Avatar

PSA: Bungie is tweaking Matchmaking logic *again* (Destiny)

by CyberKN ⌂ @, Oh no, Destiny 2 is bad, Thursday, January 28, 2016, 19:20 (3463 days ago)

https://www.bungie.net/en/News/News?aid=14284

On Wednesday of this week, we deployed some new Matchmaking settings to the Iron Banner playlist in Destiny. Since then, we’ve been listening to your feedback and crunching the numbers. We're ready to deliver updates on what we've learned and the fresh action we're taking to enhance the state of play in the Crucible.

Your results may vary (gotta love the Internet), but we’re hearing and seeing some good things: Lag is on the decline. Match-ups are still close enough to provide the level of challenge demanded by the Iron Lords. Matchmaking time between games is less than a minute, on average. By many accounts, we’ve taken a step in the right direction, but there is always more room for improvement.

The dev team at Bungie has deployed another new set of matchmaking settings. For the time being, they are exclusive to Iron Banner. Once again, our goals are to further reduce lag.

Please be advised that your field of opponents may demonstrate a wider range of skill, still within the boundaries of a good game. Matchmaking may also take slightly longer – all in the interest of delivering a quality match with a good connection. We’ll continue to monitor the conversation and keep an eye on the data.

These changes are live right now. Suit up in your Iron Banner gear, jump into the fight, and tell us all about your experience. Tomorrow, check back for more commentary on the potential for future changes to the Crucible in the Bungie Weekly Update (published one day late to get out of the way of progress).

Game on!

Avatar

Here's hoping for less lag!

by Blackt1g3r @, Login is from an untrusted domain in MN, Thursday, January 28, 2016, 19:34 (3463 days ago) @ CyberKN

- No text -

Avatar

interesting

by unoudid @, Somewhere over the rainbow, Thursday, January 28, 2016, 19:40 (3463 days ago) @ CyberKN

I'm definitely interested to see how things go tonight. I'm keeping my fingers crossed that I don't have to say bad things about other people later on.

Avatar

Remember: Try aiming this time... :p

by Ragashingo ⌂, Official DBO Cryptarch, Thursday, January 28, 2016, 19:57 (3463 days ago) @ unoudid

- No text -

Avatar

Bungie needs to fix my aim assist! :)

by unoudid @, Somewhere over the rainbow, Thursday, January 28, 2016, 20:01 (3463 days ago) @ Ragashingo

- No text -

Make sure you're using the right barrel mods

by Claude Errera @, Thursday, January 28, 2016, 20:06 (3463 days ago) @ unoudid

https://www.reddit.com/r/CruciblePlaybook/comments/3lzyc2/sniper_guide_taken_king_edition/

Smooth Barrel mod has highest aim assist - aggressive has the lowest.

Avatar

BWU posting tomorrow instead of today.

by Xenos @, Shores of Time, Thursday, January 28, 2016, 20:01 (3463 days ago) @ CyberKN

The level of communication is more important than the tweaks

by electricpirate @, Thursday, January 28, 2016, 20:47 (3463 days ago) @ CyberKN

More of that pls.

Avatar

First 2 games I played today...

by CruelLEGACEY @, Toronto, Friday, January 29, 2016, 01:46 (3462 days ago) @ CyberKN

Top player on the scoreboard in each game was red-bar. The next 2 matches were fairly laggy as well. Can't say I've seen an improvement at all, but I'll see how it goes when I jump back online later tonight.

Avatar

First 2 games I played today...

by DiscipleN2k @, Edmond, OK, Friday, January 29, 2016, 03:12 (3462 days ago) @ CruelLEGACEY

For me, yesterday was great. Matches were reasonably close and I didn't notice a bit of lag the entire day.

Today was awful. Games felt completely lopsided. Might have been the lowered priority on finding skill based matches, or it could have been the fact that every team I played against had an invincible player that kept grabbing the spark. I don't remember Destiny lag ever being this bad. And that's really saying something considering how often I play with Funkmon ;)

-Disciple

Avatar

ಠ_ಠ

by Funkmon @, Friday, January 29, 2016, 05:03 (3462 days ago) @ DiscipleN2k

For me, yesterday was great. Matches were reasonably close and I didn't notice a bit of lag the entire day.

Today was awful. Games felt completely lopsided. Might have been the lowered priority on finding skill based matches, or it could have been the fact that every team I played against had an invincible player that kept grabbing the spark. I don't remember Destiny lag ever being this bad. And that's really saying something considering how often I play with Funkmon ;)

-Disciple

Avatar

Love ya, Funk :p

by DiscipleN2k @, Edmond, OK, Friday, January 29, 2016, 14:03 (3462 days ago) @ Funkmon

- No text -

Avatar

one of my games tonight

by unoudid @, Somewhere over the rainbow, Friday, January 29, 2016, 05:24 (3462 days ago) @ DiscipleN2k

I hadn't even landed before getting the "your team lost" screen. Dropped me into a game that finished 21k to 8k. On the bright side... I did get a free medallion and a hit to my win/loss ratio :)

It's just crazy to think that join in progress dropped me into a game where I hadn't even landed or spawned in before the game was over.

