Compromise and Conceit

Having written yesterday’s post about magic and its faults, I am today not so convinced that the problem I described is a big deal; and in any case, it can be resolved to some extent by inventing magic items which improve wizard’s save DCs. Nonetheless, I think the magic system is anachronistic in AD&D, representing the accumulation of years of ideas from different versions of the game, and it can be both streamlined and redeveloped so that it a) fits the skill resolution framework and therefore b) is amenable to the introduction of a spell-free magic system.

Here I aim to present some alternative methods of spell resolution, based on skill systems similar to the combat system. I will for now avoid discussion of the problem of how many spells a character gets, or the cost of using them. Let us first consider some methods of resolving the fundamental problem of magic – the spell attack – through a couple of skill resolution methods based on different fundamental magical cosmologies:

1. More powerful magic is harder to cast and harder to resist: We imagine that more powerful magic requires greater effort by the caster to use, and more skill to handle, but unleashes greater power that is more difficult to resist. Set the DC of casting the spell to be the same as that of resisting it, calculated as for example 15+(spell level). The player rolls the spell-casting roll using spellcraft, and if successful the spell is cast at the stated DC. If unsuccessful the spell fails and/or costs more to use. If successful, the magnitude of difference between the failure and the DC determines the size of the effect.
2. More powerful magic is harder to cast and easier to resist: This framework is closest to the existing philosophy of the d20 skill system. Consider, for example, a move silently challenged skill check. The difficulty of the check is determined by the listen skill roll of the person being stalked, but the rogue’s skill roll is modified by situational modifiers related to the environment and the difficulty of the particular stalking task. Creeping up behind someone to backstab them with a dagger is harder than creeping within 30′ to ambush them using a bow, and should be modified differently. With magic, we assume the greater energies being manipulated in more powerful spells mean that the caster is less able to focus them accurately, making them easier to resist. Consistent with this philosophy, the save DC of the spell is fixed by the caster’s spellcraft check, with a penalty equal to the level of the spell. The magnitude of the effect is determined by the degree of the target’s failure. Then the target makes a save against this DC. Under this model the hardest spells to resist will always be the ones with least effect, which the mage has most familiarity with.
3. Difficulty of resistance is entirely spell dependent: This is the current system, where spells are like a device independent of the training of the mage – when read, their effects are unleashed with an objective power level over which the mage has limited control. In this model there is little  use for a skill check

So, I just made a pair of characters using the rules exactly as written down in the DMG: a 19th level wizard-hunting thief and a 19th level wizard. The wizard-hunting thief used his spare cash to buy a stash of potions and scrolls; used his 19th level special ability to gain skill mastery in Use Magic Device (so those scrolls can be used with impunity) and bought a Cloak of Resistance +5 with the spare change. This gave him saving rolls of +14 to Fortitude, +23 Reflexes, and +13 Will once he had taken potions of Bull’s Strength and Owl’s Wisdom. The Bull’s Strength also renders him immune to Power Word Kill (128 hps) but he bought a scroll of Protection From Evil and Death Ward to make sure none of the Symbol or Power Word spells can do anything (not that this is relevant, since his Composite Shortbow guarantees he is out of range of any of the Wizard’s spells). For good measure he bought some flaming adamantine arrows.

The wizard, on the other hand, has only 68 hps and is not expecting to be attacked. This is only fair, since the thief will only get to use one unique thief special ability (sneak attack) once in this battle (though I suppose he could use his spare change on a cape of the Mountebank and dimension door behind the wizard in round 2…) The mage has a contingency with Stoneskin, because what mage wouldn’t have? 

The key point of this little tale is this: the Mage’s most powerful spell has a save DC of 28 (if he has spell focus), so the Rogue only needs to roll over 5 on a reflex, or 14 on a fortitude, or 15 on a will save. Now, granted, 15 is not so great. But if we look at the wizard spell list, the only long range spell which isn’t  reflex save-based is Horrid Wilting, an 8th level spell, so the thief will resist it on a roll of 13 or more. These aren’t good odds, really, and they just get worse as we go up in levels. Because the save DC for spells increases only with the improvement in the caster’s intelligence (i.e. once every 8 levels), while the save bonus increases every level, or every three levels. So by the time the wizard’s save DC goes up by 1, the rogue’s saves  have all increased by at least 3.  

