Poll of the Day > d20's are the d6's of the tabletop gaming world

(+) Yeah! (+)

d20s are preferred over d6s because the greater precision allows the wider range of outcomes for advanced simulation
---

I have many d20s. Sometimes I can't resist picking up a cool one even though I already have enough
---

Thanks, let us know what else you come up with
---

OniRonin posted...

d20s are preferred over d6s because the greater precision allows the wider range of outcomes for advanced simulation

No, d20s are preferred because the OGL made them popular.

In the 80s and 90s, d10s were the absolute kings - many of the 80s games used them as percentile dice to simulate even more potential outcomes on a scale of 1-100, while the massive success of White Wolf in the 90s led to widespread adoption of the d10 roll/drop and "exploding" mechanics.

In the years to come, if D&D 6e (or beyond) radically changes its choice of primary die, or another system grows in popularity to the point of becoming a major rival (or outright outpacing D&D entirely the way White Wolf did for a time), then that will inspire imitators and give us a new most popular die.

The only thing that remains forever constant is that the d12 will always be the bastard stepchild of every system.


---

ParanoidObsessive posted...

percentile dice to simulate even more potential outcomes on a scale of 1-100,

My dad had a couple of really weird dice from his DnD days that each consisted of a clear plastic d10 with a second, smaller d10 inside of it, so you could do percentages with a single roll. They don't strike me as being particularly practical, but they were kind of cool.

_AdjI_ posted...

ParanoidObsessive posted...

percentile dice to simulate even more potential outcomes on a scale of 1-100,

My dad had a couple of really weird dice from his DnD days that each consisted of a clear plastic d10 with a second, smaller d10 inside of it, so you could do percentages with a single roll. They don't strike me as being particularly practical, but they were kind of cool.

Ive got a few of those. Ive also got a couple d100s which are hilarious because they roll forever
---

Stop playing DnD if you think d20 is the base dice.
---

shadowsword87 posted...

Stop playing DnD if you think d20 is the base dice.

Everyone knows its the ever popular d12
---

_AdjI_ posted...

My dad had a couple of really weird dice from his DnD days that each consisted of a clear plastic d10 with a second, smaller d10 inside of it, so you could do percentages with a single roll. They don't strike me as being particularly practical, but they were kind of cool.

I've got a d100 that basically looks like a large golfball. But I've also got a d30 and a d24 that look about as weird as possible. And an awkward d3.

My original RPG die (that I still have) is a d10, though. And it's about 35 years old now. It was from the era when you literally had to draw the numbers into the etched parts of the die with a white crayon, because they weren't made with pre-printed numbers.

I've basically got dice from four different decades in my collection.

And I roll like shit using every single one of them.



shadowsword87 posted...

Stop playing DnD if you think d20 is the base dice.

To be fair, it basically is if you're playing 5e.

Or 4e.

Or even 3e, for the most part, because it was the origin of the d20 OGL in the first place.

Sure, other dice ARE used... but the d20 is easily used the most, and has been for about 20 years now.

d20 years now


---

I mean, yeah D20s are cool. I like all dice though, even the D12. But I really liked Warhammers D10s system though
---

Mead posted...

_AdjI_ posted...

ParanoidObsessive posted...

percentile dice to simulate even more potential outcomes on a scale of 1-100,

My dad had a couple of really weird dice from his DnD days that each consisted of a clear plastic d10 with a second, smaller d10 inside of it, so you could do percentages with a single roll. They don't strike me as being particularly practical, but they were kind of cool.

Ive got a few of those. Ive also got a couple d100s which are hilarious because they roll forever

Theyre also a little hard to figure out exactly which side is up at a glance. You really gotta scrutinize it a bit more than usual...
---

ParanoidObsessive posted...

OniRonin posted...

d20s are preferred over d6s because the greater precision allows the wider range of outcomes for advanced simulation

No, d20s are preferred because the OGL made them popular.

