Mearls On D&D's Design Premises/Goals

First of all, thanks [MENTION=1]Morrus[/MENTION] for collecting this. I generally avoid Twitter because, frankly, it's full of a$$holes.

That aside: this is an interesting way of looking at it, and underscores the difference in design philosophies between the WotC team and the Paizo team. There is a lot of room for both philosophies of design, and I don't think there is any reason for fans of one to be hostile to fans of the other, but those differences do matter. There are ways in which I like the prescriptive elements of 3.x era games (I like set skill difficulty lists, for example) but I tend to run by the seat of my pants and the effects of my beer, so a fast and loose and forgiving version like 5E really enables me running a game the way I like to.

 

[Hussar]

Meh, I really don't see it as a problem. It works both ways as well. The DM gets to plan around things as well and since there's only one DM, he can implement his plans easier than if the players have to hash out ideas. Like I said, it's just not worth the extra hassle IMO.

 

[5ekyu]

Good.

My contention is that the focus shouldn't be on the metagame "turns" but rather on what's happening in the fiction in the chaos of war; and that the type of planning and co-orcination you're talking about should neither be free nor easy.

Ahead of time, sure. But once the swords and axes start flying, forget it.

Even if it's random they can still choose who goes when, but it would involve a lot of delaying while people waited for other turns to come up - in other words, their ability to pull off their co-ordinated plan comes at a cost of frequently acting late in the round.

But this isn't even the worst part. The worst part in cyclic init. is that the players also know when the opponent's turns are (or will after the first round), allowing them not ony to co-ordinate among themselves but to meta-plan around when the foes get to act. This problem - for problem it is - goes away if init's are rerolled each round and the DM rolls opponent init's in secret.

"Ahead of time, sure. But once the swords and axes start flying, forget it."

Mostly I see both, sometimes its things they discussed out of combat then work towards in combat. Add in a random roll between stages in a combat and you vastly reduce opportunities to implement plans just due to the added random screw of a new unit roll.

"Even if it's random they can still choose who goes when, but it would involve a lot of delaying while people waited for other turns to come up - in other words, their ability to pull off their co-ordinated plan comes at a cost of frequently acting late in the round. "

5e does not have delay, so if you are meaning this for some other game, not 5e, that's fine. 5e has ready and ready limits you to a rather selective sub-set of options and the if-then etc.

" The worst part in cyclic init. is that the players also know when the opponent's turns are (or will after the first round), allowing them not ony "

That's a feature, not a bug, it's not a problem, it's an option.

Really, what makes it some objective problem ?

Both sides in a cyclic init learn the order and can make choices based on it.

Why is it good or better to have a random re-order at specified times making that "break for new init" a much bigger element in how the combat plays?

"More random" and "less choice driven" are not goals I would choose to enhance combat scenes if I was focusing "on the fiction of the scene" because "random" does not care about fiction.

 

[epithet]

Meh, I really don't see it as a problem. It works both ways as well. The DM gets to plan around things as well and since there's only one DM, he can implement his plans easier than if the players have to hash out ideas. Like I said, it's just not worth the extra hassle IMO.

The reason rerolling initiative appealed to the group in the first place was that players wanted to have more control over when they took their actions. With static initiative, following the published turn order rules, you act on your turn, period. You can hold an action to take as a reaction, but you can't say "I going to wait and go after the rogue." We were letting people do that anyway, with the house rule that you can adjust your initiative down as much as you want to, but that meant that in subsequent rounds you'd be stuck down there at the bottom of the tracker. If you let players delay their turns once per round (stating on your original initiative what you want your lower adjusted initiative to be,) that means that if a player "wins" initiative it makes turn order into a tactical choice. It makes feats and items that give advantage to initiative more appealing, too.

Another thing it lets me do is create a hidden actor on the combat tracker for the lair with a +10 init bonus. Rerolling every turn means that the PCs are never quite sure when lair actions will come up, adding dramatic tension.

 

[Aldarc]

I don't think that the only options should necessarily be static initiative or re-roll per round initiative.

The Cypher System, for example, has a pretty intuitive initiative system that is a mix of static and free form. Foes have an associated Target Number or Difficulty Class. Players roll initiative. Players who roll higher than the DC can act before the foe in whatever order they prefer. The foe acts. Players who roll lower act after the foe in whatever order they choose. Having multiple foes is generally not an issue either.

 

[Hriston]

Which is why I dubbed it the non-rule rule. If that rule didn't exist, there would literally be no difference in how initiative plays out. The DM would still make the decision, which could still result in simultaneous resolution.

A strict reading of the initiative rules without the paragraph about what to do in the event of a tie would have combatants with tied initiative rolls always act simultaneously (at the same time). The 3rd paragraph allows ties to be broken.

I look at the dex of the tied individuals and high dex goes first. If dex is tied, it's simultaneous. As I mentioned earlier, truly simultaneous actions are fairly rare.

Perhaps interestingly, the Holmes Basic Set (1977) rules decide initiative by comparing Dexterity scores, which is, I think, the earliest initiative system published specifically for D&D. The original rules (1974) use Chainmail's initiative system which has two options: the "Move/Counter move" system, the ancestor to AD&D's side initiative, and the "Simultaneous Movement" system, which requires players to write out orders for their units beforehand, simultaneously take half their movement checking for unintended melee contact, complete the movement phase, and then resolve missile fire and melees in simultaneous phases. Simultaneous in a system like this means something like "happening roughly in the same one minute of time". A similar system could be devised for 5E, all action in a given round happening "simultaneously" in roughly the same six seconds as a level of abstraction.

