EVERY Initiative Method??

Not really. But maybe, with your help, we actually can. Let's give this a shot.

Every now and then I've seen a blog post attempting to cover this topic and they always just seem to fall short. "There are two methods: individual and side." Please. Not only are there dozens of ways of doing initiative, but many of them are desperately in need of a proper name at this point.

But exhaustively cataloguing every method is unfeasible. There are so many games that have nearly identical methods but with just a teensy tiny little quirk. "Instead of rolling a d20, you roll a d10." Yeah, whatever. I'm not about to list every single one of those as an individual entry. Some degree of broad categorization is needed.

But you will quickly see that my system of categorization is... feeble. I have unfortunately not been able to divine some grand system of fundamental initiative typology, easily charted along 6 key axes or something like that. You'll come to see why pretty soon, and we'll circle back to that topic in the conclusion. I instead tried to order my categories by vaguely-increasing order of general unfamiliarity / difficulty to explain and understand. We'll see how it goes.

Overview

Oh, right, I should introduce the topic. Uh, initiative? Y'know, "turn order." Whatever method you use to determine the order that actions are declared or resolve or whatever in a complicated gameplay scenario with lots of participants. Usually combat. This is an action game thing, okay? It supports tactical integrity, which is relevant to some games but not others. Most story games don't have any system for initiative because most story games aren't about tactical thinking, for example. If those are the kinds of games you play, and not the kind I'm talking about in this post, you have permission to leave and not write an annoying comment about how your favorite system doesn't even have an initiative method and you don't see the need for such a ridiculous, D&D-centric idea anyway.

I will not be talking about how different games handle surprise mechanics, because I see that as a different topic. However, I should note that in the vast majority of systems I researched for this post, the amount of text spent explaining their surprise mechanics would form the bulk of their text describing their initiative method. Kinda wonder why people seem to overthink surprise so much. But hey, that's a post for another day.

This post is also not about different action systems or types of action economy or whatever. However, many of my examples invariably require talking about those things, simply because some initiative methods are intrinsically entangled with the game's general action structure, whereas others are more modular.

A person or group's preference for initiative method is shaped by all kinds of concerns and priorities. Some people want realism, willing to make it complex and detailed if need be. Others want it to be nice and gamified, looking for ways to incorporate variables tying into character traits, meaningful choices, trade offs, etc. And some just want an option that's smooth to play, trying to prevent miscommunication and loss of momentum.

If I know of an actual game where a method is used, I'll list it. Otherwise, an awful lot of these exist only in the blogosphere and twitter threads. A lot of folks took a look at this post before it went live, but I should specifically thank Ian of Benign Brown Beast, Clayton of Explorers Design, and my own D&D group members for their contributions. I don't want to go through the effort of making this a "living post" but if you leave a comment with an addition, I'll try to find the time to add it to the post.

I'll try to keep my opinions out of it. Mostly.

Turn-Based, Individual

Individual initiative is the most intuitive method for most folks. It's how you play most board games and parlor games. One player at a time. That being said, it's also notoriously cumbersome. It's a lot of info to keep track of, it's easy to accidentally skip someone, you spend forever waiting for your turn to come, and yet it's also hard to plan ahead for your turn since the situation is constantly changing.

What kind of game benefits the most from individual turns? In my opinion, it's the kind where every character has a lot of things going on at once during their turn. Multiple actions of multiple types, lots of options to consider, complications like interruptions and conditional effects and whatnot. In those kinds of games, you need the extra breathing room afforded by, "alright, everyone else shut up. It's your turn now. What do you want to do?"

Randomized

This is not the most basic method in this long, long list... but it is by far the most commonly used, by such a staggering margin you probably cannot even imagine. This is the version used in modern D&D, and therefore hundreds of other games I'm not going to bother listing. Every individual participant in the fight rolls a die or draws a card or something, then they each take a turn from best to worst. Have a tiebreaker ready.

It's more common than not for there to be modifiers to the rolls, so it isn't fully randomized. Fast characters add a bonus, slow characters subtract a penalty. Also, because this process is such a pain in the butt, games using randomized individual initiative usually just roll for turn order once and then stick with that order for the rest of the fight.

This system can be messy to resolve. There are all kinds of techniques people use to try and make it more manageable. I really like my own DM's method, which is to call for results in groups of five. "Did anyone get a result of 20 or higher? … Alright, did anyone get from 15-20? … Alright, anyone from 10-15?" and so on. All the players shut the hell up for a moment and wait for their chunk to be called, so that the GM has the time to write things down in the proper order.

Fixed Order

This is used in 1977 "Holmes Basic" D&D, GURPS, Call of Cthulhu, Fate, Beyond the Wall, and Five Torches Deep. What's the fixed order based on? Usually something like a character's "speed value." Maybe their dexterity or agility or whatever. The 5E DMG calls this the "Initiative Score" method since everyone gains a new fixed stat, like their Strength Score or Wisdom Score.

Character with highest speed goes first, second highest goes second, etc. No dice rolling, no variation. The fast character will always be faster than the slow character. The party's order will always be the same. But there almost definitely needs to be a tiebreaking mechanic of some kind.

Subtype: Arbitrary Fixed Order

AKA "clockwise order." This is used in ICRPG, Shadowdark, and, like, most board games. Just go around the table in a circle. If you're playing online, substitute it for something like "alphabetical order" or whatever else works best. It's technically based on a metagame variable and can be manipulated accordingly, but it's mostly just favored for being really, really simple. A lot of people aren't convinced that initiative order is actually very meaningful, tactically-speaking. In a lot of RPGs, getting to go earlier in the round isn't that much of an advantage, especially after the first round of combat. So why not just accept an arbitrary turn order if it makes your life easier?

Popcorn

This is used in Marvel Heroic Roleplaying, Death in Space, and (I think) Fate Condensed, but it's also a houserule that's been frequently recommended on the internet since maybe forever? Characters take individual turns, and at the end of their turn decide who gets to take the next turn. Bob takes his actions, then when he's done, he says, "alright uhhh Alice, how about you go next?" This nearly always includes the extra restriction of giving every participant at least one turn before anyone can take their next turn.

