Came up with this for White Star (OD&D in SPACE!!) after being told Level was more important than Attribute in OD&D. Also was influenced by Lindybeige's Youtube criticism of typical D&D task resolution. My system:

Character Competence = Attribute (eg STR, normally 3-18) + Level (1-10 in White Star)

Task Difficulty is rolled:
Easy - 3d6
Medium - 4d6
Hard - 5d6
Very Hard - 6d6

If Competency equals or exceeds Difficulty, the attempt is successful!

Quote from: S'mon;976058Came up with this for White Star (OD&D in SPACE!!) after being told Level was more important than Attribute in OD&D.

So you set out to reverse that? Your system is a cool way to do "roll under" and handle higher difficulties without dealing with modifiers, but it makes Attribute twice as important as Level. If your goal was to make Level more important, you'll have to rearrange things.

Yeah, as much as I like Xd6 under TN mechanics, your version does indeed make Attributes 3-18 times more important than levels for a 1st level character.
Now, if you go multiply Level by 3 and only add a +1 to +5 adjustment from Attribute, things would be indeed different, with Level outstripping Attribute in importance at 2nd Level. You'd have to make the Easy tasks starting at 1d6 or 2d6, but that just gives you more difficulty steps!

It's similar to the Traveller 4 and 5's way of doing task checks. But those games don't have leveling characters, which can mess with difficulty scaling.

It's pretty much what The Fantasy Trip does. Works great in TFT, which doesn't use levels at all, and which uses point-buy attributes so people have reasonable scores in them.

Attributes for starting human characters range from 8-16, average 10-11 (so 3d6 succeeds 50% of the time).

1d6 - automatic success
2d6 - easy tasks
3d6 - most somewhat difficult tasks
4d6 - challenging tasks (e.g. hitting a defending enemy)
5d6 - very hard tasks
6d6 - extremely hard tasks etc
etc.

Having appropriate talents (or not, or conditions) adds or subtracts dice from the difficulty of a task. Some things may also give a +/- to your ability (e.g. a well-balanced fine sword may give +1 to DX when trying to hit things with it, while trying to hit a snake is at -3DX).

Yeah, I don't think it's a problem for level to be more important than ability score.

Updated version following feedback:
++++++++++++++
Character Competence = Attribute (eg STR, normally 3-18) + Level (1-10 in White Star)

Task Difficulty is rolled:
Easy - 3d6
Medium - 4d6
Hard - 5d6
Very Hard - 6d6

If Competency equals or exceeds Difficulty, the attempt is successful!

Edit: Following feedback, here is an extended table with Long Jump example. Current world record long jump  is 29'. I went with 1' = 1 point of Difficulty, so your DEX 3 Level 1 PC can jump ("step") a 4' gap.

Task Difficulty is rolled:
Very Easy - 2d6 - eg jump a 7' gap
Easy - 3d6 - eg jump a 10' gap
Medium - 4d6 - eg jump a 14' gap
Hard - 5d6 - eg jump a 17' gap
Very Hard - 6d6 - eg jump a 21' gap
Heroic - 7d6 - eg jump a 24' gap
Formidable - 8d6 - eg jump a 28' gap
Amazing - 9d6 - eg jump a 31' gap
King of the Impossible - 10d6 - eg jump a 35' gap

For White Star (or S&W White Box) a maxed-out PC with an 18 attribute & level 10 has a 28 on their check, same as the average result for Formidable, and on average can make a jump just under the current world record.
+++++++++++++++++++++

IME character attributes on best 3 of 4d6 or recent point buy tend to be in the 8-18 range, with around 14 most common. For a level 1 PC obviously Level is only a small boost, that's the whole point. For Level 10, a low-stat character with Attribute 10 makes the check on an 18, vs a Level 1 stat 18 character makes check on a 19. Usual comparion is more a typical 13-14 attribute vs a top end 18. So for me that makes Level more important. Compare to 5e where Attribute bonus is -1 to +5 on the d20 (6 pt range) and Proficiency +2 to +6, a 4 pt range. My system for level 1-10 has a 1-10 or 9 point range, which is akin to playing 4e (+0.5 level to checks) from 1st to 18th level. And that's the most Level-matters verion of D&D task resolution I know. Only Castles & Crusades does this similarly, and then only for Proficient checks - plus a d20 roll is far more random than rolling a bunch of d6s.

Quote from: S'mon;976796Updated version following feedback:
++++++++++++++
Character Competence = Attribute (eg STR, normally 3-18) + Level (1-10 in White Star)

Some neat stuff here, S'mon, but it still seems to have the same issue raised by Xúc xac in his post. The stat (average of 10.5) still seems to dominate the result. Maybe you could do a "half stat (round down)" plus level? Something like that to make stat important, particularly at low levels, but eventually level will take over.

Oh, and you might need to knock 1d6 off of each of your difficulties if you make a change such as I suggested.

Quote from: finarvyn;976805Some neat stuff here, S'mon, but it still seems to have the same issue raised by Xúc xac in his post. The stat (average of 10.5) still seems to dominate the result. Maybe you could do a "half stat (round down)" plus level? Something like that to make stat important, particularly at low levels, but eventually level will take over.

Oh, and you might need to knock 1d6 off of each of your difficulties if you make a change such as I suggested.

