A few months back, I wrote about our plan to use Method K for ability score generation in our AD&D game. Method K was a roll 3d6 in order, swap any two system with a seventh “bonus” roll which could be swapped into the mix if desired. The idea was to keep the 3d6 range of scores and the randomness of in-order rolling, while also giving the player a small amount of control by allowing score swapping; the seventh roll also gave an extra chance for a high score or a way to avoid a really low score. But at the time I commented:
One thing with our current method is that anything besides base cleric, fighter, magic-user, or thief is VERY uncommon. That’s not a problem in my mind, but I suspect that players may be frustrated with the fact that almost no one will ever get to be rangers, druids, etc. We’ll see how it goes.
How it went was “this isn’t what we want,” and it wasn’t just the players who ended up frustrated with it; I also came to the conclusion that the method wasn’t a good fit for how we wanted our game to play. So Method K is out.
Our new plan is allow players a choice of Method I (4d6 (drop lowest) six times, arrange as desired) or Method II (3d6 twelve times, keep six scores, arrange as desired). The scores are a little higher but still not “superhero” high, and the full control over arrangement means that, if they roll well enough, players can choose whichever class they want. Though I still believe that there’s a lot to be said for the forced creativity that results from trying to make the most of a PC you wouldn’t necessarily have chosen, player satisfaction improved when we made the change and now we’ll actually see some PCs besides the four basic classes. This is AD&D, after all.
Anyway, back when we were discussing this, I made a spreadsheet to simulate 10,000 sets of scores by each of the two methods. As one would figure, Method I results in more high or low scores than Method II at the cost of a lower overall average. I came across the spreadsheet and thought I’d post it.
Here is a screenshot of the top of the spreadsheet:
Each line is a set of ability scores arranged high to low. 17s and 18s are green, 15s and 16s are yellow, and scores below 10 are red. The red numbers near the top of each column are the averages for that column, so the average fourth-highest roll on Method I is 11.8 while the average fourth-highest roll on Method II is 12.2. At the right are total counts of how many times each score was rolled and a graph comparing the two methods.
The graph tells the story, so here’s a better look at it:
There’s probably nothing here that surprises anyone, but it’s nice to see so clearly how the curve of the two methods compare.
Personally, I’d probably go for Method II every time. But I can see how gambling for a better chance at some top-end scores at the risk of a bad score or two and a lower overall average would appeal to some, especially if a base class is the goal anyway. Selling out in hopes of a 17 or 18 in a prime requisite might be worth it.