Do you play with people or groups that oversimplify the Cube area mechanics? I think many of us do, because many of us were oversimplifying it too.
In my experience, groups tend to make one big (and incorrect) assumption when it comes to Cube area-of-effects; That they want to or must squeeze themselves into as small a grid space as they can. This is just not true. Crack open your Thunderwaves and Clouds of Daggers, you've probably been shorting yourself how much space they can hit.
The PHB gives us exactly zero information on how any Areas of Effect act with a Grid... because the assumption (and the default) is that people are not playing with one. What it does tell us is simply that wherever we put its Point of Origin, must end up being somewhere on one of the faces of the cube. "Cube -- You select a cube's point of origin, which lies anywhere on a face of the cubic effect. The cube's size is expressed as the length of each side. A cube's point of origin is not included in the cube's area of effect, unless you decide otherwise."
The DMG gives us how that interacts with a grid, and simply adds that the Point of Origin must be at a grid intersection: "Areas of Effect -- The area of effect of a spell, monster ability, or other feature must be translated onto squares or hexes to determine which potential targets are in the area and which aren’t. Choose an intersection of squares or hexes as the point of origin of an area of effect, then follow its rules as normal. If an area of effect is circular and covers at least half a square, it affects that square."
Notice that the special 50% rule only applies to cicular areas of effect... which a Cube isn't. Neither is a Cone, and most of us have had no problem understanding how those work.
So... Point of Origin at an intersection, Point of Originanywhere along one side of the cube. That is ALL we must satisfy. This illustrations shows the common configurations relative to the Point of Origin (red dot), most of us assumed were the only options [A and B]. But in fact C is entirely fine... and, this will be the shocker for most, so is D and MANY variations of D. (The orange shows the squares that would be affected).
Thunderwave is a bit of a special case, in that Cube area spells don't normally also have a Range of Self. The description of Thunderwave, however, makes it clear that it effects "Each creature in a 15-foot cube originating from you". And as we know that our Point of Origin for a cube must be a grid intersection, it must be one of the grid intersections that make up your space. And travel from that Point of Origin AWAY from you. (So no shaping it funny to include yourself). Effectively place yourself in one of the uneffected squares beside any of those Points of Origin in the illustration, and you're good to go.
You seem to write with the assumption that it is ok for you to position your spell effect off grid, but the creatures you're trying to hit must snap to grid. That is awful. Since a medium creature needs about 5' of space, there is no way that a 2x2 square shaped effect can hit more than 4 creatures on the ground, no matter how you position it (or them). If the creatures are more than 10' apart, you cannot hit them.
Since creatures can move within their spaces, you can simply say that the four creatures in C are on the edges of their spaces, and that cube would miss them all. D1 and D2 both only fully cover 1 square, so can only ensure that you only hit 1 creature (on the ground).
I would agree with this. If movement and position must snap to the grid, spells should too. I’m not saying I wouldn’t allow diagonal placement, (potentially making the area more of a tetris z block than a square?) but it shouldn’t increase the area of the spell. Off the top of my head I think saying the spell must cover at least half the square to hit it keeps the number of target squares even.
if you are playing gridless and just measuring movement in inches, by all means go wild with fitting as much in the cube as you can.
I'm not suggesting a DM couldn't choose to have things snap to grid if they wanted. Only that they would be inventing that, and many people do visualize cubes keeping locked to the Cardinal direction. But the RAW works as illustrated.
Jeremy Crawford backs up both of those points (cubes aren't locked to grid, and the DM can decide otherwise) in a tweet. "Cloud of daggers (5 ft. cube) can affect 1+ squares on a grid, unless the DM says effects snap to the grid."
If you want to rule that that cube does not snap to the grid, than neither do the people. Which means that while you THINK you are sneaking your corner into the victim's grid, there grid is actually off set so it doesn't.
In other words, the only way you gain any advantage to doing this cheesy crap is if you can do it but the enemy does not.
Screw it, K.I.S.S. and just have everything snap to the grid.
You seem to write with the assumption that it is ok for you to position your spell effect off grid, but the creatures you're trying to hit must snap to grid.
