None of the Above

In Halo’s online matchmaking the process of voting is an intregral part of the experience. If players want to play a particular gametype or map, voting is the way to go about it. Across the various online communities you’ll find colourful debates about which map is better for what gametypes, which weapon is best for a particular situation, complaints about quitters and stories of glory and defeat. The voting process goes largely ignored but, in truth, it’s the key to a great online experience.

In Halo 3 players were presented with a single combination of map and gametype and had the option to veto this selection. Once a majority of players decided to veto the current map and gametype, the second combination was automatically loaded; players had no further influence over what game they would play. This second combination was automatically loaded and the game would begin as players would have no further influence over the decision. Halo 3‘s online service was split up into two main sections, ranked and unranked (and a third for special playlists such as Double Experience weekends, 7 on 7th, etc.). Each section contained a list of modes, such as Slayer, Team Slayer, Objective and more. Each playlist had it’s own hidden weightings for certain maps and gametypes so that a particular map or a particular gametype would show up to players a certain percentage of the time. Additionally there were three filters players could layer on top of any playlist – “Good Connection”, “Language” and “Skill” to further limit the pool of potential players.

Playlists were determined by the playlist management. Regular updates adjusted how often or not a gametype or map would appear, corrected glitches and exploits, added new gametypes and maps and so on. Bungie claimed that these playlist updates, communicated to the players via a change list, were based off veto data collected by the playlist management. Some outspoken members of various online communities claimed that this data was either largely ignored or not used at all. No information about the collection of veto data has ever been released.

The veto process for Halo 3 wasn’t particularly useful for players as often they found they had to repeatedly veto certain maps and gametypes for many months to enact changes in the actual playlist structure. Sometimes this feedback process was so slow, it may as well not have existed. The veto process was also a matter of blind faith for the players; given that the map and gametype combination that loaded after a veto was incontrovertible, some players would accept what the game selected first rather than risk being forced to play a worse combination.

When playlists were changed, as part of regularly scheduled updates, we were given a list of the changes but usually no clear explanation of why those changes were being made. It seemed as if communication between the developers and players was largely a one way street. The community didn’t know how or why the playlists were managed the way they were. It could be said that there was no real choice and players had to rely entirely on the discretion of the playlist management. This certainly contributed to some of the friction that surfaced between the playlist management and the vocal online communities during Halo 3‘s tenure.

With Halo: Reach, the overal playlists and voting process has been changed dramaticly and further options have been introduced into the system. The overall playlists are split into three main sections, one for Arena hoppers, one for the majority of regular game modes and a final one for playlists dedicated to Firefight and Campaign matchmaking. Before a player jumps into any playlists, there is a “Psych Profile” that defines additional filters for match-made players. The profile provides various selectable tolerences of “Chatiness”, “Motivation”, “Teamwork” and “Tone”. Although these settings have the lowest priority in determing which players one encounters, they add an additional layer to the overall experience. At this point in time it’s been difficult to evaluate the actual effect, per person, these profile effects have had.

The actual process of playing a map in Reach has changed from Halo 3. Before the map loads the game presents three map and gametype combinations and a “none of the above” option. Each player has a vote to use, and at the end of the countdown the most-picked option wins. If “none of the above” is the chosen option, the game presents an additional list of options to chose from. The default options loads if no players cast a vote. Certain playlists modify the voting process by limiting the amount of choices available, hiding player votes once cast, or removing the option to choose altogether.

Not only does Reach give the player much more control over the specific set up over the games they will play than Halo 3 but it also provides much more concise data about player decisions, via voting data, than ever before. What we should see is the playlists in Reach being more reactive to the player base than Halo 3 – resulting in more agressive playlists updates. This has happened.

Reach gives the player control over what they want to play and addresses one of the biggest criticisms of Halo 3‘s veto system. This control assists the player-base in forming their own unique experiences and grants them more control of how they spend their time playing Halo online.

In light of Reach‘s enhanced suite of multiplayer features, one would think it’s perfect as it is. I’d argue that it is easily one of the best, if not, the best solution out there but there’s still room to manoeuvre in this space. There’s a greater amount of control over the game setup options, there’s a proven ability that players can have their votes hidden, so why not extend that into the greater population of playlists? Players are reactive. During the voting countdown votes shift between the options. Occasionally some players wait until the last moment to shift their vote from one option the next, some times players will vote for the opposite of whatever option is currently in the lead and other times votes are dictated by the player in the lobby that can shout the loudest. Whilst some issues are out of the hands of any game developer, we know that the voting process is open to a certain amount of control at the discretion of the playlist operators. Would hiding the votes of players until the voting process is over solve all of the issues and annoyances that can crop up? Certainly not, but it can ensure that voting is less reactive. If players aren’t responding directly to the choices of other players there’s the hope that they will vote for something they actually want to play rather than the option that goes against what other folks want.

