...

Past 28 day stats are what should be used. I used a combination just for test purposes.



Just to be clear, this weighting was designed specifically for HH#1. It would be pointless to merge both servers data and generate a list for the reasons you have mentioned.



Again, the data should be taken from "past 28 day" data.



As I said, I weighted the data according to what I consider relevant Comp/Deathmatch statistics. In my opinion standard comp classes have useful/pertinent accuracy stats. (And before anyone points it out, I am aware HH#1 is not 6v6. Its just a proof of concept)


My weighting doesn't take into account medics at all. A solution of the top of my head would be to replace the Kill/Death normalized value with Healing/Death normalized value for medics. Disregard that, im an idiot. There is definitely a good solution tho...



Didn't know this. This data could be added if needs be. I just took some data around 2 weeks ago to test out this idea

The wonderful thing about normalizing data is its a cinch to add new weights.

Thanks for responding, most people would avoid talking about spreadsheets

I never understood the point of having individual stats for a game that is supposed to be all about teamplay. In my opinion, the only stat that would really make sense (and isn't even shown) is the games won/lost ratio. Since you have equal chances of ending up in a great team, a terrible team or just one of 2 balanced teams, the determining factor for this stat would be how much you are helping your team to win games. Obviously, it would take a lot of games before this becomes statistically meaningful.

Of course, that is assuming you don't join the server as spectator and then join the strongest team. But nothing can prevent that, the same goes for all the individual stats we have now.

The only imbalance in the stats in my opinion is the fact that if you are in the top 200 players (only because you play regularly, not because you're good) and you get a kill, you'd be lucky to get more than 3 points. Then you happen to die to an ubered Heavy whose first map on the server is this one after a standard three-kill-streak life and you lose 18 points for dying. So you lose points whilst maintaining a 3:1 kpd ratio.

It's ridiculous. Killing any player no matter their 'total HLStatsX points' should give the same value because skill/difficulty/credit of kill to your team does NOT equal amount of time spent on server.

There's a 50-kill requirement in order to be able to gain more than 2 points for a kill. This takes care of new players who won't ever come back.

That goes to show how much attention I've been paying to the stats of late.

I don't agree with this at all. Just look at the weapon stats, and consider that there's a limit of max 2 snipers per team.

The Rifle is the weapon with the 5th most kills and 69% of the kills are headshots. This means that it's not really THAT skillful to require a headshot as sniper. All these gives you a bonus of 2 points.

There are less backstab kills then there are headshots made. Yet hs gives you an extra bonus point.


I'm still wondering why backstabs count towards crit-kills when a hs doesnt? Makes no sense.

I actually once examined the stats in a similar way, trying to think of a nice way to count "skill". The calculations were based on who you kill. If you kill a newbie ranked 10349 you get less points than if you manage to kill that bloody Heisenberg guy.

Well that was couple of years ago and in a different server and since then I quite lost my interest on Hlstatsx. They are fun to look at sometimes, but no one should care about them too much. Caring about them also decreases the quality of the teamwork imo.


I agree, the stats are purely a reflection (and not very accurate) on individual skill. I would rather see some killstreaks and such be replaced by clever algorithms that measure your level of teamwork.

Besides that, I have always thought that the killstreak plugin in HH is really needless and misleading. It encourages people to think about their own kills and not the team.

One more thing. When I play spy I use the normal watch 99% of the time and I very often sacrifice myself to kill a medic. (The same with scout play actually.) I haven't done any calculations, but I bet it is more rewarding for your stats if you use DR and just take down a sniper or a loner here and there and stay away from the combos.

I find the stats totally misleading and rather annoying at times

Ok I've got 17 000 + and about a 150 rank, mostly due to being on the servers 24/7, but do I get any points for being part of the skirmish line in front of our capping team?

The occasional point for being glued to the bomb cart really makes up for having my arse shot off by 4 enraged sollys 2 seconds later

But my real complaint is that they need to be reset a bit more often as its very intimidating when you join the server and see people with 60 000+ scores

Boris

PS . A woman I know sent me a 'dear John' , which is strange as my name is'nt John

PPS In fact what is my name..... I know it is'nt Boris... but I cannot remember it