Halo 3 Forum
This topic has moved here: Subject: In depth explanation of the Halo 3 skill ranking system.
  • Subject: In depth explanation of the Halo 3 skill ranking system.
Subject: In depth explanation of the Halo 3 skill ranking system.
  • gamertag:
  • user homepage:
  • last post: 01.01.0001 12:00 AM PDT

"I think it's totally better to use the team outcome instead of the individual outcome."
--
Why do you think that? Because its easier? Thats the opposite of progress. I say its simple because it is in fact VERY simple. I'm not sure if you read the post correctly. I deal with millions of records across thousands of variables using models 100 times more complicated than a simple linear regression. A server could knock out thousands of individual records with 15-20 vars in seconds. The depate should be how much more is a kill worth than an assist , or should you get more points for a long killing spree (spotting a sniper is obviously impossible to tabulate, just like a block downfield by a wide receiver. It doesn't go on the stat sheet).

They can do all of this and STILL make match making and the ranking system fast. Its called innovation and progress. Its logical to think an indivual rank should have something to do with an individuals performance, so why not create it.

We are talking about servers here, not your dual core pentium machine that would break in half with these calculations.

"Also, the main reason for having a skill cap is keeping searches quicker. Having no top limit on skill wouldn't make it better, only slower."
--
Don't get me started on the useless matchmaking search. The problem is much more complicated than finding someone with a close rank.

  • 12.21.2007 11:21 AM PDT
  • gamertag:
  • user homepage:
  • last post: 01.01.0001 12:00 AM PDT

Hey guys I figured I'd just put my 2cents into this thread and give you some practical examples based on my own experiences. On this account: TheMoose15 I achieved a rank of lvl 49 with a team of 4 almost a month ago, and due to our disparate schedules couldn't really play together all that much anymore. So, I decided to go into slayer by myself in an attempt to get my 50. Now don't get me wrong I'm not a pro player or anything, but I consider myself at least above average at the game. So long story short I wound up losing my 49 and dropping to a 48 where I would win maybe 2-3 games then lose one due to playing with random teammates. Now if I knew how the ranking system worked in the past I would've never bothered playing extra games in Team Slayer by myself, because playing additional games drastically reduces your leveling potential. This is because overtime the system hones in on your apparent "actual" skill level. Recently I began playing with my team again and won close to 10 games before finally losing one. The funny thing was the teams we were playing were far lower ranks than us. Our team consisted of 48.48.48.46 while the opposing teams would all be low 40's such as 41.40.42.39. Here's an example from that 10 game run of wins.

http://www.bungie.net/Stats/GameStatsHalo3.aspx?gameid=195850 892&player=THEMooSE15
These game can be found around page 3-4 on my profile with a couple more recent one's on page 1.

As you can see our opponents are far lower level than my team making a win for us almost a pointless endeavor. I'm not exactly arguing for a change in the ranking system however, I find it slightly humorous that playing more games actually penalizes you in regards to ranking. The true skill system is actually a pretty clever mechanic but being basically penalized for going in and just playing some games by myself is my only beef with the system. Is my true skill a 48, yea it may be but in order to bring the system back into functional leveling territory you need to play and win an absurd amount of games which basically requires a functional TO4. Can you get a 50 by going in by yourself, sure, but the amount of luck involved in getting matched up with capable teammates at a higher levels is one of the deciding factors. Versus opponents of lower skill level it's possible to carry your team however as you approach people who have a similar/higher inherent ability this options becomes very limited. Basically it's just a word of warning for people playing loads of TS for fun and then complaining about having a static level. If you are in there and value your rank you need to play to win almost every game.

So in summary the system definitely works like it is supposed to, the only real problem with it was the lack of education about it's inner workings that would've prevented me from wasting my time in ranked playlists only to be penalized for it with a static rank. Also I'm not sure if you guys noticed that table they have on the trueskill microsoft research website where it lists how many games it takes in a specific gametype in order to gain a skill level. I believe the reference for 4v4 was around 10 games although I might be mistaken.