And they're correct. The DMG's rules only require that the point of origin be on a grid intersection. A cube's point of origin can be anywhere on its surface, so in practice cubes don't have to snap to the grid.
In fact, there's no requirement for cubes to be parallel to the ground either. The only shape with an alignment requirement is the cylinder.
That is awful. Since a medium creature needs about 5' of space, there is no way that a 2x2 square shaped effect can hit more than 4 creatures on the ground, no matter how you position it (or them). If the creatures are more than 10' apart, you cannot hit them.
The rules even come close to suggesting that an area of effect needs to completely cover a creature's space (or one of the 5x5 spaces it occupies) to be affected, unless the area is circular.
Pg 249 Adjudicating areas of effect. This suggestion is aimed at theatre of the mind but could be applied to closely spaced creatures on a grid using minis.
"ADJUDICATING AREAS OF EFFECT Many spells and other game feat ures create areas of effect, such as the cone and the sphere. If you're not using miniatures or another visual aid, it can sometimes be difficult to determine who's in an area of effect and who isn't. The easiest way to address such uncertainty is to go with your gut and make a call. If you would like more guidance, consider using the Targets in Areas of Effect table. To use the table, imagine which combatants are near one another, and let the table guide you in determining the number of those combatants that are caught in an area of effect. Add or subtract targets based on how bunched up the potential targets are. Consider rolling 1d3 to determine the amount to add or subtract."
On a grid the DMG has the following rules for AoE. Pg 251
"AREAS OF EFFECT The area of effect of a spell, monster ability, or other feature must be translated onto squares or hexes to determine which potential targets are in the area and which aren't. Choose an intersection of squares or hexes as the point of origin of an area of effect, then follow its rules as normal. If an area of effect is circular and covers at least half a square, it affects that square."
The last sentence indicates that if an area of effect is circular then it has to cover at least 1/2 the square in order to affect that square. I use that as guidance for applying all spell shapes on a grid. If the AoE affects 1/2 or more of a square (eyeballed and usually leaning towards the players benefit) then I apply the effect to the square. Otherwise not. A medium creature occupies a 5' square on a grid, however, it is not required to remain at the center of the square and it is not required to be stationary, as a result, I personally rule that when a player decides on placement of an AoE that it has to cover at least 1/2 the square in order to affect it.
Of course, application is up to individual DM preference. I personally take the specified AoE of the spell to indicate the maximum number of squares that can be affected. A 10'x10' spell can affect up to a maximum of 4 squares depending on placement. A 15' x15' cube would affect up to a maximum of 9 squares no matter how it was positioned on a grid ... but that is just how I run it.
You seem to write with the assumption that it is ok for you to position your spell effect off grid, but the creatures you're trying to hit must snap to grid.
And they're correct. The DMG's rules only require that the point of origin be on a grid intersection. A cube's point of origin can be anywhere on its surface, so in practice cubes don't have to snap to the grid.
The problem isn't the assumption that you don't need to snap a cube to the grid, the problem is that the OP assumes that the creatures must and the cube doesn't. Thinking that you can hit 10 enemies with a 2x2 area is problematic (D1). Taking a closer look at D1, the top squares highlighted are 15' from the bottom highlighted square, even though the area of effect is only 10' diagonal -- using the distance counting rules that the "using a grid" sidebar tell us to use. There is no way that I'd rule that something 10' across can hit creatures that have 15' of empty space between them.
Brian, it ALWAYS boils down to what each DM will allow and how they interpret. As such, your comment is not useful in any way, and is in effect meaningless.
The main objection we have is that the assumption you are making that because sphere specifies "half" and Square does not, that it must be "any part of the square in the cube." We disagree with your belief that "any part" is the default.
But that does not really matter, anymore than your comment about the DM having last say.
What it REALLY comes down to is this:
One group maintains, with significant amount of evidence, that the RAW are over-powered for Cubes, and also want to use it according to their over-powered interpretation. You are not claiming it is appropriate, you are simply stating it very powerful and want it.
