# Thread: FYI: How D20s work.

1. ## FYI: How D20s work.

Having played a PnP game, I happen to know that a D20 has 20 sides given a 5% chance on average per outcome.

I currently am in possession of a magical D20 which i may roll to get cards which i may trade for stuff.

It is apparent however that a magical D20 does not give a 5% chance or either a card IX or card X which any normal D20 would give.

Not even a 5% chance that is then split in some way by another roll.

So i am curious has anyone worked out the actual probabilities here? we could with a large enough sample (if people could post what htye have received).

I have, for example, not used any cards and have accumulated the following:

I = 7
II = 3
III = 7
IV = 6
V = 8
VI = 5
VII = 9
VIII = 4
IX = 1
X = 0

So mostly these look about the same... That is 50 cards and i they were even amoung the 10 it would be 5 of each. a little higher or lower is not too weird but card X and IXis an exception.

the probability of not getting a card in 50 rolls if it was actually a 5% chance is buy the way 0.95^50 = approx 7%

but i have talked to other people and so i know it is not bad luck, it is a bad die. Clearly fixed.

Anybody else have some clean data for analysis?

2. Originally Posted by RapkintheRanger
Having played a PnP game, I happen to know that a D20 has 20 sides given a 5% chance on average per outcome.

I currently am in possession of a magical D20 which i may roll to get cards which i may trade for stuff.

It is apparent however that a magical D20 does not give a 5% chance or either a card IX or card X which any normal D20 would give.

Not even a 5% chance that is then split in some way by another roll.

So i am curious has anyone worked out the actual probabilities here? we could with a large enough sample (if people could post what htye have received).

I have, for example, not used any cards and have accumulated the following:

I = 7
II = 3
III = 7
IV = 6
V = 8
VI = 5
VII = 9
VIII = 4
IX = 1
X = 0

So mostly these look about the same... That is 50 cards and i they were even amoung the 10 it would be 5 of each. a little higher or lower is not too weird but card X and IXis an exception.

the probability of not getting a card in 50 rolls if it was actually a 5% chance is buy the way 0.95^50 = approx 7%

but i have talked to other people and so i know it is not bad luck, it is a bad die. Clearly fixed.

Anybody else have some clean data for analysis?
IX & X are rare cards the rolls are not actually based on a D20 roll.

the D20 more aesthetic than functional

4. It's normal that Turbin use features of D&D without respecting them. This game is full of this.

5. Based upon my rolling, I don't have my exact numbers, but I will go with around 1 in 250 or 0.4% chance.

99.6 % you get a card 1 through 8.
00.4% you get a card IX or X.

When you roll the 99.7 to 100 you get another d2 roll and it is 50-50 to get either a IX or X. Early in the event I thought card X was rarer, but I am now in the camp of the above chances.

In conclusion, I think it is a d1000 that is rolled.

roll 1 to roll 996 = card 1 through card 8
roll 997 to roll 998 = card IX
roll 999 to roll 1000 = card X

6. Oh boy, actually it is common practice in D&D to assign ranges to a die roll or multiple die rolls that are not necessarily symmetrical. Examples include most random treasure tables, encounter tables and especially d% or d100 tables.

Example:

1d20 roll:
1 - 2 card 1
3 - 4 card 2
5 - 6 card 3
7 - 8 card 4
9 - 10 card 5
11 - 12 card 6
13 - 14 card 7
15 - 16 card 8
17 - 18 roll again
19 roll again a roll of 19 or better gives card 9
20 roll again a roll of 20 gives card 10.

Just eye balling it, I'd guess that would be roughly teens % chance for cards 1 - 8 which jibes with the small sample sets we've seen so far reported on the forums and a less that 1% chance at cards 9 and 10, which also actually seems to be a close aproximation of the small samples we can look at.. and there are other ways to achieve a similar approximation and/or use the 1d20 like a d%.

So yeah totally D&D like and entirely plausible that a staggered range 1d20 roll may be what the code is approximating.

7. Originally Posted by MagicBlue
It's normal that Turbin use features of D&D without respecting them. This game is full of this.
In some threads this would make sense. But in this one its just silly.

8. To be fair, it is a *magical* d20, magic often usurps the laws of probability :P

9. Originally Posted by RapkintheRanger

Anybody else have some clean data for analysis?
Yeah, create 40 fake accounts and keep rolling!

:/

10. maye tur.ine needs to install another 3rd party program that knows what dice are and how they work..
it's a bad day when we need to check tur.ies ice for cheeting

11. Isn't it tiresome to just spout negative drivel all the time?

12. Originally Posted by toaftoaf
maye tur.ine needs to install another 3rd party program that knows what dice are and how they work..
it's a bad day when we need to check tur.ies ice for cheeting
BTW your kb seems to be malfunctioning.

