Classes as Skills, Part Two

Continuing on from the introduction to my Skills system, let's revisit the character sheet I included the other day:






Note there's been one change to the character sheet since I took this picture - Thief skills have been split from Scout.

Points Totals
The right margin was cut off with the scan - there should be subsection points totals on the right hand side - 32 for combat, 8 for athletics, 16 for general, 128 for spell lists - with a top right hand overall total of 188 (32 + 8 + 16 + 128). There's also a total of 128 for spell skills at the bottom, but they don't contribute to the total.

What are these spell skills? Why are they in a separate box? And why weren't their points included in the total.

Spell Skills
In addition to spell lists, there are three special spell skills - Detect Magic, Artefacts, and Runes.
But who wants to waste points on skills like these?
Well, sadly, in order to be a good spellcaster you have to study the fundamentals of the art.
The way this works is that spells cost twice as many points, but then for every point you spend in a spell list, you get to spend a point for free in one of these three special skills.

The spell skills cost does not contribute to the total, and can be anything up to the points in spell lists.

For example, Mazak had 128 points in spell lists, so can spend up to 128 points in spell skills, which just happens to be exactly the amount spent.

What about Athletics? That looks more complicated.

Athletics
Athletics is the pure build up of muscle power, and calculations are derived from this instead of directly from Strength. It gives you your Power bonus (which is used for the kill bonus, and range of missiles), Lift (in pounds) and Speed (in feet per round).

Power and Lift are on the first line - they both use stat ST but are also affected by your SZ (Size Mod).
For example, Dwarves are SZ -1 so get -1 on Power and Lift.

Speed is on the second line - it uses the stat AG but is also affected by your HM (Height Mod).
For example, Dwarves are small for their weight so have HM -1, so get -1 on Speed.

While Power is a simple bonus, Lift and Speed are real world values so are looked up on a table.

Finally, you cannot increase muscle power without limit, so beyond level 3 you stop getting any power bonus, and lift stops improving (there is no limit to speed).

The cost is split onto a new line after level 3 to remind you of this.

As you can see from the character sheet, Athletics is the only thing affected by your height and weight. Also Lift and Speed are the only things on the character sheet you need to look up when spending skill points.

What are those blank lines for?

Misc Skills
There are a couple of blank lines for other skills if anyone takes them:

Sage (ME, by subject e.g. History, Geography etc.)

Languages (ME, by language)

Music (IN Singing, or by instrument)

Arts (IN Painting, Sculpture, Dancing etc.)

Science (ME, by subject e.g. Alchemy, Maths, Architecture, Engineering)

Hey, how come there aren't any social skills such as Bribery, Intimidation et.c.?

Missing Skills
Just like there are no Logic or Charisma stats, there are no skills which should be resolved through role playing.
Comparing with D&D 5e, this does away with 4 of their 18 skills (Insight, Deception, Intimidation, Persuasion).

Hang on a second, I thought this was a skill system: why is there a level listed at the top right of the character sheet?

Levels make a reappearance
I didn't want people to have to spend points on boring skills which everyone needs to improve at: Perception, Spell Saves, and Incapacitation.
These three are not treated as skills which you take levels in, they simply use the level of the character, which is derived from the number of skill points they have to spend.
At first level you have 10 skill points to spend (i.e. 100 XP). When you get to 20 / 40 / 80 / 160... skill points you are second / third / fourth... level.

For example, in the example character sheet the character had spent 180 skill points, which makes him 5th level, so his level on the back (LHS) of the character sheet for Perception, Spell Saves, and Incapacitation is 5.

Combat Manoeuvres?I'm currently considering adding combat manoeuvres such as Disarm,Trip, Multiple Opponents. It could use the same system as Spell Skills (if I want to track bonuses in it), OR I might have a "Combat Level" derived from the number of points spent in combat skills. Either way I'm not sure if Athletics or Acrobatics should add to the total or not.

That's pretty much covered the whole skill system. In the third and final part I'll cover how this system reduces to a system of classes and levels with no need for skill points when you don't want to be bothered with them.

Classes as Skills, Part One

As I posted here and here, it is a tricky path to create any skill/class system. You can't please everyone, and my first few attempts didn't please even me - but they clarified what I liked and what I didn't! Then I tried writing my own multi-classing rules for D&D, but couldn't resolve what to do when you got a fighting or hit point bonus from two different classes. I resolved that by removing any such bonus from all classes except the Fighter, with the only way to improve in these being to multi-class in Fighting. At the same time we were playing Star Frontiers, and appreciated its points system and the concept of broad skills (though the way they are split is awful). This melting pot of ideas resulted in the following system.