EDIT: Link info midway down page: http://research.microsoft.com/mlp/apg/faq.aspx

On a side note I recently made a new account by the tag: "Moose Snacks" where my goal was to see how high I could get by just going into slayer by myself. On my ascent through the levels I played on a couple teams. The first being a group of guys who were legitimate level 12-15's and were amazed at how I would rank up 3 levels per game while they were completely static. . Anyways counting up my games from lvl 1 to lvl 47 beginning on page 7 and ending with the link provided below my record was 42-14 going in by myself with a couple spurts of playing with random parties. I was amazed at how quickly I leveled up and since trueskill doesn't take into account individual game performance I was wondering if anyone has any insight. A similar occurrence was brought up during a win steak of double team where my brother who was a lvl 7 at the time and me who was a lvl 1, won about 25 games straight, bringing his skill to a lvl 17 while mine got to around a 32. Now the question is my brother who had hardly played any games in DT to achieve his lvl 7 should've been leveling proportionally to me in the playlist, however his leveling completely stalled once he hit around lvl 15 requiring 5 or more consecutive wins in order for him to rank up. A decent example of this is shown on page 4 of Moose Snacks. Anyways I'm bored at work and thought I'd jump in here, if anyone actually takes the time to read through this feel free to fire some feedback / questions whatever at me. Thanks.

47 game, page 4
http://www.bungie.net/Stats/GameStatsHalo3.aspx?gameid=189508 878&player=MooSe%20SnAcKs

[Edited on 12.21.2007 11:43 AM PST]

  • 12.21.2007 11:40 AM PDT

Posted by: KA0S777
Posted by: KA0S777
How does a lone wolf even come close to a team player.
That's like having an assasin compared to a hit squad.
Seriously - see who has more skill.
And they aren't wolves because wolves hunt in packs...
and the lone just gets 0wn3d - natural selection - we survive as a unit.
You just can't compare them - and you can't convince me or anyone else logically.
Strategic warfare - that's what teams are.
It's not committing suicide because "Only the fools rush in" - Bungie
So why are you telling me that people that draw the other team out are what count?
Doesn't make sense in the real world teams so why apply it here.
Unless your suicidal and in that case - seek help.


Posted by: TwoFaced1680
I'm not sure where this rant comes from or what you are addressing so I'll ignore it as nonsense until you can make more sense of it.


I don't understand what you don't understand. But then again you must be a very happy person all the time.
Oh ya, the system works alright - that's why you have 10 million people complaining and asking the same questions a million times over.
Is that your measure for success?

"Your right, I get it. It all makes sense your the perfect person. So right, so wrong. Let's all live in your imaginary life" - Chevelle
10 million huh. So the number people complaining actually exceeds the number of people playing the game. If you actually tried to make a real point and didn't exagerate I might be able to take you seriously.

My goal isn't to be right. However, I think I am. I've tried to explain how the system can work using a single system for all playlists/gametypes. You either don't get it or trully think it just can't work. If you don't think it can work make your case. How hard is it to make a real point instead of writing the nonsense above.

If you've read even a few of the last pages you would realize there are a few issues with the system. You would realize I'm aware of them. The vast majority of complaints have to do with these two issues (mixed parties and people winning so many games they actually level slowly). Bungie is looking into these issues and I'm sure they'll be fixed/improved. Those issues I don't believe have anything to do with the TrueSkill system itself. It has to do with how Bungie is using people skill levels to match players.

However, your main point is that the system needs to take into account individual performance in order to work. Somehow the concept of a better player helping there team more often then a weaker player and thus winning more, completly eludes you. You don't seem to think this can work. I do. This principle was used in Halo 2. It worked for 3 years. So forgive me if I trully believe the principle works. I mean I only have 3 years experience as a user of that type of system and Bungie has 3 years of data using that type of system. Fans excepted it very well. There were very few complaints about it not using individual performance. Just like there are very few complaints now.

