Class statistics: how they rank

[ohioastro]

Well, I took the online class stats (very interesting DG) and had a look at them. I divided up class survival into a few benchmarks involving what percentage of characters made it to level 10 (typically the first 2 dungeons); level 20 (add the next 4); level 30 (far east); and level 50 (end battle.) The data is...interesting. Temporal warden is far worse than any other class; it is followed closely by reaver, rogue, and alchemist. Some classes (like archmage) start out strong and fizzle out. Others (like cursed) are actually very challenging at the start and improve at the higher levels. It looks to me as if the pet and rogue classes need love, and the mages need better survival. Cursed look as if they have a rough start, but once they reach level 30 they are golden. Slinger is unusually unpopular for an unlocked class, and corruptor is very unpopular given when in the game it gets unlocked.

class Total %10 %20 %30 %50
archmage 1119 33.3 9.7 1.8 0.2
cursed 1011 15.5 6.2 3.4 1.6
berserker 971 33.7 11.4 3.9 0.8
fighter 965 22.0 8.6 2.2 0.3
alchemist 810 15.9 4.1 0.7 0.0
rogue 705 11.6 3.1 0.7 0.0
summoner 642 19.8 5.3 1.4 0.2
arcane bl 578 16.8 5.2 1.4 0.3
archer 484 25.8 9.9 3.3 0.4
wyrmic 379 19.0 7.7 2.1 0.0
shadowbl 368 16.6 6.3 1.9 0.3
anorithil 254 26.4 11.4 5.9 1.6
reaver 249 19.7 3.6 0.4 0.0
slinger 167 15.6 6.0 3.0 0.6
temp ward 159 4.4 1.3 0.0 0.0
mindslayer 130 21.5 10.8 3.8 2.3
sun paladin 129 21.7 10.1 4.7 0.8
doomed 115 23.5 7.8 2.6 0.9
corruptor 33 33.3 12.1 3.0 3.0
median 19.8 7.7 2.2 0.3

[edge2054]

Is this just dead characters? Because there's a lot of Wardens that just didn't get played. People load up characters to look at the trees.

Anyway Wardens are getting more buffs and possibly a more melee focused tree.

[ohioastro]

It's all characters. I don't think the wardens have a higher fraction of level 1 toons than the norm; for essentially all of the classes the number of characters gets cut in half from level 1 to level 3-5. This suggests either many early deaths or many abandoned toons.

[edge2054]

I understand your point but it's off.

The vault has only been up for a few betas, new classes are naturally going to have more people creating level 1 characters to check them out then old classes simply because we've only been collecting statistics recently.

Anyway, Wardens do need a buff. They don't have as much appeal as other classes and that very well could be that they do to little damage at low levels or don't have enough defensive options. So they'll be getting a more melee oriented combat tree to broaden their appeal along with several other buffs, both defensive and offensive.

Having played a Warden up to level 32 or so the current incarnation does a ton of damage at higher levels and was very hard to get going. The worst offender in the damage department is getting a duration nerf and moved to a class tree while damage across the board is getting increased so the high damage will still be there but it won't be dependent on a single talent and should be available at earlier levels. Survivability will still be on the low side because they're mostly a ranged class (and even if you play one as melee you still have the option of ranged attacks through spells or weapon swapping) and have good mobility. They will be getting another mobility talent in the melee tree along with a displacement effect and a single attack 'bone shield' type spell. I also upped the duration on damage smearing but I don't think anyone other then myself and sandworms are using it.

[Canderel]

Does it exclude Arena? I got a *LOT* of lvl1 deaths recently in the arena...

[ohioastro]

Virtually all classes have the same starting profile: a lot of level 1 characters, and the totals get cut in half between levels 3 and 5. I think this includes all characters, live or dead. Nonetheless, it is useful to know for balance purposes that by the time you look at level 10 toons there are clear winners and losers - and this pattern amplifies as you go up in level. If there are major class changes, of course, you'd need to throw away the old data and only look at how "new" characters fare. You also have to factor other things in - for instance, if people have unlocked the Sun Paladin they've managed to get at least one character pretty far in the game, and as a result Sun Paladins will do better than classes used mostly by new players.

[Final Master]

It would be possibly much more useful if the little chart in the first post was way more readable.

[Grey]

Easily arranged - here it is in code tags.

Code: Select all

[b]Class        Total   %10    %20   %30   %50[/b]
archmage     1119    33.3    9.7   1.8   0.2
cursed       1011    15.5    6.2   3.4   1.6
berserker     971    33.7   11.4   3.9   0.8
fighter       965    22.0    8.6   2.2   0.3
alchemist     810    15.9    4.1   0.7   0.0
rogu          705    11.6    3.1   0.7   0.0
summoner      642    19.8    5.3   1.4   0.2
arcane bl     578    16.8    5.2   1.4   0.3
archer        484    25.8    9.9   3.3   0.4
wyrmic        379    19.0    7.7   2.1   0.0
shadowbl      368    16.6    6.3   1.9   0.3
anorithil     254    26.4   11.4   5.9   1.6
reaver        249    19.7    3.6   0.4   0.0
slinger       167    15.6    6.0   3.0   0.6
temp ward     159     4.4    1.3   0.0   0.0
mindslayer    130    21.5   10.8   3.8   2.3
sun paladin   129    21.7   10.1   4.7   0.8
doomed        115    23.5    7.8   2.6   0.9
corruptor      33    33.3   12.1   3.0   3.0

median               19.8    7.7   2.2   0.3

Edit: Grr, readable now.

[eronarn]

There aren't enough games played yet to make meaningful high-end comparisons, so I'd suggest ignoring the 50% column.

The percent that make it to L10 is interesting, though. It's actually much closer than I would've expected. However, I suspect that some of this might be due to difficulty levels. This will likely have a moderating effect: harder classes will get played less often on hard difficulties, and will therefore get to L10 more often than they would otherwise because of the abundance of lives.

I'd be up for running a more detailed analysis if someone gave me the full dataset (I don't feel like scraping it). I think the most interesting variables would be version, class, difficulty, level, number of deaths, and current status (alive/dead). Obviously you'd want to exclude non-Maj'Eyal games, though!

[gruevy]

From what I've heard, at least one class has a revamp already written - Temporal Warden. I've also heard of revamps or modifications to reavers, cursed, anorithil, berserkers, and possibly corrupters. Oh ya, and some minor things with Alchemists, I think. There may be more. It'll be interesting to see how different classes start doing in the next patch.

[madmonk]

Fighters have also been tweaked because of Class of the Week feedback.

{Correction]I am obviously having brain fade. Forgive me is is not a Fighter but an Archer!