Spending Points
For every 10 XP you get to spend 1 point on skills.

You start with 100 XP and 10 skill points to spend.

Skills cost 1 / 1 / 2 / 4 / 8 / 16 / 32 / 64... points for 1st / 2nd / 3rd... level.

Thus for example, third level costs you 1+1+2 = 4 points. You don't have to add this up as when you cross them off:

1 / 1 / 2 / 4 / 8 / 16 / 32 / 64...

The total cost is the first number not crossed off.

Total Bonus
Each level gives you +1 bonus.
Each skill has a single stat which gives a bonus.

All the possible skills are printed on the character sheet, with the name of the stat, and a space to fill in that stat bonus, so updating or checking the bonus is simple.

Skill Resolution
You roll 2d10 (open) and add your bonus. The standard difficulty is 11.
If you want your new character to be good at a skill, then you'll have +1 in the attribute and spent 4 points on getting +3, for a total of +4. Thus for an 11 you'll need to roll 7+, which is 85% chance of success. For a difficulty 20 this is reduced to 16%.

Combat Skills
There is Melee (AG), Unarmed (AG), Thrown (AG), Bows (IN), and Parry (REF).
Melee covers all melee weapons, Thrown covers all thrown items, Bows cover all ... bows.
Parry is about parrying with any weapon, or shield, or chair, or just plain dodging.
Note that Bows rely upon Intuition since they're all about judging the shot, whereas Thrown is more about hand-eye co-ordination so is Agility.

General Skills
There is Acrobatics (AG), Thief (IN), Scout (IN), Ranger (IN), and Healer (ME).
Acrobatics covers Climbing, Swimming, Diving, Tumble, Balance, Catch, Juggle, Escapology etc.
Thief covers Pick Locks, Disarm Traps, Disable Device, Pick Pockets, Sleight of Hand, Disguise etc.
Scout covers Move Silently, Hide, Tail, Covert Observation etc.
Ranger covers Track, Navigate, Ride, Animal handling, Survival, Foraging, Trapping etc.
Healer covers Herbs, First Aid, and Surgery.
By choosing one of these you are good at a whole swathe of loosely related things, so they are liable to be useful frequently.
These obviously map to my concept above of "class but without the fighting skills".

Spell Lists
There are 16 spell lists divided into 4 categories: Elementalist, Wizard, Enchanter and Sorcerer. You choose one as your primary area. You can learn a spell list in another area, but only if your primary spell lists are already at that level.
For example, an Elementalist can only take a level of Summoning (from Sorceror) if they already have Earth, Air, Fire, and Water at first level.
Spells cost you twice as much as other skills - i.e. 2 / 2 / 4 / 8 etc.
Details of spell lists and spell points are for another day.

That's it for the basic system. Next time I'll cover remaining skills such as Athletics and Perception. In part three I'll cover a character class system that is 100% compatible with the skill system - players in the same campaign can choose either system for their characters.

Today I Hate Class Systems Because...

A Clarification due to comments I received on these two posts:
 A lot of people fall into either the "I hate classes" camp or the "I hate skills" camp, whereas I love the good bits of both and am frustrated by the downsides of both. So before presenting my own system (where classes are just pre-packaged sets of skills), I listed the pitfalls I’m trying to avoid in two tongue-in-cheek posts heavily influenced by the fantastic http://roll1d12.blogspot.co.uk/2015/04/so-your-character-is-jackass.html.
 This classes post is obviously about D&D - various incarnations thereof!

Roll d12.

1. But I don't want to be the Cleric!!

2. If classes are so good, why in Greyhawk, the OD&D supplement that introduces the Thief, does it also introduce triple-classed Elven Fighter/Magic-User/Thieves?

3. What do you mean I can't wear any armour?

4. Getting to be a 5th level Thief and finally realising that Thieves are rubbish.

5. The eternity between first level and second level, then suddenly you're twice as good!

6. Bards. Need I say more?

7. Why can't my Fighter learn how to climb?

8. Why doesn't a level 0 character have zero hit points?

9. Random class feats. "Beginning at seventh level you have honed your memory skills to perfection. In any situation you will be able to instantly recall a rule that the DM has forgotten about."