A bit chaotic, gives players more control, rewards them for planning, but also burdens everyone with remembering everyone who's already gone in the current round. Plus, you don't know when to expect your turn, and it might come before you're ready. It's also pretty much inevitable that the players will simply keep picking each other until they can't anymore, functionally making this option almost identical to regular side initiative. How often will a player say, "actually, I think I'll pass the ball over to the bad guys first"? Still, no bookkeeping and no rolling dice.

Phased Action Monsters

This comes from a recent Patreon post by Mike Mearls, who's still cooking up new design ideas for 5E. The basic idea is that randomization only applies to the players. It's an asymmetric system. Players roll for initiative like normal, but the monster takes their turns at specific initiative spots (e.g. the monster acts on initiative count 20, 15, and 0). Then, you can designate specific actions and moves and whatnot for each of those spots. All of this is written directly into the monster's stat block during your prep, like you're scripting their behavior. This is all essentially a substitute for Legendary Monsters from 5E, but it's a neat way to reduce bookkeeping and add a pinch of structure (that the players can then exploit).

The Daniel Sell Subsection

Because Daniel has a thing for this I guess

Stack Initiative

This comes from Daniel's game Troika!, although you can also read his original blog post describing it right here. It's also used in 17th Century Minimalist. People call it a "chit-extraction" method, where you can use chips or tokens or whatever. Let's go with cards.

Every participant in the fight has some cards that then get shuffled into a deck. Each turn, the GM draws a card from the deck, determining the next person to go. So the order is totally random, but a character with more cards in the deck is more likely to take turns more often. Hypothetically, this should trend towards everyone eventually getting to go as the deck gets lower and lower... however, there's also one "End of the Round" card in the deck. When that card is drawn, the deck is re-set. Remove all the cards of defeated combatants and shuffle everything back together. So there's no guarantee that everyone will get to go at least once in every round. You could get totally bananas luck and draw the End of the Round card after only like 2 or 3 turns have been taken.

Miscellaneous

The rest of these all come from this blog post where Daniel just churns out a ton more methods, arguably pre-empting a number of other innovations described elsewhere in this post. I'll repeat a few unique ones here.

• Euro Game Initiative: Clockwise around the table, except the player who gets the first turn each round is also rotating in clockwise order. So if the first round goes Alice, Bob, Charles, Dana, and Ed, then the second round goes Bob, Charles, Dana, Ed, and Alice, and the third round goes Charles, Dana, Ed, Alice, and Bob, etc.
• Drama Initiative: The player with the least health goes first. Give a small advantage to the most vulnerable character, I guess.
• Auction Initiative: The player willing to take the biggest penalty to all rolls goes first. The GM starts at a number that would make for a very high penalty and begins counting down. Chime in when you're ready to take your turn, subtracting that number from whatever rolls you're about to make.
• Rondel Initiative: It's sort of a racing minigame? Everyone goes around and takes an action, and each different type of action also moves you up a certain number of spaces on a tracker. Then, whoever is in "last place" gets to always take the next turn, moving slowly up the tracker with each new action. Once you've reached the end of the tracker, you can't go again until everyone else has reached it and finished the "race." I dunno, I guess Daniel explains it better than me but I'm not crazy about it anyway.

Turn-Based, Side

Side initiative is usually considered much faster than individual initiative. It invites more teamwork and emphasizes the party as a collective unit. Plus, you don't have to wait long for your turn, and you're less likely to zone out. However, some groups might fall into the trap of over-analyzing their side's play, bogging the game down with discussion.

What about fights with more than two sides? Some of the following methods are ready to accommodate this, but others aren't. A term I see sometimes is to refer to each grouping as a "pod," which can be helpful.

Most of the time, side initiative methods are only partly turn-based. Actions taken by characters within the same side / pod are usually resolved simultaneously, so all the principles governing simultaneous action still come into play.

However, it's also possible to instead make it fully turn-based, and only partly side-based. That is to say, you could still have each member of a pod take individual turns, rather than acting simultaneously as a side. But they just decide amongst themselves what order they'll go. This might be the right call if the system is still crunchy enough to justify giving each player their own turn, but you don't want to bother with every participant making an individual initiative roll and needing to write down like a dozen results.

Back and Forth

This is used in Into the Odd and Cairn 2E. One side goes first, then the other side goes next, then you just keep repeating until the battle ends. One of the most simple methods possible. Maybe there's some kind of rule to determine which side goes first on the initial turn, but that's a pretty minor advantage if the battle goes on for more than one or two rounds.

Randomized

This is used in AD&D 2E, OSRIC, White Box FMAG, White Star, Marvel Super Heroes (but not FASERIP, for some reason?), Maze Rats, Knave, and MÖRK BORG. At the top of the round, roll initiative to determine which side goes first. But then, re-roll every round. Yes, this means there will be times when one side goes twice in a row. This is a feature, not a bug. Adds a bit of chaos, and can really turn the tide.

Additional note: not just for this category, but plenty of others, there's a bizarrely consistent design choice that I think is mostly just bad. That is, having both sides roll-off, and then needing a tiebreaker ready. The vast, vast majority of these methods are a simple 50/50 odds anyway, literally indistinguishable from a coin flip, and yet nearly all of them say "both sides roll 1d6, higher wins, re-roll in case of a tie." And like... I don't know who needs me to explain this to them but in case you aren't aware: you can simplify this by just having one die roll. GM rolls 1d6, 1-3 the baddies go first, 4-6 the party goes first. Ta-da! Modify to taste, obviously. But I just... find it really, really weird that in all my research I only ever saw that once or twice.

Speed Sandwich

This is used in The Black Hack, Mausritter, the GLOG, Mothership (demoted to an optional method in the new version), Grave, Black Sword Hack, The Nightmares Underneath, Cypher System, The Electrum Archive, and Trespasser. At the top of the round, every player character rolls a simple pass-fail initiative check, maybe adding a speed modifier. Those who succeed go in a pod first. Then all the NPCs / monsters go as a pod. Then all the PCs who failed go in a pod last. Thus, nearly all rounds have three phases: Fast / Enemies / Slow.

