• Increased likelihood of getting Control over Clash in Quickplay

• Javelin-4, The Burnout, Rusted Lands, Midtown, and Wormhaven all have a higher chance of being picked as the map
• Disjunction, Cathedral of Dusk, Twilight Gap, The Dead Cliffs (in Clash), Anomaly, and Cauldron (in Clash) all have reduced map weights
• Exodus Blue and Convergence are not in playlists
• The rest of the maps have "standard" map weighting

• Clash will leave Quickplay, Quickplay is becoming exclusively Control.
• Loose SBMM (QP, Iron Banner) will have its SBMM parameters modified so connection quality is more prioritized
  • As matchmaking takes longer, the range of skills acceptable will expand faster than the range of connections acceptible.
  • This means that skill closeness will be sacrificed more quickly than connection quality when matchmaking takes longer.
• The Rotator is being split into two playlists, both using CBMM with no plans to change that.
  • Crucible Relentless Rotator
    • Clash, Rift, Zone Control
    • Haven for players who want large team-based gameplay outside of SBMM, will also mean there's always a more conventional team-based playlist when Iron Banner is active
    • This is replaced by Trials when Trials is active.
  • Crucible Party Rotator
    • Mayhem, Scorched, Momentum Control
    • For fun focused modes.
• All remaining freelance nodes have been removed, with FTMM (Fireteam Matchmaking) taking their place.
• Competitive, Rumble, and Private Matches are still always on.
• Crucible Labs will start in the middle of this season (season 20) to put 3v3 Countdown and a few variations through testing
• Preview of the director screen:

• This system keeps fireteam sizes in matches mostly consistent, so it's easier to play as a solo or duo in 3v3 or 6v6 playlists.
• Their original goal was to push it to Quickplay, then test in Iron Banner by late January, then test more nodes this season, and slowly start removing freelance nodes.
• Their first testing was so successful, they quickly enabled it in Iron Banner, and now Freelance is removed across the board.

• These are the current matchup stats for Fireteam Matchmaking in the second week of Iron Banner (last season):
  • 95% of the time, if you're in a team of 6, you matched with another team of 6
  • If you ran solo, you only went against a full fireteam 0.2% of the time (1 in every 500 matches). The majority of these cases either put you with a team of 5 (60% of the time), or a team of a mix of solos and duos (31% of the time)
• They feel they can improve the system in two areas:
  • Large fireteams (4+ in 6v6) prioritize equivalent fireteam size over all other options, even over alternatives with a narrower skill disparity
  • If FTMM can't find a suitable match in enough time, it will just take the first available match, which can result in situations like a group of solos being places against a 6 stack.
• In the mid-season patch, they're making a change to FTMM in Control/Iron Banner where if matchmaking can't find an equal fireteam with a narrow skill delta, it will search for similar size fireteam with a narrow skill delta, before looking for wider skill deltas.

• Here's the stats for comp this season with Fireteam Matchmaking:
• Full fireteams match against eachother 98% of the time
• 2+1 configurations match with other 2+1 configurations 98% of the time
• For solo players in particular, 80% of comp games are against entirely solos
• When you're solo and put in a 2+1 team, 98% of your games will be against another 2+1 team.
• They feel that 2 player fireteam matchmaking is a bit long, but otherwise this system is working extremely well in 3v3

• Here is the graph of players' rank and how it relates to their internal skill value:
• Outside of a handful of outliers, no one is 1000 Division Ranks above their Skill.
• They feel that this data shows that it's a bit hard to climb to your target rank, as the average player is well below their target skill level.
• Comparatively, this chart of players who played more than 150 games last season shows that players generally land at a rank very close to matching their skill.
• The dots around medium to low skill that were in ascendant or adept ranks indicated an issue to Bungie, either a serious bug or some exploit. They've issued bans for win trading in response.
• And finally, this is a graph of how many players are in each of the ranks

• Matching purely on rank, instead of on skill, can lead to issues like:
  • Smurfing (high skill player creating a new account purely to play against lower skill players). This is technically possible still, with Skill Matchmaking, but a player's Skill quickly adjusts to their actual performance and stops it.
  • Tanking, where a high skill player loses a number of games in a row so they can de-rank and play against low skill players. Currently, it takes many losses to actually convince the game that you deserve to be matched with lower skill players, which would not be the case with rank based matchmaking.
• Most modern competitive games match based on skill, and not rank.

• Each player starts with a neutral skill value.
• After the match, each player is ranked based on a wide variety of internal stats (more than shown on the scoreboard), generally looking mostly at stats that correlate with wins.
• Once everyone is ranked, guardians that are out of place (Guardian A played better than Guardian B, but B has a higher skill than A) have their skills tweaked to better match that game.
• Over time, when the system is more confident about a player's skill level, the amount that their skill fluctuates will slow down.
• In order to raise your skill, you must play better than your skill consistently over several matches.