10. How come every Innkeeper in this city is a Fighter or a Thief?

11. What on earth's the difference between a Fighter, Barbarian, Cavalier, Berserker, Gladiator, and a Swash Buckler?

12. Bilbo was a Thief. No he wasn't, he was a Fighter. So how many locks did he pick, or traps did he disarm? Ah, but he picked a pocket and was sneaky! But he failed to pick that pocket - he was better at fighting. (Argument continues ad nauseum).

Today I Hate Skill Systems Because...

A Clarification due to comments I received on these two posts:
 A lot of people fall into either the "I hate classes" camp or the "I hate skills" camp, whereas I love the good bits of both and am frustrated by the downsides of both. So before presenting my own system (where classes are just pre-packaged sets of skills), I listed the pitfalls I'm trying to avoid in two tongue-in-cheek posts heavily influenced by the fantastic http://roll1d12.blogspot.co.uk/2015/04/so-your-character-is-jackass.html.
 The classes post is obviously about D&D, but for the skills post there is no system that has all these irritations, they change with the system you’re using. For example, Old school D&D Thieves has #9. D&D3/3.5/Pathfinder has #1, #2, #5, #10. D&D5E has #10. Star Frontiers has #6, #9, #12. Rolemaster has #1, #2, #3, #5 (it happened, I did that), #6, #7, #9, #10, #11. Runequest/BRP/Call of Cthulhu has #1, #2, #9. #4 and #8 are things I've observed in other systems on the net.

Roll d12.

1. Spending points in swimming, then it never ever, not even once, being needed.

2. Going to dive into the pool and discovering that you can't as you have "Swimming" but not "Diving".

3. Hunting around for ages looking for what to spend that last point on.

4. Character sheets which go onto a second sheet of paper.

5. Spending two weeks writing a computer program to make character creation easier, then realising that's two weeks of my life I'll never get back.

6. Being the world's greatest swordsman, but being unable to swing a mace without dropping it on your foot.

7. Realising that I spent points on something that I'll never use, and the which annoys me every time I see it on my character sheet as I could have used those points for X.

8. Weird dice rolling systems that make it completely unclear how likely you are to succeed at anything.

9. Percentage systems that make it really clear how good you are at things. That is, clearly rubbish at everything!

10. Bribing persuading the town guard being resolved with a dice roll against "bribery 55%".

11. Having a big budget of points to spend, then having to spend half of them on boring things like Perception, Body Development (!), or Armour (!!).

12. Why on earth to be Stealthy do I also have to learn "Finding Directions" and "Analyzing Samples"?

Stat Generation: Random but fair - this time for D&D

Previously I gave a random (but balanced) method for generating stats for my game Explore. This was a random method of generating three pairs of bonuses for Strength/Con, Agility/Reflexes, and Memory/Intuition where the total bonus is 0, and no pair differs by more than 2.

Unfortunately this method cannot easily be adapted to D&D stats because there are too many possibilities to list them all in a table!

So here's a method for generating random stats for D&D where the sum of all the stats is always the same:

Rolling Stats

Draw out a 3*3 table as follows:

3	4	5	2	5	2	13
4	3	3	4	5	2	12
1	6	6	1	5	2	12
	13		7		6

Roll one d6 nine times and fill in the table in order (left to right starting at the top):

3		5		5		13
4		3		5		12
1		6		5		12
	13		7		6

Then sum each row and write the sum at the end:

3		5		5		13
4		3		5		12
1		6		5		12
	13		7		6

Then for each of your 9 rolls fill in 7 minus your roll:

3	4	5	2	5	2	13
4	3	3	4	5	2	12
1	6	6	1	5	2	12
	13		7		6

Then sum these new columns:

3	4	5	2	5	2	13
4	3	3	4	5	2	12
1	6	6	1	5	2	12
	13		7		6

Your 6 stats are these column and row sums: 13, 13, 12, 12, 7, 6.

Add them up and check that they add up to 63!

Stats of Stats
So what's the chance of rolling an 18 with this system? You need to roll all 6s in a row, so 1 in 216 chance for these, or all 1s in a column, so again 1 in 216 chance. So the distribution of stats is almost identical to that of the normal 3d6 system.

It isn't as balanced as my system for Explore - using the bonus tables from B/X it is possible (just) to get as much as a total bonus of +3 or -3 (e.g. 18, 18, 9, 6, 6, 6). It works much better for the 3rd edition stat bonuses, in which case the total bonus is always between +1 and -1. Either way I would allow anyone with a total negative bonus to increase their lowest stat until the bonus is zero. For example 15, 15, 15, 12, 3, 3 is -3 total so you increase the 3 to 9 and get 15, 15, 15, 12, 3, 9.

