Sekiri hands you another slim volume bound in pristine leather. "We have previously briefly discussed how the magic system and spells work, but have held off on some of the detail. This is a thesis I wrote when I was younger, with a few details updated recently. I encourage you to read it and come back to me if you have any questions."

The Magic System
An explanation of key concepts


SPELL ORDER

Many reference books, spell help studies, and other works often refer to the 'order' of a spell. Those orders are generally either zeroth, first, or second order, and those that are currently officially unclassified for whatever reason. Here we will explain what those different orders imply about the nature of a spell.

For a listing of all currently known spells and their official or unofficial classification, please refer to the appendix at the end of this volume. [Refer to Appendix A: Order of Spells]

Zeroth order spells:
Zeroth order spells comprise two differing sub-groups. The first of these, most often seen in offensive spells, indicate that the final results of any particular casting of the said spell, will, while theoretically based on the skills of the caster, be limited at a relatively low level due to the very nature of the spell itself. That is, while it will take a certain, fairly low, amount of skills in order to reach this cap, once it has been reached, no further improvement will be seen in that spell. A non offensive example of this is the spell Sorklin's Field of Protection. These spells are effectively low level spells for students to hone their skills with.

The second of these sub-groups are those spells that do not have variable results. That is, the spell either succeeds and the expected result occurs, or the spell fails, and the expected result does not occur. While there are some offensive and defensive examples of this, it is most prevalent in spells of the miscellaneous type. In these spells, improving skills are only relevant to whether a particular spell succeeds or fails. This does not necessarily indicate that these spells are low level.

First order spells:
First order spells likewise comprise two differing sub-groups of spells. The first of these, seen in most offensive spells, are those spells where the end result of a successful spell varies considerably from spell to spell, and those results are heavily influenced by the skills of the wizard casting it.

When these first order spells are offensive, that variability in results usually takes the form of the amount of damage done to the target, but may also be related to the duration of the spell, such as in Mugwuddle's Muddling Mirage. In defensive spells, the first order classification usually means that both the strength of the shield and its duration are increased. When we come to first order miscellaneous spells, the variable results are usually the duration of the spell, but may also be the 'power', such as in Master Glimer's Amazing Glowing Thing.

Clearly, for these spells, time spent improving the skills involved is beneficial. A more detailed discussion of how the determination of power is made will follow under the Spell Power section.

The second group of spells in this category are those where the actual difficulty of the spell itself will vary depending on the target of the spell. The classic examples of this are the spells Endorphin's Floating Friend, Al'Hrahaz's Scintillating Blorpler and Fabrication Classification Identification, where the spell difficulty increases when cast at heavier shields, more expensive jewellery, and higher level artifacts respectively. If those difficulty checks are passed however, the results are always consistent.

Second order spells:
The exact nature of the second order classification has been a matter for debate for some time. As currently understood however, it indicates that while the results of the spell are variable, and increase as the skills of the casting wizard do, any improvements in the results are only seen at certain discrete points, i.e. there is no continuous improvement.

The primary example of this is the spell Kamikaze Oryctolagus Flammula. Here, for a given set of skills, and absent a partial failure of the spell, the caster should mostly get the same amount of firebunnies in each initial wave. If the caster's skills are increased, no improvement in the spell will occur until they pass a certain threshold, at which time the spell will 'jump' up to the next discrete improvement, being one additional firebunny. A similar but harder to detect effect can be seen with the number of 'layers' in Chrenedict's Calcareous Covering, and the number of charges in Collatrap's Instant Pickling Stick. It should also be noted that second order spells may also have caps in place, for instance it is not possible to get greater than 12 firebunnies in the initial wave of Kamikaze Oryctolagus Flammula.

SPELL POWER

This section of the volume concerns those first order spells where the power of a spell improves as a wizard's skills do. It aims to explain some of the factors that influences the results of the spells, and strategies for a wizard to follow to improve the results of their spells.

Under the system of magic as presently found on the Disc, the majority of the power of any variable spell is based on the relevant magic.spells subskill, being offensive, defensive or misc. In fact, leaving aside the question of methods for the moment, the behaviour of many of the offensive spells indicate that a rough estimate for their normal or mean damage is a multiple of the offensive bonus of the caster.

For instance, if we take the spell Pragi's Fiery Gaze, it is estimated that the normal damage it causes is around three times the offensive bonus of the caster. If a hypothetical wizard had 400 bonus then, we would expect that the damage of the spell for that wizard would be approximately 1200 hit points. Of course, once we start talking about any one individual spell, the actual damage would be most likely either more or less than that, given the randomness present in almost any magic spell.

What determines whether that particular spell does say, 800 or 1600 hit points instead? Firstly, of course we must give credit to the randomness of the magic fluxes, which play a significant part. However, that randomness is influenced by the magic.method skills of the caster. If the wizard has high levels of skill in the methods used in the spell, then they are more likely to generate random casts above the 'normal' damage than they are below it. I.e. over a long period, the amount of casts of Pragi's that fall between the 1200 hit point nominal amount and say, 1600 hit points will exceed the number in the 800 to 1200 range. If they have low method skills, then the opposite is likely to take place.