In the 80s and 90s, d10s were the absolute kings - many of the 80s games used them as percentile dice to simulate even more potential outcomes on a scale of 1-100, while the massive success of White Wolf in the 90s led to widespread adoption of the d10 roll/drop and "exploding" mechanics.

In the years to come, if D&D 6e (or beyond) radically changes its choice of primary die, or another system grows in popularity to the point of becoming a major rival (or outright outpacing D&D entirely the way White Wolf did for a time), then that will inspire imitators and give us a new most popular die.

The only thing that remains forever constant is that the d12 will always be the bastard stepchild of every system.

Wow, it's an honor to be "POwned" by this wall of knowledge, sir!
---

Here's another cool d20 fact -

it takes 4 d6's to simulate a d20 - simply sum the results of the d6 roles. treat a 21,22, or 23 as a 1, 2, or 3. On "24", discard and reroll.
---

That's 100% not true, just to be clear.
Here's the difference between the rolls: https://imgur.com/28cDYRb
---

shadowsword87 posted...

That's 100% not true, just to be clear.
Here's the difference between the rolls: https://imgur.com/28cDYRb

I already accounted for that. four D6's can't roll 1-3 due to a lack of zero, so I assigned 21, 22, and 23 to 1, 2, and 3. That way each number in the d1-20 space has exaclty one result assigned to it.
---

OniRonin posted...

That way each number in the d1-20 space has exaclty one result assigned to it.

Results with different probabilities of occurring, as shown by Shadow's graph.
---

OniRonin posted...

That way each number in the d1-20 space has exaclty one result assigned to it.

No. 4 and 24 have one result. 14 has about 28 or 29 ways to get there.

Questionmarktarius posted...

OniRonin posted...

That way each number in the d1-20 space has exaclty one result assigned to it.

No. 4 and 24 have one result. 14 has about 28 or 29 ways to get there.

No - only a result of 14 on the d6s returns a 14. I think you and shadow have misunderstood how my system works
---

OniRonin posted...

I think you and shadow have misunderstood how my system works

You clearly have no idea how bell curves work, unless your intent is to trend results towards the middle.
http://www.thedarkfortress.co.uk/tech_reports/4_dice_rolls.htm

1d20 has twenty possible results. 4d6 has 1296 possible results: 4 and 24 almost never happen, while 13, 14, and 15 happen quite often:
[4] => 1
[5] => 4
[6] => 10
[7] => 20
[8] => 35
[9] => 56
[10] => 80
[11] => 104
[12] => 125
[13] => 140
[14] => 146
[15] => 140
[16] => 125
[17] => 104
[18] => 80
[19] => 56
[20] => 35
[21] => 20
[22] => 10
[23] => 4
[24] => 1

If you want to approximate a d20, use a coin and a d10. Flip the coin, and add 10 to the dieroll on heads.

Questionmarktarius posted...

OniRonin posted...

I think you and shadow have misunderstood how my system works

You clearly have no idea how bell curves work, unless your intent is to trend results towards the middle.
http://www.thedarkfortress.co.uk/tech_reports/4_dice_rolls.htm

1d20 has twenty possible results. 4d6 has 1296 possible results: 4 and 24 almost never happen, while 13, 14, and 15 happen quite often:
[4] => 1
[5] => 4
[6] => 10
[7] => 20
[8] => 35
[9] => 56
[10] => 80
[11] => 104
[12] => 125
[13] => 140
[14] => 146
[15] => 140
[16] => 125
[17] => 104
[18] => 80
[19] => 56
[20] => 35
[21] => 20
[22] => 10
[23] => 4
[24] => 1

If you want to approximate a d20, use a coin and a d10. Flip the coin, and add 10 to the dieroll on heads.

I have no idea why this would matter. I'm only rolling the dice once, not 1296 times.
---

I just love tabletop dice so much in general.
---

OniRonin posted...

I have no idea why this would matter. I'm only rolling the dice once, not 1296 times.

