Chad Perrin: SOB

18 March 2009

Alternative attribute score system for D20/OGL/PRPG

Filed under: Geek,RPG — Tags: , — apotheon @ 12:23

This is part of my RPG series of entries here at SOB. See the inaugural entry in the series for more details.

I’ve come to prefer a point-buy system for attributes in D&D 3.x and PathfinderRPG, pretty much lifted directly from Mike Mearls’ revised edition of the D20/OGL Iron Heroes game originally developed by Monte Cook — yes, that Monte Cook. I’m rethinking that, though, in favor of a point-buy system using an actually different means of representing the attributes themselves. The new system is essentially a drop-in replacement for the standard D20/OGL system, without any need to alter anything else in the core rules to use this new system other than how attribute bonuses and penalties (by race or level progression) are assessed.


The reason I’m considering using this modified attribute system is that the Iron Heroes point-buy system doesn’t solve a big problem I’ve had with D20/OGL attributes all along (as well as attributes in several other systems, including OD&D): some attributes are basically “dead air”. That is to say (using D20/OGL as the example) that, for instance, increasing your Strength from 11 to 12 gives you a +1 to attack and damage rolls, but increasing it from 12 to 13 doesn’t give you anything in-game. This causes some minor problems such as skewing the motivations for players to buy up to certain attribute numbers as they get their bonus attribute points with level advancement. It changes strategies for character advancement, and even the roleplayers most dedicated to the craft of non-metagaming character development must feel tempted at times to increase one stat rather than another simply because of the stat modifier effects.


I’ve recently become increasingly interested in taking an alternative approach to attribute numbers. Instead of 3-18 attributes with modifiers based on the attribute, I could use the modifiers themselves as the attribute numbers, rather than using numbers between three and eighteen as intermediaries between character creation and in-game effects of attributes. I think this works particularly well for games that use a point-buy system for character creation.

In the following table, there are three columns: the first, “Score”, is the attribute score for the new system; the second, “Cost”, is the cost of purchasing the attribute score at character creation using the point buy system; the third, “Correspondence”, shows the equivalent 3-18 scores (actually 1-20) for standard OGL 3.5 and Pathfinder RPG rules. This chart works not only for giving you an idea of what each modifier-style attribute score means, and for determining the point-buy cost of each score, but also as a means of determining attributes based on a randomized dice-roll character creation system; if you want to roll for your stats, just roll like usual, and just translate from the number in the Correspondence column to the modifier in the Score column (which, in case I haven’t made it obvious, is equal to the attribute modifier for each number in the Correspondence column according to official D20/OGL rules anyway).

Score  Cost    Correspondence

-5     -21     01
-4     -12     02 to 03
-3     -06     04 to 05
-2     -03     06 to 07
-1     -01     08 to 09
+0     +00     10 to 11
+1     +02     12 to 13
+2     +04     14 to 15
+3     +07     16 to 17
+4     +13     18 to 19
+5     +25     20


Now that I’ve explained it all in terms of how you get from a traditional 3-18 D&D attribute system to this modifier-as-score system, I’ll let you in on a secret: that’s not really how I ended up considering this as an alternative system. It really started with a “Modern Horror Genre” roleplaying game a friend and I cooked up a couple decades ago for our own use, shortly before the first edition of the White Wolf Vampire: the Masquerade game came out, that we called Inner Sanctum. That system used a random dice roll system to generate attributes that were modifiers rather than numbers that translated to modifiers. Stats were generated using 2d6-7 to produce values between -5 and +5.

More recently, as I mentioned in passing in How do you feel about energy resistance?, I’ve been working on the design of another original roleplaying game. Its working title is currently Apotheosis RPG, though that’s subject to change. Inspired by my own previous work on Inner Sanctum, I decided to go with a basic -5 to +5 system for attributes. I was also inspired, to some degree, by the desire to avoid the problems of the traditional 3-18 system, as well as those of GURPS’ approach of rolling against the attribute numbers themselves as a kind of difficulty target.

By default, I’m looking at a point-buy system for determining attribute scores in ARPG, and the numbers above that match up with the ultimate effects of using the Iron Heroes point-buy system for values from 3-18 are numbers I was considering using, though I won’t settle on anything for sure until I know more about how the rest of the system will fit together (of course). The above table is, in essence, the result of me directly porting the working system for ARPG to Pathfinder RPG (or, in effect, D20/OGL), with the addition of the Correspondence column to give a quick point of reference for making transition from the traditional 3-18 system to using modifiers as attribute scores.


All in all, it’s not a really dramatic change. Things are basically the same as they ever were with a D20/OGL compatible system, other than notation. The minor changes, in practice, involve only adjusting attribute bonus and penalty points and skipping all the annoying “some attribute points don’t really matter” stuff.