The other group maintains that while there is some evidence that the RAW are as you believe, that such a reading is ridiculous and laughable. No reasonable DM would allow it, despite the evidence you put forth. You admit it is far more powerful than a sphere, and have put no argument forth that this was intentional.
You are of course free to play as you desire. But I would recommend that:
Do not be surprised if your DM ignores what you say
If the DM does overrule you, (whether he calls it a 'house rule', or simply says you are wrong.) then objecting that it is the 'RAW' will neither win your argument nor do anything but make you look like a power gamer..
As a DM, I won't let people exploit system granularity to enlarge an area effect, which means for case C you can only pick rounding on one side, and case D is just wrong even using the rules for circular area, he's showing the affected area as being every square touched by the effect, rather than the squares that are actually at least 50% covered.
The problem isn't the assumption that you don't need to snap a cube to the grid, the problem is that the OP assumes that the creatures must and the cube doesn't.
Ok, so show me the rule that says your area effects have to snap to the grid.
Taking a closer look at D1, the top squares highlighted are 15' from the bottom highlighted square, even though the area of effect is only 10' diagonal -- using the distance counting rules that the "using a grid" sidebar tell us to use. There is no way that I'd rule that something 10' across can hit creatures that have 15' of empty space between them.
There's only 10 feet of empty space between the bottom highlighted square and the topmost square it's aligned with. The diagonal of a 10 x 10 right triangle is sqrt(10*10 + 10*10) = 14.14 feet long.
All of the configurations in the OP are perfectly valid in theater of the mind and there's nothing in the grid rules that prevents this from working in that context either. If we want to get philosophical about this, to me it's a much bigger problem if someone's interpretation of the grid rules disallows normally valid plays.
The problem isn't the assumption that you don't need to snap a cube to the grid, the problem is that the OP assumes that the creatures must and the cube doesn't.
Ok, so show me the rule that says your area effects have to snap to the grid.
I think you misread what I said (twice). My point isn't that area effects must, it is that if area effects need not then neither do creatures. If you're spell only touches the tiniest corner of my square, then I'm just going to stand in the other corner. It seems completely fair unless you can show the rule that says that a square partially covered by a cube is counted as affected by that spell's effect.
Taking a closer look at D1, the top squares highlighted are 15' from the bottom highlighted square, even though the area of effect is only 10' diagonal -- using the distance counting rules that the "using a grid" sidebar tell us to use. There is no way that I'd rule that something 10' across can hit creatures that have 15' of empty space between them.
There's only 10 feet of empty space between the bottom highlighted square and the topmost square it's aligned with. The diagonal of a 10 x 10 right triangle is sqrt(10*10 + 10*10) = 14.14 feet long.
Nope.
To determine the range on a grid between two things—whether creatures or objects—start counting squares from a square adjacent to one of them and stop counting in the space of the other one. Count by the shortest route.
The distance between the diagonal squares in a 2x2 is 10' according to the rules for using a grid. The rules also tell us that there are 15 ft if we count to 3 before we get from the square below the top square to the bottom one.
All of the configurations in the OP are perfectly valid in theater of the mind and there's nothing in the grid rules that prevents this from working in that context either. If we want to get philosophical about this, to me it's a much bigger problem if someone's interpretation of the grid rules disallows normally valid plays.
I think David said it best. If you want to hit 3 medium creatures in a row, you simply cannot do it with a 10' cube, theater of the mind or not.
Brian, it ALWAYS boils down to what each DM will allow and how they interpret. As such, your comment is not useful in any way, and is in effect meaningless.
The main objection we have is that the assumption you are making that because sphere specifies "half" and Square does not, that it must be "any part of the square in the cube." We disagree with your belief that "any part" is the default.
There are three basic options: 'any part of the square is in the cube', 'the majority of the square is in the cube', and 'the entirety of the square is in the cube'. Spheres specify the middle interpretation, cubes are silent.
I think you misread what I said (twice). My point isn't that area effects must, it is that if area effects need not then neither do creatures.
The difference is creatures always occupy spaces that are neat increments of 5x5 squares.
If you're spell only touches the tiniest corner of my square, then I'm just going to stand in the other corner.
