The default seems obvious: use the d20 Test system as described in the 2024 D&D Player’s Handbook. It is fairly simple and most people are used to it. However, it is very random—swingy is the word a lot of people use. Your ability to do stuff is based more on chance than on chosen skills. Perhaps a system that places more emphasis on skills than luck is worth a look.

Because of this I have been thinking about using d4s as a sort of pool. You would roll a number of d4s equal to your attribute modifier plus your proficiency bonus (if proficient) and add the results to see if you succeed. That works OK for skills—where not being proficient means you won’t have a lot of dice, so your chances are a lot lower. This makes your choice of skills to be proficient in more important.

But we could even apply this to singular attack-plus-damage rolls. Say you are using a longsword or battleaxe (both use a d8 for damage). If you have a Strength modifier of +3 and a proficiency bonus of +2, then you roll 5d4 to see if you meet or beat the target’s Armor Class. But what if we also added the damage die to the same roll? Then you would roll 5d4 + 1d8, subtract the AC, and any remainder is the damage dealt.

Similarly, a spellcaster using Fire Bolt would roll their spell-attack modifier in d4s, plus the d10 for damage—totalling 5d4 + 1d10.

For leveled spells that normally roll lots of damage dice (e.g., Guiding Bolt—4d6), the caster might be allowed to turn one of those damage dice into a Risk Die to ensure a hit. The die is added to the attack total but then omitted from the damage result. Using your highest roll guarantees a hit but lowers damage; using a low roll risks missing but keeps damage high.

That might be OK, but the idea that spellcasters are simply more powerful yet limited by spell slots is fine too.

Advantage and disadvantage change. Instead of rerolling everything, add one extra d4 and then:

• Advantage – remove the lowest die.
• Disadvantage – remove the highest die.

Stacks of advantage / disadvantage add more d4s and remove multiple high or low dice appropriately.

Example (two × advantage, one × disadvantage): A 1st-level character with +3 mod and +2 prof would normally roll 5d4. They add 3 more d4s (for the three “stacks”) and then remove the two lowest and the one highest die.

For save-based spells, use opposed rolls. The caster rolls as for an attack (including optionally spending a Risk Die). Each target must meet or beat that result (the DC).

The extra die removed from damage is the Risk Die—not included in damage but improving the spell’s chance of success.

If the calculated dice pool is zero or less and it’s a player-initiated action, they simply cannot do it. If the GM calls for the roll, they always roll at least 1d4.

We can remove Action / Bonus-Action bookkeeping by reducing available d4s instead:

1. At the start of a turn set used counter = 0.
2. Each action taken within the round adds 1 to used counter.
3. Current dice pool = base pool − used counter.
4. When pool drops to 0, no more actions this round.

Example: A fighter with a 5-d4 pool attacks:

High-level characters naturally get more d4s, allowing more actions.

Powerful abilities may require a minimum pool. Casting Fireball might require at least 6 d4s; if you cast it, subtract 6 from your remaining pool and add the same amount to used counter. Spell “levels” could simply be the minimum d4s required—think “Action Cost”.

Players add their dice, call out the total, and the GM quietly subtracts the target’s evasion / DC, narrating the outcome (without exposing exact numbers).

See https://anydice.com/program/3c172 for rolling-stat probabilities.

We might want to drop simple hit points. See https://www.youtube.com/watch?v=39Zney4NVB4 for strong arguments.

Alternatives such as the threshold model in Daggerheart or the system in Tales From Elsewhere could fit. Roll damage, but classify the result:

• Stress / Strain
• Serious
• Critical

Resistance / vulnerability could shift the category up or down, à la Tales From Elsewhere.

Healing spells would need rewrite: a cantrip removes one strain, Healing Word removes one serious or two strain, and a 2nd-level spell is needed for critical wounds. Glass-cannon casters get fewer slots; fighters and barbarians get more.

We might need cards or a standard list of conditions to track these effects neatly.

The Aspect Web replaces traditional Saving Throws and Skills with 15 Attribute Pairings, each representing a core capability of the character—from Agility (STR + DEX) to Willpower (WIS + CHA). Any challenge in the world can be approached through a strand of this web.

For players accustomed to 5e’s skill list, here’s one mapping:

Replacing Saving Throws means Aspect checks handle Concentration. Use different Aspects depending on the effect:

• https://www.youtube.com/watch?v=KSoc2qFrRKE
• https://www.youtube.com/watch?v=nltDDXguhpE
• https://www.youtube.com/watch?v=lqvcgSeYL-k
• https://jfacegames.substack.com/p/health-systems-in-tabletop-rpgs-beyond

• Daggerheart has a lot of cool ideas.
• Look into Spheres of Power and Crown and Skull (already downloaded the player’s guide).
• Read DCC book.
• GURPS
• Knave 2
• Monte Cook’s Cypher System
• https://en.wikipedia.org/wiki/Magic_systems_in_games

Telegraphed actions: A player, NPC, or monster may declare next turn’s action in advance. If they follow through, they gain advantage on the d20 test. If they change plans instead, they suffer disadvantage.

Maybe this becomes the core of “called shots.”