Why Does Elite mean rolling 1's every other d20?

[Gandalfs_Ghost]

I don't believe it. Maybe there is one like this, out of the thousands of d20 generators.

Prove it's in place in DDO.

Thanks for that.
Havent needed to clean my monitor for a while but now that its covered in coffee....

[DSL]

We lined up all our dice on the concrete basement form against one of the cinder block walls. We grabbed one of the offensive dice, and and placed him in front of his fellows.

Threatening the dice remaining in line, we casually said, "This will be your fate if you continue to roll such ****!", and promptly smashed the singled-out die with a hammer into bits and pieces.

Our luck seemed to change for the remainder of the night, or so my childhood memories claim.

If only we could do something similar here.

-Vorthian

I do remember one player who had a habit of hurling poorly-rolling dice out the window, but I'm not sure if that encourages or dissuades the rest (I mean, they are being "set free", aren't they?) - I would estimate somewhere between 10-12 d20's were cast out into the wide world this way to meet their fate, for good or for ill.

[Thelic]

I know that there is one thing that will always happen. When your in the worst spot you could be and need to roll anything but a 1, you roll a ******** 1....

[Twerpp]

I hear the next raid loot trinket makes you immune to 1's by automatically upgrading them to 2's. Too bad raid loot will be harder to get next mod! xD

[stockwizard5]

I was getting a little concerned over the number of times I was failing a UMD check on heal scroll. I have a 36 UMD so basically I fail on a 1,2 or 3 (that is about 15%). So just for fun I fired off 100 heal scrolls standing inside the Reaver door (results follow):

1: 26
2: 10
3: 6
4-20: 58

The chances of the d20 UMD check being truly random appear to be low

[Mad_Bombardier]

So just for fun I fired off 100 heal scrolls standing inside the Reaver door (results follow):

1: 26
2: 10
3: 6
4-20: 58

The chances of the d20 UMD check being truly random appear to be low

Ouch!

P.S. Maybe you should check with another, cheaper type of L6 scroll, such as CMW:mass. Then, you'd have a better sample.

[stockwizard5]

cheaper

100 clicks takes the same amount of time - plat is meaningless

ps: mod 5 - blah blah - not the point

[stockwizard5]

I was getting a little concerned over the number of times I was failing a UMD check on heal scroll. I have a 36 UMD so basically I fail on a 1,2 or 3 (that is about 15%). So just for fun I fired off 100 heal scrolls standing inside the Reaver door (results follow):

R1: 26
R2: 10
R3: 6
R4-R20: 58

The chances of the d20 UMD check being truly random appear to be low

Some quick math: Sum k=27..100 (M k)p^k(1-p)^(M-k) - Looks right but may be wrong ...

Shows that P(R1 = 0.05) -> 0.0000000000025

I conclude that P(R1) != 0.05

[Mad_Bombardier]

100 clicks takes the same amount of time - plat is meaningless

Ok, then use Heal scrolls. What do I care!?

My point was to test with a bigger sample; 500 or 1000+ attempts. Anyone who has payed attention should know how incredibly streaky DDO's RNG is.
The question is whether it truly balances with time.

[In_Like_Flynn]

I have a fellow in my PnP group who rolls only 19s and 20s, and throws in an occasional 1 so that nobody gets suspicious. Oddly, the monsters he fights have equally good rolls for hit points.

[Okita]

I was getting a little concerned over the number of times I was failing a UMD check on heal scroll. I have a 36 UMD so basically I fail on a 1,2 or 3 (that is about 15%). So just for fun I fired off 100 heal scrolls standing inside the Reaver door (results follow):

1: 26
2: 10
3: 6
4-20: 58

The chances of the d20 UMD check being truly random appear to be low

I guarantee that anyone with a UMD character will absolutely agree with you. I have NO doubt that the dice are rigged as far as UMD rolls... I've got a rogue that fails 50% of the time when I should only be failing about 20% of the time... matter of fact I may start keeping track....