Encountering Difficulties?

I've read many discussions on how lots of low level opponents compare to a single high level opponent. For example recently at Roles, Rules & Rolls, in Overkill and Monster Experience Roger discusses eight 1HD monsters compared to a single 8HD monster and says they are much easier to kill - he calls it overkill.

"a group in control of tactics is going to have a much easier time dealing with 8 x 1 HD orcs than 1 x 8 HD giant, for any number of game mechanical reasons."

In contrast at Hack & Slash, Courtney Campbell complains that in D&D5e

"An encounter with 12 CR 1/2 Kobolds is more threatening than an adult dragon."

So how does the number of monsters and PCs affect the difficulty of the encounter?

Let us imagine 1 monster versus one PC and that it takes the PC three rounds on average to kill the monster. Then, on average, the monster gets in about 2.5 attacks before it is killed.

What about 3 monsters versus one PC? It takes nine rounds to kill all the monsters on average, but initially there are 3 attacking you until you kill one, then 2, then 1. So on average each round you would imagine you get hit 3, 3, 2.5, 2, 2, 1.5, 1, 1, 0.5 times, or 16.5 in total; about 6.6 times as many times as one monster would (roughly a triangular number progression).

When I ran a combat simulation it actually came out slightly higher than this – the strength was increased by the square of the numbers. I previously measured strength of opponents via a combat simulation to be how many of them you could kill on average (facing one after another) before they kill you. When you repeat this with multiple opponents at a time, the number you can kill is roughly divided by the number of opponents.

So the effect of changing numbers in a combat is to alter the difficulty of the combat by the square of this ratio: halving the number of PCs makes it twice as hard, treble the number of monsters and you make it nine times as hard, double the number on both sides and the difficulty remains the same.

We have previously seen that an 8HD monster is about 25 times as powerful as a one HD monster, this is the same as meeting twenty five 1HD monsters one after another. But this rule shows that the group of eight 1HD monsters as a single group is 36 times as powerful as a single 1HD opponent. Hence eight 1HD monsters in a single group is a more serious challenge than the single 8HD monster - the opposite of what you might expect!

Similarly the 12 kobolds in reality may be poorly armed and armoured, and you may be able to use an area attack such as fireball, and there are odd rules about multiple attacks specifically against 1HD opponents, but they certainly have the potential to be lethal against a lone high level fighter.

This corroborates with my experience - single high level foes are surprisingly easy to kill when they are outnumbered, there is a lot of truth in the adage “safety in numbers”. To make a challenge from a high level monster you need to give them something more than just better chance to hit and more hit points.

So how can we design our game / scenario / monster to mitigate against these issues - stopping single high level monsters from being a push over, and keeping hordes of low level monsters a challenge, and giving appropriate XP for both types?

Monster Tactics
As Roger points out, tactics can change the difficulty of a combat markedly. We don't want our mass of low HD monsters to be a push over, a meaningless slog fest, or easy XP. Monsters are not mindless zombies (unless, of course, they are mindless zombies!) - if it is obvious that they are individually outclassed but have greater numbers they should not allow themselves to be bottlenecked. If on the attack, they should look to ambush or sneak around behind and surround the PCs. If on the defensive they should retreat until they have bottlenecked the PCs.
Similarly, single monsters should bottleneck opponents, it's especially good for them to block the exit! If in danger of being surrounded, they should retreat so their back is to the wall.

Appropriate XP
In White Box D&D XP was linear with HD (100XP per HD) leading to the "too much XP for low level monsters" problem discussed above. In Greyhawk this was replaced with the familiar roughly exponential system, but this means that in Greyhawk you get 65 times as much XP for the 8HD monster, which is over generous (or, rather, the XP for 1HD is stingy). Killing a single high level monster gives you  inordinate amounts of XP, so should probably be reduced down to only 25 times as much (in line with how hard they are). Here is a suggested revised XP table:

Level	White Box XP	Greyhawk XP	Revised XP
1	100	10	25
2	200	20	50
3	300	35	100
4	400	75	175
5	500	175	250
6	600	275	350
7	700	450	450
8	800	650	600
9	900	750	750
10	1000	900	900

This is a balancing act. In Explore I've reduced the amount of fighting required to progress significantly, hence there is a real danger of massive improvement from a single encounter, so I reduce the XP even further (to only 12 times as much XP).