I like this because I like side initiative and I like a process that's fast, but I also see the value in incorporating character speed as a variable as well as the undesirable swinginess of a full party that always acts together (potentially "ganging up" on your bad guy and curb-stomping him). The GM doesn't have to roll anything or write anything down. The players don't have to note their exact result, only whether they passed or failed (which I also like because that makes it basically the same as every other kind of roll you normally do in an RPG, rather than being a weird exception).

The game Trespasser has a neat variation where you can roll a critical success on your initiative check, allowing you to take a turn during both the fast/opening phase and the slow/closing phase.

Sam of Dreaming Dragonslayer has a cool idea: make your initiative bonus equal to the number of empty item slots you have. Well, sort of. It's like 90% the Speed Sandwich method. We can nudge it there. The point is, it's a cool idea to make the weight of your equipment affect your "speed" without decreasing your movement distance, and gives you a reason to not fully load up like a pack mule all the time.

Guerilla Initiative

This comes from a blog post by Homebrew Homunculus. Their main thing is just re-discovering the Speed Sandwich method described above, but they also talk about a very interesting idea they call "guerilla initiative": whichever side has fewer members always goes first.

The players storm the lair of a solitary dragon? The dragon reacts first. The lair has a dozen kobolds in it? The players go first. The rule communicates something about the world: it really emphasizes the stealth in small numbers and the overhead of organizing as a larger group.

I assume this would be re-determined each round. At the beginning of the fight, the party has the initiative. But kill enough kobolds, and eventually the few survivors will take the initiative, gaining a chance to maybe slip through your fingers.

Tank Initiative

Similar to the previous, this comes from a blog post by Alchemist Nocturne. They also have a whole lot of ideas swishing around in that post, but I just have one interesting thing to extract: initiative goes to the side who has the single participant with the highest max HP. This means that having a beefy fighter in the party is an automatic initiative bonus. It also means that there's an advantage to fighting a big group of weak guys rather than a solo tough guy. Most interesting of all, it incentivizes you to take out the biggest adversary first, so that way your side can gain the initiative in each subsequent round.

The Patrick Stuart Subsection

Because Patrick has a thing for this I guess

Lamp Initiative

This comes from Patrick Stuart's Veins of the Earth, and is thus very tied to dungeoncrawling / exploration in the Underdark. In this method, every light source in the battlefield is an initiative pod. Each pod includes the person holding the lamp (who makes the initiative roll with their own bonus) plus everyone within the lamp's range of light. Patrick says that the lamp-holder chooses the order that their pod's members take their turns, but they're going to talk it over anyway so it may as well be one turn.

Any character that has no light source automatically loses initiative and has to take their turn after every light pod has finished going. This puts Underdark creatures with darkvision at a weird disadvantage, but that just means they'll work that much harder on trying to secure a surprise round.

Lamp initiative could end up just being simple side initiative if the party all clusters together around one lamp and the baddies all cluster around their own lamp. But why not split the party into multiple groups that each have their own lamp?

Physical Initiative

Comes from this blog post. "The Initiative" is a literal physical object somewhere on the battlefield, obvious to everyone. Whichever side currently possesses that object wins initiative each round, up until they lose it. An actual tangible MacGuffin. I assume this was inspired by watching sports, since this sounds to me like it's basically soccer or basketball or something.

Patrick offers a list of weird initiative objects, and even suggests an even more bonkers idea: if the initiative is broken / killed / destroyed (likely accidentally. Be careful with those attacks!) then you switch to not using any initiative method instead, like you're playing a PbtA game I guess.

Query Initiative

Comes from the same blog post as the previous. This method gives the players a tradeoff, where the chance to go first is exchanged for having more information at the start of the fight. This makes the most sense in the context of an ambush or a surprise run-in in the darkness of a dungeon corridor, where you don't have full knowledge of the situation and may not even realize who you're fighting initially.

The enemy has an initiative number. The players can start the fight by asking questions. "What type is it?" "How many are there?" "Are they armed?" "Where are they?" Etc. If they ask a number of questions less than the enemy's initiative number, then the party retains the initiative and will get to take the first turn. But if they go over the enemy's initiative number, then they spend too long sizing up the situation and lose the chance to go first.

It sounds to me like a lot of these questions would inevitably be answered after the first round of the fight anyway, but I can imagine that could make the difference between victory and defeat. It's a half-baked method with some interesting potential, at least. I think this could make an excellent method for submarine combat (or something comparable. Submarine-like spaceship combat, maybe).

Davide of Daimon Games elaborates on these previous two methods and incorporates item slots. Each player has an initiative number equal to their empty item slots. They can go first if this number is greater than the enemy's initiative number, but the player can first spend empty item slots to ask questions. So your speed is a triple tradeoff of how fast the enemy is, how weighed down you are, and how much time you spend analyzing things. ¡Ay, caramba!

Simultaneous

This is often mistakenly referred to as "no initiative system" but that's actually something different. Simultaneous resolution is very much its own family of distinct procedural methods, and it tends to really freak people out who've never encountered it before. Folks like having a firm, authoritative order of operations to rely on. Simultaneous resolution introduces ambiguity, requires GM discretion. Scary stuff! But proponents of simultaneous methods often praise it for being more, well, realistic. Real-life combat isn't turn-based, after all. It also adds a bit of fog of war to your tactics, since you have to pick your actions without knowing exactly what the situation will be when those actions play out.

Written Orders

This is the original standard found in old-school wargaming. Here's a blog post describing it. The basic method goes like this: each round, every participant writes down their actions. They pass their actions over to the referee, who then decides their results. All actions execute simultaneously. The referee might have some rules or even loose guidelines for figuring out how actions should resolve, especially if they conflict with one another. But they might just have to use their best judgment instead.

This is the most straightforward form of simultaneous resolution, but writing down your actions can be a bit of a pain. Everyone needs paper, everyone needs to pause to write, lots of scrap is going to get thrown out. A digital setup makes this easier, as everyone can just type their actions. Works well for play-by-post games.

Turn-Based Declaration

Because writing down actions is cumbersome, maybe just verbally declare them. But because you can't declare simultaneously (no sense in everyone talking over one another), you'll have to take turns declaring, which itself can be gamed. There's an inherent advantage in declaring later, because you can base your answer on what other people said previously. It's like turn-based Rock, Paper, Scissors. Wouldn't be very fair, would it? Thus, there are different ways to account for that.