Avatar

No way it was as bad as some of the Crota raids.

by ProbablyLast, Friday, January 29, 2016, 07:11 (3462 days ago) @ DiscipleN2k

- No text -

Not as much lag as you, I guess...

by Claude Errera @, Friday, January 29, 2016, 05:38 (3462 days ago) @ CruelLEGACEY

...but definitely lopsided games.

We played 6:

Game 1 - we lost, 20k - 17.5k. Felt pretty close.
Game 2 - we lost, 20k -17k. Felt like we should have won.
Game 3 - we won, 12k - 3.8k. Mercy'd.
Game 4 - we won, 20k - 18k. Really close, felt good.
Game 5 - we lost, 20k - 12k. Should have been mercy'd, we clawed back to kill the Decisive Victory.
Game 6 - we won, 20k - 7k. Should have been mercy'd, I think.

So: 3 close games, 3 blowouts. Definitely not the pattern we've had since September.

Avatar

Not as much lag as you, I guess...

by Robot Chickens, Friday, January 29, 2016, 18:31 (3462 days ago) @ Claude Errera
edited by Robot Chickens, Friday, January 29, 2016, 18:57

So: 3 close games, 3 blowouts. Definitely not the pattern we've had since September.

Last night was my first IB experience this week, but I played a lot. I ran my numbers for fun. For the stats I'm defining a blowout as a spread of more than 6000 points.

------------------------
Data:

26 total games
9 wins
10 good matches
16 blowouts

  • 6 wins
  • 10 losses

------------------------
Percentages

Win rate: 35%

Good match rate: 38%

Blowout win rate: 38%

------------------------
Analysis

  • A win rate of 35% means I was was redeeming about 2 medallions every 3 games. That feels a lot better than previous IB experiences but I haven't run any data to compare them. Overall I can live with that.
  • The rate of getting good matches was just over 1 in 3. That seems unacceptable.
  • On the other hand, my win rate in blowout matches was pretty much the same as my overall win rate. I think blowouts in riff happen more easily because it just takes about two unanswered scores to produce this spread. If you have a relatively close game where you're only down by one detonation, it quickly becomes a blowout if the other team manages to score. Perhaps these numbers are a bit misleading as a result.
Avatar

Follow up with comparison data

by Robot Chickens, Friday, January 29, 2016, 22:54 (3461 days ago) @ Robot Chickens

I went back to the IB last year held in July to compare data. This was when I was playing Destiny better than I am now and it is arguably near when Year One had its best networking code due refinements over time. I played those 90 games towards the beginning of the week so the boon was week.

One significant difference is that scores in Rift swing a lot faster than in Control. As such I am defining a blowout (somewhat arbitrarily) as a difference of more than 4000 points in Control. This means a blowout happens when the losing side was under 15999 points. I used a 6000 point spread in for Rift because a close match where you're only down by one score, quickly can become lopsided if the other team scores just one more time. I'm open to critique about my reasoning.

------------------------
Data for IB July 2015:

90 total games
41 wins
40 good matches

  • 18 wins
  • 22 losses

50 blowouts

  • 23 wins
  • 27 losses

------------------------
Percentages

Overall Win rate: 46%

Good match rate: 45%
Good match win rate: 46%
Blowout rate: 56%
Blowout win rate: 46%

------------------------
Analysis

  • Once again, my win rate is the same in the blowouts and in the good matches where it is closer. At least destiny is consistent.
  • There is a significant change in my personal win percentage. Either I have gotten worse (totally possible because my pvp k/d went from 1.2 to 1.1 between the two years), or matchmaking is doing things differently to my detriment. My gut says it is a combination of both.
  • The number of blowouts was significantly less during Year One IB. However, as with Year Two, my win rate was consistent in the good games and the blowouts.
  • I redeemed a medallion almost every other round. Last night I was redeeming two every three games but it felt more fair. Goes to show that memories of those long stretches without redeeming medallions loom larger in my mind than the ones where I was winning every round.
  • Assuming each round takes 15 minutes of my life when turnaround time is included, 90 games is 22.5 hours of my life. I take on the equivalent of a part-time job during the week of Iron Banner.

------------------------
Data for last night:

26 total games
9 wins
10 good matches
16 blowouts

  • 6 wins
  • 10 losses

------------------------
Percentages

Win rate: 35%

Good match rate: 38%

Blowout win rate: 38%

------------------------
Analysis

  • A win rate of 35% means I was was redeeming about 2 medallions every 3 games. That feels a lot better than previous IB experiences but I haven't run any data to compare them. Overall I can live with that.
  • The rate of getting good matches was just over 1 in 3. That seems unacceptable.
  • On the other hand, my win rate in blowout matches was pretty much the same as my overall win rate. I think blowouts in riff happen more easily because it just takes about two unanswered scores to produce this spread. If you have a relatively close game where you're only down by one detonation, it quickly becomes a blowout if the other team manages to score. Perhaps these numbers are a bit misleading as a result.

uhh

by TheeChaos @, Friday, January 29, 2016, 12:38 (3462 days ago) @ CruelLEGACEY

yea that was some BS, although I did really well!

Back to the forum index
RSS Feed of thread