Post

Roll 2d10. Those are the arrows you grab. If you roll the same number on both dice, you fumbled the 'grab two without looking', and only got one. Arrows 6-10 are cursed.

I simplified that part with the 'you only got one arrow' bit (reasonable, I think). The alternative would be to reroll the second die if they tie. The fun bit would be if there were 13 arrows or such. You'd need to roll d20 and reroll anything 14+.

I had the same thought but that's actually off slightly. You could roll one D10 for the first arrow, but then you'd need to roll a D9 for the second.

Randall, I am , well kinda old. My husband was a mathematician. And seeing this post brought so many memories. soooo remember him studying combinatorics. Never thought would see that word again..and have daughter into D&D. So this I have to follow.

No. The correct odds are 2/9 and that's what the 3d6 + 1d4 dice roll yields. Here are the odds for the 3d6 (out of 216): 18 - 1 17 - 3 16 - 6 15 - 10 14 - 15 * .75 = 11.25 13 - 21 * .5 = 10.5 12 - 25 * .25 = 6.25 Add them up and it's 48/216 = 2/9 The .75, .5, .25 are for the d4.

Mine's the much easier approach when you haven't taken probability in 31 years. I did write out the permutations of dice rolls by hand (e.g., 14 - 662 653 644 544 -> 3 + 6 + 3 + 3 = 15). I found it interesting that the combinations for a 18 thru 13 followed the triangular numbers.

there's also whatever this is. basically there's a 1/3 chance of each outcome in increments of 0.1 (because there's 10 arrows total). you can divvy up the numbers however you see fit btw, assuming truly random dice, i just did that bc it's easier to write down.

Roll for humor. Rolls 5. Silence with a distant cough. Rolls 14. Polite, consistent chuckling. Rolls 1. Rotting fruit is flung at you in rapid succession. Rolls 36, dies from laughter. Ghost reappears and also somehow dies from laughter.

If the arrows are cursed if you touch them, how did they get into your quiver? If you have to fire it to bring about the curse, couldn't you just not fire it if you saw it was one of the cursed ones? Why are you walking around with cursed arrows in the first place?

Maybe the arrows are on the temple table under a sign - To gain my favour fire 2 of these identical arrows at the target. If you hit with a cursed arrow, the guardian will awaken. Chose wisely adventurer. Oh, and if you miss with a non cursed arrow they will awaken as well.

Correct. Because I was curious, I calculated the number of permutations of getting a 12 through 18 on the 3d6. I then reduced the permutations for 12 thru 14 by 25%, 50%, and 75%, respectively, to account for the d4. It add up to 48. There are 216 (6³) possibilities for the 3d6, 48/216 = 2/9

I had a student say she "wasn't good at math". Then I saw her at lunch in the DM role for several different groups she was running at school. Totally flabbergasted her when I told her that calculating the odds of the party's survival was the kind of math called probability and statistics.

Love that. I just imagined there could be a puzzle that can only be solved by rolling a prime number, and letting the player roll any dice. Many of them would go straight to d20, not to d4 or d6, without thinking about it.

Rolling 3d6+1d4 there are 6^3*4 = 864 possible outcomes. Enumerating out the number of ways to get each sum from 4 to 22 reveals that in 192 cases, the sum is 16+. 192 / 864 reduces to, you guessed it, 2/9.

I don't know anyone who would solve that in their head for more than two dice; there's just too much to keep track of. I also don't know of any analytic solution for the inverse problem - figuring out what dice to use to achieve a given probability; 3/

Of course, you could use 2d6, but you can't use the top of the bottom range saying you need at least an X. It would be you need "a 5 or 9" or "a lucky 3 or 7," with everything else failing. I don't know how to do an elegant way of finding 2/9 like Randall did without trial-and-error.

It is precisely 2/9, which you would expect from Randall. Since I'm old and out of practice, I had to manually figure out the # of combinations for 12 thru 18 on the 3d6 and then multiplied the odds for 12, 13 , and 14 by ¼, ½, and ¾, respectively. 48/216 = 2/9

I'm not much of a math or numbers person. Would it also work to say "roll 2d10. If you roll doubles, you only get one arrow. Any odd numbers mean you grabbed a cursed arrow."

YES!! This is the D&D I grew up with in the 80’s when we had a substitute English teacher with a PhD in Mathematics as a DM. It was a little annoying at first, but peeled my brain like a grape and let so much in. This really takes me back.

for first arrow make a flat DC10 roll, on fail you drew a cursed arrow, could carry this over to roll 2, but to make it more accurate you can make the second roll with DC9 (if you failed first check) or DC11 (if you succeeded), the math's not 1:1 but it's fairly close

Fun. 3 other ways to resolve: 1. Roll a d10 and a d9. (Love an excuse to use weird dice.) 2. Use dry-erase item cards. Shuffle up and deal. 3. Use a quiver and 10 arrows. Mark the tips of the arrows so the player knows which ones are cursed when they fire.

technically you can do a "roll 2d6 and you have to get at least 9, but you cannot get the same number on both dice" if we want minimum number of dices (although, idk if rules like this in DnD are too complex or not)

Combinatorics, merging, fusion, everything everywhere all at once hahaha, now do you see why it's no wonder everyone went a bit nuts during production? The roll-out is always the stressful part but once you take care of the technical debt -- all dat debt talk -- you truly have a so much easier path

you have to consider the frequencies of each combo. some will be more common bc of the D4. so the possibilities are 6x6x6x4 = 864. (2/9)x 864 = 192 successful rolls. i dont wanna calculate the winning number though lol

That would work and be fairly easy to figure out on the fly, but it lacks the elegance of "you need at least an X for success." I don't play roll playing games so I don't know if they ever use disjointed rolls like you suggest. I suspect not though I suppose a DM can do what they want.

But this roll isn't considering that there are 3 combinations of outcomes and he could just use two seperate rolls of a d10 (rerolling 10's on the second dice. but like... bro what kind of adventurer puts their cursed arrows in the same quiver!

Damn, I wish I'd seen this prior to giving a test on counting and probability to my students earlier this week. Or maybe I should say "Crap", given the probabilities involved...

when your expertise in something forces you to apply a complex logic when the simplest answer does sufice. I hate when i do that at work lol. "just roll 2d10" 😆

With Monte Cook's cypher system I might handle this with a GM intrusion, you don't roll to pick your arrow, I just decide to press a "fuck you" button and you get 2 XP if you take it, one to give to someone else. It costs one XP to reject it. Or I just decide not to press the button.

Further nerd-snipe: Can you construct arbitrary rational (i.e. n in m odds) probabilities using just an arbitrary number of standard tabletop dice and a single target score? If not, which odds are constructable in this way?