One of my biggest frustrations with Wizards of the Coast versions of D&D is the time it takes to run a combat encounter. Fifth edition trimmed some of the issues with 3e, so many bonuses to walk through every attack, and 4e, so many off-turn actions added to the combat round. Still, in our sporadic Monday night game, a couple of smallish combats are all we can squeeze into one session and do any social or exploration play at all. I've heard this same complaint from other old, grumpy gamers, like myself; but, interestingly (to me anyway), in a fairly brief discussion in a different forum, a friend of mine who isn't much older than the release of 3.0 wrote the following:
"You hit the nail on the head regarding 5e combat being slow and frustrating. I think I've isolated the key sources of why that's the case.
1: initiative - whenever it isn't someone's turn, they check out of the game. This system also discourages teamwork as it focuses on one person at a time.
2: action bloat - there is no restriction on how many special powers you're allowed to use in a turn, so most players try to maximize the number of things they can do in their turn. This slows the game down exponentially, and also contributes to the teamwork problem as players spend more time staring at their sheets while disregarding the rest of the table
3: hp bloat - this will be the controversial one, but ascending hit dice is one of the worst ideas ever conceived for the hobby. The game offers a wealth of tactical options but then punishes any choice that doesn't focus on big damage numbers. Plus, every enemy having a large hp pool makes combat feel less like a fight and more like chopping down a tree. This also leads to players checking out mentally as battles slog on."
Now Owen has played a lot of other games and I know he's played some older versions and OSR/retro-clones so his opinion is probably shaded by those experiences as mine is. Still, I think he has masterfully encapsulated the major issues with combat speed in 5e. I'm going to dig into each one and then add one more factor of my own.
Initiative, and its related system Surprise, at first glance, seems pretty slick and it is pretty easy to explain to even new players. Basing it on the d20 mechanic and unifying it into the check system gives it a built-in set of levers for character customization. Abilities, feats, and bonuses from Inspiration of various spells can all be used to game the initiative roll. Here's the deal though, Initiative, like Hit Points, is a complete game concept. A pure holdover from D&D's war game roots. It is pretty important to decide who goes first but when you add all the levers and possible bonuses and the need to step all the players and the NPC/monster roles and put them all in order, it becomes a time sink and the combat hasn't even really started yet. As Owen rightly stated, there is an emergent problem with each player's initiative being unique and bulky to arrive at, it gets in the way of character coordination and diminishes player engagement.
I am going to use 1e's initiative system. A d6 for the party and a d6 for the NPCs/monsters. The high roll goes first. More importantly, no ability scores modify the initiative roll, except in one corner case. On the player's turn, play will go around the table but if a player wants to wait for someone else to go then it's simple enough to adjust the order the party goes for that round. Initiative is re-rolled every round which gives the added excitement of making the order of play a little more dynamic which I think helps keep the players engaged and reduces the feel of repetitiveness that arises from the 5e system of rolling once at the start of combat and then walking the checklist in order however many times it takes to complete the encounter.
The corner case is when initiative is tied, which will be 1/6th of the time, characters who start the round in melee range with an opponent will use their Dexterity modifier as a tie-breaker. If the modifiers are the same, then actions will be simultaneous. For those not starting the round in melee range, there is no tie-breaker their actions will be simultaneous. Every once in a while this creates a "death blow" situation where a character or monster is rolling an attack as they are taking a mortal wound.
AD&D has some special case handling of initiative. I am still uncertain if I will stick to all of them. The most common are mid to high-level fighters with multiple attacks per round. AD&D has those fighters always getting the first attack, then the opponent attacking with the fighter getting his additional attacks at the end. This rule essentially sets aside the initiative role for those combatants. It is also somewhat inconsistent in that a monster with a multiple-attack routine, say claw, claw, bite, always makes all of those attacks as one action. The inconsistency slightly favors players so I am sure it was a deliberate and subtle way to increase fighter power by guaranteeing their first attack would go before an onslaught, but it also keeps a fighter from winning initiative and delivering his attack combo before his opponent can respond.
I think in this instance I am going to follow the 5e concept (and the 1e monster concept) and consider all of a character's attacks as a single combat action. It's simpler and creates less head-scratching at the table. Speed of handling is still one of my primary concerns.
The other fairly common special case is combatants under the effect of magical haste or slow. I'm going to slightly depart from AD&D here as well. A hasted character will get their first attack before any non-hasted combatants but their second attack will come on their side's initiative, instead of always coming last which is the AD&D method. Slowed combatants always go last in the round, in initiative order if there are slowed combatants on both sides.