One approach I've seen is to simply grant the turn-based advantage to the GM. Players declare first, then the GM reacts to this and decides what the NPCs / monsters do, and then it all resolves simultaneously. The GM always has the "advantage," and they're just trusted not to abuse it. Which, I mean, is true of literally everything else the GM does so it's probably fine.

A different approach is to grant the players a slight advantage. In the above Spells and Steel blog post, Charles recommends that the GM give a hint at the monster's actions before the players declare. So not quite the full benefit of getting to pick last, but also not totally blind.

And of course, a third answer is to embrace the tactical potential here and directly make the "declaration advantage" a point of contest. Participants / sides may not be rolling initiative to determine the order that their actions resolve, but they could still roll for the chance to win the declaration initiative. Have each side roll each round, and whoever wins gets to declare their actions last.

Some folks mix this with other methods in this post. I haven't seen this idea explored much but I'd be very interested.

WEG Action Segments

This is used in the D6 system by West End Games. The generic rules don't really dig into this in much detail, but their Star Wars RPG from the 80s lays it all out.

First, the GM decides what the NPCs will be doing while the players declare what they'll be doing. Everyone must describe their actions in detail. In fact, they can declare multiple actions. "I want to run to cover, draw my blaster, and fire at the stormtrooper" is 3 actions. "The stormtrooper is going to shoot at you three times and then reload" is 4 actions.

Then, the round is split into segments for each action. In the first action segment, everyone's first declared action executes simultaneously. In the second action segment, everyone's second declared action executes simultaneously. Continue with as many action segments as needed until everyone has performed all of their declared actions.

What's to stop you from just declaring 50 actions and running around like the Flash? Well, the D6 system uses dice pools for every roll you make. And for every additional action you declare beyond the first, every single action suffers a penalty. So if you would normally roll 4 dice for a single blaster shot, then attempting two shots in one round will force you to roll both with only 3 dice. And if you attempt three shots in one round? All three are rolled with 2 dice. So the more stuff you try to squeeze into one round, the less effective everything will be. Your effort gets spread thin, dividing your attention between a bunch of tasks.

Burning Wheel

Luke Crane's Burning Wheel RPG uses a "written orders" method, but with some specific constraints.

Each round ("exchange") is split into three chunks ("volleys"). Everyone writes down their actions for the whole exchange ahead of time, split across the three volleys. But you only reveal and resolve each volley one at a time. All the actions in the first volley execute simultaneously, then the second volley, then the third. It just provides a wee bit more structure than the basic method, but is still ultimately a pretty freeform, writing-heavy method.

Interestingly, the game still finds a way to incorporate "character speed" as a factor. Your character's Reflexes value determines how many actions you can perform per volley. A character with Reflexes 3 can do 1 action per volley, Reflexes 6 gets 2 actions per volley, Reflexes 4 gets 2 actions in one volley but only 1 action in the other two, etc.

I see this as almost the inverse as the WEG Action Segments method. With that system, the same amount of stuff takes place during each chunk of time, but the amount of time chunks can vary. But with this system, there's always exactly 3 time chunks, but the amount of stuff that takes place during each chunk can vary.

Phased Initiative

By now you've noticed the conspicuous absence of what was possibly the most common initiative method historically. Including in, like, most versions of D&D. How do you write three and a half thousand words listing different initiative methods before finally getting to this one? I dunno, man. I warned you that I'm bad at categorization. I wasn't sure where to place it. Some versions of this method incorporate side-based turns, others do staggered sides, others do simultaneous, etc. So I figured I should explain all of those ideas first.

The best way I can describe this method in general is thusly: the round is divided into set phases that correspond to different types of action. All action within a phase is resolved simultaneously, but the order of phases is fixed. Every game or GM who uses this method has their own specific list of phases in their own order, with their own rules for a gazillion other questions that'll come up.

A classic example is the one found in B/X D&D and BECMI D&D.

1. Each side rolls off to determine who wins initiative. Ties are either rerolled or, if the DM is feeling spicy, they can try some of that unpredictable simultaneous resolution stuff.
2. The winning side resolves their actions in this order:
1. Morale rolls
2. Movement
3. Missile fire
4. Magical spells
5. Melee attacks
3. The losing side then resolves their actions in the same order
4. Go back to step 1, re-rolling each new round.

So it's kind of like randomized side-initiative, but each side has a strict order of operations they'll have to follow.

There's an optional rule for "Pair Combat" where combatants can pair off in a fight and then simply roll initiative against one another. Old School Essentials helpfully provides some extra details, including 1) intentions to either cast magic spells or to attempt a retreat must be declared before initiative is rolled, since those are highly telegraphed to your opponent and could potentially be disrupted, and 2) Characters attacking with slow weapons (two-handed melee or some missile weapons) always act last, as if they lost initiative. See how many little details we can add into this stuff?

Compare this with the method found in Swords & Wizardry:

1. Spellcasters declare spells
2. Roll for initiative, side-based
3. Members of the winning side may either move or fire missiles
4. Members of the losing side may either move or fire missiles
5. Members of the winning side may either make a melee attack or cast a spell
6. Members of the losing side may either make a melee attack or cast a spell
7. Go back to step 1, re-rolling each new round.

Notice that this has two major differences: 1) this method isn't fully side-based, but instead staggers each side's actions back and forth across the phases, and 2) it has blended phases, where you can choose one of several comparable actions. You can imagine creating a wholly new method of phase initiative with whatever combination of these traits you prefer.

When examining these methods, the order of the phases may have some logic to it ("arrows are fast! sword swings are slow!") but it's ultimately pretty arbitrary and contrived. In a 6-second span, a character can move and then swing an axe... but they can't swing an axe and then move? Even if they managed to kill their foe? Would using a big polearm be a bonus or a penalty? According to OSE, they're so unwieldy that they break the phase order entirely and force you to act last, even if you had won initiative. But in Basic Fantasy RPG. they have such long reach that they break the phase order entirely and allow you to attack anyone approaching you before they can close the distance, even if you had lost initiative. Classic Traveller essentially has three phases: 1) movement, 2) attacks, 3) morale. Yup, morale comes last. Why? Because it's all arbitrary.