Not your call to make. "Space" is an abstraction. You don't get to pick where you are in that space. The rules clearly say you need all of that space to fight effectively. Any part of it is fair game for targeting you. This is akin to arguing you can increase your cover from Half to Total by staying out of the side of your square that's exposed.
It seems completely fair unless you can show the rule that says that a square partially covered by a cube is counted as affected by that spell's effect.
It's your job to show a partially covered space is excluded. Combat on a grid works exactly the same as combat without a grid except for a short list of exceptions and that's not one of them. Here's the relevant rule:
Choose an intersection of squares or hexes as the point of origin of an area of effect, then follow its rules as normal.
A cone is a cone and a cube is a cube whether you're playing on a grid or not. There's absolutely nothing in those rules about aligning cubes, cones or lines to the grid. If you stick so much as one finger inside the cube for Thunderwave, you are in the cube and suffer the spell's effects.
The distance between the diagonal squares in a 2x2 is 10' according to the rules for using a grid. The rules also tell us that there are 15 ft if we count to 3 before we get from the square below the top square to the bottom one.
The distance between two things and the empty space between two things are not the same thing. The squares are 15 feet apart. There's only 10 feet of empty space between them. Something that's 14 feet long will obviously span a 10 foot gap and then some. The diagram shows that.
The "template method" for determining who's in an area of effect in Xanathar's Guide to Everything is just as loose with regards to targeting:
The template method uses two-dimensional shapes that represent different areas of effect. The aim of the method is to accurately portray the length and width of each area on the grid and to leave little doubt about which creatures are affected by it. You’ll need to make these templates or find premade ones.
To use an area-of-effect template, apply it to the grid. If the terrain is flat, you can lay it on the surface; otherwise, hold the template above the surface and take note of which squares it covers or partially covers. If any part of a square is under the template, that square is included in the area of effect. If a creature’s miniature is in an affected square, that creature is in the area.
Actually, I realize that the problem is not limited to cubes, nor is it limited to playing on a grid. The key question is really "If a creature's space intersects part of an area of effect, how is the creature affected". If you assume any part of the creature's space overlapping the area means the creature is affected, you get the results from the OP, whether or not you're using a grid. The DMG rules for spheres require half the creature's space overlaps in the case of S/M creatures, but are silent about how it applies to creatures with a space that's larger than 5' or for non-spheres (it would probably be easier to resolve if you did away with the 50% requirement for circles and just said any overlap, but it results in significantly larger numbers of creatures affected for small AoEs).
I think you misread what I said (twice). My point isn't that area effects must, it is that if area effects need not then neither do creatures.
The difference is creatures always occupy spaces that are neat increments of 5x5 squares.
They do, but those increments need not align to your battlemat either. All that means is that 5x5 is the minimum space that a creature can take, but you can be 6 or even 7' from your closest neighbor. You cannot be 4', that would put you within the threatened area of another creature.
I also like the idea of just requiring that it is the other person's job to come up with rules when I can't come up with a relevant one. Show me a rule that says I can't have 8' between two creatures!
Show me the rule that says that the range between two objects is different from the amount of distance between them. We're really mincing words if we're trying to say that being 15' away from another creature doesn't mean that there are 15 feet between the two.
And sure there are some optional rules that say that when using a template, any square partially covered counts; but there are just as equally valid optional rules (in the same section) that say you have to have a token for each square included. Its hard to put tokens in 10 squares when you only start with 4. The default state of the game makes no such remarks about partially covered squares for cubes. (As an aside, I find that it is problematic that there is such a dramatic difference in the ways that the game offers in its optional rules for adjudicating areas of effect.)
And finally, the most important part: Whether its RAW or RAI, it isn't RAF to metagame the sh*t out of a 10'x10' area of effect spells (that covers 100 sqft) to hit 10 creatures (spread over 250 sq ft) when the DMG actually suggests that it should probably hit 2.
They do, but those increments need not align to your battlemat either.
Then you're not using the grid rules and there's nothing further to discuss. There's no point in talking about snapping to grid if you're only using the grid as a ruler.