(ps, I can't afford doing a 100 heal scroll test, but I may take off items and such for a lower UMD and work with some lower lvl scrolls...)

[Starcloud]

It's been obvious to me ever since DDO launched that the dice-rolling functions in DDO are streaky.

The developers claim they tested it over tens of thousands of rolls and came out with the expected probabilities... they refused, IIRC, to even consider the fact that the problem might exist in smaller sample sizes.

I've been keeping track of the number of times my Warforged fails to cast a repair spell, and so far the percentages are around 32%, session over session... but the *actual* chance of spell failure is supposed to be 25%.

Interestingly, he never fails to cast Featherfall.

[PaintHorseCowboy]

I'm not talking about the DDO number generator. I'm talking about real dice.
It's a documented fact that dice have a certain number of 20s that is replenished by rolling 1s.

I think this explains everything.

Argue all you want about streaky RNG's in DDO.

Sword smacked that nail right on the head.

DDO is supposed to resemble PnP as closely as possible, so why shouldn't "Hot" dice and "bad" dice exist also?

Got a problem, look into changing your die color. *grin* Might be a silly PnP superstition, but come on, talk to a PnP geek and many will tell you this actually does appear to work.

[Sue_Dark]

Well, it has to do with the way a d20 works; or rather, the magic that works upon it.
There is a little-known fact that a d20 is created with a random number of natural 20s built into it. As it scores more and more 20s, it adds to its counter until it is depleted of all its d20s. At this point, to compensate for its lack of 20s, the d20 shifts the rules of probability so that 1s appear more often, about 7.5% of the time. Once a 1 is rolled, an extra 20 is added. This keeps going until the die has around ten or fifteen 20s stored up, at which point probabilities normalize once more. Until you run out of 20s again.

So you aren't rolling more 1s on elite, you're rolling more 1s because you're doing elite quests on depleted dice. If you use a brand-new die every time, you should be more lucky with your rolls, especially due to the Beginner's Luck charm on all new dice that slant probabilities of a 20 toward 7.5% and a 1 to 2.5%.

Hope this has been educational.

That was great, thanks for the grin early in the morning.... usually i'm just reloading from my drive to work....

[Sue_Dark]

I was getting a little concerned over the number of times I was failing a UMD check on heal scroll. I have a 36 UMD so basically I fail on a 1,2 or 3 (that is about 15%). So just for fun I fired off 100 heal scrolls standing inside the Reaver door (results follow):

1: 26
2: 10
3: 6
4-20: 58

The chances of the d20 UMD check being truly random appear to be low

Hate to mention, but without recording all of the actual numbers rolled, you arent providing an accurate indication of the randomness. Please remember that EACH ROLL has that ~15% failure rate, not 15% of 100 rolls. If you think about it long enough you will understand. Based upon the info you provided, it looks like you expected to see exactly 5 each of rolls 1-3 or some other nonsense.

As another poster mentioned, grab a hundred or so of a lower spell, reduce your UMD to the same level as to have the same success fail NUMBERS on that scroll, strip naked (to remove ANY arcane failure, cause it does affect scrolls as my perma dead bard could attest, were she still alive) and record every roll. By recording all of the actual numbers, you can verify that a 1 and a 20 and everything in between have 5% chance of landing, (which isnt really the case) and you'll also waste more of your time worrying over something that is so trivial as to be a non-issue.

[stockwizard5]

DDO is supposed to resemble PnP as closely as possible, so why shouldn't "Hot" dice and "bad" dice exist also?

Not so much a problem as a desire to understand -

The chances of my results are less than 1 in a trillion if random
The psuedorandom generator is assuredly reasonably random

Therefore I conclude that I am getting appropriate random rolls that are being translated into the results that I recorded.

Could this be like the 1d4 roll for MM - if they are willing to modify the results (or the base) for one roll - then I am interested in where else they have perhaps intentionally or unintentionally modified the outcomes.

The determination of DDO spell efficiencies was greatly affected by the altering of the underlying math. I am interested if target UMDs are likewise affected?