Note that when the wizard was first level the thief had to make a roll of 13 or more to resist the wizard’s spells if they required a fortitude save, or a 9 or more if they required reflex. So the 18 intervening levels have led to no benefit for the wizard in attacking the rogue’s weakest defences using the wizard’s supposed special ability, and he has actually gone backward against the rogue’s strongest defences.

This entire problem arises because the wizard has a “special ability” (magic) which is not at all special – everyone in AD&D gets magic items – but the power of the wizard’s “special ability” is hampered within the rules to prevent it being “too powerful”. Of course, the entire idea of magic is that it should be too powerful (it’s meant to be scary) but in AD&D its mundane use, and this deliberate weakening of its power, means that in reality it is not actually very frightening at all. By taking the wizard’s power out of the context of the basic framework of the system (the level-based skill system) the designers have essentially weakened the wizard relative to all other characters.

And his saves are themselves shit – so he is his own worst enemy!

Having described the imbalances between the combat and skill systems in the d20 system, I here present my proposal for an alternative combat system which combines them properly. This description relies on use of my proposal for ability scores which are represented entirely in the form of the ability modifiers; and it should be assumed that all skill checks are conducted with 2d10 (much better probabilistically; I may have something to say about this at the end).

The requisite skills: Under this system saving throws are developed as skills, labelled Fortitude, Reflexes and Will, and the Fortitude skill total also indicates the total number of wounds a PC can sustain (lethal or non-lethal). The base attack bonus is also developed separately as two skills (Melee and Missile). 

 Wounds: Health is measured not in hit points but as a total number of wounds which can be sustained, equal to the Fortitude skill total. Exact handling of wounds can be campaign-specific, but I envisage a character becoming incapacitated and capable only of limited movement when wounds received equal total wounds, then unconscious when wounds increase further; and dead after a failed Fortitude skill check (i.e. a failed Fortitude save).

Armour Class: Armour class is determined using the Reflex skill, with armour penalty applied and no armour bonus. In combat armour class is considered to be a Reflex skill save, which is usually done by taking 10, so armour class becomes 10+Reflex total. However, characters with a shield or the dodge can choose to make a skill check to improve their armour class, and take the better of the skill check or taking 10. This turns the combat roll into an opposed skill check. The armour bonus of the armour worn counts as damage reduction.

Example skills and wounds: Depending on the skill development system few characters will have rapid advancement in all these skills. A first level fighter might have 3 ranks in Fortitude and melee, and perhaps 1-3 ranks in reflexes. Let us assume for now 3 ranks in Fortitude and 1 in reflexes, giving skill totals of 5 and 3 respectively, assuming above-average ability scores of +2 in Constitution and Dexterity. Further suppose a melee skill of 3 with a +3 in Strength, giving a total of +6. This means a first level fighter can sustain 5 wounds. Wearing a chain shirt, the fighter will have a total reflex skill of 1 (due to armour penalty of -2) and damage reduction of 4. Usually this will give an armour class of 11 and damage reduction 4, but with a shield the PC can opt to make a challenged skill check, raising armour class as high as 21 depending on the roll.

The attack roll: The attack roll is a combat skill check challenged by the defenders reflex skill, which is usually resolved with the defender taking 10. The difference between the attacker’s roll and the target DC is damage done, which is reduced by damage reduction. The result is the number of wounds of damage done. The weapon used sets the maximum damage done after damage reduction (which may differ by armour types), and if the damage is reduced to 0 or below it becomes simply 1 non-lethal wound. For example, a dagger may have a maximum damage of 1 wound, while a greatsword has a maximum of 5. (We will see later that this rebalances spells so that they are not as dangerous as the bigger weapons). Criticals occur automatically when the critical range is rolled (which happens much less frequently on 2d10) and do not automatically double damage; rather, they increase the wounds done by the roll, and the maximum number of wounds, by some figure. A rapier, for example, may have a critical effect of +2; it increases the damage done on a critical roll by 2, and increases the maximum damage by 2.