  • 12.21.2007 11:44 AM PDT
  • gamertag:
  • user homepage:
  • last post: 01.01.0001 12:00 AM PDT

Posted by: TheMoose15
Hey guys I figured I'd just put my 2cents into this thread and give you some practical examples based on my own experiences. On this account: TheMoose15 I achieved a rank of lvl 49 with a team of 4 almost a month ago, and due to our disparate schedules couldn't really play together all that much anymore. So, I decided to go into slayer by myself in an attempt to get my 50. Now don't get me wrong I'm not a pro player or anything, but I consider myself at least above average at the game. So long story short I wound up losing my 49 and dropping to a 48 where I would win maybe 2-3 games then lose one due to playing with random teammates. Now if I knew how the ranking system worked in the past I would've never bothered playing extra games in Team Slayer by myself, because playing additional games drastically reduces your leveling potential. This is because overtime the system hones in on your apparent "actual" skill level. Recently I began playing with my team again and won close to 10 games before finally losing one. The funny thing was the teams we were playing were far lower ranks than us. Our team consisted of 48.48.48.46 while the opposing teams would all be low 40's such as 41.40.42.39. Here's an example from that 10 game run of wins.

http://www.bungie.net/Stats/GameStatsHalo3.aspx?gameid=195850 892&player=THEMooSE15
These game can be found around page 3-4 on my profile with a couple more recent one's on page 1.

As you can see our opponents are far lower level than my team making a win for us almost a pointless endeavor. I'm not exactly arguing for a change in the ranking system however, I find it slightly humorous that playing more games actually penalizes you in regards to ranking. The true skill system is actually a pretty clever mechanic but being basically penalized for going in and just playing some games by myself is my only beef with the system. Is my true skill a 48, yea it may be but in order to bring the system back into functional leveling territory you need to play and win an absurd amount of games which basically requires a functional TO4. Can you get a 50 by going in by yourself, sure, but the amount of luck involved in getting matched up with capable teammates at a higher levels is one of the deciding factors. Versus opponents of lower skill level it's possible to carry your team however as you approach people who have a similar/higher inherent ability this options becomes very limited. Basically it's just a word of warning for people playing loads of TS for fun and then complaining about having a static level. If you are in there and value your rank you need to play to win almost every game.

So in summary the system definitely works like it is supposed to, the only real problem with it was the lack of education about it's inner workings that would've prevented me from wasting my time in ranked playlists only to be penalized for it with a static rank. Also I'm not sure if you guys noticed that table they have on the trueskill microsoft research website where it lists how many games it takes in a specific gametype in order to gain a skill level. I believe the reference for 4v4 was around 10 games although I might be mistaken.

EDIT: Link info midway down page: http://research.microsoft.com/mlp/apg/faq.aspx

On a side note I recently made a new account by the tag: "Moose Snacks" where my goal was to see how high I could get by just going into slayer by myself. On my ascent through the levels I played on a couple teams. The first being a group of guys who were legitimate level 12-15's and were amazed at how I would rank up 3 levels per game while they were completely static. . Anyways counting up my games from lvl 1 to lvl 47 beginning on page 7 and ending with the link provided below my record was 42-14 going in by myself with a couple spurts of playing with random parties. I was amazed at how quickly I leveled up and since trueskill doesn't take into account individual game performance I was wondering if anyone has any insight. A similar occurrence was brought up during a win steak of double team where my brother who was a lvl 7 at the time and me who was a lvl 1, won about 25 games straight, bringing his skill to a lvl 17 while mine got to around a 32. Now the question is my brother who had hardly played any games in DT to achieve his lvl 7 should've been leveling proportionally to me in the playlist, however his leveling completely stalled once he hit around lvl 15 requiring 5 or more consecutive wins in order for him to rank up. A decent example of this is shown on page 4 of Moose Snacks. Anyways I'm bored at work and thought I'd jump in here, if anyone actually takes the time to read through this feel free to fire some feedback / questions whatever at me. Thanks.