The first part of changing the way attribute score bonuses and penalties are handled is dead easy. In both the D20 3.5 system and the Pathfinder RPG system, all race-based bonuses and penalties come in twos. Divide the numbers by two, so that a +2 becomes a +1 and a -2 becomes a -1, and you’re golden.

The second part, which consists of dealing with attribute bonuses as a character advances in level, is a little less obvious. In the default system for D20/OGL compatible games including both D&D 3.5 and Pathfinder RPG, a single discretionary point of attribute score bonus is acquired at every level divisible by four. It may seem arithmetically obvious that those points should just be divided by two as well, but that results in no attribute score increases except at 8th and 16th levels. Another option might be to split the difference, going with a bonus point every level divisble by six (6th, 12th, and 18th levels). I think a better option would be to divide by two, but round up, giving a bonus point at 4th, 12th, and 20th levels.

edit: As John points out in the first comment response to this SOB entry, spells and similar effects that temporarily alter attribute scores might also need some conversion, which is a fact I managed to completely fail to notice when I first wrote this. That complicates things significantly for me, if I want to explain all the necessary changes, but I think that for the most part such conversions should be very easy for GMs to sort out on a case-by-case basis.

John also suggests just going with a +1 discretionary attribute score bonus every four levels, exactly as in the standard character progression charts, even though the (theoretical) arithmetic effect in attribute bonuses is doubled. The argument is that the effect in play is not the same as the effect on the attribute modifiers themselves, which is true. Still, it changes the rate of attribute advancement, so I’m reluctant to just jump off the ledge on this one.


I think that about covers it. Have I forgotten anything? What you do you think of the idea? Is there anything I should do differently?


  1. You would also need to adjust spells that modify ability scores, too.

    I would argue that there really isn’t a 1/2 point increase every 4 levels. Based on your earlier comments, you say that people tend to tailor their character to take advantage of odd value stats for the increases. Sometimes players will focus on a single stat or an even numbered one, but in general people will in fact plan things based on even/odd ability scores for increase. Why not just stick with a +1 at every 4 levels? Attribute modifiers become less and less meaningful as characters go up in level in 3.x/OGL stuff anyway. At 1st level having a +4 Str makes a significant difference to a to-hit roll. It ammounts to 80-100% of the bonus to hit (dependant on class), at 20th level having a +4 Str will only ammount to 16% of the bonus to hit (at most, given magic and other modifiers). So, if we go with your idea of 4,12,20 (or 6,12,18) we end up with +7 Str (a 25% modifier) vs +9 Str (a 31% modifier). There is at most a 6% increase with the extra two points. While the extra +2 becomes a little more significant for things like Spell DCs, It is not likely to significantly change ability to make these DCs….

    Comment by John — 18 March 2009 @ 02:29

  2. Good point about spells. Somehow, I forgot about that.

    As for the rate of stat increase due to level advancement, the recommendation I offered is intended to approximate the rate of advancement in the core rules. If you want to increase it faster than that, go for it — and please let me know how that works out. I may test that myself at some point.

    Comment by apotheon — 18 March 2009 @ 03:18

  3. I agree that it won’t break anything to have a revised attribute point every four levels.

    For spells and other temporary effects, just cut the modifier in two, rounded up. Another option is to make the modifier the actual change. That is, a -3 to CON from poison would subtract three points from every constitution-based roll until healed. That does double the results of modifying effects, but it is still equitable and balanced as both bonuses and penalties would be affected.

    Comment by Golgotha Kinslayer — 18 March 2009 @ 04:32

  4. That reminds me of what Green Ronin has done in True20. If you haven’t checked that out yet, you should.

    Comment by Stargazer — 18 March 2009 @ 10:46

  5. I haven’t read the True20 books. I saw something online where someone mentioned they did away with hit points, though.

    I’d like to take a look at it, at some point, but I’m not in the mood to spend money on it right now.

    Comment by apotheon — 19 March 2009 @ 10:01

  6. What Stargazer said. I loved the idea when I read it in True20, and it still gets my thumbs up. The 3-18 ability scores are a remnant from 2nd Edition, no need for them to stick around.

    Comment by amoeba — 1 April 2009 @ 05:42

  7. What happens to death/paralysis from ability score loss?

    Comment by tom frome — 12 July 2009 @ 04:37

  8. There are a few options for that sort of thing, I guess. One that appeals to me is a saving throw against attribute losses. I’m only just now about to start using this attribute score system for real, so we’ll see how it plays out when the characters run up against attribute losses. I frankly haven’t thought much about attribute loss, though — in part because it doesn’t come up much. I guess, for now, my answer is “wing it”, but I might come up with something a bit more concrete.

    If you have any great ideas, I’d love to hear about them.

    Comment by apotheon — 13 July 2009 @ 10:12

RSS feed for comments on this post.

Sorry, the comment form is closed at this time.

All original content Copyright Chad Perrin: Distributed under the terms of the Open Works License