Loading...
  OR  Zero-K Name:    Password:   

ELO Distribution

17 posts, 1306 views
Post comment
Filter:    Player:  
sort


Only including players that played at least 10 games with one of them being in 2016 (To not include all of the 391'462 ZK accounts). Average is 1545 ELO (Would be 1500 without filtering). FFA and increased sensitivity is not accounted for, 1700 in the graph equals to about 1800 in actual ZK ELO.

I did the same analysis about a year ago and the "gap" from 1600 to 1800 has remained but is much less pronounced than it used to be. (Just from the gaussian distribution one would assume more players slightly right of the center)
+2 / -0
7 years ago
Thank you.
+0 / -0
7 years ago
pls do kernel density instead of histogram XOXO
+0 / -0
Skasi
Does this show real elo or does it include fake modifiers?

quote:
the "gap" from 1600 to 1800 has remained but is much less pronounced than it used to be

I like saying "1600+ elo is when players really grasp all the basic concepts". Could it be that at one point in their learning experience people enjoy improving at something and thus (...) start soaking up a lot of knowledge/skill?
+0 / -0
No modifiers to neither value nor sensitivity.

I've had the theory that once you understand the basic unit interactions and the way economy works, you can easily make your way to the 1800 league, without knowing exactly what you're doing there's no way to get even past the 1600 mark. (ZK is very punishing for casual play/ignoring the eco numbers)
+0 / -0
Skasi
Could you expand the graph to reach from lowest-elo-player to highest-elo-player? I'm kinda interested in the very-low-elo numbers and even more so in comparing them against very-high-elo.

Or have you already do that and all the 2100+ elo players not visible on the ladder just really didn't play a single game this year?
+0 / -0
I have felt for a long time that pretty much all elos under 1600 are typically a meaningless predictor of a player's skill.

As I recall from when I started playing, getting past around 1800 was probably the largest difficulty spike.

[Spoiler]
+1 / -0
My elo is constantly declining. Highest on team elo was rank 8 or 7 lol and look at me now, can barely keep 1700... What am I doing wrong... I am not trolling or playing experimental... Maybe the game has changed too much and I failed to adapt.
+0 / -0

7 years ago
As I've written in the OP, these values are generally tighter than you'd expect. All active players are visible, but 2200 corresponds to 2400 on the site.
+0 / -0
7 years ago
DErankAdminmojjj maybe people make less skuttle targets nowadays...
+0 / -0

7 years ago
Nah, skittle just got serious decloak radius nerf. (For me it's owled, but some can still get kills with it)
+0 / -0

7 years ago
Didn't that happen aaaaages ago?
+1 / -0
Skasi
Indeed. 29 May 2012, over four years ago.
quote:
Ignore previous balance changes since stable.
(...)
Skuttle
  • Decloak distance 150 -> 160
  • Jump speed 6 -> 5.2
  • Jump height 200 -> 120
  • HP 320 -> 250

This also pretty-much-ruined the fun of using Skuttles for me. Among other things it was a massive nerf to AA-Skuttle. No more exciting and funny surprises against Rapier-ball and Krow.
+0 / -0

7 years ago
Skuttle decloak distance is 180 nowadays. You missed a patch. Or somebody forgot to put up patch notes like it happens every second update :/
+0 / -0
Skasi
I didn't miss any patch since then (see previous link). GF just posted wrong numbers in the commit message. Decloak range has been 180 since 2012.
+2 / -0

7 years ago
Skuttle is next to impossible to use right now. If it has to decloack before it reaches its target, any turret including few offenders or stinger kills it on spot.

1/8 of my skuttle reach its target. Also it does seem to fail to auto-detonate when ordered to attack. I end up self-d when in radius.

I'd suggest increasing its hp to around 500-ish and make it explode as normal when killed.
+0 / -0
Skasi
quote:
increasing its hp to around 500-ish

Probably not needed. Skuttle was very much usable before the nerf listed above and not really op considering it requires a lot of energy and attention and (presumable) didn't receive anywhere near as much metal per game as Defenderspam, Scythespam or Commmorph. It had 320hp and some other stats higher back then. The additional 180hp towards 500 could be too much.

quote:
make it explode as normal when killed

It probably already does. All death explosions are (supposed to be) the same as selfD. With the exception that selfD leaves less reclaim. This was part of a mod-wide change back in CA days.
+1 / -0