47 game, page 4
http://www.bungie.net/Stats/GameStatsHalo3.aspx?gameid=189508 878&player=MooSe%20SnAcKs


You bring up a few very valid points:
1) Dont play ranked until you are happy with your skill level...you will be penalized with a slow grind to your true skill. Basically don't learn how to play this game in the ranked lobbies.
2) There seems to be something additional to number of games played within a playlist that Bungie uses (your doubles example). A friend of mine and I started TS at the same time...maybe 10 games total seperated us. We have played togetehr ever since and were almost the same rank when we started (I think 1 skill level diff), and now we have an 8 point skill rank disparity on that playlist. It's possible Bungie does some other "off the books" analysis on your gamertag...maybe things like EXP, total games played across all playlists etc, but only applies minor importance to those numbers.

  • 12.21.2007 11:58 AM PDT
  • gamertag:
  • user homepage:
  • last post: 01.01.0001 12:00 AM PDT

It's possible Bungie does some other "off the books" analysis on your gamertag...maybe things like EXP, total games played across all playlists etc, but only applies minor importance to those numbers.

That may be the case because the trueskill explicity states it doesn't take into account in game performance. Me going +20 versus some of the lower level teams while I was leveling up should have no effect on my level progression.

  • 12.21.2007 12:14 PM PDT
  • gamertag:
  • user homepage:
  • last post: 01.01.0001 12:00 AM PDT

Posted by: TheMoose15
It's possible Bungie does some other "off the books" analysis on your gamertag...maybe things like EXP, total games played across all playlists etc, but only applies minor importance to those numbers.

That may be the case because the trueskill explicity states it doesn't take into account in game performance. Me going +20 versus some of the lower level teams while I was leveling up should have no effect on my level progression.


I'm not talking about individual performance...I definately understand the True Skill system. I'm talking about aggregated win/loss data from non related playlists that could be used to tweak the system for match quality purposes. If they do this (and I am not stating they do...just a theory), it probably comprises of only a very minute portion of the Skill Calculation. I know they officially say they keep playlist win/loss data exclusive from other playlist rank data...but some of the qualitative observations seem to point that it isnt 100% accurate. It would also make sense they throw in little tweaks here or there to improve match quality that are over and above the plain vanilla True Skill calculations.

I guess the only way to test would be to send a team of two guys into a new playlist (no rank yet) that had wildly different skill levels and/or EXP in other playlists. Never seperate them until they have 50-100 games played, and make sure the ranks are exact. Any of the regulars on this thread ever seen this test performed?

  • 12.21.2007 2:03 PM PDT

Adigalo
You actually have a rank gap of 10 levels that the hopper tries to pair you with. If you look at the left side of the screen where it is searching for players, it will tell you the ranks the server is looking to match you up with. It starts with your rank (or the highest ranking member in your party) and looks for players 10 levels lower than you or the highest ranked member of your party and 10 levels higher than you or lowest ranked member of your party. If your party is not within these parameters, the party is considered "mixed" and will try to find another "mixed" party with the same highest ranked member.

Great post El Kafungus!! It really helps answer some questions.

[Edited on 12.21.2007 3:57 PM PST]

  • 12.21.2007 3:54 PM PDT
  • gamertag:
  • user homepage:
  • last post: 01.01.0001 12:00 AM PDT

I have a question about the ranking system, I read some parts of the topic, but could not find my answer.
Today I played about 30 matches team slayer ranked with 3 friends. My rank when I started was 28, and the rank of a friend was 25.
During those matches...my friends' level increased much faster. After the 30 games...his level was 35, and mine was 32.
Now...that can be logical...but... I was playing almost every game better than my friend...he was mvp 3 times...me 25 times... I scored running riots...k/d spreads of +15 and even the 'perfection' medal.
How is it possible that my friends level increased so fast...while I was playing the same games...only made more kills?

