Destiny Tracker Network
Yin
Yin

309 Credits 356 Posts
39 Credit(s) Received


DestinyTracker TrueSkill is an important aspect of a Guardian's Destiny career. This thread will explain roughly how TrueSkill works and what you can do to increase your level.

TrueSkill is an image-defined metric that appears on each playlist that a Guardian competes on that gauges that individuals performance in that mode. The range of levels begin at level one and max out at rank fifty. TrueSkill is measured by your individual performance over time (much like Site Score), but prioritizes wins above all else. TrueSkill is collection-based, that means that your rank will change only based on the statistics from your collections, it is not based on averages like Bungie's Combat Rating statistic. Players are never locked to a specific-rank and can always improve to a higher TrueSkill no matter how many games they have played in the past.

Due to the nature of our formula and its calculations based on the data acquired from a Guardian's collections, it becomes difficult to grade new visitors to the site. So, in order to prevent every Guardian (even ones with hundreds of games) from beginning at a level one TrueSkill on their first visit to DestinyTracker, a second formula evaluates your existing statistics and places you based on its results. These are by no means perfect, but it helps save new members from beginning from level one.
//content end


__________________________________________________






__________________________________________________





How is TrueSkill calculated?
TrueSkill increases or decreases based on the stats from your collections (the games you've played between the last time you updated your stats and the newest update). This information is evaluated and points are given or taken away from your TrueSkill value. Once you pass a threshold for a specific rank, your TrueSkill will increase. As you level up, the threshold becomes more difficult to attain. Consider your TrueSkill similar to other games with ranking structures.

What factors influence my TrueSkill?
Winning your games is a big indication of skill level, but individual performance in your games is also considered. If you are scoring high consistently, you can expect to increase your skill level. Just make sure you keep winning your matches!

Where is TrueSkill calculated?
We provide a TrueSkill value for each playlist that exists in the Crucible; your highest TrueSkill value will also be displayed on your player page as a sign of what you have accomplished.

Why is my TrueSkill so much lower compared to my friends, but my stats are better?
We usually get these questions from members who are new to the site. Remember that when you visit for the first time, we place you based on the summary of your performance thus far, so you were likely graded lower than what you are capable of playing at. As you continue to play and visit DestinyTracker, you should notice that TrueSkill value change rapidly until you reach your expected level.

Can my TrueSkill decrease?
Yes, your level can drop if your performance does not meet the requirements of the level you are currently at. But never fear! You are never locked in to a specific TrueSkill, so improvements in your skill will always lead to an increased TrueSkill.

Why am I matching this person when their TrueSkill is so much higher/lower than mine?
Our TrueSkill value does not provide insight to the other players you will be matched with in the Crucible. Bungie likely uses a specific algorithm to match players that we are unaware of.

If I delete my character, will my TrueSkill change?
No, since TrueSkill uses stats from your collections, the number of characters you use or compete will not affect its calculation. However, deleting a character will affect your DTR score. Be aware of this before you delete your characters.

What about Rumble? Do I lose TrueSkill if I don't come in first?
Bungie technically considers only the top player in a match of Rumble to be considered the winner. To accommodate our free for all player base, we have a specially modified version of TrueSkill that handles the unique nature of the mode.

My question doesn't appear to have an answer, what's next?
If you have any questions about TrueSkill that you'd like answered, feel free to contact Darkened Night or post a reply with your question on this thread. Your feedback is important!

Danymac89
Danymac89

12 Credits 12 Posts
7 Credit(s) Received
Stupid system seeing as I keep getting put into losing matches of 18000-7000 point control matches

hatebreedsrage
hatebreedsrage

7 Credits 1 Posts
7 Credit(s) Received
Danymac89 said:Stupid system seeing as I keep getting put into losing matches of 18000-7000 point control matches


Yeah, its essentially tailoring your score to whatever random group of people you've been paired with. I've seen some truly amazing people with crazy high scores on the losing team because they got paired with players who had 4 kills and 17 deaths. I dont see how they can call this a TrueSkill score when it has nothing to do with your skill alone.