Since I did the research and took all these notes anyway, I may as well offer a few more examples.

• Runequest keeps it simple. 1) Declare actions, 2) Movement, 3) Everything else. Uh oh, that third one could get messy. Melee, missiles, and spells are resolved in order of "strike rank," a number modified by the character's dexterity, size, weapon length, etc. Sounds like more composite speed design (see below).
• Pendragon is likewise easy. 1) Declare actions, 2) Attacks, 3) Movement. Here's the catch: you can either attack or move, but not both. But attacks are always contests anyway, so you still may win an attack initiated by the other guy.
• Under Hill, By Water has four phases. 1) Missile attacks, 2) Catch-all category for "monstrous creatures doing monstrous things," 3) miscellaneous (all non-missile, non-monster, non-melee actions and maneuvers), and 4) melee attacks.
• Forlorn Encystment wrote a blog post where he came up with a Phased Initiative method for 5E D&D. It is one of the most excruciatingly detailed systems I've ever seen. He includes fucking everything. It's really impressive.
• I think there's like a gazillion blog posts lost to the ages where each grognard lays out their own personal variation on Phased Initiative. Here's a neat one someone on Reddit once shared with me, which I saved the link to. I really like the gimmick with spell levels.
• Doctor Who Roleplaying Game by Cubicle 7 is pretty neat. 1) Declare actions, 2) Talking, 3) Movement, 3) "Doing" (miscellaneous actions), 4) Fighting. If two characters are trying to do the same thing at the same time, you can break the tie by comparing the stats they're using. I think this is a nice example of the granularity being applied to something other than different types of attack.
• The game Advanced Fantasy Dungeons (by the ever-charming Nova / Idle Cartulary of Playful Void fame) cleverly disguises its phases, but they're there. All non-attack actions are resolved first (ties broken by Dexterity), then all attacks are resolved in order of "speed," which is just their damage die. Low to high, 1d4 then 1d6 then 1d8 and so on. E.g. daggers, then short swords, then long swords, then great swords.

Composite Speed

These are all variants on the basic modern D&D method of turn-based, individual, randomized initiative incorporating speed modifiers of some kind. The reason it bears an entire category is because complicating that basic method creates lots of possibilities very fast. The main distinction is that these methods all incorporate a variety of modifiers in any given roll, usually situational ones that change from round to round.

(The 5E DMG calls this method "Speed Factor" but that's also the name of an unrelated mechanic from 1E AD&D. That one is used for breaking initiative ties and sometimes granting bonus attacks. Not well remembered by most, and not really an actual initiative method for the purposes of this blog post.)

A simple way to do this is by treating each "speed factor" as an arithmetical modifier. Roll 1d20, then add or subtract as necessary. Here's some examples:

Example from the 5E DMG, page 271

Another, more elaborate example from the 2E AD&D PHB, page 125

Thus, everyone has to decide what they're going to spend their turn doing before rolling initiative. This has the advantage of avoiding the dreaded "oh shit it's my turn uhhhh what do I wanna do?" thing, but it has the disadvantage of making you commit to an action which might be obsolete by the time your turn arrives.

Additionally, you now have to reroll your initiative every round, since you probably take different actions each round, after all. In order to spare the GM from the burden of bookkeeping, they forgo writing down initiative order and instead begin each round simply counting up from 1, with each player speaking up when their number is reached (or counting down from a high number, ascending vs descending, whatever).

All in all, this method is kind of like if you took phased initiative and just added a lot of wiggle room. Missiles will usually be shot before melee attacks are made... but not always. Your choice of action will affect how quickly you can act, but it's not set in stone.

Most other composite speed methods instead change the die size itself based on your chosen action. The most famous example is Greyhawk Initiative, based on a house rule that Mike Mearls once shared online which later got released as a full Unearthed Arcana for 5E D&D. It's got plenty of crunchy details and answers for all your edge case situations.

About four years before this, my cunning and crafty friend Nick over at Papers and Pencils came up with the same basic idea. However, he still retains a way for a character's personal speed to be a factor: you roll your initiative die a number of times equal to your Dexterity and then take the best one (or, if you have negative DEX, take the worst one). Interestingly, this predates the popularization of advantage / disadvantage from 5E D&D!

Sam from Dreaming Dragonslayer iterated on the Greyhawk variant. 1) He makes your initiative roll also your damage roll for that round, so that at least if you roll slow then you also hit hard, and 2) he offers two types of retreat options, one fast but unreliable and the other slow but sure.

Notorious bird Prismatic Wasteland seems to have independently come up with nearly the same idea. I always thought he and Mike Mearls were much alike. Some differences from the Greyhawk version:

• If you're doing multiple actions in one turn, Mr. Wasteland has you roll all dice and simply take the highest result, rather than having you add them together. Not quite as punishing against multitaskers.
• If you're casting a spell, then you add the spell level to the result. The more powerful the spell, the slower it will (probably) be.
• He has a rule allowing you to make a change of plans mid-round, which I'm surprised I don't see more often. It's slightly punishing, but at least you won't be screwed over if your declared action becomes irrelevant by the time your turn arrives.

And you know what? Just because I'm so gosh darn thorough when I research these posts, I'll throw in a bonus thing I found. Dungeon Crawl Classics incorporates the most minor version of composite speed dice I have yet encountered. While the main initiative method is the mainstream modern D&D option (turn-based, individual, randomized), it has one exception: whereas almost everyone rolls a d20 for initiative, anyone using a two-handed weapon rolls a d16 instead. That's literally the only example in the game, I believe.

Hybrids

AKA I couldn't figure out which of the previous categories to put these into, but I also don't think any of them constitute a new category?