Extra lethality
Instead of making the monster harder to kill or harder to hit, make the attacks more lethal.
In D&D you find that Giants (for example) get extra dice damage, which makes them tougher, but most of all that makes them capable of killing you, and a single blow becomes potentially fatal. For example, if a party of PCs meets a large number of Orcs or a single Giant, the Orcs may do more damage in a combat, but they're liable to spread that around as a few injuries which are easily healed, whereas the giant may take out one of the party with a single hit.
Poison or paralysation or other nasty attacks (level drain anyone?) also make the lone monster “feel” tougher as the consequences of being hit.

Individual Initiative
My experience with side based initiative is that if you win initiative against a single foe you can often wipe it out without it making a single attack, which is a massive anti-climax. So I’d recommend individual initiative in such situations, even if you don’t use my initiative system the rest of the time.

Henchmen
Give the monster henchmen who soak up the PCs: E.g. this Giant has a bunch of annoying Orc lackeys who engage the PCs in combat and stop them all attacking the Giant at once.

Engaged In Combat
Note that henchmen work best if you insist on people being engaged in combat with an opponent – if someone attacks you, you have to engage them in combat else they get a rear attack bonus, and you can only attack someone you’re engaged in combat with. In general this also helps stop people all ganging up against a single opponent.

Area Attacks for Monsters
Dragons get their toughness from their area attack. A bunch of kobolds attacking a Dragon would be wiped out. Area attacks stop people emasculating your nasty monster purely by virtue of numbers. I'd either restrict such monsters to always be on their own, or to disallow multiple area attacks of the same type against an opponent in a round  - e.g. if two dragons breathe on you in the same round then you only save against one.

No Area Attacks for PCs
On the other hand, Area Attacks for PCs are very problematic - they turn low level monsters into easy XP, and they can wipe out all the henchmen - which is one reason why I've removed mass spell attacks from Explore. Fireball comes from Chainmail where it makes sense, but it's massively overpowered in D&D. Gygax nerfed it in OD&D firstly so that you have feet indoors versus yards outdoor (which is random), and secondly so that it blows back and kills the PCs, which is not the sort of game I like to play - in fact I think this sort of nerfing is always a bad idea (magic user armour restrictions and super low hit points, demihumans level limits).
You can't really remove Fireball from D&D though and still call it D&D...

Multiple Attacks

Give the monster multiple attacks against different opponents.
E.g. allow a giant to attack up to three man-sized opponents so long as they’re not behind him (one swing at several opponents) – one attack roll for each of them. A Dragon could do the same with its tail and attack everyone behind it. Using engagement rules again, this advantage only comes into play when the giants are outnumbered.

Squaring the Cube

Doubt Sets In
In Sizing Things Up, I presented a method for working out the weight of a creature when you modify its size. The reason behind wanting such a system is that I like things to be consistent, and to work out a rule once that can then be applied everywhere, instead of having to make lots of little decisions.

Based upon what I read about the inaccuracies of BMI, I made weight go up with the 2.5^th power of the height of a creature. That is, if you make the height 4 times as much, the weight goes up 4^2.5 = 32 times. This seemed to give realistic values, so I didn't worry about it any further.

The other day on Semper Initiativus Unum, Wayne blogged about the weight of Giant Spiders, and he invoked the Square-cube Law. to say that it goes up with the 3rd power. That is, if you make the spider 4 times the leg span, the weight goes up 4³ = 64 times. This would indeed hold if the spider kept the same proportions, but does it? (Everyone seems to think so). Or does it only go up by the 2.5^th power? Do different families of animals have different rules, or is it the same power for all families? This got me thinking hard about the veracity of my rule...

There is a lot on the internet about Allometry studies, and all the ones I saw proved the cube rule - except - they all had something else in common. They all dealt with a specific species of animal, not a family of animals. That is you might find that a particular species of fish obeys the cube law, but there is no data about fish in general. The variations within a species are often purely due to age or environmental factors or sex. What we're interested in is extrapolating from a small animal to a large one in the same family and then applying the same rule to a giant one, so these findings are not necessarily applicable.

Finding Some Data
Spiders seemed like the obvious place to start. Unfortunately, although the leg span of spiders is fairly easy to find, their weights are elusive. People don't bother to weigh small spiders! On wikipedia I could find the Black Widow Spider (1.5'', 1 gram) and the Goliath birdeater (11'', 6oz). It is 170 times heavier. My rule predicts that as it is 8 times bigger, it would be 8^2.5 = 181 times heavier, whereas the cube rule suggests 8³ = 512. Hence the 2.5 power rule seems good in this case, but this limited data hardly constitutes as proof!