Danymac89
Danymac89

12 Credits 12 Posts
3 Credit(s) Received
Well I mean all good if I get to play the game out I can still get some good kills and have a chance of getting 1-4% positive to my current rank but when I search and put into a match that is already 19000 points done walk in get superd 4 times, get a KD of 0.25 and then a loss ontop of that and lose 50% on my true skill level it just sucks so bad lol.
But yeah I definitely understand where you are coming from wins are definitely not the measure of a players true skill you could have 2 amazing players on one team, match them up with 4 new players against 6 average players and they will lose

Yin
Yin

309 Credits 356 Posts
6 Credit(s) Received
Danymac89 said:Stupid system seeing as I keep getting put into losing matches of 18000-7000 point control matches


Those losing matches don't seem to be affecting you. You've been consistently going up in Control and you have a solid 33 in Trials of Osiris. The system is working as intended, and hundreds of our members have 50s. Remember that even the best player won't earn a 50 overnight, it takes time.

There's a level of difficulty for us when it comes to creating skill-based metrics as a fan website. We do not have the resources (both numbers and the power to process them) that Bungie does. We do not have the ability to avoid getting you into games through join in progress, nor do we have the ability to pair you up with players of similar skills. This makes it significantly more difficult for us to create an accurate system, despite our best efforts. If you are intent on insulting the features that we provide, we ask that you provide constructive feedback for the sake of improving the features for everyone to use or simply dismiss them entirely. Our features attempt to prolong the longevity of the game, and lots of care goes into them to give Destiny players more ways to enjoy the game.

My hands are ultimately tied with this method of progression. If our TrueSkill value was based on averages, players would be stuck at a specific rank and then no one would have a chance to improve their scores, which would mean little for those who could start low and work their way up.




Danymac89
Danymac89

12 Credits 12 Posts
1 Credit(s) Received
Darkened Night said:Those losing matches don't seem to be affecting you. You've been consistently going up in Control and you have a solid 33 in Trials of Osiris. The system is working as intended, and hundreds of our members have 50s. Remember that even the best player won't earn a 50 overnight, it takes time.

There's a level of difficulty for us when it comes to creating skill-based metrics as a fan website. We do not have the resources (both numbers and the power to process them) that Bungie does. We do not have the ability to avoid getting you into games through join in progress, nor do we have the ability to pair you up with players of similar skills. This makes it significantly more difficult for us to create a system that works. If you are intent on insulting the systems that we provide, we ask that you provide constructive feedback for the sake of improving the features for everyone to use.

My hands are ultimately tied with this method of progression. If our TrueSkill value was based on averages, players would be stuck at a rank and then no one would have a chance to improve their scores. I would predict that your new TrueSkill ranks would actually be lower than they currently are were I to change the system to this method.


I understand that it doesn't happen overnight, when I'm at 50-60% towards my next level and then get put into a couple of these games then I'm back at 2% it's frustrating. And yeah I went from 28-27 last night.
Constructive criticism? Is there a way to not include games that you are in for less than 5 minutes into the statistics? I think having games like this affect true skill rating do not give a very good indication of anything. I wouldn't expect you guys to be able to affect the matchmaking or to put me in games with people of similar skill. But these games im in for 2 minutes that affect my score are extremely frustrating and I'm sure im not the only one.

Yin
Yin

309 Credits 356 Posts
1 Credit(s) Received
hatebreedsrage said:Yeah, its essentially tailoring your score to whatever random group of people you've been paired with. I've seen some truly amazing people with crazy high scores on the losing team because they got paired with players who had 4 kills and 17 deaths. I dont see how they can call this a TrueSkill score when it has nothing to do with your skill alone.


I can assure you that if someone was getting wins while going 4 and 17, they would never be able to get a 50. The same can be said for the truly amazing players who still lose, their stats can mitigate any possibility of being deranked.

Think of TrueSkill like playing ranked matches on any game. Players who want to have the highest score will be playing with a team. If we didn't base our system on this, getting a 50 would be meaningless task and wouldn't offer anything to Destiny players.

Danymac89
Danymac89

12 Credits 12 Posts
And trials is all good due to a fresh game everytime I don't have to worry about this issue but I've been playing control a lot lately to boost my total score and watching my true skill as I go.
Because in the game modes other than trials you always (95% of the time) get put into the losing side to fill someone who has rage quits place on a team that is getting slaughtered

Yin
Yin

309 Credits 356 Posts
Danymac89 said:I understand that it doesn't happen overnight, when I'm at 50-60% towards my next level and then get put into a couple of these games then I'm back at 2% it's frustrating. And yeah I went from 28-27 last night.
Constructive criticism? Is there a way to not include games that you are in for less than 5 minutes into the statistics? I think having games like this affect true skill rating do not give a very good indication of anything. I wouldn't expect you guys to be able to affect the matchmaking or to put me in games with people of similar skill. But these games im in for 2 minutes that affect my score are extremely frustrating and I'm sure im not the only one.


