Dice Roller — D4 to D100, 3d6+2 Notation, Advantage/Disadvantage

Real-world use cases

• Run a 5e combat round without reaching for physical dice

  You're GMing online and the fighter swings with advantage: type 1d20, tap Advantage, roll. You see both d20 faces — the kept one and the struck-through one — so the player trusts the call. Hits, so you roll damage: switch to 2d6+4 for the greatsword, roll, read the total. The whole round is two notation edits and two clicks, and the history list keeps the last few rolls visible so nobody argues about what the cleric rolled three turns ago.

• Roll up a D&D character with the 4d6-drop-lowest method

  Classic ability-score generation: roll 4d6, drop the lowest die, sum the top three, repeat six times. Type 4d6, roll, and read the four faces — you do the "drop lowest" by eye since the dice are all shown. Six quick rolls give you a STR/DEX/CON/INT/WIS/CHA spread. Because the RNG is a fair CSPRNG, you can't accidentally bias the array the way a tired hand fudging physical dice might.

• Settle a board-game dispute or decide who goes first

  Catan, Monopoly, Risk — half of tabletop is "highest roll goes first" and the other half is "did the attacker actually beat the defender." Hand everyone the same link (notation travels in the URL), each person rolls 2d6 on their own device, compare totals. Nobody can claim the dice were loaded because the generator is the same fair CSPRNG for everyone, and the history shows their own rolls in order.

• Pull a fair name / number for a giveaway or class

  Numbered your raffle entries 1 to 100? Roll 1d100 and read the winner. Need to cold-call a student from a 30-seat class? 1d30. Because each face is exactly equiprobable (rejection-sampled CSPRNG, no modulo bias), the draw holds up if anyone questions it — unlike Math.random() pickers that can skew slightly toward low numbers on certain ranges.

• Demonstrate probability in a stats or math class

  Teaching the bell curve? Have students roll 2d6 fifty times and tally the totals — 7 should dominate, 2 and 12 should be rare. The history list gives them the raw data to chart. Then roll 1d6 a few hundred times conceptually and point out that each face is 1/6, which the tool's fair RNG actually delivers, so the empirical counts converge on the theoretical ones instead of fighting a biased generator.

Common pitfalls

• Stacking percentages in your head — wait, wrong tool. The real one here is thinking 30% + 10% works for dice; it doesn't apply, dice are summed. But the analogous dice mistake is adding the modifier to EACH die instead of once to the total. 3d6+2 is (d6+d6+d6)+2, not (d6+2)+(d6+2)+(d6+2). This tool always adds the modifier exactly once, to the final sum, so you get the right number.

• Treating advantage as "roll the modifier twice too." Advantage/disadvantage only doubles the DICE roll and keeps the better/worse total — the flat modifier is added once at the end. Rolling 1d20+5 with advantage is "best of two d20" then +5, not "best of two d20+5 rolls" (which would be the same number anyway, but the mental model trips people up on multi-die rolls).

• Assuming any online dice are fair. Many roller widgets use Math.random(), which is a non-cryptographic PRNG and can show slight bias over certain ranges. If fairness matters — a paid tournament, a graded class draw — check that the tool uses a CSPRNG. This one uses crypto.getRandomValues with rejection sampling; a generic Math.random() picker does not guarantee uniformity.

Privacy