Chainmail

The 1971 medieval fantasy wargame by Gary Gygax and Jeff Perren, and one of the key predecessors to D&D, offers not one, but two different initiative methods for you to pick from. One of them is a hybrid of phased initiative + randomized side initiative, whereas the other is a hybrid of phased initiative + simultaneous initiative. How the fuck do all these old-ass games somehow have the most complicated methods? Anyway, here's the basics (according to my best interpretation. You know how Gygaxian rules are):

1. The Move / Countermove System: Both opponents roll a die; the side with the higher score has the choice of electing to move first (Move) or last (Counter-move). First side moves and does missile fire, second side moves and does missile fire, then both sides suffer the effects of any artillery fire, then both sides suffer the effects of missile fire, then all melee is resolved. Notice that 1) it's only side-based for the first phase, but then becomes simultaneous for the rest of the phases and 2) there's a substantial delay between the moment when arrows are shot and the moment when arrows land.
2. The Simultaneous Movement System: Both sides write down orders for each of their units, including direction of movement and facing. Then, both sides move according to their written orders. At halfway through each movement, check for passing melee attacks. Then, take artillery fire, then take missile fire, then resolve melee.

In both cases, morale rolls can happen at any point.

Lamentations of the Flame Princess

Roll for which side goes first. Then, separately, every member of that side rolls to determine their individual turn order. Then, if there's a tie, they resolve simultaneously. Thus, combining the first three categories of this post all into one method, forcing you to contend with all of their various issues.

No, I don't really get why you would go for this option, either.

Tortoise and Hare

This is used in Shadow of the Demon Lord, Errant, and Wandering Blades. At the top of the round, everyone chooses to either act quickly or act slowly. Anyone who chooses to act quickly will only get to do 1 action, but it will come before everyone who chose to act slowly. Anyone who chooses to act slowly will get to do 2 actions, but they risk getting stabbed in the face first.

These chunks are staggered by side / pod. In SotDL and Wandering Blades, the players' side always wins initiative. In Errant, each side rolls. In either case, resolve the four chunks like so:

1. Characters on the winning side act quickly
2. Characters on the losing side act quickly
3. Characters on the winning side act slowly
4. Characters on the losing side act slowly

There's an additional rule in Errant only allowing you to make either one attack or cast one spell per round, even if you're acting slowly. Balance reasons, you know how it is.

So in this version, players have some control over who gets to act first. Most of the time, it's probably better for you to act slowly by default, since it's better to do 2 things instead of just 1. But if you're in a situation where it's really really important for you to act before your opponent, you have the option to sacrifice your second action if you want to guarantee that.

Bowling Initiative

This is used in Lancer, Star Trek Adventures, Fabula Ultima, and Four Crystals. Initiative is rolled off between each side, but characters take individual turns. It's just that those individual turns alternate back and forth between each side.

For example, let's say the party wins initiative against the goblins. Therefore, the party gets to have one of their members go first. They pick the wizard to take a turn. Then, the goblins get to choose one of their own to go. Now the party again, who picks the rogue. Then the goblins, who pick a new guy. Then the party, fighter goes. Etc.

What happens if one side has more members than the other? Common solution is that the side with greater numbers just has all their leftover members take their turns all in a row at the end of the round. So if, for example, it's a fight between 5 guys and 8 guys, then you only alternate for the first 5 guys on each side. The last 3 guys on the bigger side just take the last 3 turns.

Miss-Initiative

I don't think this is used in any published games. The earliest description of it I'm aware of is this blog post by Justin Alexander (from which I'm taking the name), but there was also recently a Twitter thread by M.T. Black where they described the exact same method and called it "Failure Switching."

Here's the idea: turn-based, with individual turns for each participant. The party decides their own turn order, the GM decides a turn order for the baddies. When a battle begins, the players start taking their turns one at a time, in the order they decided. BUT the moment someone fails a big action (e.g. an attack roll, ability check, saving throw), the other side seizes the initiative and gets to start taking their turns instead. They keep taking turns for as long as they can until one of their members misses an attack or check or whatever, at which point the players seize the initiative back and resume where they left off. Continue acting turn-by-turn until either everyone on your side has taken their turns or you lose the initiative again.

Of course, if your side manages to go through all their turns without anyone missing at anything, then your side is done for the round and now the other side can just take their turns without fear of interruptions. But if your side managed to pull that off, that probably means they just whupped a whole lotta ass.

Dynamic

I know that is also a meaningless name, and a lot of these would probably fit into a previous category, but I wanted a section just devoted to the most tactically wacky options I've found.

Active-Time Battle

This comes from a blog post by Pink Space where Lino speculates about a possible method inspired by the active-time battle (ATB) system found in Final Fantasy and Chrono Trigger. This is more of a theory post, which ends with a spitballed idea rather than a rigorously-tested houserule, but we can definitely include it.

On the table in front of you there is a gauge with ten or so segments. Each character has a token starting at one end of the gauge. 

In this model, each character has their speed reflected with a single die; its size varies depending on the character speed, with a larger die size for speedier characters. Each turn, all characters simultaneously roll their speed dice and move that number of spaces along the gauge. 

When a character’s token reaches the end of the gauge, they act. After their turn, they reset their place on the gauge.

Turn order is constantly changing. Maybe more importantly, this method is one of the few that actually interprets the meaning of speed a little more realistically: being fast doesn't mean that you go first, but rather that you go more times in a given time interval. In this case, a high-speed character will likely get more turns overall, but at a fairly unpredictable rate.

Shot Clock

This is used in Feng Shui, 7th Sea 2E, and Hollowpoint. They all use different terminology, and use slightly different processes, but it's the same idea.

Roll for initiative. This value is called your "shot." The GM starts the round by counting down shots from the highest result. When the GM reaches your shot, you get to take an action. Every action has a cost, reducing your shot by that amount.

For example, let's say you roll 1d20 and get 16 as your shot. The GM starts counting down: 20, 19, 18, 17, 16. Now it's your turn, and you choose to attack. That costs 3 shots, so your number is reduced to 13. Once your turn is over, the GM resumes counting down. 15, 14, 13. Now you get another turn. Maybe this time you go rummaging around a dead enemy's body to find their MacGuffin, and the GM decides this costs 6 shots. Your shot goes down to 7. Meanwhile, there are other characters taking actions during some of those other shots in between.

Like ATB, this method creates a totally chaotic turn order, and allows some characters to take more actions in a given round than other characters. But instead of the unpredictability coming from the randomization of dice rolls, it comes from regular fog of war: if you get less turns than your foe, it's because you chose to take more costly actions. Rolling initiative is like generating your "action budget" for the round.