  • 12.21.2007 8:18 PM PDT

Posted by: LochNess22
That sucks when you can't go up in rank if you are playing with lower ranks on your team. The ranking should be based on individual effort in combination with a victory. A VERY simple regression equation could have been developed and tweaked in Beta to make the ranking an actual ranking. Not this (Mu) and (Sigma) second rate math BS!
Bungie is working on fixing the issue with mixed parties so hopefully ranking will improve in that type of situation. However, it doesn't make much sense for me to move up the ranking ladder when I beat lower level ranks. This is essentially what happens when someone plays in a mixed party with a much lower rank. I mean should a 40 gain skill levels because they beat 30's all day?

Posted by: LochNess22
Simple: assign weights for kills, deaths, wins, losses, headshots, assists, etc. and bam you get your rank, with an UNLIMITED ceiling. (I work with the types of models daily, and they take hardly any computing power to calculate)
It may in fact be possible to create a system to evaluate and measure individual performance. However it just seems like it wouldn't be worth the effort. I think it would be fairly dependent on trial and error. There would be so many variables to consider it would be a pain to tweak and get right. Each gametype would probably need to have different weights. CTF games should also include who scored the flag or grabbed the flag. King of the hill would include hill time etc.

The current system works on the principle that someone who is better is more likely to help their team win. Therefore it reduces to essentially wins and losses (of course it takes into account skill level differentials as well). Such a system works for all gametypes and doesn't rely on trial and error at all. Do you really think the current concept works so poorly that a system that doesn't work the same from gametype to gametype and is seemingly more error prone needs to be developed. From my experience the current concept works very well. The only thing it lacks is that it can't seperate players who only play together. But in general it does do a very good job. I just don't see any major benefit to including individual skill.

Just out of curiosity how would the weighting of kills and deaths work for a game of Team Slayer? For example who would be evaluated as better in this scenerio. Player A goes 18-20 for his team and Player B goes 3-0. I personally think Player A was potentially a much better asset. I know there are people who agree with me and I know there are others that don't. What do you think? If you agree with me could this be weighted properly. Also whether you agree with me or not dont you think the fact that this is subjective is a problem for such a system?

Posted by: LochNess22
There is no need in putting an arbitrary highest rank possible. Why not find out who the best really is and how bad I really am.
The skill levels are to group players of similiar skill. The ranking system is not a leader board. There is no point to having and unlimited ceiling. You don't need to know if you are the 5th best in the world or the 2nd worst. That's not what it was designed for. Maybe that's what you want (I can respect that opinion), but that's not why it was created.

  • 12.21.2007 9:04 PM PDT
  • gamertag:
  • user homepage:
  • last post: 01.01.0001 12:00 AM PDT

Thanks for the info man. I thought the system was based on those myths you said aren't true like mvp and kill/death ratio. Looks like I thought wrong.

  • 12.21.2007 10:30 PM PDT
  • gamertag: [none]
  • user homepage:

"Pain heals, Chicks dig scars.. Glory lasts forever"- Shane Falco
Huge Man of War!
RIP Johnson :(

Ok, can you help me out with this, i am a level 31 and i played last night and won 9 team slayer games in a row. And i only ranked up to a 32!?

http://www.bungie.net/stats/PlayerStatsHalo3.aspx?player=Elpo on

  • 12.22.2007 8:43 AM PDT

etc etc/glaringly obvious/and so on, and such <=Not redundant!
Posted by: Cr4ne Style
Taxes do nothing to affect the share of wealth, since taxes are only applied to income.

So that's not even a part of the conversation at all, so it's pointless talking about it....