Wounds as a penalty: Wounds received are applied as a penalty to the attack roll but not the reflex save, so that as combatants suffer wounds there is a simple game mechanic for reducing their effectiveness. In character development this leads to a trade-off between focussing on toughness, defensive ability and offensive ability. Too much focus on offense means that the PC will collapse rapidly in combat; too much focus on wounds means that the PC can keep fighting through damage, but will rapidly lose efficacy; too much focus on defensive ability means that the character is hard to hit but dies quickly (and will likely be at their best in light armour with low penalties). Note also that much damage sustained by heavily armed fighters will be non-lethal wounds. Making non-lethal wounds easy to heal will mean that they can be seen as shock, bruising and stunning effects rather than actual wounds.

Multiple attacks: As a fighter gains combat skill they can pick up multiple attacks. Rather than resolving these as multiple dice rolls, these multiple attacks increase the maximum damage the fighter can do by a multiple, enabling the single combat roll to have a more effective maximum ceiling. For example, a fighter with 2 attacks wielding a weapon with maximum damage of 3 will be able to do a maximum damage of 6 in a single round. This maintains the old description of combat as an abstract process, with the round containing multiple feints and attacks (under such a description there is no justification for multiple attack rolls); it also allows a fighter’s lethality with a weapon to increase beyond the weapon’s basic statistics.

Initiative: can be resolved as a challenged reflex skill check, though I am considering introducing a new skill, Presence, which covers initiative, fear, and Cleric’s turning checks. If initiative is resolved using reflex, we can incorporate character’s dodge/shield defensive skill checks in the roll.

Example attack: Imagine our example first level fighter fighting himself, with a longsword whose maximum damage is, say, 4, with a critical effect of +2. First our antagonists roll initiative, and because they have shields we use the result also as their reflex skill check to set the DC of attacks. Fighter 1 rolls a 4, so his initiative is 5; Fighter 2 gets a 12, for a total of 13. Since Fighter 1 rolled less than 10 he takes 10 on his reflex skill check, which sets the DC of attacks against him as 11 for this round; Fighter 2 uses the initiative check result as the DC of attacks against her. Fighter 2 then rolls to attack Fighter 1, rolling 7 for a total of 13. This beats the DC by 2, doing 2 damage, which is absorbed to 0 by Fighter 1’s damage reduction. Fighter 1 therefore suffers a single non-lethal wound. This acts as a penalty on Fighter 1’s melee skill, whose total reduces to 5. Fortunately Fighter 1 rolls a 13, giving a total of 18. This beats Fighter 2’s hit DC of 13 by 5, doing 5 wounds; after damage reduction this becomes 1, i.e. 1 lethal wound.

We now commence round 2. We retain the initial initiative roll but reroll reflex for the shield defense checks. Both players roll 9, so the combatants take 10 on their defense rolls, therefore setting attack DCs of 11. Fighter 2 attacks first with a penalty of 1 (for her 1 lethal wound) for a total adjustment of 5. She rolls 19 – a critical! The total of her attack roll is 24, which does beats Fighter 1’s defensive DC of 11 by 13, doing 13 wounds, which is reduced to 9 wounds after damage reduction. Usually the maximum damage Fighter 2 could do with a longsword is 4, but the critical has increased this to 6, so Fighter 2 does 6 wounds to fighter 1 after damage reduction. Since Fighter 1 only has 5 wounds remaining, Fighter 1 goes unconscious. Game over.

Rationale: This system merges combat and its results entirely within the skill system, and introduces some elements of the Rolemaster system, specifically the idea that heavily armoured fighters are easy to hit but hard to damage. It encourages the development of combat styles by establishing clear tactical differences between agile, tough or offensive combatants, and making it difficult for PCs to be all 3. It removes the damage roll and (by dint of using 2d10) the threat roll for criticals; at higher levels it removes all rolls for extra attacks. It introduces a single roll for the challenged combat skill check, but only when a combatant is parrying, has the dodge feat or is wielding a shield. It increases the lethality of heavy weapons, and gives a direct comparability between magical and physical attacks, both in damage done and the skill checks to be used. Under this system combat should require a maximum of 3 dice rolls in the first combat round and a maximum of 2 thereafter. Finally, it introduces an armour penalty on reflex saves (now handled by the reflex skills) but allows the possibility that armour protects against traps and magical attacks through its damage reduction. It also makes non-lethal damage a more important part of adventuring.