Looking further I discovered a paper on the web about flying bird wingspan versus weight, and it gives a graph with a 2.42 power rule:

(It's actually got the axes reversed in the paper). The outlier (top left) is the Black Throated Loon.

This is hopeful, so I thought I should try a few different families of animals.

There's a whole load of data on snakes, which I imported:

So that's a 2.61 power rule based upon length. The outlier (at the top) is the Burmese Python. A lot of other snakes are far thicker than you'd expect, until you see that they are snakes like the Boa Constrictor.

For horses and ponies I went for an ideal weight chart based upon the height of the animal's back, (which looks to be about proportional to the overall height) plus zebras and donkeys from Wikipedia:

That's a 2.55 power rule based upon height.

So what about cats? I found data from wikipedia for Wildcat, Jaguar, Bengal Tiger, Leopard, Asiatic Lion and Feral Cat. Not a great sample size but anyway:

There's a 2.63 power rule. Now for this one I had a choice of minimum and maximum height and weight, so I choose the maximum given for each species, and also I has the data for both the length and height - the length has a 2.33 power rule, whereas I've given the height, in keeping with the horses.

For whales I used wikipedia again for Hector's dolphin, Dwarf Sperm Whale, Pilot Whale, Killer Whale, Beaked Whale, Sperm Whale, Blue Whale:

That's a very nice match for a 2.64 power rule.

For deer I used Wikipedia and the Scottish Forestry commission for Southern Pudu, Northern Pudu, Moose, Elk, Roe Deer, Fallow Deer, and Red Deer (scottish):

Which is a 2.72 power rule.

I also tried collecting data for snails, but had the same problem as spiders for getting reasonable weights. I got three values and a power rule of 2.83, but I spent more time stumbling across disturbing experiments about the force to crush a snail shell, or their weight loss when left to dessicate in hot air... 

Crocodiles looked promising, and I got five values and a power rule of 2.31, but it seems rather tricky to measure the length and weight of a crocodile accurately for some reason. I can't imagine why. The weights and lengths are often estimates from a distance!

Crabs seem to be all different shapes so their weights are all over the place; and I thought I'd have the same problem with fish, so I didn't try with those. If you could find a family of similar shapes you could try the rule, but you're in danger of being selective with your choices just to get the rule to work. Hence you can only really choose families where there's size variation but they all look similar.

I had no luck for scorpions, or for centipedes, both for the same reason as spiders.

The last family I successfully got data for was the Tortoise/Turtle family. The Leatherback Sea Turtle, Alligator Snapping Turtle, Giant Tortoise, Speckled Padloper Tortoise, Marginated Tortoise, and Red Footed Tortoise:

Which is a 2.78 power rule.

So the power rules I've found are 2.42, 2.61, 2.55, 2.63, 2.64, 2.72, 2.78. The average of this is 2.62.

So for the length/height of a family of creatures of similar shape, the 2.5 power rule seems a pretty good approximation - far closer than a cube rule, and close enough for my purposes. So we're on pretty solid ground to use this  rule for the weight of a giant version of any animal. There can be big variety of forms in a family of animals - so for a giant boa constrictor choose a boa constrictor as a starting point. With dinosaurs for example there are three basic shapes of dinosaur - 2 legged, 4 legged, and 4 legged with a long neck - so you'd have three starting archetypes for deriving all the weights.

Squaring This Observation With "The Accepted Wisdom"
So what explanation can there be for why this rule seems to hold so widely instead of the cube rule? I know my data's not great, but the results are certainly very consistent.

That the cube rule holds for deriving the weight of giant animals seems to be accepted wisdom. It's the rule everyone seems to use, and it's even used for disproving the possibility of giant animals. Elephants are used as an example of how big creatures have to become stocky and stumpy to cope with their great weight and they have to have big ears to cope with their low surface area to volume ratio, whilst Giraffes are studiously ignored.

The 2.6 power rule simply says that as you get bigger, you become thinner or elongated. Take a look at photos of a Blue Whale compared to a dolphin, or the body of a cat compared to the body of a lion. It seems clear that, precisely because of the Square-Cube law, your shape has to change to maintain a reasonable surface area. That is, as animals get bigger they don't size up according to the cube rule, precisely because of the Square-Cube law!

So does any of this matter? Perhaps not, but it was fun looking up all these different animals, and now I know that some centipedes look really mean. I'm going to have to show the players a photo next time they meet a giant centipede!