If you need to roll 17+ on a d20, you've got a 15% chance of doing so. Doing the same with 4d6 (actually 21+) has less than a 3% chance of happening.

Questionmarktarius posted...

OniRonin posted...

I have no idea why this would matter. I'm only rolling the dice once, not 1296 times.

If you need to roll 17+ on a d20, you've got a 15% chance of doing so. Doing the same with 4d6 (actually 21+) has less than a 3% chance of happening.

There are 8 numbers 17 or higher that can come up on 4 d6's -- 17, 18, 19, 20, 21, 22, 23, 24. 21, 22, 23 are subtracted by 20 to give 1, 2, and 3. 24 is ignored. So you have a 4/23 chance of rolling 17+ on 4 d6's, which is 17.3%, very close to the d20's 15-20% depending on how you do the math.
---

OniRonin posted...

Questionmarktarius posted...

OniRonin posted...

I have no idea why this would matter. I'm only rolling the dice once, not 1296 times.

If you need to roll 17+ on a d20, you've got a 15% chance of doing so. Doing the same with 4d6 (actually 21+) has less than a 3% chance of happening.

There are 8 numbers 17 or higher that can come up on 4 d6's -- 17, 18, 19, 20, 21, 22, 23, 24. 21, 22, 23 are subtracted by 20 to give 1, 2, and 3. 24 is ignored. So you have a 4/23 chance of rolling 17+ on 4 d6's, which is 17.3%, very close to the d20's 15-20% depending on how you do the math.

Your weird 'wraparound' gives an even weirder distribution than just subtracting 4.
And why ignore 24 instead of just subtracting 20? All that changes below is that 4 has two occurrences, instead of a 1/1296 chance of a wasted roll.

[1] => 20
[2] => 10
[3] => 4
[4] => 1
[5] => 4
[6] => 10
[7] => 20
[8] => 35
[9] => 56
[10] => 80
[11] => 104
[12] => 125
[13] => 140
[14] => 146
[15] => 140
[16] => 125
[17] => 104
[18] => 80
[19] => 56
[20] => 35

You've still got results clustering around 14, but now you've got a smaller peak on the low end.
This method is only useful if you want to intentionally skew the probabilities into a large amount of moderate success, with an elevated chance of abysmal failure.

Questionmarktarius posted...

And why ignore 24 instead of just subtracting 20?

Because 1+1+1+1 = 4

Your weird 'wraparound' gives an even weirder distribution than just subtracting 4.

A 17+ includes 17,18,19,20. Because there are 20 numbers between 1 and 20, that corresponds to an ideal of a 20% chance. Your method gives a 15% chance of a 17+, whereas mine gives a 17.3% chance, so it's actually more accurate than yours.
---

OniRonin posted...

There are 8 numbers 17 or higher that can come up on 4 d6's -- 17, 18, 19, 20, 21, 22, 23, 24. 21, 22, 23 are subtracted by 20 to give 1, 2, and 3. 24 is ignored. So you have a 4/23 chance of rolling 17+ on 4 d6's, which is 17.3%, very close to the d20's 15-20% depending on how you do the math.

Let's put this as simple as possible, a d20 gives you 20 possible roles, all having the same 5% chance of occuring, the numbers from 1 to 20.

Rolling 4d6 has 1296 possible outcomes. the two extremes, 1+1+1+1 and 6+6+6+6, both only have a 1/1296 chance of being rolled. Whereas the peak of the bell curve, 14, has 146 ways of it being rolled.

In your 4d6 system, chance to roll a 4 is 0.077%. The chance to roll 14 is 11.27%. Both quite a bit away away from the 5% of a d20.

And yes, systems like yours exist. Some use 3d6 instead of 1d20, to intentionally skew the results in the favour of the average. As they give more power to statistical character bonuses, while depending less on luck.

OniRonin posted...

A 17+ includes 17,18,19,20. Because there are 20 numbers between 1 and 20, that corresponds to an ideal of a 20% chance. Your method gives a 15% chance of a 17+, whereas mine gives a 17.3% chance, so it's actually more accurate than yours.