Any simultaneity at all requires some level of abstraction. For example, if you're accustomed to resolving combat in Planck time units, simultaneous events are going to be far more rare than that in which your above method would result.

 

[Maxperson]

A strict reading of the initiative rules without the paragraph about what to do in the event of a tie would have combatants with tied initiative rolls always act simultaneously (at the same time). The 3rd paragraph allows ties to be broken.

Nah. A strict reading shows that only one person can go at a time. "The DM ranks the combatants in order from the one with the highest Dexterity check total to the one with the lowest." That necessitates having to break ties somehow, and without the non-rule rule, it falls to the DM to decide just like with the non-rule rule.

Perhaps interestingly, the Holmes Basic Set (1977) rules decide initiative by comparing Dexterity scores, which is, I think, the earliest initiative system published specifically for D&D. The original rules (1974) use Chainmail's initiative system which has two options: the "Move/Counter move" system, the ancestor to AD&D's side initiative, and the "Simultaneous Movement" system, which requires players to write out orders for their units beforehand, simultaneously take half their movement checking for unintended melee contact, complete the movement phase, and then resolve missile fire and melees in simultaneous phases. Simultaneous in a system like this means something like "happening roughly in the same one minute of time". A similar system could be devised for 5E, all action in a given round happening "simultaneously" in roughly the same six seconds as a level of abstraction.

Any simultaneity at all requires some level of abstraction. For example, if you're accustomed to resolving combat in Planck time units, simultaneous events are going to be far more rare than that in which your above method would result.

I don't want to get too complicated with initiative. A lot of those old 70s and 80s games had rulebooks that were thicker than Game of Thrones books, and more complicated than trying to figure out the tax code.

A quick easy dex check with ties going to highest dex, and further ties being simultaneous is as complicated as I want it.

 

[5ekyu]

I don't think that the only options should necessarily be static initiative or re-roll per round initiative.

The Cypher System, for example, has a pretty intuitive initiative system that is a mix of static and free form. Foes have an associated Target Number or Difficulty Class. Players roll initiative. Players who roll higher than the DC can act before the foe in whatever order they prefer. The foe acts. Players who roll lower act after the foe in whatever order they choose. Having multiple foes is generally not an issue either.

Indeed. While I see rolling randomly every turn as a bad option, I certainly don't thing rolling once is a good option, just better.

In my homebrew scifi 5e gsme, I use first or last.

On the opening turn, players quickly decided whether a PC goes first or a PC goes last.

If they choose first, a singlenpc goes first, their choice, then a foe (FM choice) alternate until you get yo the end where an npc is always going last.

If they choose last, a foe goes first alternate etc but a PC goes ladt.

That turn order is then set.

Newly arriving characters are added into the middle but font change ghd first/last.

So its entirely choice driven with alternating play side hy side.

Tactical aspects of the situation drive the choices, not the die rolls.

Works great and plays quick.

 

[Maxperson]

Indeed. While I see rolling randomly every turn as a bad option, I certainly don't thing rolling once is a good option, just better.

In my homebrew scifi 5e gsme, I use first or last.

On the opening turn, players quickly decided whether a PC goes first or a PC goes last.

If they choose first, a singlenpc goes first, their choice, then a foe (FM choice) alternate until you get yo the end where an npc is always going last.

If they choose last, a foe goes first alternate etc but a PC goes ladt.

That turn order is then set.

Newly arriving characters are added into the middle but font change ghd first/last.

So its entirely choice driven with alternating play side hy side.

Tactical aspects of the situation drive the choices, not the die rolls.

Works great and plays quick.

That's an interesting system and I can see where it would have some appeal. For myself, though, it doesn't make sense that sides alternate like that. It makes sense to me that there would be clumps on one side or the other going before someone on the opposing side. If three PCs beat my fastest monster, one PC loses to him, then two monsters go, I have no problem with that.

 

[5ekyu]

That's an interesting system and I can see where it would have some appeal. For myself, though, it doesn't make sense that sides alternate like that. It makes sense to me that there would be clumps on one side or the other going before someone on the opposing side. If three PCs beat my fastest monster, one PC loses to him, then two monsters go, I have no problem with that.

Absolutely. Every init system seems to embody some degree of compromise vs realism in order to reach whatever its goals are.

Part of my major focus in many games as they have developed over the years have been to focus on two keys - character capability and player choice (the former being a manifest coding of the latter.)

One of my biggest gripes with most diced knit system has tended to be what rolls init? Mostly if it were me, it would be a combo of perception and quick thinking or tactics - more situational awareness - but also discipline could play a tole etc etc etc...

All in all I see it as hard to tie to specific character traits in a simple way and my go to for "not character trait" is "choice".
So we tried it a while back and loved it.

That said, with new groups of players where I want yo minimize house rules, I just go with regular init by RAW and let advantage and disadvantage play a significant role.

 

[Lanefan]

Absolutely. Every init system seems to embody some degree of compromise vs realism in order to reach whatever its goals are.

Part of my major focus in many games as they have developed over the years have been to focus on two keys - character capability and player choice (the former being a manifest coding of the latter.)

One of my biggest gripes with most diced knit system has tended to be what rolls init? Mostly if it were me, it would be a combo of perception and quick thinking or tactics - more situational awareness - but also discipline could play a tole etc etc etc...

All in all I see it as hard to tie to specific character traits in a simple way and my go to for "not character trait" is "choice".
So we tried it a while back and loved it.

That said, with new groups of players where I want yo minimize house rules, I just go with regular init by RAW and let advantage and disadvantage play a significant role.

What did you think of Merals' variant initiative system that he put out - was it last year? earlier this year? There was a big thread in here about it.