Action bloat and the action economy, here I think Owen miss-states the issue, slightly, while touching on the gist. There is a limit on the special powers a player can use in a round, but the three-action structure, especially the bonus action, does encourage players to scour their character sheets so as not to "waste" their round. The class-build game can create a very busy set of powers, and for casters, spells add to the list of options. With a lot of the options allowing you to swap in powers in place of actions, it becomes a complex decision loop for most players in each round of combat. This creates a classic analysis paralysis situation. Even for experienced players who don't hesitate during their turn, it takes some time to get through the action resolution, especially when you start throwing in all the ways other players can affect an ally's turn with bonus dice and reactions.
I admit this adds a lot (and I mean a lot) of tactical depth to combats. This tactical depth adds a lot (and I mean a lot) of handling time to a round of combat. When it takes a few minutes, each, for all of the other combatants to take their turns, players get distracted. The tension and flow start to suffer and the distraction tends to add to the time when players have to be reminded that it is their turn.
Again, AD&D has a quick-paced action economy in a round. In essence, you get one action. You can move or attack or cast a spell or (every fighter's favorite) charge, where you can move and then attack if you started the round outside melee range. AD&D also was very generous in how characters can interact with stuff, with only one action to resolve it can afford to be. So a character can quaff a potion, switch weapons, use a wand, or open a door while performing their action.
The action economy was heavily house-ruled at most tables, at least to the extent that a lot of tables allowed a move and an attack for unengaged combatants without requiring a charge. Later editions have added a lot of rules that amounted to codifying common house rules. Understandable, but in this case, it is pretty easy to look back over the history of the game and see how this common house rule evolved into the action bloat of 5e. I'm going back to the one action a round.
This also lets me re-introduce another concept that has been gutted because of the time to handle during combat, companions and pets/summons that do not use a player's action. I think I can hear faint cheering from Druid and Ranger players.
HP bloat is a great way to characterize monster design in 5e. The designers created a very elaborate Chinese puzzle box of character powers, actions, and spells, greatly increasing the damage output of a player party. With initiative being manipulable and damage dice being stackable, if monsters don't have huge honking stacks of hit points or punitively high defenses it is very possible for a party to go first and nova blast down an encounter before the monsters get to act at all.
Occasionally, this is fun, and the players certainly feel awesome when it happens, but it grows old quickly when all sense of challenge disappears. So monsters have huge bags of hit points and do large amounts of damage because even with their huge bags of hit points they won't live long so they have to seem a creditable threat while they last. This is the best possible outcome for a given monster, too many of the monsters miss even this ideal. A good many of them have defenses just a little bit too good and a huge bag of hit points and then the party is stuck chopping down Owen's tree.
This idea of monster design took away a classic encounter type, the party being heavily outnumbered but able, through smart play, to overcome. Slaying a dragon is an awesome feeling, but so is wading through a cave full of hobgoblin warriors and defeating 3 or 4 to 1 odds (or more). This issue isn't a combat rules issue. Class design and monster design play a bigger role. So do the magic system and spell designs. Even equipment and magic item design play a part. I will say this before moving on, lair actions and legendary actions in 5e monster design are a very cool addition to the game. They will be included in the modified monster designs.
Here's my last factor before I close this post. How in the world is adding so many ways to add additional die rolls as a bonus to checks and damage rolls an improvement over adding a bunch of static bonuses?
The design team came up with a very slick idea with the advantage/disadvantage system to take out a bunch of finicky bonus fishing and weird chasing of circumstance bonuses and neatly wrapped them into an elegant construct. They rightly touted how this speeds up play and reduces the amount of mental arithmetic required for each attack or use of a skill.
Then they added back in a broad array of bonus dice from spells and class powers that bypass the advantage/disadvantage rule and reintroduce not just bonuses, but variable bonuses, so you still have to do the extra math, but only after you roll some extra dice. It is even more annoying because some of them happen on other players' or monsters' turns as reactions. I'm looking at you, Silvery Barbs, you bastard spell. So now I have a reaction spell that disadvantages an opponent, but doesn't use the elegant disadvantage rule, oh no, it makes the opponent stop, roll an extra die, and subtract that roll from the already calculated result; and it's a fucking cantrip so it can be used on every single combat turn, it doesn't even burn an action of any kind on the casters round, its a reaction to an opponent attacking.
There are a whole series of these kinds of design decisions. Each one looked at by itself seems like a neat idea, but the net result of all of them is slowing play and adding complexity and bloat in combat handling. The handing out of extra dice to be used as bonuses or penalties to future rolls will not make the cut. Most powers and spells that are simply about adding numerical bonuses (other than direct damage) will not make the cut either. There is very little of that in classic D&D and the introduction of them by Monte Cook wanting to force Rolemaster bullshit into D&D during the 3.0 design has been a bane to modern D&D.