Are you serious right now? He used 17+ as meaning, 18, 19, 20, that leads to 15%. The way you use the number, a d20 gives a 20% chance. Because that is the entire point of a d20, 20 outcomes that are equally likely to come up.

Also, I am only getting to a 13.13% chance in my calculations for rolling a 17 or higher in your system.
---

Nichtcrawler X posted...

He used 17+ as meaning, 18, 19, 20, that leads to 15%.

Well... too many numbers.
17/18/19/20 are 20% of a d20's distribution, but 21.2% of this weird 4d6&wraparound(relroll24) system.

Nichtcrawler X posted...

He used 17+ as meaning, 18, 19, 20, that leads to 15%. The way you use the number, a d20 gives a 20% chance. Because that is the entire point of a d20, 20 outcomes that are equally likely to come up.

17+ always means 17 or more ----- when was the last time you saw someone use 100+ to mean "greater than 100"? It always includes the number itself.
---

Him meaning only higher than 17 was obvious from the other numbers. Chance calculations for 1 d20 are very simple, much less so for 4d6.
---

Let's just make this glaringly simple then:
What's the chance of getting a 10 with a d20? 5%
What's the chance of getting a 10 with your system? 6.17%

What's the big deal? Those numbers are different. That means you are getting uneven dice rolls, which is fine if the system is built for that, hell, GURPS still exists and used 3d6. But, what your suggesting is changing the percent chance, which every single d20 system isn't really built for that.

If you wanted to argue that d20 should have a smaller dice or different percent success chances, that's totally fine and a reasonable argument. But, what you're doing isn't that.
---

45+11=56
---

OniRonin posted...

I have no idea why this would matter. I'm only rolling the dice once, not 1296 times.

The reason it matters is because there are 1296 possible results from rolling 4d6, while there are only 20 results from rolling 1d20.

Rolling multiple dice skews the average roll towards the middle (which on 4d6 is 14), whereas rolling one dice gives you an equal chance of rolling every number.

I'm not going to double check questionmark's numbers, but if they're correct, you'll roll between 10-18 roughly 80% of the time instead of 45% of the time.
---

PMarth2002 posted...

I'm not going to double check shadowsword's numbers, but if they're correct, you'll roll between 10-18 roughly 80% of the time instead of 45% of the time.

I just got the numbers from anydice.com
I trust it enough with it's percentages.
---

shadowsword87 posted...

PMarth2002 posted...

I'm not going to double check shadowsword's numbers, but if they're correct, you'll roll between 10-18 roughly 80% of the time instead of 45% of the time.

I just got the numbers from anydice.com
I trust it enough with it's percentages.

damn you're fast.I actually meant questionmark's, not yours and edited my post right after you posted that.

But its a moot point since onironin apparently doesn't care about how probability works, and definitely doesn't understand it.
---

PMarth2002 posted...

I'm not going to double check questionmark's numbers

Well, the "17+" thing got screwed up.
The rest comes from a brute force simulator with lots of nested for loops.

Questionmarktarius posted...

PMarth2002 posted...

I'm not going to double check questionmark's numbers

Well, the "17+" thing got screwed up.
The rest comes from a brute force simulator with lots of nested for loops.

Its cool. Shadowsword's image that onironin already discounted works with the point i was trying to make to onironin as well.

Seriously though, Onironin, your method is not the same thing as rolling 1d20. It will achieve very different results than rolling a 1d20. If you wanted to use it as a house rule, you could, but it will change how the game plays out. Characters will succeed more often because the most common results will be 10-18. As I said, those numbers will show up about 80% of the time.
---

I mean, you could roll

1d6 initially, even = 1-10 for the next roll, odd = 11=20

Next roll 1d6 even = 1d5 for the last roll odd = 6-10

Last roll is normal but 6's are re-rolled.

