Friday, July 28, 2017

Initiative Count Up, Simplified

In my last post, I talked about initiative, taking my cue from Brandes Stoddard's and Mike Mearls' coincidentally presented alternative initiative systems. I mentioned Hackmaster's Count Up system as the most innovative and promising alternative to traditional initiative systems that I've encountered, but that it suffers in my opinion from being too fiddly. So here's my promised attempt at simplifying the system.

Some theorycraft

There are two basic weapon stats in Hackmaster that affect the Count Up. Speed says how many seconds it takes to attack with a weapon after attacking an opponent with it for the first time. Reach says who goes first in melee if two attackers try to attack each other at the same time. The unnecessarily complicated part, in my opinion, is that every weapon has its own unique speed (measured in seconds) and its own unique reach (measured in feet, which is especially odd since tactical movement is measured in 5-foot squares). So a dagger has a speed of 7 and a reach of 1', a longsword a speed of 10 and a reach of 3.5', and a spear a speed of 12 and a reach of 13'. These stats aren't even based on predictable series; it seems like any possible score within a certain range is represented in these stats. It's all rather extreme, complex for the sake of simulation, I presume, but seemingly for the sake of complexity.

We could eliminate this complexity by just making all weapons have the same speed and reach, but that's pretty boring. There should be a middle ground between extreme simplification and the extreme simulationism that Hackmaster favors. Thankfully, there's an answer: combinatorics!

All I mean by combinatorics here is that by combining qualities from two or more small sets, you can get the same sort of complexity as one longer set, but with much less to keep track of. So if one set has numbers from 0-9, and another has letters from A to J, and you combined them so you have a new set from A0 to J9, that gives you as many items as a single set of 100 arbitrary symbols. But keeping track of two smaller sets is much easier than one longer set. A familiar application of this principle in d20 games is the alignment chart:

There are nine alignments, but no one really thinks of them that way. They think of the two axes, with each alignment being a combination of two points. Two sets of three is much easier to internalize than one set of nine, plus there's a pleasing symmetry that other alignment systems, like 4e's unloved 5-alignment remix, lack and suffer for. This is a good general trick for simplifying complexity in RPG design, by reducing a long list (like the list of weapons in Hackmaster) to two simple dimensions. And for weapons, the qualities I like to think of are Weapon Type and Weapon Size.

Type, Size, and Speed

By Type, I mean categories of weapons like swords, bows, axes, polearms, etc. Each Type would have its own simple rule to make it unique, but the particulars aren't terribly important for now. Size obviously refers how big the weapon is, but the frame of reference is not the size of the weapon itself, but the size of the character wielding it. That is, a Medium weapon is designed to be wielded one-handed by a Medium-sized character, like a spear or longsword. A Small Sword might be a dagger to a human, but a longsword to a gnome. A Large Sword has to be wielded two-handed by an orc, but a Large troll could use it with one hand. (In 5e, Small weapons are basically "light," and Large weapons are "two-handed", for those wanting to adapt these rules to conventional D&D.)

Like with Type, weapon Sizes have their own rules, but since we're talking about Hackmaster initiative and Count Up rules, we only have to worry about rules related to speed and reach. Here they are:
  • A weapon the same Size as its wielder has a base speed of 6. If it's larger than than the wielder, its speed is 8. If it's smaller, 4.
  • If two characters attack each other in melee on the same count, the one with the larger weapon has the larger reach and goes first.
  • Polearms count as one Size larger than they are for the purposes of determining reach.
  • An unarmed melee attack usually has the same Size as the character making it.
I'll talk about other rules for weapons in another post, but that's everything we need, and it's much easier to keep track of than Hackmaster's fiddly weapons tables.

The rest of these initiative rules are similar to Hackmaster.

Rolling Initiative

At the start of combat, characters roll a d20, subtracting their Dexterity and Wisdom scores from the check. The GM begins the Count Up, with characters acting when the result of their initiative check is called. Before their turn is called, they are flat-footed; they continue to do what they were doing and can only act to defend themselves. Afterwards, they can move, attack, cast a spell, make a skill check, or do anything else so long as they have time. A character is no longer flat-footed and may act two seconds after taking damage from an attack, unless her initiative count would allow her to act sooner.

Surprise Attack

Instead of rolling a normal initiative check, players may opt to roll a skill check to make a surprise attack. They may roll Stealth if they are hiding in ambush, or roll Deception if they are out in the open and feigning helplessness or passivity. This works like a normal skill check, except you subtract from your d20 roll as if you were making an initiative check. If your check is lower than any other hostile characters', you act on Count 1 and gain advantage on both attack and damage* checks against any flat-footed opponents. If not, however, you are flat-footed until your turn on the Count Up comes up, as though the result of your skill check were your initiative check. In addition, you have disadvantage on both defense* and toughness* checks until your turn on the Count Up comes up. (*These rules are discussed in this article. Adapting these initiative rules to 5e D&D should not require them.)


