I changed my mind. spawwn's ranking system is perfect. It has me as the #1 ranked player. ‡|JPL|‡ pickled_heretic 350 Aegy`IcE! 326 Mashav 301 sauce
This would work too, but the devs nerfed short games for 2.24d (i.e. no starting APC spawns and the comm overheat period). So I got the impression that the devs wanted to move towards long match times vs short ones.
Fixed. Giving the losing team more makes sense on paper, but in reality, people can and will abuse it (as several people already pointed out).
Would have to play more on the house of rage, or wait for the G4TC cluster to try it out. Thought you would like that Pickled, made me lol at your argument that you'd adjust playstyle to be #1
Doesn't matter. If I ever fall from the top position I'm going to minmax my way back to #1. Looks like the best way to do it is to command since you never die and get just as many kills as anyone else.
Correct, commanding has a few key benefits, but most importantly you'll have to win. Also, do whatever is needed to rack up points, because your score relative to the rest of the team determines how many rating points you get upon winning. Seriously man, did you have a candy bar or something? One post ago you're all 'bite me' and shit, now you seem calm and reasonable. Then again maybe I'm making up your tone as I go
I was wondering how that worked out because I'm #6 on that list and I haven't played on the BSID server that much (except for DD). Are you going to release how those stats are calculated?
For some reason, I specced a few rounds because I was busy IRL, and then my ranking went down by like 50.
Yes, and your raping during DD is the primary reason for your current slot. Thing is the server's basically been empty since the last session, so stats haven't moved much if at all. The thing is that your score doesn't decrease over time, you have to play to lose rating. Catch is, if others are playing, they are statistically likely to rise above your rating with time, so stopping at the top won't keep you there unless the server remains empty.
Should not have happened. Were you using sm_ratings in console to check, or watching the MOTD? I realized more recently that the MOTD is caching so you don't get realtime stats at the beginning of each round. HSM was wondering about that previously, said he won a few games and his rating didn't change. We refreshed and in fact it had moved but the MOTD was old. Possibly you lost 50 points in previous games and didn't refresh until after a few spec matches? Otherwise it's a bug, you should not lose rating by staying in spec.
Gotta win too though. The points gained from a handful of kills is nothing compared to the points you stand to lose if you are defeated. Yes though, Lvl3's would help provided you can justify/get away with the research.
So. Crossroads. Honestly, Spawwn, it's nothing personal. It's just that it genuinely can't work without just serving to piss people off. The best way to do it is this: Played for under 10 hours? worth 0.5 players. Played for 10-100? worth 1 player. Played for 100-1000? worth 1.5 players. Played 1000+? worth 2 players. The 0.5-2 might be extreme but either way it's far simpler, far less annoying, will prevent stats whoring except idling. So don't count spec/unassigned. Problem sorted.
hours played sounds ok. but there need to be some sort of decrease if ppl dont play for a good while ...
@Trickster Then you may as well just say everyone is worth 1.5 or 2. Better yet just say everyone is worth 1.75 and then you have the current balancer. So I got caught up last night but will get the video up tonight (6-8 hours from now), it will explain things a bit more. In the first example I used real stats and the game came out 9v11. I then had the team of 9 absolutely steamroll the team of 11, like one death on the whole team, average score on the team of 9 was something like 30. Then I tried to stack a team as much as possible (with the same guys from the last match) by putting the highest ranked on one team and the lowest on the other... it came out 7v13. Now this isn't totally accurate because it's unlikely that the 9v11 would have been that lopsided, but it illustrates the point that if it had been then the numbers team needs a few more players. I mean such a steamroll that partway through I had a player from the 9 teamswitch over making it 8v12.
If you say everyone is worth the same amount of points (or within a very small margin) and then try to balance the teams based on those points, all you have is "5v5 is balanced" "10v10 is balanced" "10v11 is slightly imbalanced" "10v12 is totally unacceptable". You see, it doesn't matter what number you assign them, if they are all assigned the same number, it's effectively just mp_autoteambalance. Saying 1.5 or 2 (and given time you'll go from 1.5 to 2) will just basically not do anything. With the system I have in place, the actual "Ability to win a game and score points" determines your rating, and then the balance based on rating ends up with interesting (and I would say totally agreeable) scenarios where "10v13 is balanced". From there the only step is to assign points to players in a way that promotes "the right way to play" so instead of players adjusting their style to abuse the system, they can play intuitively and score at the same time... something well within my means. Okay, so here's my goal, point blank. I want to see 28v36 games, I want to see 10v12, I want to be able to join a losing team even they currently have a player or two more... shouldn't matter if they are losing or are the inferior team, I should be able to help them. I want to know that a clan can stack (they already can then it's a matter of catching them etc) and instead of it being this huge game-breaking deal, instead the other team just outnumbers them enough that the game is very close. I want all games to be closer matches without just artificially making the winning team suddenly lose tickets or something. So how do we let the server decide if a team can have a numbers advantage? Quantify player skill in a way that the server can determine "okay, this won't lead to a landslide game". Again, this isn't 10v50, this is 9v11, 8v12. What's wrong with an 8v12? the 12 get more numbers, the 8 have more talent... I don't understand the resistance here.
