Skip to Content
 

1d6 is all you need

4 replies [Last post]
X3M
X3M's picture
Offline
Joined: 10/28/2013

I often play with dice pools. Where each die represents a projectile or an event.
I hardly ever use 2d6 or 3d6 or #d10 or whatever is possible.

No! All I need is a d6. And all the tricks you can do with it to get to a certain chance.

Limits
Before you apply tricks to a d6 to get to a certain chance of succes. You need to understand the limits of a d6. The d6 offers the following chances for succes:
16.7% = 1/6
33.3% = 2/6 or 1/3
50.0% = 3/6 or 1/2
66.7% = 4/6 or 2/3
83.3% = 5/6
100.% = 6/6

What you see here are chances with a base number. We got the 6. But also the 2 and 3.
This means, we can build up smaller portions of chances with more rolls to come.
2x2 = 4, 25% is possible with 2 rolls of 3/6.
3x3 = 9, 11% is possible with 2 rolls of 2/6.
Of course, with a tripple roll, we also have the bases 8, 12, 18 and 27.

Additions?
Are not really possible. So we use 1 re-roll. Which can be different. This re-roll is only applied to the failures. With this, we can add to a certain roll.

If we want a chance of 4/9 for example. We roll 3/9 instead.
What is remaining is 6/9 being a fail. This 6 is a fail on itself. But can be divided by 6. Thus adding 1 more.

Let's look at the base 9 series.
1/9 = 2/6 x 2/6
2/9 = 2/6 x 4/6
3/9 = 2/6
4/9 = 2/6 + R1/6

We aren't there yet.
The next one in the series is 5/9. Which is the opposite of 4/9. So instead of rolling for succes. You roll for fail.

5/9 = F(2/6 + R1/6)
6/9 = 4/6
7/9 = F(2/6 x 4/6)
8/9 = F(2/6 x 2/6)

X3M
X3M's picture
Offline
Joined: 10/28/2013
Stacking Re-Rolls on succes

Before I continue. I should share the list of base 8.

1/8 = 3/6 x 3/6 x 3/6
2/8 = 3/6 x 3/6
3/8 = 3/6 x 3/6 + R1/6
4/8 = 3/6
5/8 = F(3/6 x 3/6 + R1/6)
6/8 = F(3/6 x 3/6)
7/8 = F(3/6 x 3/6 x 3/6)

How to interpret the 3/8?
First the player rolls each die 2 times, 3/6. If both times are a succes, it counts. But if even one of the 2 fails, it is a fail. In that case, the player may roll once more, 1/6. If the roll is a succes. The die counts as a succes.

If the player has 8 dice like this. The first roll would put 4 of them in the fail pool. The second roll would put 2 more in the fail pool, totalling to 6. Then the fail pool is rolled and 1 gets put back in the succes pool. We have 3 in the succes pool and 5 in the fail pool.

If the player needed the 5/8 instead. The whole roll can be done as described above. But at the and, the fail and succes pool switch sides.

***

Cumulative chain rolls
When we roll for a 1/9 chance. We roll 2/6 x 2/6.
Both rolls need to be a succes for the total roll to be a succes.

But we can also count each succes.
The first roll on itself can be one. And both rolls can be 1.
2/6 is a 1/3 chance.
If this counts towards the total chance and we roll twice again, we get:
1/3 + 1/9 = 4/9

Is this the correct calculation?
While the answer is correct, the calculation should have done differently. Since the second roll depends on the first roll. And a chance on having 2 succesfull rolls should count twice. But the chance on having only 1 succesfull roll is 1/3 x 2/3, since a portion would become 2 successes.

1/3 x 2/3 + 2x1/9 = 4/9

I put that correct calculation in there because confusion often happens.

Explosive dice
Once the second roll is a succes. Whatever comes after. We just doubled the number of dice. Hence I call these explosive dice. And we can set a limit to how many rolls are allowed. Which doubles as a maximum possible number of successes.

_7/36 = L2E1/6
16/36 = L2E2/6 this is the 4/9.
27/36 = L2E3/6 which is another 75%.
40/36 = L2E4/6
55/36 = L2E5/6
72/36 = L2E6/6 which is indeed 2 succes.

_43/216 = L3E1/6
104/216 = L3E2/6
189/216 = L3E3/6 which is 7/8.
304/216 = L3E4/6
455/216 = L3E5/6
648/216 = L3E6/6 which is indeed 3 succes.

