Destiny 2 Tracker Network
Shane King
Shane King

6 Credits 6 Posts
Simple question I do not think will be to hard to answer. If one plays 500 games straight - each ranging to an average of 0.5 kda, then plays 500 more games with a 2.5 kda AVG, what will his final trueskill and other stats be?
DELETED
DELETED

0 Credits 242 Posts
It's extremely hard to answer. Only 2 people know the trueskill formula and neither of them are going to let the secret slip. Also trueskill is primarily based on wins so k/d is not as important.

However after 1000 games with the numbers provided the player in question would have a k/da of 1.5.
Shane King
Shane King

6 Credits 6 Posts
See that is what I find disturbing about the trueskill system and how the community rates players. It feels like even if you get better, unless you make a new account, you will never get anywhere.
DELETED
DELETED

0 Credits 242 Posts
As long as you keep winning your trueskill will improve. If you are playing better than you used to your overall stats will improve over time.
UGD
UGD

0 Credits 1 Posts
Sengir79 said:It's extremely hard to answer. Only 2 people know the trueskill formula and neither of them are going to let the secret slip. Also trueskill is primarily based on wins so k/d is not as important.

However after 1000 games with the numbers provided the player in question would have a k/da of 1.5.


While I agree that winning is the important thing, that being the primary measure means that someone of lesser skill that plays in a party will in most cases have a higher rank than someone of equal or better skill that plays solo. If you are going to call winning a skill measure (and I agree it tells something), then you should be comparing similar circumstances.

In a perfect environment, solo players of equal skill will eventually end up with a .500 win/loss record. When you insert a higher level player into the mix, he will tilt the odds in his team's favour. So far that agrees with the skill measure of wins theory. However, many pvp matchmaking algorithims will compensate for skill levels and balance out solo lobbies so that all the better players on not on the same team. The better a player you are, then, the more of the good players that will get put on the other team. The best player could then be left with the worst group, which will have a negative effect on winning percentage.

Of course I have no idea what matchmaking systems Bungie uses, so I may be talking out of rear.
DELETED
DELETED

0 Credits 242 Posts
UGD said:
Sengir79 said:It's extremely hard to answer. Only 2 people know the trueskill formula and neither of them are going to let the secret slip. Also trueskill is primarily based on wins so k/d is not as important.

However after 1000 games with the numbers provided the player in question would have a k/da of 1.5.


While I agree that winning is the important thing, that being the primary measure means that someone of lesser skill that plays in a party will in most cases have a higher rank than someone of equal or better skill that plays solo. If you are going to call winning a skill measure (and I agree it tells something), then you should be comparing similar circumstances.

In a perfect environment, solo players of equal skill will eventually end up with a .500 win/loss record. When you insert a higher level player into the mix, he will tilt the odds in his team's favour. So far that agrees with the skill measure of wins theory. However, many pvp matchmaking algorithims will compensate for skill levels and balance out solo lobbies so that all the better players on not on the same team. The better a player you are, then, the more of the good players that will get put on the other team. The best player could then be left with the worst group, which will have a negative effect on winning percentage.

Of course I have no idea what matchmaking systems Bungie uses, so I may be talking out of rear.


Unfortunately Bungie do not have a publicly viewable trueskill system in place (and it's debatable if they have one at all). With the api as it is we cannot adjust scores for the skill of the opposing players in any particular game which means that our system must assume that players are equally matched, which in many cases is not true. Winning is not the only metric that is taken in to account, but it is the primary one, though performing well will have a small effect.

The only other way we could adjust for player skill was if every single destiny player in the world was registered with us, but that is not likely to happen sadly.