I took a brief look at the system. I'm not sure I fully understand every aspect of it, but I will say from my own experience that one of the major flaws presented in many D&D works is jumping to a quadratic model. I realize that in your system you do say that any formula could be adopted and you just used quadraticas an example. But having gone through this kind of thought often, I would offer up the help that the more you stick to a linear model the longer game balance will be preserved over the course of the levels.
The quadratic formula is only for XP so it does not affect the power-difference between levels and I only use it because I have to make up my own example XP system in order not to be in violation of product identity. XP systems are not OGL to the best of my knowledge. But as you point out later in your post you realized that upon rereading the document - I am just typing this to confirm that your new understanding of it is correct.
EDIT: I reread your work again ... I was troubled by my understanding of how you were using XP in relation to the points. ThenI realized that there is no correlation. So I understand a little better now.
There is no correlation at all. You can plug in any XP system you want to and the system will work just the same.
Just out of curiosity, why did you choose to increase each level by 1 in the base group? I'm curious because of the implication it leads to. It is inherently saying that classes generally are increasing faster in power at the beginning than the end. For example, an incease from 8 build points to 9 build points is a jump of 112.5%. However, an increase from 20 build points to 21 build points is only 105%. Just something to think about. I'm not saying yours is wrong, of course. I'm just lifting up an alternate model to help you reflect upon your own model. /EDIT
The formula is a compromise system. Some aspects of classes are frontloaded to first level (or first few levels), other aspects increase linearly and others increase exponentially. Hit points and saving throws, for example, increase linearly, so although the increase per level is the same (on average), in relative terms there is a decline in the power increase. So a first level character might have 8 hit points and at 2nd level he might have 16 hit points (depending on rolls and all that), which is a jump to 200% of 1st level hit points. By contrast a 10th level character having 80 hit points advancing to 11th level gains another 8 hit points, but it only moves him to 110% of his 10th level hit points.
In some cases there is even a decline in real terms of the power increase. For example, a Wizard at 1st level will usually be able to cast 2 fist level spells (one of them due to high intelligence), but advancing to level 2 only gives him 1 extra 1st level spell, so 1st level is actually worth more for the Wizard in absolute terms (not to mention in relative terms) than the second level.
Yet, there are also powers that accrue faster than linearly. For example, higher level spells, or higher order feats that require other feats as pre-requisites.
It would perhaps be possible to account for those factors, but it would make the system very complicated and each class would have to have its own point cost for each level (and occassionally the point cost would decline at a higher level). I decided to settle on a compromise system. Because of the above-mentioned factors, there couldn't be a uniform cost per level. On the other hand, an exponential cost would be too much. So I settled on an arithmetically increasing cost, which is a compromise solution.
In the many evaluations of races that have been done on the web by many people, base races are generally said to be worth slightly less than 1/2 of the first level of a class. Since first level is given as 11, giving 5 points for race, which is just under 1/2 the point cost of level 1 seems reasonable. The actual costs for base races would probably range from 3-5, so some points might be saved for later if a weaker race is chosen. Indeed, the system would allow for playing very weak races, such as Kobolds, without artificially making them stronger to make them on par with the core races. These weak races would simply cost even fewer level points.
The base of 10 points (11 at level one [10+level]) is arbitrary as would indeed be the case with any other number chosen. It would be just as possible to chose a 100 or whatever else, with the proportional change in all other numbers.
I already explained why I chose to use an arithmetic increase of level point cost. Now on to the reason why I chose it to be 1 point per level rather than say 5 points per level. The reason for that is the desired rate of discounting. The standard system has no discounting, but I think that is wrong and wanted to have some, but I wanted to be conservative about it too, so that LA races do not suddenly become the height of popularity. The numbers are designed in such a way, that using the base/low cost progression LA loses 1/2 its value after 10 levels of advancement and 2/3 of its value after 20 levels of advancement in relative terms. For example, if a character is of a race that is currently classified as LA +2, the race would normally be translated to approximately 23 level points. Initially, the character would be 2 levels behind the rest of the group in his fighter class, but 10 levels later this would decline to only being 1 level behind. LA +3 character would be only 1 level behind after 20 levels, LA +4 character would be only 2 levels behind after 10 levels and so on. The numbers were chosen to ensure that this happens.
There were, of course, also other design considerations at hand. For example, most analyses show that wizards et all only really start to pull ahead of fighters and such at late single digit/early double digit levels. As a result, the cost of all levels is the same until level 9 at which point it begins to diverge. Before that point, Wizards, for example, might even be weaker than fighters, which would ostensibly call for their levels being cheaper, but I did not want to overcomplicate things and in any case other classes that are powerful later, such as Clerics, are not exactly bad even at these low levels.
Another thing I kept in mind was that Clerics, Druids and Wizards ought to be able to attain ninth level spells by the time they reach ECL/General Level 20. So I had to chose the class numbers in such a way that they reach level 17. If this is to hold for the Sorcerer too, than the Sorcerer should probably be moved into the medium cost group.
If you are looking to get away from automatic imbalance in class levels and multiclassing issues (especially in referance to spellcasting ability and its availability/unavailability to allclasses) check out my sig. Dreamscarred Press recently put out Complete Control, which is a system designed to refigure the leveling process. I offer it up as an example, since you are interested in redefining balance in the game. Since you seem to enjoy the math behind character design, I'll add a caveat that there is one chapter within the product that is designed precisely for people that like to see the math. The mathematical explanations for the tables/progressions within the work are all within that particular chapter. I would encourage you to look it over.
I also know that looking at other people's systems can really help inspire your own. I looked heavily at Buy the Number before refining mine. I wanted to see what that work did that I liked and what I disagreed with. In looking at that work (and a few others) it helped me refine my own.
Anyway, I hope that as you look at character design you are inspired to always tweak your system to represent the model you want. Tinkering is a wonderful thing!
Thanks, I might give these a look. I do love tinkering!