• It can feel too hard to climb to your target Division if you don't get enough wins, especially if you're in the bottom half of players
  • If you're below your target rank and have yet to reach it this season, they're working to add adjustments to combine your rank and skill in matchmaking, so you will climb easier early on (because your rank is lower than your skill).
  • This system will turn off once you are near your target division, and deranking won't turn it back on again.
• It can feel too hard to break out of your target Division once you reach it.
  • The amount of skill that you gain/lose after your skill has been extensively calibrated will increased, so it's easier to increase/decrease skill even after playing a lot.
• The actual skill system is separate to the system that calculates the scoreboard, so currently a rough estimate of skill adjustment is made in the scoreboard so that players can see how their rank will shift after a match.
  • This estimate isn't very accurate, so they're going to work to make it better, and align it with what the actual skill server is doing.
• Sometimes the rank adjustments from match outcomes can feel arbitrary, such as gaining 5 points when you win, or going down in rank even when your skill goes up.
  • This can happen for multiple reasons, most commonly even though you played well, the other team's average skill was well below yours, so your gains were minimal.
  • They're working to show the average skill of the other team on the scoreboard, so you can understand why your rank is changing the way it is.
• If you win, your rank will go up, even if the skill system thinks you should go down (another way you can gain 5 points after a match). Conversely, if you lose, your rank goes down, even if the skill system thinks you should go up (5 point loss matches)
  • They're looking at these values, but ultimately they feel that changing them would feel worse and not better. Increasing the minimum rank changes in a match would lead to scenarios where, despite playing very well, a loss would translate to a bigger drop in rank.
• If a player is well below their target rank, it scales down gains, and scales up losses. The opposite is true when they are well below their target rank.
  • They feel this system is working well right now, given that the graphs show most players are close to their target rank.

• Currently:
  • Having a high competitive rank increases crucible rep across crucible
  • Get one roll of the Rose Hand Cannon a week per character for playing 3 matches
  • Find out where you stand against the rest of the community
  • Earn the Glorious Seal by reaching Platinum, or gild it by reaching Adept, among other required triumphs.
    • Important note: there was a bug at Lightfall launch where players could claim triumphs in this seal that they did not earn. In the mid-season patch, this is being fixed such that player who did not earn the triumphs will have them taken away, but players that have legitimately earned them will be able to reclaim them again.
• In the Mid-Season Patch
  • Shaxx will offer an emblem to you if you are Ascendant 3 or higher
  • PvP emblems will gain a stat tracker that will show off your current Competitive rank.
• At the start of Season 21
  • Rose will be replaced with a new competitive-only sniper rifle
    • They plan on adding a new weapon to comp every season.
    • Rose will be available again in a later season, but unavailable during Season 21
  • Shaxx will have a new Season 21 Ascendant emblem.
    • They're replacing the earnable ascendant emblem every season
  • Shaxx will have a transmat effect for players who reach Silver 3 or above, which will change appearance based on your current competitive division (see it in action here)

• Mid-Season
  • The overshield that you get for a couple seconds when respawning was tuned for Destiny 2 launch, and hasn't been updated since. The shield values are currently too low, and will be increased from 52 hp to 150 hp
    • The shield you get when being revived is unchanged.
• Future Seasons
  • Spawning
    • Their spawn algorithms haven't kept up with the TTK and pace of the game as it evolves. They are committing to a wide-range spawn retune across many maps and modes such that it gets players into the fight quicker, ideally near a friendly player, but always in a safe spot.
    • To start, they are targeting specific maps (Disjuction, Convergence, and Cathedral of Dusk) and modes (Control, Clash, and Rift).
      • More spawn locations are being added to each map and mode
      • They are investigating disabling backfield spawns for Control and Clash, and adjusting the influence of objectives (zones) and allies so that players spawn nearer to allies and ally objectives.
      • They are investigating alternative objective locations to make gameplay more central.
      • For 3v3 and non-objective modes, they are looking to make laned maps feel more "round" by keeping objectives and spawns more central. They are even experimenting blocking off portions of the map with barriers.
      • Once laned maps are in a better place, they will work through other maps and modes.
  • Lobby Balancing
    • Now that they've seen how SBMM will affect PvP in Quickplay and Competitive, they are shifting attention to lobby balancing. Currently, lobby balancing tries to make matches fair by making the average skill of both teams be a similar as possible. This ends up with situations where the highest skill players are matched with the lowest skill players, and they fight all the medium skill players.
    • They are investigating a new style where team compositions feel more even in terms of skill allocations, while keeping the balance of teams as close as possible.