Show me the rule that says that the range between two objects is different from the amount of distance between them. We're really mincing words if we're trying to say that being 15' away from another creature doesn't mean that there are 15 feet between the two.
If you have two creatures on adjacent squares they're 5 feet apart but there is obviously 0 free space between them. A third creature could not pass between them.
And sure there are some optional rules that say that when using a template, any square partially covered counts
Optional or not, it's still an accurate representation of the DMG's rules.
The default state of the game makes no such remarks about partially covered squares for cubes.
Which is precisely why it's on you to prove that a creature's space has to be fully inside an area of effect. As far as the core rules are concerned, if you're in a spell's area the spell applies to you, and very few of them require being fully inside the area.
Not only is there no rule that says a creature's space has to be fully covered by an area, it's practically impossible for a 5 foot wide line to do that. For one thing it's a flat shape, but even if you handwave that little detail you're basically saying that on a grid Lightning Bolt can only be used straight north, south, east or west against enemies that are perfectly aligned with each other and with you because otherwise it'll only partially clip their squares.
And finally, the most important part: Whether its RAW or RAI, it isn't RAF to metagame the sh*t out of a 10' area of effect spells to hit 10 creatures when the DMG actually suggests that it should probably hit 2.
The DMG does not suggest a 10 foot cube should "probably" hit 2 creatures. That table's there for when you've completely lost track of who is where and need a conservative ballpark figure. In fact you're first told to go with your gut if you're unsure That table will also tell you a Fireball should "probably" hit 4 creatures but we both know a Fireball can hit far more than that with ease.
I will admit that your reading is technically fine using optional rules (while obviously wrong using the other optional rules presented in the same section that still fit with the rules for combat and spellcasting, which by the way are in the PHB). That doesn't make it any more fun or less powergaming.
I'd rather have a working Lightning Bolt spell than quibble about powergaming. The rules cut both ways. Whatever the players can do, monsters can do as well.
Do you play with people or groups that oversimplify the Cube area mechanics?
I think many of us do, because many of us were oversimplifying it too.
In my experience, groups tend to make one big (and incorrect) assumption when it comes to Cube area-of-effects; That they want to or must squeeze
themselves into as small a grid space as they can. This is just not true.
Crack open your Thunderwaves and Clouds of Daggers, you've probably been shorting yourself how much space they can hit.
The PHB gives us exactly zero information on how any Areas of Effect act with a Grid... because the assumption (and the default) is that people are not playing with one.
What it does tell us is simply that wherever we put its Point of Origin, must end up being somewhere on one of the faces of the cube.
"Cube -- You select a cube's point of origin, which lies anywhere on a face of the cubic effect. The cube's size is expressed as the length of each side. A cube's point of origin is not included in the cube's area of effect, unless you decide otherwise."
The DMG gives us how that interacts with a grid, and simply adds that the Point of Origin must be at a grid intersection:
"Areas of Effect -- The area of effect of a spell, monster ability, or other feature must be translated onto squares or hexes to determine which potential targets are in the area and which aren’t. Choose an intersection of squares or hexes as the point of origin of an area of effect, then follow its rules as normal. If an area of effect is circular and covers at least half a square, it affects that square."
Notice that the special 50% rule only applies to cicular areas of effect... which a Cube isn't. Neither is a Cone, and most of us have had no problem understanding how those work.
So... Point of Origin at an intersection, Point of Origin anywhere along one side of the cube. That is ALL we must satisfy.
This illustrations shows the common configurations relative to the Point of Origin (red dot), most of us assumed were the only options [A and B].
But in fact C is entirely fine... and, this will be the shocker for most, so is D and MANY variations of D.
(The orange shows the squares that would be affected).
Thunderwave is a bit of a special case, in that Cube area spells don't normally also have a Range of Self. The description of Thunderwave, however, makes it clear that it effects "Each creature in a 15-foot cube originating from you". And as we know that our Point of Origin for a cube must be a grid intersection, it must be one of the grid intersections that make up your space. And travel from that Point of Origin AWAY from you. (So no shaping it funny to include yourself). Effectively place yourself in one of the uneffected squares beside any of those Points of Origin in the illustration, and you're good to go.
