(Un)Lucky with the dice - How can it happen?

[SpringPlum]

I have the absolute worst luck rolling d20s ever. I'm not kidding. I once rolled 6 1s on 6 different d20s in a row. And I rolled different ways. Some I rolled from my right hand, some my left, one I spun, and another I dropped. The odds of this happening are astronomical. If you don't believe me, ask shalewind or falanor, they were there.

This same game, I only rolled over 10 twice, and it was a pretty combat-rolling intensive. My character nearly died two times.

Weirder still, I was describing this phenomena to another friend. She was skeptical until I picked up one of her d20s, said "watch this" and rolled another 1 on command. I haven't rolled well since. Occasionally I actually wish this were 2e when rolling low could be good.

On the flip side, I've seen my DM roll 4 or 5 natural 20s in a row.

 

[Chriskaballa]

*Bump*, because I feel the topic of bad luck deserves more attention than it's getting....

~Chris

 

[Saeviomagy]

6 1's in a row is not astronomical odds. Astronomical odds tend to approach infinity.

6 1's in a row is merely

1/64000000

Which, given the amount of die rolling which goes on during the average players life, isn't that much.

Also - to the individual who misses with touch attacks: you realise that a touch attack which misses doesn't discharge, right? And you can keep trying it round after round until you hit?

Finally - statistical clumping (ie - runs of a certain result) is a sign of a good random number generator.

 

[Guilt Puppy]

I think we've all run into this. One of my players has a tank Dwarf who, can't roll above of a ten in combat. Not during the last session. During the whole campaign. He manages to get along, seeing as there's a variety of things he can hit with a low roll. But he sure never uses Power Attack.

I, on the other hand... I have the dreaded futon/clear d20/little wooden bowl combo going on. It adds +1 to the CR of any encounter. My players expect a critical every session. I worry they might suspect I cheat against them: In reality, of course, it's only ever the opposite.

Of course, none of this should be surprising. Truly random numbers do not produce predictable results, and "an even distribution" is a predictabe results. This isn't to say probabilities don't come into play (anything about 10d6 I'll usually take average, as in my experience it rarely deviates by more than three or four), but remember: They don't govern the dice, the dice govern them.

This is why I don't like electronic dice-rollers. They never seem to mimic the "feel" to me of real dice, even if they do satisfy the statistical requirements.

Also worth noting: Don't discount the human ability to discern patterns, even in chaos. The little clumps of similarity which are inevitable in randomness can appear quite strange -- you can easily pick out a pattern, and isn't random supposed to mean there is no pattern?

In short, I wouldn't be amazed if somebody rolled 100 consecutive 20s. If they did so after saying they would, maybe -- but no more than if they'd picked out any other exact sequence of 100. Probabilities don't mean much in hindsight -- if something randomly happened, then the odds are 100% that it happened, and 0% that it didn't (+-5% error in reporting.)

Something else to take into cons

 

[TeeSeeJay]

Outlandish probabilities are only outlandish in hindsight. In the end, each spot on the die has an equal chance of coming up each and every time you roll.

I've spent enough time (and money) at the poker tables, and luck -- the perception of luck -- is an important thing. Sometimes, it's just a matter of sitting to the left of a player who never folds a hand, soaking up all the cards YOU were supposed to get. Cards and dice run "cold," but it's all equal in the end.

 

[s/LaSH]

I've always been fascinated by probability, especially when (over the course of an entire campaign) one player discovered that, roughly 50% of the time, he'd roll a 3 or a 5 on the d20. Good thing he developed a good BAB. This is a sustained phenomenon, using any d20 the group supplied him (and when I borrowed his, to roll massive monster attacks, I always rolled high). We started to attribute it to negative vibes in the end.

In related news, I tried something back in school for a statistics project. I took a d6, rolled it 20 times and recorded the numbers it came up - while concentrating on '1', then repeated with '2' etc. I wasn't really suprised that I got something like 20% higher results on the number I was concentrating on at the time for every single number.

I'm not convinced that it's psychokinesis, but I really do try to roll as randomly as possible. Maybe it's subconscious manipulation of the dice on a microscopic scale? If the human mind can calculate the parabolic trajectory of a thrown ball, it might be able to pick up on the subtle 6-degree trajectory of a polyhedron... just guessing, of course, but every second time I looked at my watch in high school it read 42 seconds (+/-1sec), indicating that something about the brain is outside our current comprehension.

And I'm not making any of this up. I'm a believer in the scientific method, after all; the statistics seem to indicate that something's going on that we don't understand.

 

[Umbran]

And I'm not making any of this up. I'm a believer in the scientific method, after all; the statistics seem to indicate that something's going on that we don't understand.

*Shrug*. Whether you believe in the scientific method or not, your experiment was highly flawed. With 20 rolls, you are only expecting each result to come up about 3 times, on average. But our expected error is very high. It is very likely that one of the numbers will show up only twice, and another 4 or 5 times.

In order for a statistical trial to be meaningful, the number of entries in the sample set must be very large compared to the number of possible outcomes. 20 isn't really much larger than 6. In order to provide evidence of statistical corelation to what you were thinking, you'd want to show the trend holds up after rolling the die a couple of hundred times.

 

[NewJeffCT]

We do all of our dice rolling in front of the group in the middle of the table. Even the DM makes his combat rolls there. We also do this for character creation.

But, we did have some incidents of both terrific and bad dice rolling. The one female member of our group was always incredibly lucky in the character generation process. We used to use 3d6 twelve times and keep your top six scores. Well, her first six rolls, in order, were 18, 17, 16, 15, 14 and 13.

And, in an old 2E Kalamar campaign in the late 90s, I had terrible luck early in the campaign. Even as a second or third level ranger, I would have trouble hitting anything – even a lowly goblin. But, later on, I bought new dice and landed a key critical hit on one of the party’s arch-nemesis’…

Our DM has a set of purple dice that the group has nicknamed “The Player Killers” for his penchant for rolling 20s. I think he got four 20s in a row once.

 

[ichabod]

On the subject of random events not having a pattern, that's not really true. When you get into the statistics of large data sets, you find that you can take any data set and find a pattern in it. The real goal of statistics isn't to find patterns in the data, it's to find patterns in the generation of the data.

On the subject of rolling the number you think about, 20 rolls of a d6 is meaningless. That's 120 events with a probability of 1 in 6. The expected value is np, or 20. 20% more than expected would be 24. Your variance is going to be np(1-p), or almost 17,or a standard deviation of a little over 4. Using the normal approximation, your odds of something at least that uncommon happening are about 1 in 6. Statistical significance doesn't occur until somewhere between 1 in 20 and 1 in 100. The statistics say nothing more than random chance is happening.

 

[megamania]

After reading some of the comments I feel like I found home. I have days of 8 being my highest and days of 12 being my lowest. It's just the way it goes. LOL 'course I do have certain dice I pull out when I'm rolling poorly.