As explained above, there's nothing necessarily wrong with the die rolls in this game. Every thread where this subject comes up and actual data is collected tend to confirm that DDO's RNG is working well and doing a good job of simulating randomness.

As usual it's simplistic expectations of fairness (even distribution) within small random samples and plain old selection/confirmation bias that are whats wrong. I.e. Because: Humans.

Frankly it's EASY to prove Turbine's RNG cheats... If it did cheat we would have proven it thousands of times already. They would have to be idiots to think they would get away with it even for a few months. So of course they don't bother doing that. And every sampling I've seen on the foums supports that. Despite the Dunning-Kruger effect that is all pervasive on internet forums, it is safe to say they are not idiots.

The card distribution is entirely consistent with a d20 roll and a typical D&D table. You guys sound like you've never cracked a DMG (or rolled dice for that matter).

13. Originally Posted by IronClan
1d20 roll:
1 - 2 card 1
3 - 4 card 2
5 - 6 card 3
7 - 8 card 4
9 - 10 card 5
11 - 12 card 6
13 - 14 card 7
15 - 16 card 8
17 - 18 roll again
19 roll again a roll of 19 or better gives card 9
20 roll again a roll of 20 gives card 10.

<snipped>

So yeah totally D&D like and entirely plausible that a staggered range 1d20 roll may be what the code is approximating.
That is a very D&Dish die table - very reminiscent of the random treasure generation tables in the old DMG. I would actually be quite impressed if this is indeed what the Turbine folks were trying to do.

Originally Posted by enochiancub
Isn't it tiresome to just spout negative drivel all the time?
On the schoolyard, bullies and jerks often do what they do because of their own inadequacy issues - and these forums too often feel like the DDO schoolyard.

14. ## FYI: How MAGICAL d20's work

You tell the DM you're rolling the die. The DM pretends to roll a die. The DM gives you whatever the heck he feels like.

15. Originally Posted by PermaBanned
That is a very D&Dish die table - very reminiscent of the random treasure generation tables in the old DMG. I would actually be quite impressed if this is indeed what the Turbine folks were trying to do.
In addition to using Old School DMG multiple rolltables like that provided by Ironclan, it is also possible that each our rolls of the d20 is in fact simple multiple rolls.

The output of rolling d20 two times can be interpreted as either range between 2-40 (i.e. 2D20) but also as a percentile giving a range of 1-100 if you ignore the first decimal place-holder.
Rolled 3x, a d20 can be 3-60 OR 1-1000.
I mean using polyhedrons in different combinations and different weightings was common place.

16. Originally Posted by Alfhild
In addition to using Old School DMG multiple rolltables like that provided by Ironclan, it is also possible that each our rolls of the d20 is in fact simple multiple rolls.

The output of rolling d20 two times can be interpreted as either range between 2-40 (i.e. 2D20) but also as a percentile giving a range of 1-100 if you ignore the first decimal place-holder.
Rolled 3x, a d20 can be 3-60 OR 1-1000.
I mean using polyhedrons in different combinations and different weightings was common place.
Or multiply the two to get 1-400 with 1 = IX and 400 = X. The rest can be assigned to ranges.

17. Originally Posted by BOgre
You tell the DM you're rolling the die. The DM pretends to roll a die. The DM gives you whatever the heck he feels like.
Not in my games. I roll a die and look the result up in a table. *Then* i say "whoa you survived this time!" and give you whatever the heck i fele like :-)

18. Originally Posted by BOgre
You tell the DM you're rolling the die. The DM pretends to roll a die. The DM gives you whatever the heck he feels like.
Ah, yes, the hallmark of the bad DM.

OP, looks to me like you just have terrible luck.

19. Originally Posted by General_Gronker
Ah, yes, the hallmark of the bad DM.
The hallmark of a bad DM was someone who unimaginatively stuck to the letter of the rules, despite the fact that every DMG ever written encouraged not doing that at times because it tends to devolve into a unfun game full of irritated and bickering players.

Fudging die rolls for the sake of not ruining a very fun session ending it early, or making a whole bunch of great fun suddenly come to a grinding halt in one anti climactic bit of bad luck; is the hallmark of a DM who values fun over single minded following of procedures and rules.

Bad DM's also tended to be the kind who got ****ed off when you didn't follow what they planned out for you, and who couldn't wing it when the party went off script and decided to turn around at the gates of the keep and go check out the wilderness to the south... the direction you hadn't even drawn borders for yet (how dare they!). Imaginative DM's don't lead the party by the nose, they aren't scared of winging it, and they have a DM's screen for the purpose it was invented for; to hide the real die rolls from the players.

20. Originally Posted by Oxarhamar
the D20 more aesthetic than functional
Like in the rest of the game, more or less.*

* May include sarcasm and hyperbole.

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts
•

This form's session has expired. You need to reload the page.