We can see then, that while the potential, or normal power, of a first order spell is based on the spells sub-skill, the methods also increase the 'average' damage of a spell over the long term.

It should be pointed out here that when using the term high method skills, this is not necessarily talking about an absolute number, but high in relation to the required level of the spells. For instance, having 300 bonus may be high when casting a spell that only requires 100 bonus to succeed in a spell, whereas a bonus of 350 may be on the low side if the spell requires around 300 bonus to cast.

In fact, in trying to improve the average power of casts by improving method skills, one sees the greatest improvement if they advance the method that is lowest relative to the requirements of the spell. This effect often erroneously leads to some wizards believing that one particular method 'determines' the power of a spell, because they see so much improvement. Usually in this situation, one stage of the spell requires a significantly higher bonus to succeed than the others, and so the wizard's skills are lower relative to it. The classic example of this is Grisald's Reanimated Guardian, where the healing requirement for the spell is around a full hundred bonus points above the others. Given how much higher it is, some often think the strength of the warrior summoned is based on the healing bonus, whereas like all other first order spells of this type, it is based on the defensive bonus of the caster, with the individual results based on randomness influenced by all of the methods used, relative to the requirements.

Other than one or two spells where there is an additional check of one method, after the spell has already been successfully cast, no one method is any more important to the results of the spell, assuming that they are all 'equal' relative to the spell requirements.

So, we have seen how improving methods can help improve the chances of any one spell having a better result than the normal amount, and accordingly improve the long term average. What happens then if we advance offensive, holding the methods constant?

This simply seems to move the normal damage and the 'ranges' above and below it upwards on a sliding scale. For instance, continuing our example, if we improve offensive to 500 bonus, our normal damage is now 1500 hit points, with the methods then helping to determine whether a spell lands in the 1100 to 1500 range, or the 1500 to 1900 range. The impact that improving offensive has on the damage of the spell is therefore far more significant than the methods, once the methods are to a reasonable level.

Note that the ranges discussed here are for illustrative purposes only, and the actual numbers relating to the spell are likely to be different.

INHERENT RANDOMNESS

Magic is, by its very nature, fairly unpredictable. This means that from time to time, we will see very unlikely results from our spells, and even manage to fail casting spells that we have mastered for many years. Naturally, improving our skills, particularly the method or items sub-skills used in any magic spell can help to reduce this variability, but it will never be eliminated.

In addition, some spells are more susceptible to this than others, such as Jogloran's Portal of Cheaper Travel, where random results are common even for the most skilled wizard. In particular, some of these spells exhibit what we term partial failures of the spell. In these partial failures, often a wizard will see their spell succeed, but have the power of the spell halved. For instance, half the normal firebunnies on a cast of Kamikaze Oryctolagus Flammula, or a cast of Jogloran's which goes to a location other than that intended. It is not currently known exactly what causes most of these partial failures.

On the other side of the ledger, very fortunate wizards sometimes encounter 'fluke' casts of spells. These provide spell results of almost unimaginable power, such as an area effect offensive spell that completely destroys all present, or a cast of Transcendent Pneumatic Alleviator that is all but unbreakable for an extended period. Again, it is not known exactly what causes these or how to generate them. Simply count yourself fortunate should you achieve one.

It is suspected that there is another influence on the randomness of magic, and that is the very subtle touch of the Demon Murphy. He seems to take delight in making us fail spells, or have defensive magics fall at the worst possible time for a wizard. The path to living to a ripe old age as a wizard then, means we simply must take randomness into account when we are planning our activities. Relying on everything to go right is a sure way to an early grave.

BACKGROUND THAUMS

A final component of the magic system is the presence and creation of background magic levels where any spell is cast, and, naturally found in certain high thaum areas of the Disc, such as l-space and near the Hub. Certain locations in our guilds, such as the Unseen University, have been built to funnel this background magic away, but for the most part, it will be present where spells are cast.

Any spell, when cast, will add a number of thaums to the area based on the size of the spell. When repeated casting takes place in a room, these thaums can build up very quickly. While they will drain away, and in fact do so faster the more there are, for the unwary wizard there are many dangers to casting in the presence of high thaums.

This is because high levels of magic attracts the Thyngs from the Dungeon Dimensions. This is of course, not a good thing. When the levels are reasonable, there is no danger, but cast enough spells or blow up a magical artifact, and those Thynges may visit for the thaums, but stay for the wizard! There have been a number of recorded cases where wizards have suffered broken limbs, missing arms, and even death when casting further spells in a high thaum area. The wise wizard is a cautious wizard then, and always pays attention to the warning signs of gathering Thyngs.

Aside from the potential for unwelcome visitors, background magic thaums have an impact on any spells cast in the room. With the greater availability of thaums in any location, magic spells cast there tend to have the potential to be more powerful than when cast with only the internal energy available to the wizard. This can be quite a significant effect.

In addition however, it is thought that the presence of so much magic can not only increase the randomness in casting a spell, but can also raise the chances of those partial failures mentioned earlier.

********************************

URL: http://gunde.de/discworld/sekiri/system.html