"for a "best" moral to exist, there must exist the "best" moral base. If the base of morality varies from location to location, culture to culture...then there can't be an absolute moral..

Posted by: elpoon
Ok, can you help me out with this, i am a level 31 and i played last night and won 9 team slayer games in a row. And i only ranked up to a 32!?

http://www.bungie.net/stats/PlayerStatsHalo3.aspx?player=Elpo on


here is what happened: you won some games, and you ranked up. moving up one rank after 9 games is reasonable.

  • 12.22.2007 9:22 AM PDT
  • gamertag:
  • user homepage:
  • last post: 01.01.0001 12:00 AM PDT

Urg, sorry my vocabulary is a bit weak, anybody mind telling me how to get a good sigma?

consistently playing lowers sigma right, and consistently means playing alot and winning?

I understand winning games increases Mu, I just don't know how to increase sigma I kind didn't understand it.

  • 12.22.2007 12:56 PM PDT

[Everything is working fine. Just crunch the numbers to prove it. You don't need to guess.

I checked your stats at 6:37 PM last night. You had an experience penelty of -104. You now have a penelty of -113 so that's 9 more penelties then the last time I checked. Looking at your games played since then you've played at total of 27 matchmaking games. You should have earned 13 experience during those games. 13 - 9 = 4. So your gain in experience since last night when I checked should be 4. Looking at your experience now you have -54 which is exactly 4 more then -58 which is where it was last night.

The math all checks out. You are getting all the experience you deserve. If you are quiting games then stop. If you are having internet difficulties then try to work them out. There is nothing wrong with your experience.] from twofaced. dont know how to blakbox it





okay thta was a while ago but i still dont enderstand what is wrong with my achievements and everyting. and plus, i dont quit. not my fault i get disconnected. i dont think its fair that i lose exp b/c i lose my connection w/ other players. it should be like COD4 since you dont go down. its easier and earning points in-game is cool. anyway i tried something. i have 1month free gold xboxlive membership to use. so i made a new account. go and check its cvillelax23. anyway i played 2 games and you think i go down rite? its still at zero. not a recruit yet but still. reply. WITH AN EXPLANATION FOR MY ACHIEVEMENTS!

  • 12.22.2007 1:47 PM PDT

Posted by: LochNess22
"I think it's totally better to use the team outcome instead of the individual outcome."
--
Why do you think that? Because its easier? Thats the opposite of progress. I say its simple because it is in fact VERY simple. I'm not sure if you read the post correctly. I deal with millions of records across thousands of variables using models 100 times more complicated than a simple linear regression. A server could knock out thousands of individual records with 15-20 vars in seconds. The depate should be how much more is a kill worth than an assist , or should you get more points for a long killing spree (spotting a sniper is obviously impossible to tabulate, just like a block downfield by a wide receiver. It doesn't go on the stat sheet).

They can do all of this and STILL make match making and the ranking system fast. Its called innovation and progress. Its logical to think an indivual rank should have something to do with an individuals performance, so why not create it.

We are talking about servers here, not your dual core pentium machine that would break in half with these calculations.

"Also, the main reason for having a skill cap is keeping searches quicker. Having no top limit on skill wouldn't make it better, only slower."
--
Don't get me started on the useless matchmaking search. The problem is much more complicated than finding someone with a close rank.

I think using the team outcome is better when you are playing a team game. I gave examples of why I thought it was better in my post. I didn't really say anything about being easier. Why would you like it changed and what current problems would it fix? TS is actually the easier team list to "individually" score. Team objective lists become even more complicated on how you define what each stat is "worth". I say it again, it then turns in to individuals working towards getting the stats they need to go up without worrying about the team. Maybe I am just of the "no man left behind" mentality where you live and die together. Not hiding while my team mate gets hiss ass blown away because I don't want to go negative.