There are a few things about the d20 system which I think are unnecessarily cumbersome or burdensome and serve no real purpose, or are a historical artifact. Two which I think definitely need adjustment – one because it is just so silly, and one because it is so clumsy – are the stats system, and the way power increases unevenly in different areas of the system.

Stats: It is completely unnecessary to have a system of basic stats (strength, dexterity etc.) in which the ability modifier (i.e. the bonus) is separate and determined entirely by the ability score, particularly when the relationship between the two is formulaic – i.e. a +2 increase in score gives a +1 increase in modifier. Just write the stat as a bonus or penalty, and make ability modifying items and spells half as powerful or half as frequent. Level-based ability increases occur every 8 levels instead of every 4; +1 ability score magic items are half as frequent; Bull’s Strength gives a +2 to strength; and so on. I have DM’d for beginners many times, and they always say the same thing – this is stupid.

Uneven levelling: The disconnect between the hit point and spell systems and the skill system means that there is a fundamental difference in power increases between the two. Consider, for example, the move from 1st to 2nd level for a well-powered character. This character will have a roughly +6 bonus in their main skill (which for rogues, particularly, is an essential skill) and a +4 to their main save or base attack. They will have a single die of hit points with (if a fighter) a small constitution modifier. When they go to 2nd level the main skill will increase by 2, i.e. 30%; the main save by 1, i.e. 25%. But if they rolled 2 hps at 1st level, the chances of an at least 100% increase in hit points are very high – 3 in 4 for a mage, 9 in 10 for a Fighter. Sure, if they rolled high at level 1 and then roll poorly at level 2 the opposite may occur, but chances are they will experience essentially a 100% increase in their survival ability; but their opponents will become only slightly more dangerous. Meanwhile, the wizard’s spell DC does not change at all, and in fact only goes up by 1 at 3rd level. In fact in general the main save for any class increases at twice the rate of spell DCs, and gets many more magical enhancements over time. The weak save of all classes increases slower than a wizard’s spell DC, but after adding in a level-appropriate magic item it will always be equal to the spell DC. For example, at 3rd level a weak save is +1; the Wizard gains 2nd level spells, so the wizard’s spell DC increases by +2. So any character with a +1 to saving throws due to magic has essentially equalled the wizard’s spell DC, with the only difference occurring if the wizard has high intelligence or feats.

This problem of differential powering occurs because of the disjunction between the skill system (which is secondary for all non-rogue/non-ranger characters) and the save/combat system, on the one hand, and the hit point and spell systems on the other. Were character health and spell power handled as skills, they would level up in the same general way as combat/saves and skills. Because weapon damage doesn’t increase with skill, and because spell DCs are weak compared to saves, character hit points and resistance to magic increase rapidly compared to weapon and spell power. There is no way for a fighter to kill another fighter of more than 3rd level with a single, well-timed hit except by means of a  critical, so no fight carries serious risk if the odds are in a PC’s favour. Similarly there is no way that a character at higher levels will regularly fall prey to spells except by very bad luck, since they only need to roll over 10 on their weak save, or 5 or 6 on their strong save, to beat most magic. Hence the phenomenon of parties tending to have opponents of much greater power than their individual levels.

By comparison, in rolemaster every fight was potentially deadly. Low-level opponents were always dismissed eventually, but there was always a risk of a very nasty accident. Sometimes rolemaster errs a little on the side of deadliness, and certainly games like Traveller and WFRP are way too dangerous to encourage swashbuckling. But the d20 system as it stands hardly encourages fear or caution… unless you are that ultimate in useless creatures, a wizard with no back-up…