures Blizzard tuned the encounter for. Not how much DPS was actually done (too large a data pool to aggregate accurately anyway). But then, what is the point of such a graph when Blizzard won't be following their current boss design pattern?
We could use the theoretical maximum as well, as it is similarly unrelated to player skill. And would probably be better suited, since this is about the potential damage per time that can be done, not about the amount needed, which is usually lower in lower tier raids, since they are supposed to be easier. Naxx was undertuned, but Ulduar was still easier than ICC.
The point of this post is that DPS is increasing like a rocket. Just reply to that.
Not that the dps-sheet is not correct, that doesnt matter and isnt the point.
It's not the point, but it does matter quite a lot.
The sheet is the OP's "proof" that DPS is increasing exponentially, so the entire discussion is based on it.
My Recount is gonna take 150 MB per raid night. Yikes.
Bigger numbers shouldn't increase memory usage by all that much.
I don't know how lua handles it's variables, but a 32bit integer can handle numbers up to at least 2.2 billion.
If you double the allocated space for the integer, you can go up to 9 quintillions. 9*10^18.
Maximum size of an integer grows exponentially.
Bloodlust never been just for one group. The guild i'm talking about chainused drums bls etc. And believe me he did pull thoose numbers. I think his character is still untouched from the 70 days. And no he did not have the legendary bow. He did however have the gruul trinket which gave him a huge advantage.Originally Posted by Temples
Bloodlust/Heroism was group limited till LK.
Also, bad necro, bad.