I never said it couldn't be done or that "servers" couldn't do it. I just reminded you it is not "easy" to track stats for the volume of games and users we are dealing with. Since you seem to have experience with statistical analysis and database management, how "easy" do you think it will be crunching 50 or 100,000 transactions a minute?

What is the problem with MM? What exactly is more complicated than finding someone with a close rank? Spreading the skill out to no top limit serves what purpose?

  • 12.22.2007 7:11 PM PDT
  • gamertag:
  • user homepage:
  • last post: 01.01.0001 12:00 AM PDT

http://www.bungie.net/Stats/Halo3/Default.aspx?player=T+And+C rump3ts

Those are my stats.

I have played around 43 double team games and not lost a single game.

I am a level 15 in this playlist.

Surely i should be leveling up faster than this, so I can play with people of skill levels closer to mine. Me and my teammate are consistantly playing against low level players and its not fun!

The account I am using is a one month free trial. My team mate is a level 45 in rumble pit. Does this have anything to do with the fact I'm not leveling up faster?

Thanks.


  • 12.23.2007 9:13 AM PDT

Posted by: T and Crump3ts
http://www.bungie.net/Stats/Halo3/Default.aspx?player=T+And+C rump3ts

Those are my stats.

I have played around 43 double team games and not lost a single game.

I am a level 15 in this playlist.

Surely i should be leveling up faster than this, so I can play with people of skill levels closer to mine. Me and my teammate are consistantly playing against low level players and its not fun!

The account I am using is a one month free trial. My team mate is a level 45 in rumble pit. Does this have anything to do with the fact I'm not leveling up faster?

Thanks.


I totally understand your frustration! Level 14/15 seems to be where teams that ONLY play together get stuck. There was a team of 4 who won like 120 games and were stuck at 14/15 who were also undefeated. Bungie is aware that this is an issue and the exact resolution is unknown. One resolution seems to be to play some games with randoms. I am not really clear if it is because you haven't lost or because you two only play together. The only other suggestion has been to lose but if it were me, I would play 4 or 5 games with someone else. Also, skill levels from different playlist have nothing to do with each other. Each playlist has its own level and is tracked separately.

  • 12.23.2007 9:43 AM PDT
  • gamertag: [none]
  • user homepage:

UNICORN POWER!!!
______________(__||__M__||__)___
|-------------___-.~,~~:_____:_:_____|================[__]==
|_______/ /_/```-------|__|

wow great post!!!
thanks for the explanation

  • 12.23.2007 10:52 AM PDT
  • gamertag:
  • user homepage:
  • last post: 01.01.0001 12:00 AM PDT

what ever. Its still ridiculous. I can understand 10 games and not ranking up. I played 20. Thats sad

  • 12.23.2007 11:29 AM PDT

Posted by: LAXICE23
okay thta was a while ago but i still dont enderstand what is wrong with my achievements and everyting. and plus, i dont quit. not my fault i get disconnected. i dont think its fair that i lose exp b/c i lose my connection w/ other players. it should be like COD4 since you dont go down. its easier and earning points in-game is cool. anyway i tried something. i have 1month free gold xboxlive membership to use. so i made a new account. go and check its cvillelax23. anyway i played 2 games and you think i go down rite? its still at zero. not a recruit yet but still. reply. WITH AN EXPLANATION FOR MY ACHIEVEMENTS!
The experience penelty is there to discourage players from quiting. I think this actually does work to some degree all though I don't know exactly how well. The problem is you can't penalize people for quitting but not for being disconnected. First, getting disconnected hurts your team and in a way is just as bad as quiting. I understand it's not your fault but it still hurts the team. Secondly and most importantly, if the system didn't penalize people who get disconected individuals would just pull there ethernet cable when they wanted to quit to avoid the penelty. The only options Bungie has are to penalize everyone for leaving or penalize no one. Since the penelty helps discourage quiters (at least I think it does to some degree) I think it's worth having.