So first roll determines if the result will be between 1-10 or 11-20

Second roll determines if the result will either be A: 1-5 or 6-10 or B: 11-15 or 16-20 (based on the first roll)

Last roll gives the final result within the set parameters.

Every number should be equally as likely with this setup. It's a 50/50 chance for the first roll, the same for the second, and the third roll is a 20% chance for each possibility.

IF even is low and odd is high you need to:

Roll even/even/1 to get a 1

to get a 20 you need to roll odd/odd/5

To roll a 10 you need to roll even/odd/5

To roll a 14 you need to roll odd/even/4

Every outcome is just as likely as every other.
---

wolfy42 posted...

I mean, you could roll

1d6 initially, even = 1-10 for the next roll, odd = 11=20

Next roll 1d6 even = 1d5 for the last roll odd = 6-10

Last roll is normal but 6's are re-rolled.

So first roll determines if the result will be between 1-10 or 11-20

Second roll determines if the result will either be A: 1-5 or 6-10 or B: 11-15 or 16-20 (based on the first roll)

Last roll gives the final result within the set parameters.

Every number should be equally as likely with this setup. It's a 50/50 chance for the first roll, the same for the second, and the third roll is a 20% chance for each possibility.

IF even is low and odd is high you need to:

Roll even/even/1 to get a 1

to get a 20 you need to roll odd/odd/5

To roll a 10 you need to roll even/odd/5

To roll a 14 you need to roll odd/even/4

Every outcome is just as likely as every other.

And thats way more complicated than just rolling a d20.
---

PMarth2002 posted...

wolfy42 posted...

I mean, you could roll

1d6 initially, even = 1-10 for the next roll, odd = 11=20

Next roll 1d6 even = 1d5 for the last roll odd = 6-10

Last roll is normal but 6's are re-rolled.

So first roll determines if the result will be between 1-10 or 11-20

Second roll determines if the result will either be A: 1-5 or 6-10 or B: 11-15 or 16-20 (based on the first roll)

Last roll gives the final result within the set parameters.

Every number should be equally as likely with this setup. It's a 50/50 chance for the first roll, the same for the second, and the third roll is a 20% chance for each possibility.

IF even is low and odd is high you need to:

Roll even/even/1 to get a 1

to get a 20 you need to roll odd/odd/5

To roll a 10 you need to roll even/odd/5

To roll a 14 you need to roll odd/even/4

Every outcome is just as likely as every other.

And thats just way more complicated than rolling a d20.

The point is if you don't have a d20 but you need to roll one:)

It's more complicated, but you can make a d6 out of paper way easier (and more accurate) then a d-20, for instance, so if your trying to simulate a d20 in prison for your DnD buddies.....you could use this method.

Also it's not really that complicated if you have different color dice and each dice is determined in advance. You could basically roll them and get the response just about as fast as rolling 1 20 sided dice.

Blue/Green/Red dice....blue determines 1-10/11-20, green 1-5/6-10 and red is the actual number. After doing that for awhile you would read the answer just as fast as a 20 sided when you roll.
---

If I was in prison and all I had were d6s, i'd probably just use 3d6 instead of 1d20 and house-rule a few things, but thats just me.
---

PMarth2002 posted...

If I was in prison and all I had were d6s, i'd probably just use 3d6 instead of 1d20 and house-rule a few things, but thats just me.

The dungeons & dragons monster manuals and many other supplementary materials are banned in prisons in my state so this may also require a workaround!
---

Questionmarktarius posted...

PMarth2002 posted...

I'm not going to double check questionmark's numbers

Well, the "17+" thing got screwed up.
The rest comes from a brute force simulator with lots of nested for loops.

Can I see the source Code? I'm not sure that I trust your results and would like to re-produce them
---

OniRonin posted...

PMarth2002 posted...

If I was in prison and all I had were d6s, i'd probably just use 3d6 instead of 1d20 and house-rule a few things, but thats just me.

The dungeons & dragons monster manuals and many other supplementary materials are banned in prisons in my state so this may also require a workaround!

