next »

[Please Support Us!]

Please Support Us!

Please help keep (=WDG=) alive

• Server rules
• Forum rules
• Join us
  
• Clan members
• About the clan
• Server stats
• WDG YouTube Channel
• WDG Downloads
• IRC Webchat

Steam groups:

• WDG community
• WDG clan group
• WDG Battlelog Platoon

Our allies:

• [NuB] TF2 clan
• Velocity Gamers Hub
• Destination Gamer

GoG Links:

• Wix Discord
• Wix IRC channel

TF2 Links:

• Official TF2 Blog
• TF2 Steam Forum
• The Fort 2

October 2025

Mo	Tu	We	Th	Fr	Sa	Su
		1	2	3	4	5
6	7	8	9	10	11	12
13	14	15	16	17	18	19
20	21	22	23	24	25	26
27	28	29	30	31
	Birthdays
Chaos (40)
	Upcoming Events
Upcoming Birthdays: Chaos (40)

[Teatime]

Yesterday I exited ranking #7 on the Stats.
Epic achievement, but I guess those guys being just a few points behind me will correct that mistake ASAP 

But of course this rank says almost nothing about my skill. Even when you would equate skill with points in HLStat (which is - at best -  a distorted view) it would have little meaning.
There a people behind me which make me wet my pants when I see them coming towards me on the battlefield

Since points are accumulated they show, for the most part, the perseverance and frequency rather than skill. I have so many points because I play a lot, not because I'm good. So basically the rank in HLStats favours the frequent players.

So I thought: How can I take out the amount of gaming for a more realistic judgment of the HLStats-point-system.
The answer is easy: Take the points and the time played and break both down to an average points-per-minute value which allows comparison between those players who play much and those who spend less time on the server.

Taking the Top100 players and disregarding everybody on the resulting list who has played less then 12 hours (an arbitrary line, created with the reasoning to have enough playing time for people to display a wide range of of action avoiding the influence of extreme lows and highs).

That gave me a list of 51 players which, sorted by points-per-minute, ranked me at #47.
Which, I might add, is far more realistic.
Justice is served.

Oh, and when I narrow down the list to people having played more than 24 hours the list holds 35 players with me being #32. 

[Hzza]

Do you still have the lists? May make an interesting read.

[Lister]

Je agree with Hzza

maybe you could keep your own league tables (if you have the time) ^^

[millhawk]

I agree with all said above, post it if you wouldn't mind

[Millicent Bystander]

I like looking at the HLstats just to see how I've done, who to watch (ie who plays the same classes as me and is getting more kills) and who to specifically target because they are killing me way too much.

[claws]

I've thought of creating a more sensible ranking system too, but a) I'm far too lazy and b) there isn't the data to do it properly. To pluck an example from thin air, both Mr Chug and Teatime played on GR defence with me last night, and I imagine neither would have had a terrific session 'stats' wise - it was a tough game, both were doing hard jobs - but they each did great work for the team. You'll never see those who die so the medic can escape, or the people harassing the back building and preventing the engie nest. I can think of one kind of tracking I'd like to try, but I'd need to rip the HLStatsX event list and they probably wouldn't like that

So it can definitely be better, but for me the best judge is just to watch people play.

[Junta]

"Unsung heroes points" - yep, I like it

I think that Valve should speed up unlockables for people who aid the team but don't necessarily get the kills. For example, as a heavy I benefit from the suicidal team mate pyros setting the enemy on fire (doing massive damage) and then I get the kill. Not how I intended, but I'm not going to ignore a burning enemy if I have a chance am I?

In this scenario, we're talking damage dealt by a player (even when the result isnt a kill). Although this could mask other types of skill measurement (kills per death being a good indicator within a specific class).

[Mr Chug]

I'd like to see 'double assists'- I'm not going to disagree with the medic getting points for assists when healing since it's all we got, but it's a bit galling to push someone to the brink of death only for a stray bullet from a heavy-medic pair to whizz across and take full points.

My stats are never pretty, thanks to a combination of lack of play time and medic kpd crushing.

[Teatime]

My stats are never pretty, thanks to a combination of lack of play time and medic kpd crushing.

True, but similar to me, where my extensive gaming makes me look better than I am, you look worse than you are.
In my list you are one place below Millhawk!

[CruelCow]

EVERY ranking system is broken by definition POINT

I had similar thoughts like you, when I was in the top10, but the problem with the time system is, you punish people who fool around in between. Heck, playing when the server is nearly empty would be punished, since you can hardly keep your kills rolling if there are less targets in the field.

Edit: Ugh, ugly spelling!

[Zarf]

I've thought about doing this too, but i'm too lazy.
Points per Minute is a much better relative skill ranking.

That said, I expect i'm last on your list since i started playing spy. I can only get better than 1:1 kpd on Badwater (do much better on all maps with non-WDG servers, but they don't count)

Claws - maybe a points per round system?

[Junta]

I have a better idea. You all pay me to adjust your points

[Teatime]

Okay, here's a Top35-list based on the players with more than 24 hours on the server.
(This excludes a number of excellent, but rather irregular people)

Alternative Ranking

Congrats, Cheez!

And in a number of cases it appears that playing much might push your points forward, but your average ppm goes down the gutter (Pest vs. Cholera - ROFL)

@Claws:
What is the interface between TF2 and HLStats? I assume the events are registered by TF2 and than passed on towards HLStats, right?

[DeadlyAvenger]

EVERY ranking system is broken by definition POINT

I had similar thoughts like you, when I was in the top10, but the problem with the time system is, you punish people who fool around in between. Heck, playing when the server is nearly empty would be punished, since you can hardly keep your kills rolling if there are less targets in the field.

Edit: Ugh, ugly spelling!

this is why hlstatsx only records points when there are a certain number of players (i think on our server it is 6)

I definitely agree with most of the comments said here though. Looking at just the points is a pretty poor indication of the actual skill of the player, definitely the time spent on the server contributes quite a lot.

Although...last night...awesomeskills.
Kill Assists - Medic   (=WDG=) DeadlyAvenged (135 medic kill assists)
Ubercharger   (=WDG=) DeadlyAvenged (53 ubercharges)
Ubersawer   (=WDG=) DeadlyAvenged (10 kills with ubersaw)


@claws: ok, ouch - you win :/

@teatime: i'm probably a good case in point - lots of time on the server, not brilliantly skillful normal rank: 15. teatimes rank: 29

[millhawk]

Wow, cheers for that. I'm not as good as i thought Lol . But yea, thanks loads to teatime... that must have taken a long time...