I think the "countdown clock" is kind of unintuitive. Feng Shui wants you to track it on an actual numbered counter displayed somewhere on the table. Seems a bit much to me. Another, simpler way this can be done is by having everyone get piles of tokens. Actions cost tokens, and whoever currently has the most tokens takes the next action. So if Alice has 15, Bob has 13, and Charles has 11, then Alice goes first. She spends 3 tokens, meaning that Bob now has the most. Bob spends 3, so now Alice once again has the most. But after her next action, Charles will finally have the most. And so on. No counting down, no keeping track of everyone's place on a tracker. Just compare the size of everyone's pile after each turn.

Shadowrun 5E has a fairly similar system. Everyone rolls a huge initiative number, which will form an action budget. But it's divorced from when you take your next action. Just do a normal round table where everyone takes their first turn, then their second, then their third, etc. in a regular order. But as soon as your initiative number reaches 0, you've run out of turns and won't get to act in the next go-around. Thus, a character with a huge initiative number will indeed get to take more turns than the other characters, but only after the other characters have run out of turns. Something else I find really interesting about this system is that your initiative amount is also reduced when you suffer damage. So in addition to harm bringing you closer to defeat, it also means your less-damaged foe might get an extra turn or two over you this round.

His Majesty the Worm

The long-awaited upcoming Tarot-based RPG by Josh of Rise up Comus fame, which I get to read before any of the rest of you (nyah-nyah). Josh explains the initiative rules himself here, but for my readers who practice strict Twitter abstinence, here you go:

At the start of combat, you draw 4 cards. You choose 1 to be your Initiative. Your Initiative determines when you take your turn AND it determines the target number to hit you. Low cards act first but you're easier to hit.

Individual turn-based, but with some hidden choice and a built-in juicy tradeoff. Slow and steady is also sturdy, while leaping into action means throwing caution to the wind. Mmmmmm risk and reward. Tasty!

Street Fighter: The Storytelling Game

Everyone has a hand of cards representing their fight maneuvers. At the start of a round, everyone secretly selects a card. Then, everyone declares what the speed on their card is. The person with the slowest speed goes first, getting to move and then use their maneuver. Sounds backwards, right? But at any time during that person's movement, any other player with a faster maneuver can interrupt and take their own turn. Moving, then maneuvering, then returning to the interrupted player to finish their turn. But interrupters can be interrupted by players with faster maneuvers, who themselves can be interrupted by even faster maneuvers. You resolve all of them in order of fastest to slowest.

So having a slower speed means you "go first," but really just means you declare yourself first, and therefore will likely be interrupted several times before you get to resolve your turn. Once the slowest character has completed their turn, go to the next slowest (assuming they didn't interrupt).

Dynamic Yet Narrative Action Melee Order (DYNAMO)

This comes from the 2017 game Paranoia: Red Clearance Edition by Mongoose, which I've heard is pretty different from all older versions of the game. This is the one I have access to and it has a unique initiative system, so that's what you get to hear about.

Every player starts the session with a handful of Action Cards given by the GM, and can earn more throughout their adventure. Cards have written on them an "Action Order" number, a Reaction symbol, or both. If it has an Action Order number, then you can play it on your turn during combat when the number indicates. If it has a Reaction symbol, you play it when the card's trigger happens.

During each round of combat, follow these steps:

1. Everyone picks one card and puts it face-down. You have 5 real-time seconds to do this, or else you don't get a turn this round.
2. The GM counts down the Action Order numbers from 10 to 0. When they reach your number, interrupt and claim your turn.
3. Before you take your turn, other players can accuse you of lying about your number.
4. Resolve turn.
5. Return to step 2. GM resumes counting down until they reach 0.

The key is that players can and should lie about their Action Order number. When the GM starts counting down from 10, you can interrupt and take your turn when they reach 8 even though your card says 2. Hell, your card might be a Reaction card with no number on it at all, but you can still lie!

After you interrupt the GM to claim your turn, but before revealing your card, any other player can call you a liar. Each player can do this once per round. If the challenger is wrong, they lose a card. But if they're right, then they immediately get to take a bonus turn (in addition to their normal turn they get this round!), while the exposed liar has to discard their card and instead take a boring standard action at the end of the round.

If two players have the same Action Order number, the first one to say the name of the other player’s character in full goes first. If neither can say the other character’s name, neither act this round.

This system fucking rules.

Momentum

This comes from a blog post by One Zero where V.V. shares their homebrew method, which is like popcorn initiative but with a twist.

Enemies have a metacurrency called "momentum" they can spend to interrupt the popcorn-passing, to take extra turns at the end of the round, or to do other bad stuff. Each enemy starts with some momentum points, and there are a few triggers which allow them to gain more. The most important one: whenever an enemy is attacked by multiple PCs in succession.

Basically, the players are passing the ball around to each other like you would expect from popcorn initiative, but it might get intercepted by the bad guys at any moment. They can only do this a few times total, but if you commit the grave sin of "ganging up" then they'll get to do it even more times. This incentivizes the players to divide their attention as much as possible.

And of course, they add in lots of other bits and bobs on top. Spend momentum to do special actions, gain momentum by getting criticals or killing PCs, etc. It's an involved enough system that I imagine basically all major decisions in combat must revolve around it.

Hazard Die Stuff

You all love the Hazard Die, so it's no surprise that some folks have tried making an initiative method out of it somehow.

We'll first look at Brendan's method, since he did invent the Hazard Die after all. As with all applications of the Hazard Die, the idea is to fold in lots of procedural variables and bookkeeping items into one die roll that you simply make at the beginning of each turn. So for this "Tactical Hazard Die," you would roll a d6 once at the beginning of each combat round, consulting these results:

1. Setback: opponents act first or reinforcements arrive
2. Fatigue: combatants engaged in melee suffer 1 point of damage
3. Expiration: some or all ongoing effects end (such as burning oil)
4. Locality: the battlefield changes in some way
5. Percept: players gain some clue to opponent strategy
6. Advantage: players choose extra action or forced morale check

Presumably, this implies that the players get to act first each round by default.

