Fast movement, weaker aim assist, with a deep and interesting metagame revolving around control and knowledge of the map and strong situational awareness rewards. Balance for the hardcore, not to introduce new players - a good high-level game will bring players of all levels, a game sloped for onboarding newcomers will drive away the players at the top as they reach the mechanical barriers. It results in a game that is either mindless and boring, or wildly imbalanced. Get players invested by rewarding team wins, encourage a competitive community to rise up around well-balanced, thought-out maps and gametypes. Let players choose their rifle if they want as long as those rifles are balanced accordingly (I'm thinking more BR vs. Carbine, not BR vs. DMR just so you get the idea), but plasma grenades, armor abilities, and semi-power weapons like the Plasma Pistol and Boltshot should never be customizable and available to the player on every spawn via a loadout system. Your core competitive game shouldn't operate on the same principles as Regicide did. On that note, disable Join In Progress across the board and treat each premature quit as a harsh black mark against the offending player. This dramatically increases the importance of each individual match. You want each game to be a full experience, not a series of throwaway moments in between lobbies. Players get invested in the outcome of a match by playing the game
when the match starts all the way to the end. Take a page from DOTA's book.
Return the investment system to cosmetic-only. It worked for Reach, it works for DOTA, it's the right move. Completely rethink your challenge and commendation systems, because they were prettttty terrible last time around. Follow the rules of good online rewards:
-Be consistent with your expectations and rewards.
-Be reasonable with your challenge parameters.
-Don’t encourage your players to be assholes.
-Avoid repetition of specific challenges.
-Make the rewards worth working for.
Make the maps free (even if there's a few months delay), don't split the community again. Find another revenue source if you want, just don't be gross about it.
Reconsider every single gametype decision made with Halo 4 - they are the single biggest reason I lost interest in playing. Look to Halo 2 and Reach for inspiration on your core objective gametype settings. Return proximity contesting to CTF and Assault (also, don't cut Assault this time only to bring back a half-baked version that contains none of its nuance a year down the line - Assault needs to be a hybrid of the arming and planting mechanics of 2 and Reach, leave 3's awful variant to rot). Return the ability for stealth runs to objective carriers by ditching the waypoint. If a player is spotted, it should be due to their action or the action of the opponent, not because the game just decided it's time for everybody to know where that player is now. Oddball throwing killed that gametype in the competitive scene. It was an interesting idea that didn't work out very well, so figure out where the middle ground is. Dominion is an extraordinarily poor replacement for Three Plots, and isn't even as interesting as the last 6v6 mode Invasion was. Dump it. People want those classic, tense objectives. Don't try to please people who don't like objective modes in the first place by ignoring the people that use those gametypes as their bread and butter.
Be risky with the campaign. That's where you can do some weird stuff and say that you're trying something new, because it's frankly easier to execute on things like loadouts and on-the-fly abilities in a PVE environment - the AI is never going to call bullshit when someone armor locks. Do your open world thing or whatever, as long as the enemies are fun to fight and the encounters are well-designed, you'll probably be okay. Even if it doesn't work out in the long run, there should be a solid, core multiplayer component to fall back on that adheres to the tried and true pillars of Halo design that were simply absent with Halo 4.
More than anything, work on your playlist management. Nobody wanted the Big Team list to be Slayer-only. Nobody wanted paywalled Extraction for months, or a repeat of that with Ricochet. Nobody wanted King and Oddball to be completely unavailable to play for the majority of the game's life because of some asinine "one gametype per playlist" policy that has since been broken over and over again, a policy that resulted in huge numbers of secondary, "impulse buy" gametypes dropping off the face of the earth almost immediately because guess what? Oddball can't sustain a population by itself. Grab bag playlists work when they're managed properly, and yes, that isn't easy, but it's not impossible. When there are core problems, fix them before working on Action Sack. It gives the impression that you just don't give a shit when Fiesta gets into the game before huge playlist issues are addressed. Your sustain team needs a serious increase in their available resources. They need to be agile and respond to issues at an incredible pace, and that takes manpower. We need updates for broken things pushed in a matter of days, not weeks. Gaps in the gametype roster or gametype/map combinations that don't work need to be addressed
quickly - you've already lost me a month down the line. That's just the nature of running an online game.
Figure out what vehicles work in what situations and if they don't, why not? Stop putting the Scorpion into matchmaking. Stop putting the Banshee into matchmaking with its fuel rod intact. The Mantis encourages extremely passive play, ditch it. The same policy goes for weapons. The Forerunner weapons were pretty much universally reviled, rethink every single one (and axe most of them). The Spartan Laser has been a consistently awful piece of design, bring back the Plasma Launcher to fill that role.
Make maps that fit your gametype pool, and decide early on what your player counts are going to be. Maps built for 6v6 are not satisfying in either the 4v4 and 8v8 pools, stop using them as a one-size-fits-all solution. "Functional symmetry" is bullshit, stop saying that to deflect poor map/gametype pairings like Daybreak CTF. Here's what's better: plain ol' symmetric maps for symmetric gametypes. You want to make asymmetric maps, great! Put asymmetric gametypes on them instead of shoving a square peg into a round hole, because that hole ends up being where fun goes to die.
Make a return to elegant player feedback systems. Reach's damage-dependent shield brightening in combination with the shield pop sound effect was a piece of brilliant A/V design, and so were the loud, distinct grunts from player characters that had been hit by a grenade. By comparison, hit markers for both shots and for grenade hits are blunt instruments shouting in your face. That information can be conveyed better. As a whole, the UI needs serious work. There are a ton of oversights like objective instructions overriding player gamertags, meaning you never actually get to know who is carrying the ball because it just shouts ESCORT in your face. It's a serious usability problem.
(I missed out on this thread the first time.)
Also with dedicated servers go crazy with big team. Make it 16v16.
I'm almost certain this was always a design decision, not a technical tradeoff. More players mean fewer hero moments for each player. You want to achieve a middle ground between maximizing each player's impact on the match and achieving that feeling of scale. The fact that discrete units of success in Battlefield are usually on a squad level, not individual players, shows this. As you introduce more and more players, the less you can do as an individual player to affect the outcome of the game.
Assault should go back to Halo 2/3 style.
Assault in these two games was very different, and it was markedly worse in 3.