Partly because stats are clustered around a bell curve, I don't think they dominate the result over the course of a 1-10 level game. Obviously in my current level 1 game the stats matter more though the +1 from level is a nice edge over normal non-classed NPCs. I think there will be a huge difference between a level 10 character and a level 1 character in competency - if you look at my Jump table you'll see that an 18-stat level 10 character is basically Captain America, they can routinely come close to the world record long jump (where a 10-stat level 1 character on average makes an 11' jump, this guy on average makes 28'), with 8 pts from higher stat and 9 pts from higher level.

I think maybe people are looking at the total stat number and seeing it's usually bigger than the total level number, and being misled, because they're not considering the likely variation among PCs.

Looking at say INT, in my best 3 of 4d6 in order 5-PC group they have 10 13 14 15 16 - a six point spread, smaller than the game's level range. For DEX they have 6 8 9 12 14, an eight point spread. Other stats look similar. CHA has a 9 pt spread 9-18.

Full stats here - http://smonstats.blogspot.co.uk/2017/06/white-star-toshe-station.html

Not a bad resolution system, but not my cup of tea.

Others have mentioned how you are heavily weighting attributes here - I don't have a big problem with that.  If one is to be realistic, the balance between natural talent and skill is going to depend on the task in question.  Some tasks rely ENTIRELY on training and attributes shouldn't matter much or at all.  So preference may be more important than realism here, since realism is unattainable.  Lately, I've had a preference for games with minimal character progression, so this would work for that.  But it seems a better fit for a game with skill levels instead of character levels, since character levels are supposed to be a bit more meaningful.

My main objection would be to the relatively steep bell curves that these rolls give you.  Again, in real life, the distribution is going to be all over the place, and may in fact be extremely steep in some situations (i.e. minimal deviation from the mean).  But for roleplaying?  As a GM, I only call for a roll when it would be significant and interesting.  Flat distributions are more interesting, to me, because they are more variable; i.e. "Anything can happen."  We get players rolling 1's and 20's every session.  18 on 3d6?  Not so much!

Personally, I'd favor a single die roll with a multiplier over rolling multiple dice, for this reason.  Also, if you changed the attribute scale (1-10, say), then you could use different dice for different levels of difficulty.

Quote from: S'mon;976806Partly because stats are clustered around a bell curve, I don't think they dominate the result over the course of a 1-10 level game. Obviously in my current level 1 game the stats matter more though the +1 from level is a nice edge over normal non-classed NPCs. I think there will be a huge difference between a level 10 character and a level 1 character in competency - if you look at my Jump table you'll see that an 18-stat level 10 character is basically Captain America, they can routinely come close to the world record long jump (where a 10-stat level 1 character on average makes an 11' jump, this guy on average makes 28'), with 8 pts from higher stat and 9 pts from higher level.

The average on 3d6 is 10.5 so attributes will usually be more than level. If you are using best 3 of 4d6, the average stat is even higher. Your Captain America character isn't getting 8 points from his stats. He's getting 8 more than average. His "competence" is 18 from stats and 10 from level. You seem to be comparing him to a level 1 character with a 10 stat. You should compare him to himself at level 1. At level 10, he has 18+10, but he's had that 18 since level 1. After 9 levels, his competence has increased 9 points which is less than an average 3d6 roll. Which means his average competence has gone up less than three dice steps on your 1 to 10d6 scale.

If you're playing a 10 level game, then level will always be less important than your stats except for your unusually low stats. Even if you managed to roll a 4 on your 4d6 roll to end up with a 3 in something, your attribute will still contribute more until you hit level 4, which is almost halfway to the top of your level scale. This system doesn't meet your stated design goal of putting more emphasis on level.

Quote from: S'mon;976806Partly because stats are clustered around a bell curve, I don't think they dominate the result over the course of a 1-10 level game. Obviously in my current level 1 game the stats matter more though the +1 from level is a nice edge over normal non-classed NPCs. I think there will be a huge difference between a level 10 character and a level 1 character in competency - if you look at my Jump table you'll see that an 18-stat level 10 character is basically Captain America, they can routinely come close to the world record long jump (where a 10-stat level 1 character on average makes an 11' jump, this guy on average makes 28'), with 8 pts from higher stat and 9 pts from higher level.

Clearly there is no single right answer for the relative value of level and stat.

But for comparison - let's say you had a level 1 military type with Strength 17, and you pitted him in combat against a level 6 military type with Strength 12. In most versions of D&D / OSR games, the level 6 would wipe the floor with the level 1 despite the attribute difference.

However, in your non-combat resolution system, these two are equal in terms of ability.

Ok I guess I do like attributes to be important.

BTW this system is supposed to work alongside standard d20 attack and save rolls.

Quote from: S'mon;976909Ok I guess I do like attributes to be important.

Nothing wrong with that. We were just saying that that's not what the title said.
If you are having fun, roll on!

One quirk of the system is how broadly character's become proficient in it. My 10th-level Magic-User with Strength 9 is going have the equivalent of a 19 Strength on his checks. If I were to do a system like this, I might limit adding your level to your class prime requisites. That's just me. If you want players levelling up all their attributes and it suits the feel of your campaign, go for it.