There’s something I’d like to see in matchmaking that borrows elements from the existing feature set and combines those elements into an additional filter or search criteria for matching players and it is something that maybe I’d like to see in a future Halo game as it could easily be built on top of what’s already there without taking anything away. Map Preference.

Map Preference would probably be considered an additional filter in the “Psych Profile” area. However I am uncomfortable putting into this category because the results of the profile’s introduction in matchmaking are completely unknown to me. Ideally the feature needs to have a meaningful long term impact on the players. What Map Preference would do, in theory, is track the games a players finishes on each map and in each individual playlist. It would track voting data to see how often a player actually plays on a map and gametype combination that they voted for in a each instance. This would be a “success” rating, maybe a percentage based statistic.

For example, a player has 300 games in Team Slayer. Out of those 300 games, the player played on the gametype and map he voted for in 160 of those games. The “success” rating in this case would be around 53%. Would this be good for the player’s experience? Is it fun to play on the map and gametype you want only 53% of the time?

The purpose of the Map Preference would be to increase a player’s “success” rating by either overhead playlist chnages or personalisied map selections in each playlist for the player. By matching players with similar preferences for specific maps and gametypes you descrease the chances of conflict in the voting process and speed up the time it takes for players to get into an actual multiplayer game. In this fashion, players would be more likely to play the maps or gametypes they want (or both simutaneously) and thus increase the “success” rating – players would be playing the game types and maps they want to play more often.

The obvious drawback is that any additional filter in place to match players up will inevitable decrease the potential pool of players that can be matched. In smaller games this would be an issue, but Halo has a massive active player base and a consistently healthy player-base. The minor impact this additional feature presents would be, I think, outstripped by it’s success in proving players with a better overall, long term experience. Players like to play solo Slayer but not Objective could easily co-exist in the same playlist as a Slayer player that likes the occasional Oddball game. Players would get more of what they want and what they like.

Such a feature would also give us more statistical information about player behaviour. You could easily see what maps or gametypes an individual player votes for the most or least in a particular playlist or overall. You could see the maps or gametypes the players plays on most often, which maps and types a player is most likely to lose or win on when playing. This information could then be fed back into the player’s online identity by making their favourite or best maps and gametypes part of the player’s public portfolio.

Map Preferences is also a potential tool that I would like to see used against an age old issue with online games; quitting. Quitters has been a nusiance since the bith of online gaming and their presence is in no way exclusive to the Halo series. Quitters have their own reasons for doing what they do from the innocent folks that have their attention drawn away by necessity or the bored tweens that sulk around and quit just to annoy other players. One of the main contributors to quitting, particularly in Halo, is players not getting what they want. If teams are unfairly balanced, the loosing team is likely to start quitting out and they don’t want to face a game where they have no chance to win. If players dislike a certain map or gametype, they do sometimes quit out around the beginning of the match in protest. Map Preference wouldn’t have a significant impact on quitters but I believe it would have impact – a small but measurable one. Giving folks what they want is one small step in placating negative actions.

Map Preference won’t ever see the light of day. It’s an example of what I think is still possible within the existing framework, an additional feature that would take away from the existing ecosystem. It’s another use for the hundreds of thousands of votes cast each day by Halo players. Our votes are being used to influence decisions but there exists more untapped potential.

Voting is about representation. When we get asked to vote we want out vote to mean something or have an impact in the result. Like in a real life, a single vote can change the an outcome but with Halo voting has a more immediate impact. If you want that Slayer DMR game on Reflection you had better vote for it least lose out to one of the other choices.

With Reach, we now have the playlist management reacting quickly to player behaviour and making swift, drastic changes to the playlist environment over a very short period of time. Clearly they are able to analyse the data being feed to them behind the mysterious curtain and enact the changes that they think the player base wants. Additionally Reach itself is much more equipped to facilitate playlist changes and updates. Additional information can be displayed on player’s HUD during a game, the voting process can be tweaked and glitches and exploits patched and corrected. Despite these substantial improvements and refinements, the player-base is still haunted by relics of the past. With Reach’s copious amounts of playlist changes in such a short amount of time, I think it’s entirely reasonable to have a much more in-depth explanation about the entire process. The player-base wants to know the decisions behind the bullet points. If we can understand the reasoning behind the decisions then we would gain an opportunity to engage in a fair and equal dialogue about the changes, not just amongst other members of the community but with the playlist management itself. When decisions are made, the players are affected the most.