By combining one of these with another roll. You can get many other chances. But keep in mind that the result is cumulative. And the chance is an average result instead.

No limits?
While I don't like infinite rolls. Like for example discarding each 6. Just to get to that base 5. Which is regressive.
I do like the progressive way, where the number of succes only climb up. It is just that after a dozen successes, you pause that roll and see how they fair later on.

And we will be meeting a very interesting number here.

1/5 = L∞E1/6 Base 5?
1/2 = L∞E2/6
1 = L∞E3/6
2 = L∞E4/6
5 = L∞E5/6
∞ = L∞E6/6

While this roll with base 5 is an option. There are several things to keep an eye on:
1. The 1/5 is an average result. Meaning that it is not a true base 5.
2. Explosive rolls cannot be imbedded into the other 2 rolls. We cannot have a FL#E or RL#E. But, we do have L#EF and L#E(+ R).

With L#EF, we roll and the failures count as succes. Then this number of dice are rolled again with the same rules. Adding more and more to the pool of succes. While you won't see L∞EF1/6, it can be used when the base is for example 8 and the chance for each die has to be for example, 7/8th. We can only get 7/8th by having a fail roll imbedded. Example:
105/64 = L2EF(3/6 x 3/6 x 3/6)

L#E(+ R), is simpler to understand. We are still rolling for seeing how much succes the current roll has. The failures get another chance here.
33/64 = L2E(3/6 x 3/6 + R1/6)

Fake base 5:
1/5 = L∞E1/6
2/5 = L∞E2/7 Imposible...
3/5 = L∞E(3/6 x 3/6 + R1/6)
4/5 = L∞E(2/6 + R1/6)

Since the explosive rolls can multiply the number of dice. You might as well use multiple dice to begin with for that fake base 5. Thus 2/5 could be 2x L∞E1/6.

X3M
X3M's picture
Offline
Joined: 10/28/2013
We stil want those d5 and d10's

Let's do it. Let's see what happens if we allow re-rolls.

So, what if we set a limit to the number or re-rolls on a d6. When rolling a 6?
the chance that we get NO result, gets smaller and smaller. And eventually, it counts as a fail.

When we include a "d2" for the d10. That counts as one extra roll.

So, if we allow a total number of rolls of 6.
We get a chance of 1/46656 that it is still a fail for the d5.
And for the d10, this is 1/7776.

The true chances for 6 dice would be:
20% would be 19.9996%
10% would be 9.99871%

***

I might as well set up a list here. To see how many re-rolls sound plausible.

The true chances for 1 die would be:
20% would be 16.6667%

The true chances for 2 dice would be:
20% would be 19.4444%
10% would be 8.33333%

The true chances for 3 dice would be:
20% would be 19.9074%
10% would be 9.72222%

The true chances for 4 dice would be:
20% would be 19.9846%
10% would be 9.95370%

The true chances for 5 dice would be:
20% would be 19.9974%
10% would be 9.99228%

The true chances for 6 dice would be:
20% would be 19.9996%
10% would be 9.99871%

While this roll is regressive. It gets close enough to a d5 and d10. That it is acceptable.
1/_7776=0.012860% inaccuracy.
1/46656=0.002143% inaccuracy.

1/5 = 1/6D6L6
2/5 = 2/6D6L6
3/5 = 3/6D6L6
4/5 = 4/6D6L6
5/5 = 5/6D6L6
1/10 = 3/6 x 1/6D6L5
3/10 = 3/6 x 3/6D6L5
5/10 = 3/6
7/10 = F(3/6 x 3/6D6L5)
9/10 = F(3/6 x 1/6D6L5)

We can have fun with setting a limit to the discards. More of that, later.

X3M
X3M's picture
Offline
Joined: 10/28/2013
Other dice?

A d6 can easily be rolled multiple times in order to form other bases. Not going to say what the roll has to be exactly for a certain chance this time. But I am going to tell you wheter it is possible or not to get a list.

3/6th is a base of 2. And rolling this one multiple times, counting or discarding a roll. Will lead to the other bases:
4, 8, 16 etc.

2/6th is a base of 3. So is 4/6th. The other bases:
9, 27, 81 etc.

Of course, you can combine the 2 and 3 into the 6. But other bases are also possible, probably you come from that base 6 instead:
12, 18, 24 etc.