Again, this goes back to the resources we have at hand. Bungie provides information to us through their API, we use that information to operate our website. The point at which you are placed into a game that is join in progress is not an item provided in the API calls. There just isn't any way to do so; believe me, I would have had this put into the formula months ago had I been given the option.

I get it. Joining a match in progress stinks. It happens to me all the time as well. But in the long run, the number of games that stick you in at the end won't matter in the grand scale of things. No matter how many games you get stuck in at the end, they will never leave a permanent stain on your TrueSkill. Also, keep in mind that playing with a team will decrease your chances of joining matches in progress.

Danymac89
Danymac89

12 Credits 12 Posts
Ok I understand but essentially what you're saying is without consistently playing with a fireteam reaching true skill 50 will be an almost impossible task?

Danymac89
Danymac89

12 Credits 12 Posts
don't get me wrong I like this site and I appreciate the effort that you guys out in, sucks that that information isn't part of what you guys get but I get it. I was just frustrated after getting dropped into 3 of these matches in a row and losing the progress Id spent an hour or two getting, lol. Keep up the good work and thanks for your replies :)

Yin
Yin

309 Credits 356 Posts
1 Credit(s) Received
Danymac89 said:Ok I understand but essentially what you're saying is without consistently playing with a fireteam reaching true skill 50 will be an almost impossible task?


Not necessarily. Before releasing the formula to Destiny, we had a similar variant on Halo 4 and I was able to achieve a 50 playing solo for all of my games, despite losing a significant number of them. It's certainly not as easy a task, but it's certainly doable, and I'm sure a number of members have earned their 50s this way.

There's no doubt that getting a 50 in Control, which currently has a horribly unbalanced meta, with 12 players will make it unquestionably more difficult to attain, but certainly not an impossible task. Again, we can't make it too easy, or it becomes a meaningless statistic. Playing with less friends will knowingly make your trek more difficult.

Let's hope the Taken King addresses some of these concerns. Joining in progress should only occur before halfway to the score limit (and that's still a stretch) and players should only be matched against groups that are +/-1 of the fireteam they enter with. This would resolve a whole lot of problems for a matchmaking system that doesn't match based on skill.

Danymac89
Danymac89

12 Credits 12 Posts
Darkened Night said:Not necessarily. Before releasing the formula to Destiny, we had a similar variant on Halo 4 and I was able to achieve a 50 playing solo for all of my games, despite losing a significant number of them. It's certainly not as easy a task, but it's certainly doable, and I'm sure a number of members have earned their 50s this way.

There's no doubt that getting a 50 in Control, which currently has a horribly unbalanced meta, with 12 players will make it unquestionably more difficult to attain, but certainly not an impossible task. Again, we can't make it too easy, or it becomes a meaningless statistic. Playing with less friends will knowingly make your trek more difficult.

Let's hope the Taken King addresses some of these concerns. Joining in progress should only occur before halfway to the score limit (and that's still a stretch) and players should only be matched against groups that are +/-1 of the fireteam they enter with. This would resolve a whole lot of problems for a matchmaking system that doesn't match based on skill.


Ok awesome I appreciate the information :) thank you

Daniel Smith
Daniel Smith

0 Credits 1 Posts
I'm really glad this system grants you a specific score for each playlist as I crumble under the pressure of trials :( , however, I was wondering whether or not the ranking system took into account weapons used and headshot percentages; as It's all very well giving a player a '50 TrueSkill' rank, but if all that player does is point at someone with a thorn, pull the trigger twice and sprint around a corner, they shouldnt be given the same 'TrueSkill' score as someone who has the same amount of kills using a sniper rifle with a %100 headshot rate.

I know its a bit vague what I'm asking but to put it simply: I don't think Kills, KDR and win rates are a very good representation of a persons 'skill' and I really am just wondering if you do, actually, take into account any other factors when deciding someone's score

-Dan

Darkened Khaos
Darkened Khaos

26 Credits 58 Posts
Daniel Smith said:I'm really glad this system grants you a specific score for each playlist as I crumble under the pressure of trials :( , however, I was wondering whether or not the ranking system took into account weapons used and headshot percentages; as It's all very well giving a player a '50 TrueSkill' rank, but if all that player does is point at someone with a thorn, pull the trigger twice and sprint around a corner, they shouldnt be given the same 'TrueSkill' score as someone who has the same amount of kills using a sniper rifle with a %100 headshot rate.