Slightly Live

Addendum

Below are the playlist changes for the first two months of Halo 3 and Halo: Reach.

Halo 3 Playlist Changes From Launch Until Two Months After

  • Social Doubles removed from social matchmaking.
  • Rocket Race added to the Social hopper – 8 players, join in parties of up to 8 players.
  • Multi Team changes:
  • Hammerzeit game variant has been added
  • The frequency of VIP gametypes has been reduced
  • Swords gametypes now be to 25 kills, rather than 50
  • Lone Wolves changes:
  • Player count was increased to 6 players
  • Swords gametype was added
  • Oddball games on Snowbound and Epitaph were reduced considerably
  • Team Slayer added Guardian and Valhalla to the playlist.
  • Social Skirmish flag gametypes appear more regularly.
  • Rumble Pit added both Construct and Isolation to several gametypes.
  • Big Team Battle was made 16-players.
  • In the Team Slayer playlist, the appearance of Shotty Snipers was greatly reduced.
  • When players vetoed Shotty Snipers, the game that results from the veto will not be Shotty Snipers.
  • In one-sided VIP matches, the attacking team now got a waypoint for the VIP they are hunting.
  • In Lone Wolves, Slayer gametypes were made to appear more frequently.

Halo: Reach Playlist Changes From Launch Until Two Months After

  • Team Snipers playlist added
  • Living Dead (Infection) playlist added
  • Infection, Safe Havens, Race, Rally, Headhunter Pro, and all Classic gametypes removed from Rumble Pit playlist
  • All Snipers gametypes removed from Team Slayer playlist
  • 1 Flag Classic and Headhunter Pro removed from Team Objective playlist
  • Multiflag Classic, 1 Flag CTF, 1 Flag Pro, and 1 Bomb Assault will no longer be playable on Hemorrhage in the 8 player Team Objective playlists due to the size of the map.
  • Oddball and Hot Potato will no longer be played on Boardwalk in Team Objective playlist
  • SWAT, all classic gametypes, and The Cage map will no longer appear in Multi Team playlist
  • Rocket Race will no longer appear only in slot 3 of map voting
  • Multiflag removed from Boneyard and Spire in Big Team Battle playlist
  • Invasion playlist now has 3 voting options
  • Friendly Fire removed from Firefight
  • Firefight playlist gametypes will be a full set rather than just 1 round.
  • Crash Site removed from Score Attack
  • Mythic Score Attack added
  • Other weapons can no longer be picked up in Sniper Attack
  • Glitch on Zealot that would allow you to shoot from space and back down into the map has been fixed
  • Softkill zones added to top of sniper buildings on Asylum
  • Respawn zones reworked for slayer on Boneyard
  • Respawn zones reworked on Hemorrhage
  • Warthog respawn timer changed to 120 seconds on Hemorrhage
  • Respawn timer in CTF is now 10 seconds
  • Classic CTF gametypes now require the flag at base to score, and flag returns happen with just a touch
  • Flag no longer required to be at base to score in non-classic CTF gametypes
  • Flag return time changed to 10 seconds (with 30 second reset) on small maps for 1 Flag gametypes
  • Flag return time changed to 30 seconds (with 45 second reset) on all other CTF gametypes
  • Respawn timer in Assault is now 10 or 15 seconds depending on the size of the map

Rumble Pit

  • Added Crazy King
  • Added Juggernaut
  • Removed Oddball on Asylum

Team Slayer

  • Removed SWAT
  • Removed Classic
  • Weights adjusted based on voting data

Team SWAT (New!)

  • Based on existing Team Slayer SWAT offering
  • Removed Hemorrhage
  • Replaced Zealot with Arena Zealot map variant

Team Objective

  • Added Crazy King
  • Added several gametypes on Powerhouse (CTF, Assault, Stockpile, Crazy King, 3 Plots)
  • Removed heavy vehicles from all 4v4 Team Objective gametypes

Multi Team

  • Added Rocket Race

Big Team Battle

  • Added Crazy King
  • Removed SWAT
  • Removed Snipers from Boneyard
  • Weights adjusted based on voting data

Doubles Arena

  • Removed Slayer Pro for Season 2
  • Removed Boardwalk
  • Replaced Zealot with Arena Zealot map variant

Team Arena

  • Removed Slayer Pro for Season 2
  • Replaced Zealot with Arena Zealot map variant

MAPS

Arena Zealot (New!)