Characters can move one or (if their speed is greater than 1) two spaces at any time. They may also move up to their speed, but only if they moved two spaces on the previous second. To slow down, they must move two spaces. Note that moving while engaged in melee may provoke reaction attacks from your opponents. Some actions, like drawing or picking up a weapon or other item, can be done instead of moving.


Ranged attacks can be made at any time. If two characters attack each other with ranged attacks at the same time, they may resolve their attacks against each other as normal. For melee attacks, an aggressor must get in range of a defender. When they are in range, the character with the greatest reach (usually the largest weapon, see above) goes first (unless the defender is using a ranged weapon, in which case the attacker goes first every time). If their reach is the same, they may attack at the same time and resolve those attacks against each other as normal. Afterwards, these two characters are engaged in melee, and are so engaged until one is incapacitated or retreats. Moving away while engaged may provoke reaction attacks from your opponent; see below.

After making an initial attack, a character must normally wait a number of seconds equal to his weapon speed before attacking again. When a melee engagement a character is involved with ends, this time is reduced to zero, meaning he can freely engage in a new melee at any time.


Certain actions may leave an opening for opponents to attack. This may include moving out of a melee without properly withdrawing, attempting to cast certain spells next to a melee combatant, or critically failing an attack roll. This is called provoking a reaction. If a character adjacent to you provokes a reaction, you can take the reaction to make an immediate attack. You can only make a reaction attack if a number of seconds equal to half your weapon speed has gone by since your last attack. Making a reaction attack also resets your count in the Count Up, meaning that you have to wait your weapon speed to make your next attack, unless someone provokes another reaction, as per the above rules.

Some final considerations

These rules are pretty rough, but should give some idea how this could work in game. One big change I could make is to turn the Count Up to a Count Down, in which case characters add their initiative score when making an initiative check. The Count Down would then start at the highest result and proceed down. This would have the advantage of making initiative checks more like other d20 checks in contemporary D&D. The disadvantage is what to do when the Count Down reaches 0. Bounce back up to 1? Continue into negative numbers?

If you're adapting these rules to 5e, I wish you luck! I would consider turning Extra Attack into a -1 enhancement to speed every time it's earned. The same goes for each attack to a monster's Multiattack if it's making the same attack more than once. For spells, a round of casting converts to 6 seconds, so you could just make that the default casting time for such spells. Recharge times for monster powers could be done a few ways. It takes an average of 6 rounds for a Recharge 6 power to recharge (and 3 rounds for a Recharge 5-6, I think), so you could give such powers a speed of 36 and 18. Or you could let the powers recharge as normal whenever the monster makes a "normal" attack.

The default weapon speeds in Hackmaster seem really slow to me. The speeds I picked are much quicker, based on the standard of one attack per "round" (which equals 6 seconds) in normal d20 games. But maybe there's a good reason for such slow speeds?

I haven't dealt with two-weapon fighting yet because there are still some things to consider. But one key will be that dual-wielding means that weapons count as one size larger than they are for purposes of calculating speed. Shields count as weapons here. Whether both weapons attack at the same second or one after the other, and which weapon takes precedent if the latter; all that needs to be worked out.

There are probably some things I haven't thought of yet. Please ask questions so I can consider them, or comment if any of the above seems weird or confusing!

Saturday, July 15, 2017

Take the initiative, please!

I've been a bit stymied about what to discuss next in my series of posts that I've tagged "me-20;" my self-indulgent hack of d20 gaming. I've been inspired by a coincidence: First, Brandes Stoddard publishes an article about some alternative initiative systems on his blog. Then, shortly after, Mike Mearls publishes an article on WotC's Unearthed Arcana blog about the initiative system he's been using for his 5th edition group. So suddenly talking about initiative seems like a good idea. Apart from kismet, it actually makes sense: I've talked a lot about how combat and other systems could work in general, but not offered a specific example of them in practice. And what better way to begin talking about encounters than by starting at their beginning? It's a very good place to start, as the sages say.

RPG veterans may skip this paragraph, but perhaps the reader may wonder what I'm talking about. Basically, in an RPG encounter, like other sorts of games, there must be a way of figuring out who goes first. In American football, it's determined by a coin toss. In baseball, by who the visiting team is. In many RPGs, it's a special roll called initiative. Every player rolls a d20 at the start of combat or other encounter, taking into account special modifiers (faster players often react faster) and situations (being ambushed usually means going last). The GM does the same thing for individual monsters or groups of monsters. The character with the best initiative result goes first, whether it's the PC's ranger, the GM's archvillain, or a mob of NPC orcs. The next highest initiative goes next, then the next highest, until the lowest-scoring player goes and the top-initiative roller goes again.