You seem to write with the assumption that it is ok for you to position your spell effect off grid, but the creatures you're trying to hit must snap to grid. That is awful. Since a medium creature needs about 5' of space, there is no way that a 2x2 square shaped effect can hit more than 4 creatures on the ground, no matter how you position it (or them). If the creatures are more than 10' apart, you cannot hit them.
Since creatures can move within their spaces, you can simply say that the four creatures in C are on the edges of their spaces, and that cube would miss them all. D1 and D2 both only fully cover 1 square, so can only ensure that you only hit 1 creature (on the ground).
I would agree with this. If movement and position must snap to the grid, spells should too. I’m not saying I wouldn’t allow diagonal placement, (potentially making the area more of a tetris z block than a square?) but it shouldn’t increase the area of the spell. Off the top of my head I think saying the spell must cover at least half the square to hit it keeps the number of target squares even.
if you are playing gridless and just measuring movement in inches, by all means go wild with fitting as much in the cube as you can.
I'm not suggesting a DM couldn't choose to have things snap to grid if they wanted.
Only that they would be inventing that, and many people do visualize cubes keeping locked to the Cardinal direction.
But the RAW works as illustrated.
Jeremy Crawford backs up both of those points (cubes aren't locked to grid, and the DM can decide otherwise) in a tweet.
"Cloud of daggers (5 ft. cube) can affect 1+ squares on a grid, unless the DM says effects snap to the grid."
https://mobile.twitter.com/jeremyecrawford/status/672590371835088896
I agree with Wolfof the Bees.
If you want to rule that that cube does not snap to the grid, than neither do the people. Which means that while you THINK you are sneaking your corner into the victim's grid, there grid is actually off set so it doesn't.
In other words, the only way you gain any advantage to doing this cheesy crap is if you can do it but the enemy does not.
Screw it, K.I.S.S. and just have everything snap to the grid.
And they're correct. The DMG's rules only require that the point of origin be on a grid intersection. A cube's point of origin can be anywhere on its surface, so in practice cubes don't have to snap to the grid.
In fact, there's no requirement for cubes to be parallel to the ground either. The only shape with an alignment requirement is the cylinder.
The rules even come close to suggesting that an area of effect needs to completely cover a creature's space (or one of the 5x5 spaces it occupies) to be affected, unless the area is circular.
The Forum Infestation (TM)
The DMG provides some options ...
Pg 249 Adjudicating areas of effect. This suggestion is aimed at theatre of the mind but could be applied to closely spaced creatures on a grid using minis.
"ADJUDICATING AREAS OF EFFECT
Many spells and other game feat ures create areas of effect, such as the cone and the sphere. If you're not using miniatures or another visual aid, it can sometimes be difficult to determine who's in an area of effect and who isn't. The easiest way to address such uncertainty is to go with your gut and make a call. If you would like more guidance, consider using
the Targets in Areas of Effect table. To use the table, imagine which combatants are near one another, and let the table guide you in determining the number of those combatants that are caught in an area of effect. Add or subtract targets based on how bunched up the potential targets are. Consider rolling 1d3 to determine the amount to add or subtract."
On a grid the DMG has the following rules for AoE. Pg 251
"AREAS OF EFFECT
The area of effect of a spell, monster ability, or other feature must be translated onto squares or hexes to determine which potential targets are in the area and which aren't. Choose an intersection of squares or hexes as the point of origin of an area of effect, then follow its rules as normal. If an area of effect is circular and covers at least half a square, it affects that square."
The last sentence indicates that if an area of effect is circular then it has to cover at least 1/2 the square in order to affect that square. I use that as guidance for applying all spell shapes on a grid. If the AoE affects 1/2 or more of a square (eyeballed and usually leaning towards the players benefit) then I apply the effect to the square. Otherwise not. A medium creature occupies a 5' square on a grid, however, it is not required to remain at the center of the square and it is not required to be stationary, as a result, I personally rule that when a player decides on placement of an AoE that it has to cover at least 1/2 the square in order to affect it.