I really didn't understand what you wanted me to check on the new account. I saw a -2 experience penelty which is correct. Since you don't have any experience you don't have a military ranking. Your achievments page is empty which is also probably correct. Did you expect to have earned any achievments?

  • 12.23.2007 3:51 PM PDT
  • gamertag:
  • user homepage:
  • last post: 01.01.0001 12:00 AM PDT

I dont have a problem with the exact ranking system. the problem lies in the matchmaking system. earlier today i was dominating in lone wolves and decided to switch to slayer to get my lvl up to 45. I played 6 games in a row, and out of them my team lost all, i was mvp with the most kills for 5, and placed 2nd for 1. The problem with this is why i constantly get matched up with horrible teammates. I tried my best to pull the teams weight but after ranking down twice i gave up. i want to know why this happens, all the time, and what can i do to get around it. I just want to win like everyone else, but team games are stating to turn into 1 on 4 on my part.

  • 12.23.2007 3:52 PM PDT

Posted by: Chaos Blood 034
Urg, sorry my vocabulary is a bit weak, anybody mind telling me how to get a good sigma?

consistently playing lowers sigma right, and consistently means playing alot and winning?

I understand winning games increases Mu, I just don't know how to increase sigma I kind didn't understand it.
There is no known way of influencing sigma. Sigma should decrease after just about every game, with a few exceptions. Before every calculation is made Sigma is raised some. The developer controls by how much. My guess is the amount it's raised would be constant, however, in theory Bungie may raise it more under certain conditions like a win streak. There is absolutly no proof to support or refute that statement. Long story short...don't worry about Sigma.

  • 12.23.2007 4:01 PM PDT

Posted by: Darkllama
I dont have a problem with the exact ranking system. the problem lies in the matchmaking system. earlier today i was dominating in lone wolves and decided to switch to slayer to get my lvl up to 45. I played 6 games in a row, and out of them my team lost all, i was mvp with the most kills for 5, and placed 2nd for 1. The problem with this is why i constantly get matched up with horrible teammates. I tried my best to pull the teams weight but after ranking down twice i gave up. i want to know why this happens, all the time, and what can i do to get around it. I just want to win like everyone else, but team games are stating to turn into 1 on 4 on my part.
The truth is this isn't always happening to you. Your mind latches on to those games you do well but your teammates don't so you think it's happening more then it is. If you aren't at your level you will probably go positive more often. However, your teammates are more often at there level. That means your teammates performance will deviate around even. Some games they'll do well others they won't. Everyone experiences the same thing. There is nothing you can do except play with a full party. This will certainly give you more consistency as you are leveling up. But once you hit your level you actually should experience the same phenomenon you are now, even with a full team. That's because when you are playing people around your skill level your performance can't be consistenly positive. I can't help but notice in your last game you went -6. I bet the person who went positive 11 was thinking the same thing as you...Boy my teammates always suck. Don't let 6 games get you down. It all evens out. If you can't handle uncertainty then find teammates or play lone wolves. You aren't getting it any worse then anyone else.

  • 12.23.2007 5:14 PM PDT
  • gamertag:
  • user homepage:
  • last post: 01.01.0001 12:00 AM PDT

thanks for this so much it answered so many of my questions.

  • 12.24.2007 7:44 AM PDT

yeah I can win all day and not gain rank but I lose one match and I drop level. I think this needs to be fixed I have only lost like 3 games in double team and win well over 50+ games still I am only a level 14 so I don't get how this is far. this is the only change that I see that needs to be fixed because there are a lot of players in this game that are not that good but are high level I do report all people that suck on the game but I don't know if any of this is being changed. I can win all day in double team and get no where how is the game going to be fun. It was addictive in Halo 2 cause if you won a lot you got level and if you sucked you got no where for the most part until you played with really good people. I will say the boosting is gone in Halo 3 but the mult accounts are running around like crazy I know that is the hardest thing to stop.

  • 12.24.2007 1:25 PM PDT