  • Added a softkill zone to the space area
  • Modified initial spawns so enemies cannot see each other

Zealot

  • Objects can no longer be dropped into the boot_base

Asylum

  • Removed multiple duplicate Headhunter score zones

Boneyard

  • Changed the base object for territories from a hill marker to a flag stand
  • Objects can no longer come to rest on the out of bounds cliff behind the scaffolding where the Invasion core is captured
  • Players can no longer sit in the out of bounds area under the ship scaffolding

Pinnacle

  • Health Packs should not respawn after 15 seconds

Hemorrhage

  • Landmine removed from Red team’s west cliff
  • Respawn time for all Mongoose vehicles is now 45 seconds
  • Scorpions replaced with Wraiths (4 minute respawn)

Paradiso

  • Respawn timer for Concussion Rifle at Red base is now 30 seconds to match Concussion Rifle at Blue base

Reflection

  • Changed the base object for territories from a hill marker to a flag stand

Boardwalk

  • Changed the base object for territories from a hill marker to a flag stand

Powerhouse

  • Respawn timer for all Needler weapons is now 45 seconds

Global Changes

  • Slayer DMR now has motion tracker enabled
  • Evade has replaced Hologram for all gametypes in which players can hold objects (CTF, Assault, Stockpile, Oddball)
  • Evade has replaced Dropshield for gametypes which create concentrated gameplay in specific areas (Crazy King, Territories, Oddball)
  • All Classic gametypes should now include Sprint as default equipment

King of the Hill

  • Teams are now always able to correctly score points
  • Players standing in the hill no longer receive extra points upon returning from a host migration

Rocket Race

  • Players now attached to vehicles after being on foot for 10 seconds
  • Flipped vehicles will be detected to prevent players from attaching to a non-upright vehicle
  • Added functionality for tracking and reporting rotations
  • Fountain of Mongeese should no longer be possible (don’t ask)
  • Players are no longer able to jack another teams Mongoose
  • Players are no longer forced into a specific role of driver or gunner
  • All players start with a Rocket Launcher
  • Players move at 50% speed, do significantly reduced damage and are marked with a nav icon
  • Players on foot are now able to score points in Rocket Race

Juggernaut

  • The Juggernaut no longer sees other players as allies upon returning from a host migration
  • New clients after a host migration will no longer lose their Juggernaut status

Territories

  • All teams now have separate capture timers for each territory
  • Locked territories will no longer prevent players from using equipment
  • The flag in the territory is now attached if the territory is a flag stand object
  • Added a HUD_Widget to display contested status to players inside a territory
  • The progress bar for a territory will now show as full when that territory is being contested (capture progress is saved and will appear again once the territory is no longer contested)
  • Random flag clothes will no longer disappear after a round transition

Oddball

  • Bombs in Hot Potato now properly report carry time
  • Multiple hot potato bombs will no longer spawn in the same location
  • The carry time reported stat will now always be correct after a host migration has occurred
  • Players carrying the oddball during a host migration will no longer receive extra points

Stockpile

  • All stockpile flags are now attached to their spawn location if it is a flag stand

Invasion

  • The game score is now set per frame tick based off of the current phase
  • A failsafe game end timer has been added to catch any possible bad cases of Invasion games not ending properly

Invasion: Boneyard

  • The core will now reset whenever it is thrown through a shield door into a spawning location in Phase 3

Invasion (Assault) and Assault

  • Added sudden death to Invasion (Assault) and Assault gametypes

Race

  • Players who die before the first checkpoint will no longer spawn in a random location on the map
  • 2 Comments

    1. November 17, 2010
      Reply

      Just thought this was worded a little funny, seemed to repeat the automatic loading portion:

      Once a majority of players decided to veto the current map and gametype, the second combination was automatically loaded; players had no further influence over what game they would play. This second combination was automatically loaded and the game would begin as players would have no further influence over the decision.

      Feel free to delete my comment if you choose to edit the paragraph.

      Overall, an interesting read 🙂 Seems Bungie is more proactive in making changes in Reach!

    2. spartan_111
      December 4, 2010
      Reply

      FFA is my only goal …

    Leave a Reply

    Your email address will not be published. Required fields are marked *