Of course, application is up to individual DM preference. I personally take the specified AoE of the spell to indicate the maximum number of squares that can be affected. A 10'x10' spell can affect up to a maximum of 4 squares depending on placement. A 15' x15' cube would affect up to a maximum of 9 squares no matter how it was positioned on a grid ... but that is just how I run it.
The problem isn't the assumption that you don't need to snap a cube to the grid, the problem is that the OP assumes that the creatures must and the cube doesn't. Thinking that you can hit 10 enemies with a 2x2 area is problematic (D1). Taking a closer look at D1, the top squares highlighted are 15' from the bottom highlighted square, even though the area of effect is only 10' diagonal -- using the distance counting rules that the "using a grid" sidebar tell us to use. There is no way that I'd rule that something 10' across can hit creatures that have 15' of empty space between them.
Brian, it ALWAYS boils down to what each DM will allow and how they interpret. As such, your comment is not useful in any way, and is in effect meaningless.
The main objection we have is that the assumption you are making that because sphere specifies "half" and Square does not, that it must be "any part of the square in the cube." We disagree with your belief that "any part" is the default.
But that does not really matter, anymore than your comment about the DM having last say.
What it REALLY comes down to is this:
You are of course free to play as you desire. But I would recommend that:
As a DM, I won't let people exploit system granularity to enlarge an area effect, which means for case C you can only pick rounding on one side, and case D is just wrong even using the rules for circular area, he's showing the affected area as being every square touched by the effect, rather than the squares that are actually at least 50% covered.
Ok, so show me the rule that says your area effects have to snap to the grid.
There's only 10 feet of empty space between the bottom highlighted square and the topmost square it's aligned with. The diagonal of a 10 x 10 right triangle is sqrt(10*10 + 10*10) = 14.14 feet long.
All of the configurations in the OP are perfectly valid in theater of the mind and there's nothing in the grid rules that prevents this from working in that context either. If we want to get philosophical about this, to me it's a much bigger problem if someone's interpretation of the grid rules disallows normally valid plays.
The Forum Infestation (TM)
I think you misread what I said (twice). My point isn't that area effects must, it is that if area effects need not then neither do creatures. If you're spell only touches the tiniest corner of my square, then I'm just going to stand in the other corner. It seems completely fair unless you can show the rule that says that a square partially covered by a cube is counted as affected by that spell's effect.
Nope.
The distance between the diagonal squares in a 2x2 is 10' according to the rules for using a grid. The rules also tell us that there are 15 ft if we count to 3 before we get from the square below the top square to the bottom one.
I think David said it best. If you want to hit 3 medium creatures in a row, you simply cannot do it with a 10' cube, theater of the mind or not.
There are three basic options: 'any part of the square is in the cube', 'the majority of the square is in the cube', and 'the entirety of the square is in the cube'. Spheres specify the middle interpretation, cubes are silent.
DMs should say 'Yes, you can try that'. That does not mean 'Yes, you can try that and it works'.
The difference is creatures always occupy spaces that are neat increments of 5x5 squares.
Not your call to make. "Space" is an abstraction. You don't get to pick where you are in that space. The rules clearly say you need all of that space to fight effectively. Any part of it is fair game for targeting you. This is akin to arguing you can increase your cover from Half to Total by staying out of the side of your square that's exposed.
It's your job to show a partially covered space is excluded. Combat on a grid works exactly the same as combat without a grid except for a short list of exceptions and that's not one of them. Here's the relevant rule:
A cone is a cone and a cube is a cube whether you're playing on a grid or not. There's absolutely nothing in those rules about aligning cubes, cones or lines to the grid. If you stick so much as one finger inside the cube for Thunderwave, you are in the cube and suffer the spell's effects.
The distance between two things and the empty space between two things are not the same thing. The squares are 15 feet apart. There's only 10 feet of empty space between them. Something that's 14 feet long will obviously span a 10 foot gap and then some. The diagram shows that.
