Just a reminder/disclaimer. This OD&D series is based on me, a longtime gamer and amateur game designer, reading, reading about, and encountering OD&D for the first time. I've found the whole thing fascinating, and I thought my relatively naive reactions might be interesting to someone else. I claim no expertise.

Two columns ago, I talked about the slim chances of getting to play a Paladin or Psychic character if you roll by-the-rules, and then the even slimmer chances of getting such a character past the first level meat grinder.

Old Geezer had an insightful comment about Paladins specifically. I hope he won't mind if I quote him here:

"In 42 years I've had 2 Paladins. And it's a damn good thing, too, because they are every bit as powerful as they seem to be.

Some of us consider this a feature, not a bug."

My question to Old Geezer, and to anyone else with a lot of OD&D experience, is "Is this ideal, or would it have been better/more fun to have a more accessible Paladin class that was more in line with the power level of the other classes?"

I can see how this could be answered either way, but I don't really have to context to do more than guess. Only a longtime player and DM could know.

My strongest instinct regarding Psionics is that if you're going to include an entirely different magic system, you should make it accessible. But my second thought is "only if you want it to be a central part of your setting." If you want it to be a rare thing that still holds surprises for the players, make it super-rare for PCs to be psychic, and save the psychic monsters back and something the GM can bring out once the players think they have everything figured out. In other words, do exactly what Gygax & co. did back in the day.

Skipping subjects slightly, the rarity of Paladins and psychics (roughly 1.85% and 2.5%, respectively) got me thinking about some of the other longshot chances in older D&D.

Here's a basic one: the Attribute rated 18. This traditionally gives a +3 attribute bonus and/or a big bonus to experience points earned (depending on edition, of course). Assuming you're using the standard method for creating characters (rolling 3d6 six times, in order), your chances of getting an 18 in any particular trait are 1 in 216. Your chances of getting an 18 at all during character creation are 1 in 36 (six chances at 1 in 216). These are the same chances you have of getting a 3 in an attribute.

Of course, you can sometimes get to 18 by swapping Attribute points, if you roll close to it, and roll well in the "reducible" abilities. (this varies by edition, and changed between White Box and Red Box). But the point is, you won't end up with many characters with the coveted 18 Prime Requisite if you're rolling by-the-book. And that's a good thing.

Here's one that has less effect on the game, but stunned me by the sheer magnitude of its improbability. Remember the 30d10 bandits I wrote about a few columns back? There is a chance that they will have a magic-user and/or cleric with them, as well as their fighting-man leader. And if there are exactly 300 of them, they will definitely have a magic-user, and the chance of them having a cleric with them doubles to 50%.

If they have exactly 300. Let's unpack that. If they roll thirty ten sided dice and every one of them comes up 10.

And I know they don't mean for you to roll 3d10x10 (which is my first instinct), because the example has 183 bandits appearing. Even rolling 3d10x10, only one bandit group in 1,000 would have exactly 300 members. But rolling 30d10... one bandit group in 1x10^30 will have exactly 300 members.

That's 1 in 1,000,000,000,000,000,000,000,000,000,000.

I don't actually know the name of that number. Fortunately, UNC does (https://www.unc.edu/~rowlett/units/large.html). It's a nonillion in American notation and a quintillion in European.

So, how long would it take to roll 3d10 a nonillion times? If you had very fast hands, and could roll 30d10 once per second, it would take you 31,688,087.814 years (There are 31,557,600 seconds in a [365.25 day] year, which doesn't sound quite as catchy as the song from Rent). The good news is that you'd be finished long before the sun dies (1 billion years). The bad news is, you'd have the worst case of carpal tunnel syndrome in history.

As best I can tell, OD&D sold about 9,000 copies of various printings before being split into Basic and Advanced in 1977 (Acaeum puts the number at 29,000, but Wizards puts it at 9,000). Let's round up and say 30,000. Then let's say that on average, 3 1/3 people play for each set sold. Let's say we have 100,000 players, all of whom have mad dice skills and can roll 30d10 once per second. Those players could get in a nonillion rolls in just 316.881 years.

In the 80?s, D&D was easily selling 1,000,000 a year. It's slowed down a lot since then, but a grand total of 20,000,000 sets isn't completely out of the question. With that level of saturation, it's unlikely that very many people played regularly but didn't buy sets. So let's see how long it would take for those 20,000,000 people to roll the dice a nonillion times if they all rolled once per second: 1.584 years. In addition to being a particularly batty world record, it would also cause a worldwide shortage of wrist braces and ibuprofen.

Okay, so was there a point in all that? Yes, actually. Although it seemed like nonsense at first, I'm really coming to appreciate the way old school game designers put extremely rare events into their games. Even though you may not really see it until you start calculating the percentages (or when you play the game for a long time), it has an effect. It allows the world to be predominantly one way, filled with "normal" fighters, mages, clerics, and thieves, while still allowing the rare weird stuff to creep in (crazy-powerful paladins, the weirdness of psionics, etc.). It's something modern game designers could do well to keep in mind.