1/6th with rolling a 6 meaning a re-roll. Gives the base of 5. Combining this one with 2 and 3 gives us:
10, 15, 20 etc.

Base 7

We arrived at a base of 7. But how to get a base of 7? We cannot discard a 6 this time. So, what other option do we have?

We simply get a bigger pile to discard from. And whatever the roll will be, it will be a d7 in the bigger picture. The simplest form that also will give us 1/7th till 6/7th as hitchances. Is simply rolling 2d6.
There is a catch!
We can no longer do a bucket roll here. Since we need to track the d6 in what it does.

Your roll the die, subtract 1 and multiply the result by 6. Then you roll the die again. And add this result to the previous one.
Example: you roll a 5, this is 4x6=24, then you roll another 5. The total is now 29. 29/36 that is.

The agreement here will be that getting 36 as result means a discard. Then, any other result divided by 5 and rounded upwards, means a result with a base of 7.
29 would be 5.8. Rounded upwards is 6, we rolled a 6 on this d7.

So, we want to roll a...

1 on d7: 1 then 12345.
2 on d7: 1 then 6, or 2 then 1234.
3 on d7: 2 then 56, or 3 then 123.
4 on d7: 3 then 456, or 4 then 12.
5 on d7: 4 then 3456, or 5 then 1.
6 on d7: 5 then 23456.
7 on d7: 6 then 12345.
Discard: 6 then 6.

A bucket'o'dice for 6d7 is decently possible with 12d6.
We simply need to roll the dice in a box and sort them blindly. A box often has a lid. You roll the dice behind the lid. Others can watch though. And you push them in a 2x6 or 1x12 format into a corner.

However, if you fear that the pips can be felt and the player might be cheating. Another option is to have the dice just fall into a corner. Then remove them layer by layer. Going from top to bottom. While each layer has the dice removed from top left to bottom right.

As for the other bases, like 14, 21 etc. You need to make a plan. And go for the simpler step first. This safes trouble with the number of dice and the discard event occuring too often.

***

Somehow, d5 and d7 aren't that problematic, compared with d8.
Perhaps there is some merit in having more rolls being treated with discards instead of a fancy roll rule.

d11 is next?

X3M
X3M's picture
Offline
Joined: 10/28/2013
d11 and further

For a d11, we can do the same as the d7.

We roll the dice. And the result is again 1 to 36.
Instead of dividing by 5. We divide by 3.
We then discard 34, 35 and 36. Which is 1/12th of a chance.
Which is lower than having to simulate the simpler d5.

You might be thinking to have 2d6 and then consider the total. This gives a false result. The chance distribution is parabolic.

***

d13...
The portion of discard is huge with this one.
We can only divide the results by 2. Thus 27 or more is a discard. 10/36th as a discard? No, that is too much. This is a 27% discard chance.

We can look at a tripple system. Where we get to 216. And divide this by 13 to see what the divider should be.
We need to have the first d6, subtract 1, multiply by 36.
second d6 gets the same treatment, but we multiply by 6.
And the last d6 is simply added.

216/13=16.6
So, we divide the result by 16.
We discard when we roll 13x16+1= 209 or more. This time, the dicard chance is only 3.7%.

Remember that the discard chance for a d5 is at 16.7%.
But having to discard 3 dice after all those calculations is more annoying.

***

d17
Goes much better through 2d6. With a discard chance of 1/18th or 5.6%

d19
Needs 3d6 again.
The divider is 11 this time.
We discard at 210 or higher. Which is actually better than d13.
It sits at 3.2%.

d23
3d6.
Divider is 9.
We discard at 208 or higher.
Discard sits at 4.2%.

d29
3d6.
Divider is 7.
We discard at 204 or higher.
Dicard sits at 6.0%.

d31
2d6. (3d6 is a bad option here)
Divider is 1.
We discard at 32 or higher.
Discard sits at 13.9%.
With 3d6, the discard is at 13.9% as well, so, take your pick.
We can get a 1.9% with 4d6

Oh, why am I even going this high? A d31 is just like a d30. And these chances aren't even worth designing into a game.

The d7 and d11 and even the d13 will be usefull, no doubt. Certainly for the translations that I mentioned in the Factions topic. But that would be about it.

Cheers, X3M

Syndicate content


forum | by Dr. Radut