• Rift isn't super friendly to solo, PvE, or new players, which isn't ideal for an Iron Banner mode. There will be an Iron Banner specific twist to address these issues in a future season, as well other twists being created.
• They're looking at ways to decouple the "Earn a Daily Reputation Bonus" system from the "Complete matches as specific subclasses to earn a pinnacle" challenge, but don't have a solution yet.

• In Season 20, FTMM was enabled for Trials.
• They also plan on using Trials Labs to gather feedback on a new matchmaking system that will remove the flawless pool and ticket-based matchmaking for a new behavior.
• The goals:
  • Allow players to play with friends without worrying about who has gone flawless
  • Remove the motivation to reset cards in order to farm or carry
  • Better protect players who are struggling to find success in the playlist
• There are two constantly active "soft pools" (matchmaking will initially prefer to keep players in their pool, but will pull from the other if necessary, will not be as dramatic as the flawless pool)
  • Challenger Pool, where players who play Trials week-to-week and want to play Flawless will play.
    • Should be an experience more like the original iteration of Trials in D1, without the focus on making games more difficult as the card progress / after being flawless
    • This pool matches solely on connection, there is no matchmaking based on tickets, wins, or skill.
    • Players who have any card with no losses (including after a reset), or those playing with someone in their fireteam that has a card with no losses, will be in this pool
  • Practice Pool, for new or less experienced players (players who aren't ready to give going Flawless a shot), lets these players dip their toes into Trials, without being thrown straight into the deep end.
    • The pool matches based on connection and weekly performance (resets weekly)
    • Only players with a flawed card, or those on their first game of the week can play in this pool (everyone must either have a flawed card or be on their first game)
    • This pool has stomp protection, reducing your weekly performance rating if you consistently are losing
    • Also has farming protection, where it detects if a player is potentially trying to farm the practice pool, and will slow down matchmaking times, and eventually lock them into the challenger pool for the rest of the weekend, regardless of card state.
• These changes will be tested this season, with the intent to push it live in Season 21.
• Later this season, there will be a more in-depth discussion about Trials, including additional changes coming in Season 21:
  • Passage of Wealth will reward extra Trials ranks on every win, increasing as your major rank increases
  • Passage of Mercy will forgive two losses if you haven't been flawless for the weekend, reverts to one loss when you have.
  • Changes to gilded flawless requirements, bringing it more in line with glorious, will require more dedication and individual/team skill in Trials
  • A new flawless emblem, earnable for going Flawless without ever being down in round wins.
  • Experimenting with changing Trials' game mode to Zone Capture Elimination (renamed Dominion)
  • A new introductory quest to better onboard new players

• Countdown will be in the Crucible labs this season with multiple variants. The base mode is the one used in Trials of the Nine, at Destiny 2 launch. There are several Cabal Charges on the map, once a charge is armed by the attackers the defenders have 35 seconds to defuse it before it detonates. If a whole team is eliminated (and the bomb is defused, if active), then it also counts as a win
• Countdown Classic
  • Attackers: Arm one of two sites and defend until it detonates, or eliminate all defenders, to win the round.
  • Defenders: Preventing the attackers from arming, defusing an active charge, or eliminating the other team (if no charges are armed) to win the round.
  • First to 4 wins, revive tokens enabled, revives cost 1 token, 1 token per round, each way of winning a round gets 1 point.
  • Maps: The Fortress, Radiant Cliffs, Pacifica, Midtown, Eternity, Meltdown (Season 21)
• Countdown Respawn
  • Same as Classic, but players can respawn after 7 seconds. Revive Tokens are still enabled and can be spent to immediately revive an ally.
• Countdown Rush
  • Same respawn and resurrection rules as Countdown Respawn, but with multiple Cabal Charges per round. When one charge is defused or detonated, the other charge opens up to activation.
  • Once both charges are out of play, the round ends and sides swap.
  • Earn points by detonation, defusal, and defense. Defense earns points based on sites defended, if both charges are up they get 2 points.
  • Same maps as previous modes.
• One variant of Countdown will be available a week starting Week 5.
• Checkmate
  • Bungie wants to explore new modes that branch out from the current Destiny sandbox. Checkmate is a new set of rules intended to create a slower-paced experience where players have more opportunity to react to encounters, and where power spikes are earned instead of guaranteed.
  • There will be changes to how special ammo is acquired, ability recharge rates, weapon TTK, and player health settings.
  • The mode is a significant shift from the core experience, so Season 22 is the target timeframe.
  • Checkmate is built as a modifier, so it will be able to be applied to other standard modes, to give them as much flexibility as possible.
  • They hope to convert other modes to modifiers over time.
• There will be more new and returning modes this year as well, but that's all the previews for now.

Did we miss anything? Let us known on Twitter or in the #tips channel on our Discord.

Sources:

• Lightfall Crucible Update