I know its a bit vague what I'm asking but to put it simply: I don't think Kills, KDR and win rates are a very good representation of a persons 'skill' and I really am just wondering if you do, actually, take into account any other factors when deciding someone's score

-Dan


Seeing as anyone can have access to virtually any gun or combinations of weapons given time and effort, I personally see no reason that the current system in play should be segregated based on weapon choices. That being said, the amount of work necessary to even attempt to pull (if the data is even in the api at all) the statistics necessary to weigh in weapon choices would be tremendous.

That being said, players have a choice in what weapon they use, choosing an optimal weapon for the situation shouldn't be a detriment to their
Trueskill (no matter how cheap it is :\ ).


I asked my brother for a forum sig... I got one.
Yin
Yin

309 Credits 356 Posts
Daniel Smith said:I'm really glad this system grants you a specific score for each playlist as I crumble under the pressure of trials :( , however, I was wondering whether or not the ranking system took into account weapons used and headshot percentages; as It's all very well giving a player a '50 TrueSkill' rank, but if all that player does is point at someone with a thorn, pull the trigger twice and sprint around a corner, they shouldnt be given the same 'TrueSkill' score as someone who has the same amount of kills using a sniper rifle with a %100 headshot rate.

I know its a bit vague what I'm asking but to put it simply: I don't think Kills, KDR and win rates are a very good representation of a persons 'skill' and I really am just wondering if you do, actually, take into account any other factors when deciding someone's score

-Dan


To add to what Khaos has said, the problem lies in the complexity of accounting for such things and the capabilities we have on our end. The biggest concern being that Destiny's weapon offerings are by no means limited. The game has a constantly evolving meta and the number of guns that can be found on the game is excessively large. As the game continues to expand, so to will weapon selections. Weapons such as Thorn will be dethroned from updates, new weapons will take their place. Having to moderate such a system that will inevitably increase further in complexity would be too difficult to manage.

When you look at blockbuster titles that debut with multiplayer modes that use ranks and other methods of skill-based matchmaking, developers don't punish players for camping with a shotgun. The developer fights back with updates, preventing unwanted elements by taking them out of the game entirely or reducing their impact to a more appropriate level. This is the challenge with a game such as Destiny, which by nature, is a much more casual game than many of its players would like.

With that, the responsibility remains with Bungie, who are in control of the game's meta and balancing those unfair weapons. When weapons like Thorn fade from the spotlight, those who relied on it to win their matches will find themselves losing their TrueSkills in due time if they cannot adapt to these changes.

Neo-N8
Neo-N8

0 Credits 1 Posts
"Prioritizes wins above all else"

This makes no sense. Me and some other guy with almost identical scores (with me beating him in both things like K/D, kills in general, W/L ratio, etc) and he has like maybe 20 more wins than me. I'm at rank 26 while this dude who I'm clearing beating is RANK FORTY SOMETHING?

Broken.

Yin
Yin

309 Credits 356 Posts
Neo-N8 said:"Prioritizes wins above all else"

This makes no sense. Me and some other guy with almost identical scores (with me beating him in both things like K/D, kills in general, W/L ratio, etc) and he has like maybe 20 more wins than me. I'm at rank 26 while this dude who I'm clearing beating is RANK FORTY SOMETHING?

Broken.


Please read the FAQ for the reason behind this.

We also had an issue with TrueSkill where players were favorably awarded points that jumped their scores. This was fixed a while back, and we didn't reset scores because it typically helped players, not hurt. Might be another reason for the difference. The system isn't broken, I assure you. :)

PSlayer
PSlayer

0 Credits 1 Posts
chill man, the only true measure of skill is in clutch performance. I play with guys that may have equal or better skill than me, but when we are playing trials or a heated pub match, I either finish or go down fighting last. I'm no god tier player, seeing as connection won't allow it in pub matches, but I play smarter than majority of players. (most times) out play your opponents, because stats mean nothing. Unless it's a game like MKx where your skill directly correlates with your stats. anything above rank 45 wrecks in that game.

Piazza03
Piazza03

0 Credits 5 Posts
When I check my stats for each crucible game type, it appears that none of the games I've played recently are registering for true skill.

I know I've played control, skirmish, and trials games within the last few days, but Skirmish isn't showing an update in games for 10 days, 15 for control.

Is this a bug, or am I doing something wrong? I always do the refresh and/or update stats, but that doesn't seem to help.