Michael of Sheep and Sorcery takes it further, committing fully half the results to some kind of initiative effect. His version instead goes (paraphrased):

1. All PCs act first, then all enemies act last
2. Staggered into "fast" turns first (first all fast PCs, then all fast enemies) then "slow" turns last (first all slow PCs, then all slow enemies). "Slow" here is flexible. Heavy armor, over-encumbered, wielding big heavy weapons, being short, etc.
3. All enemies act first, then all PCs act last.
4. Ongoing effects end or allow for a Save.
5. Enemies call for back up!
6. Environmental effect!

I'm not sure what the "default" turn order would be on any given turn when you don't roll a 1-3. Do you just re-use the same initiative as you did on the previous turn? Roll a 3 once and now the enemies always get to act first each round until you roll a 1 or 2?

Miscellaneous Stuff

There's a lot of stuff I found in my research that isn't exactly a whole new method, but it was worth including anyway. Weird variables to account for, or just totally different approaches.

The Talking & Analysis Phase from Righteous Blood, Ruthless Blades

This is technically separate from the initiative rules but I think it's really cool and worth including. RBRB is a gritty wuxia action game with lots of bloody kung fu goodness. During each round, before re-rolling turn order, all characters get to act in the talking & analysis phase. During these breaks from the action, characters trade words and study their opponents. Mockery, threats, jokes, interrogation, flirtation, admiration, whatever. You could use this time to attempt to psych out the enemy, or to read their body language and gain a bonus, or even to gain new details about their signature kung fu moves.

Most RPGs allow characters to speak freely during combat scenarios, but it's still very uncommon for players to take advantage of this. For most groups, the beginning of combat is the end of roleplay. So I think it's a clever idea to set aside a moment during combat specifically designated for the purpose of roleplay, prompting PCs to say something when they otherwise probably wouldn't have. Even better, it's both thematic for the genre and it's incorporated into the mechanical design of the combat rules.

Approach-based initiative modifiers from Pathfinder 2E

Pathfinder obviously does turn-based, individual, randomized initiative, just like modern D&D. But instead of just having a Dexterity-based initiative bonus, you actually roll a skill check, using whatever skill is most appropriate for that moment. Perception is the default initiative skill, but it could just as easily be a Stealth check, Deception check, Diplomacy check, Survival check, whatever.

Flexible turn slots from FFG Star Wars

In the Star Wars RPG by Fantasy Flight Games, the party can sort of "trade" their initiative numbers as much as they want. It uses turn-based, individual, randomized initiative like in modern D&D (albeit much more difficult to explain the exact rules, since it uses those funky story dice. There's lots and lots of threads on the subreddit of people asking for initiative explanations). But whatever results everyone rolls, that just buys those slots for the party as a team. The party chooses who to assign each slot to on a round-to-round basis.

For example, let's say that there are 3 PCs and 5 bad guys. That means there'll be a total of 8 turns each round. Everyone rolls, and the players got the 1st turn, 4th turn, and 5th turn out of the total 8. But the PC who got the 1st turn on their roll doesn't have to be the one to take the 1st turn, and the PC who got the 4th turn doesn't have to be the one to take the 4th turn. They could swap if they want to.

Stop showing up late you fuckers

In Robin Laws' Og: Unearthed Edition, the initiative tiebreaker rule is to go in the order that each player arrived at the session, which I think is a pretty funny idea.

Chō-Han in Errant

In addition to the Tortoise and Hare method for ordering actions, Errant also has a unique method for determining the winner of initiative: the GM rolls a d6; the player also rolls a d6, but before they do, they announce whether the sum of the results will be odds or even. If they call it right, they win initiative; if they call it wrong, they lose initiative.

This is mathematically identical to a coin flip. Why make it so elaborate? Well, as Ava explains:

For anyone even vaguely familiar, they'll realise this is a variation on side initiative inspired by the popular Japanese street gambling game Chō-Han. I don't have much to say about this except for the fact that, despite being mathematically equivalent to regular side initiative, this feels so much better because players are invested in the outcome of the die because they made a decision. Victory is so much sweeter and defeat extra bitter; a player who makes an incorrect call is lambasted, while the player who calls it correctly is a hero. I also love the weird superstitions players develop around it: "always odds/evens" or "keep calling odds/evens until it happens."

Concluding Thoughts

Something I find interesting is how different games try to capture the idea of "speed" and what it means in conversation-based action gameplay. Going earlier in a round vs going more times within a round, being able to interrupt, having more actions, having greater movement distance. And a lot of games will include multiple of these! Marcia once toyed with the idea of making one's movement speed and initiative bonus derive from the armor you wear. Why not?

And if you can somehow determine how to simulate speed, you still must then decide what variables affect speed. The intrinsic speed of an action vs the intrinsic speed of the person attempting the action. Personal size versus the team's size. An attack's reach versus its unwieldiness. Heaviness of armor, heaviness of other gear, heaviness of heart. Personal confidence, recklessness, leadership ability. Luck. Some of these methods try to incorporate as many of those as possible. Others incorporate none of them.

When it comes to categorization, my colleague Ian offers some insights (which I've polished up below).

First, two big questions:

1. Structure. On a "continuum" from synchronous to phased to alternating.
2. Who. On a "continuum" from individual to pod to full side.

For non-synchronous methods, three more questions: 

1. Random or arbitrary. A spectrum from pure dice-rolling to pure reproducibility. This point is unconcerned with "meaning" or "interpretation," that's point 3.
2. Fixed or fluid. The more these numbers change the more fluid it is, from between rounds to between combats to between players or characters (completely fixed).
3. Diegetic or metagame. "Diegetic" is influenced by character or game concerns, while metagame is influenced by players. Most will fall between the two, if there are strategic concerns, for example, "popcorn" is more metagame. "Clockwise" is pure metagame.

I think that's a pretty dynamite attempt, but there are always weird edge cases. If you've somehow gotten this far, I hope you can now understand why I struggled a bit with this post.

I don't normally do "call to action" stuff at the end of my blog posts but I sincerely would love to hear people's thoughts, preferences, suggestions, etc. As you can imagine, after going through all this work I now have some very developed opinions on what makes for a good or bad initiative method. But that's a conversation for comment sections and maybe future posts. I hope this post stands for quite some time as the semi-definitive catalogue of most noteworthy initiative methods used in TTRPGs.

-Dwiz