Pretty simple, right? Everyone goes in the order they roll, like people waiting in line with numbered tickets at the deli. Why would anyone want to futz with this? Well, as Stoddard point out in his article, it's actually a bit of a slog trying to figure out who goes in what order. Yes, there are apps and more analogue hacks like index cards that make it easier, but you still have to go around the table, figure out how everyone scored, then record it somehow. It's a bit of dreary data-entry that kills the dramatic momentum of what should be an exciting event.

It's a cosmic coincidence that Stoddard and Mearls took up the subject of initiative systems at around the same time, but their approaches to the problem couldn't be more different. Stoddard offers a few radical simplifications: the first is by adapting the "first player button" that changes hands in many board games and doing away with initiative rolls entirely, the second by replacing initiative rolls every encounter with a single initiative roll every session (i.e. once every afternoon or evening of play). Mearls, on the other hand, attempts to make the tedium of calculating initiative more interesting by adding more tactical variety to the roll. So characters roll different dice depending on the actions they want to take, with a ranged attack being quicker than a melee attack, and either attack being slower if you want to both attack and move. Mearls makes no bones about the added complexity here, but the hope is that this complexity makes the process of determining initiative more relevant and engaging.

Mearls' system has some interesting concepts, but overall I agree with Stoddard's take: that the added complexity is kind of arbitrary seems less likely to engage and more likely to frustrate. That said, Stoddard's systems seem a little bland. The "First Player Button" has little to do with character skill, which is something I want expressed in any RPG activity. And while initiative once per session by necessity reduces bookkeeping, it's a pretty ruthless, cut-off-your-leg-to-stop-a-cramp solution. It seems that both designers are kind of stuck on the idea of what initiative and taking your turn actually means in RPGs, and that we could benefit from a more fundamental reimagining of the idea. And that brings us to Hackmaster.

Hackmaster has kind of a weird history, finding its conceptual origins not in any real gaming group, but in the world of the cult RPG parody comic "Knights of the Dinner Table," as the old-school-D&D stand-in the protagonists play. But as is often the case with fictitious games, eventually someone had to design and market the real thing, in this case the publishers of the comic, Kenzer and Company. The style is true to the gritty old-school feel of the comic, with lots of granular combat mechanics like facing and individual weapon speeds and "threshold of pain checks." It's all a bit too fiddly for my tastes, but the one unique idea that I love is "the Count Up."

With the Count Up, players and the GM roll for initiative as described above, with lower rolls being desirable. But instead of the GM killing the excitement by going around the table and recording the results, he revs the excitement into high gear. "One! Two! Three!" he counts, each number announcing a second of time going by. At this point, the players are waiting for their initiative result to come up, hoping the monsters don't come up first. That's because any character is considered surprised before his or her initiative count and can't do anything but rather weakly defend themselves. But after their count comes up, the players can act as they wish, whenever they want, so long as they have the time to do so. Walking one space takes a second, or one step of the count, and you can move whenever you want. Attacking takes a certain number of seconds, depending on a number of finicky factors, your weapon's speed being principle among them. The same is true for casting spells. The result is that attacks and spells sort of adhere to the traditional round-by-round structure of many combat RPGs without being bound by it. And it seems like it can be very exciting: check out the combat example in the free basic version of the game, starting on page 136.

A lot about this is very brilliant, even beyond the way jumping right into the Count Up reduces the annoying bookkeeping of most initiative systems. The absence of traditional turns means players are always engaged even as they wait for the next chance to attack or cast a spell. The option for any player to move at any time makes play more dynamic and reactive. The various "tie-breakers" (ranged-vs-melee-vs-spells, weapon reach for melee attacks) for simultaneous attacks make a lot of sense. Making surprise part of the core initiative system is much more elegant than the rather clumsy "surprise round" rules found in many d20 games*.

Sadly, there's a lot about the system that's pretty kludgey. Each weapon has its own unique speed and reach. Each spell has its own unique casting time as well. Even something as simple as moving depends on a bunch of factors like the character's size, race, and preferred pace (There are four different movement rates!). But this is all a result of Hackmaster's old-school philosophy, where complicated tables are part of the look and feel of the game. The basic idea of the Count Up system works even when other parts of the system are greatly simplified. Someone just has to do it.

This post is running long, but next time I'll take a crack at exactly that!