This is RAW and RAI whether you like it or not:
"If you're playing without a grid, distances in the D&D rules are meant to be read in their natural English sense. For example, I'm within 5 feet of you if any part of me is within 5 feet of you. We don't mean for you to mentally project a grid onto the action."
"AoE on a grid: You can target someone if any of their space is in the AoE. Spheres are the exception (DMG, 251)"
"The intent is that a circular area of effect is circular whether or not you're playing on a grid."
The "template method" for determining who's in an area of effect in Xanathar's Guide to Everything is just as loose with regards to targeting:
This is just the DMG's rules with a visual aid.
The Forum Infestation (TM)
Actually, I realize that the problem is not limited to cubes, nor is it limited to playing on a grid. The key question is really "If a creature's space intersects part of an area of effect, how is the creature affected". If you assume any part of the creature's space overlapping the area means the creature is affected, you get the results from the OP, whether or not you're using a grid. The DMG rules for spheres require half the creature's space overlaps in the case of S/M creatures, but are silent about how it applies to creatures with a space that's larger than 5' or for non-spheres (it would probably be easier to resolve if you did away with the 50% requirement for circles and just said any overlap, but it results in significantly larger numbers of creatures affected for small AoEs).
They do, but those increments need not align to your battlemat either. All that means is that 5x5 is the minimum space that a creature can take, but you can be 6 or even 7' from your closest neighbor. You cannot be 4', that would put you within the threatened area of another creature.
I also like the idea of just requiring that it is the other person's job to come up with rules when I can't come up with a relevant one. Show me a rule that says I can't have 8' between two creatures!
Show me the rule that says that the range between two objects is different from the amount of distance between them. We're really mincing words if we're trying to say that being 15' away from another creature doesn't mean that there are 15 feet between the two.
And sure there are some optional rules that say that when using a template, any square partially covered counts; but there are just as equally valid optional rules (in the same section) that say you have to have a token for each square included. Its hard to put tokens in 10 squares when you only start with 4. The default state of the game makes no such remarks about partially covered squares for cubes. (As an aside, I find that it is problematic that there is such a dramatic difference in the ways that the game offers in its optional rules for adjudicating areas of effect.)
And finally, the most important part: Whether its RAW or RAI, it isn't RAF to metagame the sh*t out of a 10'x10' area of effect spells (that covers 100 sqft) to hit 10 creatures (spread over 250 sq ft) when the DMG actually suggests that it should probably hit 2.
Then you're not using the grid rules and there's nothing further to discuss. There's no point in talking about snapping to grid if you're only using the grid as a ruler.
If you have two creatures on adjacent squares they're 5 feet apart but there is obviously 0 free space between them. A third creature could not pass between them.
Optional or not, it's still an accurate representation of the DMG's rules.
Which is precisely why it's on you to prove that a creature's space has to be fully inside an area of effect. As far as the core rules are concerned, if you're in a spell's area the spell applies to you, and very few of them require being fully inside the area.
Not only is there no rule that says a creature's space has to be fully covered by an area, it's practically impossible for a 5 foot wide line to do that. For one thing it's a flat shape, but even if you handwave that little detail you're basically saying that on a grid Lightning Bolt can only be used straight north, south, east or west against enemies that are perfectly aligned with each other and with you because otherwise it'll only partially clip their squares.
The DMG does not suggest a 10 foot cube should "probably" hit 2 creatures. That table's there for when you've completely lost track of who is where and need a conservative ballpark figure. In fact you're first told to go with your gut if you're unsure That table will also tell you a Fireball should "probably" hit 4 creatures but we both know a Fireball can hit far more than that with ease.
The Forum Infestation (TM)
I will admit that your reading is technically fine using optional rules (while obviously wrong using the other optional rules presented in the same section that still fit with the rules for combat and spellcasting, which by the way are in the PHB). That doesn't make it any more fun or less powergaming.
I'd rather have a working Lightning Bolt spell than quibble about powergaming. The rules cut both ways. Whatever the players can do, monsters can do as well.
The Forum Infestation (TM)
I mean, there is a token version of lines too.. having a working version of a spell doesn't mean that it has to cover 2.5 times its area.