Luckily most DM's have most of it memorized, I wouldn't have it all, but more then enough to run campaigns. Then again, I'd die of boredom in prison in a week, even with DnD to play.....so it wouldn't matter.
---

OniRonin posted...

Can I see the source Code?


<?php
header('Content-Type:text/plain');
$count = 0;
$results = array();
for($a=1; $a<=6; $a++){
for($b=1; $b<=6; $b++){
for($c=1; $c<=6; $c++){
for($d=1; $d<=6; $d++){
echo $a . $b . $c . $d;
echo "\r\n";
$total = $a + $b + $c + $d;
if($total > 20 && $total < 24) $total -= 20;
$results[$total]++;
$count++;
}}}}
echo "\r\n";
echo "\r\n";
echo $count;
echo "\r\n";
print_r($results);
?>


OniRonin posted...

I have no idea why this would matter. I'm only rolling the dice once, not 1296 times.

Actually, you're rolling it four times. It's easier to visualize with smaller numbers, so we'll look at rolling 2d4 instead of 1d8 (not a perfect substitution, since you can't roll a 1 on 2d4, but it illustrates the point).

Possible 2d4 outcomes:
https://imgur.com/tpss6I2

2: 1 combination
3: 2
4: 3
5: 4
6: 3
7: 2
8: 1

This means you're going to get 5 four times more often than you'll get 2 or 8. On a d8, however, a 5 is just as likely as any other roll.

Any combination of dice that can be rolled to have 20 total possible results will be equivalent to rolling a d20. That can pretty much be done with any dice, or even a coin (although it's very easy to cheat with a coin....unless you let it hit the floor each time.

A d6 though is the easiest to make/create yourself, so I would suggest using them to simulate any other dice you want to roll.

A d3 = 1-2=1 3-4=2 5-6=3

A d4 = Even =1-2 odd =3-4 then Even =1 or 3 and odd 2 or 4

A d6 = duh

A d8 = same as d4, but with 1 more dice (1-4 or 5-8), then same as d4

A d10 = Odd/Even, then roll as a d5 (reroll 6's)

A d12 = odd/even then roll as d6

Etc etc....all rolls can be done easily with a d6...and you can make a d6 easily with paper.
---

wolfy42 posted...

I mean, you could roll

1d6 initially, even = 1-10 for the next roll, odd = 11=20

Next roll 1d6 even = 1d5 for the last roll odd = 6-10

Last roll is normal but 6's are re-rolled.

So first roll determines if the result will be between 1-10 or 11-20

Second roll determines if the result will either be A: 1-5 or 6-10 or B: 11-15 or 16-20 (based on the first roll)

Last roll gives the final result within the set parameters.

Every number should be equally as likely with this setup. It's a 50/50 chance for the first roll, the same for the second, and the third roll is a 20% chance for each possibility.

IF even is low and odd is high you need to:

Roll even/even/1 to get a 1

to get a 20 you need to roll odd/odd/5

To roll a 10 you need to roll even/odd/5

To roll a 14 you need to roll odd/even/4

Every outcome is just as likely as every other.

By the logic everyone else is using in the thread, this is different than rolling a d20 because there are 180 different results
---

OniRonin posted...

By the logic everyone else is using in the thread, this is different than rolling a d20 because there are 180 different results

Only if the last roll can reroll indefinitely instead of starting over.

Nope there are not, because you literally are forcing there to be only 20 results in every case, other then when you re-roll 6's (making them not count.....even though it's technically a result).

There are 2 results for odd/even rolls....odd or even, the numbers don't matter.

Then the final rolls have a set amount of results (5 in most cases).

So you end up with exactly 20 results, a 5% chance for each result, which is exactly the same as rolling a 20 sided dice.

In fact, my suggestion is based on doing exactly that, creating the exact same 5% chance for each result, or whatever the chance is if your simulating other dice. All you are doing is creating a way to get that chance/number of results.
---