Attribute scores are irrelevant in OGL systems

[Necron 99]

Something I noticed while picking apart rules in various games systems based off of the OGL, is that, post AD&D 2E, the actual attribute score for characters is essentially a useless number. Every WotC edition of D&D falls under this, along with games like DCC and C&C, not to mention many OSR games.

In original D&D and AD&D, the score gave you several specific modifiers or target numbers for accomplishing a variety of tasks like opening doors or the number of languages a character knows, reaction adjustment, etc. In addition, the number may also give modifiers for attacking in combat or additional HPs.

Sometime in Basic and then later in AD&D, the score itself became useful in that it gave the DM a number by which characters could resolve additional tasks related to that attribute, using the roll-under method. The roll-under mechanic was later polished up and became the system used in Non-Weapon Proficiencies and Secondary Skills. Rolling under the score was an attempt to determine how well that character performed using that particular attribute. Want to remember some obscure knowledge concerning history? Roll under INT or WIS to succeed. Want to try and persuade the innkeeper to give you a free meal with your room? Roll under CHA to see if it happens or fail and get tossed out, maybe charged double.

The attribute score played a direct part in determining how well a character succeeded, or not, at a particular task.

All of this changed when D&D 3E came out. In the later d20 system, attribute scores denoted a modifier which was then applied to every d20 roll. Rolling a skill check? Roll a d20 and add your modifier, beat a target number. In the modern d20 system, the modifier is the number that then becomes the number directly affecting the roll, not the actual attribute number. The same d20 system then also added in another mechanical number by way of "target number". Now, the character no longer rolls against their own personal score, but a seemingly made up series of target scores, each denoting the difficulty of the challenge faced (5 Easy, 10 Average, 15 Challenging, 20 Impossible, etc.).

In AD&D, a character with an 18 CHA would have a base 90% chance to succeed if they roll equal to or less than their score. In the d20 system a score of 18 lands you a +4 modifier, the Average target number to beat is 10. On a d20 roll without any DM modifiers rolling a 10 or greater is a 55% chance, with the +4, the character has an assured modifier of 20% so essentially the character has a 75% chance to succeed at the task. It is actually harder to succeed in D&D 3E than it is in AD&D, when looking at the baseline percentages. In order for a character to have a similar chance in D&D 3E, they would have to roll with a +7 modifier, which imo, is ridiculous.

When the D&D system flipped to using a d20 roll over system, it introduced (in my personal opinion) an unnecessary additional mechanic by requiring the addition of more modifiers on top of modifiers on top of modifiers, etc. With AD&D, you roll the d20, if it's under the attribute score, it succeeds and while the DM could impose situational modifiers, I'd argue that it is still a more simple mechanic for task resolution.

Daniel of the YT channel, Bandit's Keep has a couple pretty good videos discussing the changes for skill checks from TSR through WotC era D&D. I'm intrigued by the notion of using d6s as a way to determine attribute checks, first introduced by Frank Mentzer back in Dragon Magazine. I want to look into this resolution system more and possibly implement it into future games.