I have made five previous attempts over the course of my Dungeon Mastering to produce a working weather table. The impetus was the AD&D Wilderness Survival Guide, on the whole an extremely disappointing book. I'm sorry to disparage - it is only that the gap between what was hoped for and what was received was immense.
See, next to my computer, within arm's reach, I've been keeping a book called Mountaineering - the Freedom of the Hills. I have the 6th Edition. It weighs about 1.1 kilos, has 528 pages and costs $32.
On the other hand, the Wilderness Survival Guide weighs about 0.25 kilos, has 130 pages and set me back $20 in 1986 money. In fact, this inflation calculator says I paid $42.50 in today's money. But I quibble.
The AD&D WSG was a game attempt. It is far more content-deep than any of the splat books put out since the 1990s, but most of the rules weren't practical - or rather, they depended upon poorly defined variables that made them unusable. For example, the rules on page 36 for "chance of stopping a fall or tumble" depend on the DM judging whether the slope is non-slippery, slightly slippery or slippery, versus gentle, moderate, severe and cliff-like slopes. Such attempts to discriminate, to make the rules more 'detailed,' nearly always end in arguments over gray areas, while in fact making the rule hard to memorize (a principle that is ignored throughout the book), made worse by the designers insisting on using fractions instead of percentages.
On the other hand, my mountaineering book spends 55 pages describing surfaces in excessive detail, in every way possible, right down to the right equipment for the right surface. No, there are no tables; but tables I can create.
All this has a point. I and others recognized the failure of the WSG right off. We had many talks about how to improve various points. Viewing it from the present vantage point, it did serve to inspire us to produce something better. We saw what was there and recognized a need for details that we had never properly considered. It opened our minds in a hundred ways.
A pity that the book didn't solve those problems in addition to pointing out that these were problems that needed attention. But there are many, many early books in the sciences that fell short in the exact same way. Freud, for example, became the Father of Modern Psychology that way.
Much of what I've written on this blog does this also. Oh hey, look at this problem. Here's an attempt to fix it. Oh, shit, that didn't work.
The longest lasting influence of the WSG has been, without a doubt, those damn weather tables that began with page 107.
For those who have never seen the guide, the weather table was ingenious. Or so I thought in 1986. The world was broken down into Arctic, Subarctic, Temperate, Subtropical and Tropical regions. Temperature was broken down into 26 grades of temperature ranging from extremely cold to extremely hot. These grades were each assigned a letter, A to Z.
Then, for each month, for each climate and for each type of terrain (desert, forest, hills, mountains, plains and seacoast), the book gave three letters. For example, for temperate plains in January, the letters were B H M. 'B' had a high of -15F and a low of -30F. 'M' had a high of 65F and a low of 45F. 'H' was supposed to be a kind of average between these, with a high of 30F and a low of 15F.
According to the rules, the temperature would swing back and forth between B and M throughout January, based on a random system the designers proposed. It sounded very exciting and interesting, and I threw myself into it full bore.
Only, I had climate data for different parts of the world even then, in my World Almanac, so I ignored the table provided in the WSG and spent a lot of time making my own letters for real places. Then I tried to use the system and made my first big discovery.
Weather is not temperature.
Oh, the WSG tried to include random tables for rain and wind as well, but those tables were absolute shit. I tried using them. Then I tried fixing them. Then I tried a bunch of my own efforts that failed. All this was done in the dark ages, for me, without Excel and without a better computer than a Commodore 64. 64 whole kilobytes of RAM.
When the Mac OS at my university came available to me (I didn't have money to buy my own) I tried again using Excel. That would be my second attempt. It was, I'm afraid, no better. The problem was a lack of data, a lack of experience, a lack of knowing what was important for the game, a lack of perspective and definitely a lack of detailed, well-designed world to graph any climate system onto.
In other words, without the strong platform provided by my maps, which did not even begin to exist in their present incarnation until 2005, any attempt at creating a weather system to match was doomed to be hopelessly and uselessly random. That's what I found with both those first two forms; no matter how much work I poured into them, the constant randomness is what broke the system.
Weather has to shift back and forth within a very small box - the two cats I spoke of in my previous post. In turn, those small shifts have to matter, or else they're so small they can be ignored. I wasn't getting anywhere with the WSG's idea . . . though the temperature scale that was proposed in 1986 still survives today in my world; adjusted in both degree [zing] and purpose, but still there.
My 3rd attempt began on this blog. On the whole, I couldn't really call it a system. It was an idea, at best, one that didn't make it out of the experimental stage. I never applied it to my world. It was too vague and non-specific. It was part of the process that got me here, however, as I realized that details really did matter. At the time that I posted that, back in October of 2008, I was just starting to play with the idea of doing something.
I am amazed, sometimes, at how much my world has changed since 2008. One would think that after 28 years of game-play that I would have calcified my design by now, but that is anything but the case. My world is vastly different from anything I played before 2000, quintessentially redirected from what it was when my daughter started playing around 2007 and it continues to shift and change, both in design and in function.
My 4th attempt can still be found on my wiki. If the reader likes that file, better save a copy because within a few months it will be going bye-bye. I don't use it anymore, but I'm fine with leaving it up until I'm ready to put up its replacement.
My 5th attempt was going to be an in-depth rework of the 4th system, but in the middle of working on it (and nearly finishing it) I had an epiphany. That is when I started working on Mark 6, about six months ago. The first four months of that work was done all in my head, but that's not important.
The 4th system that has been on the wiki was the closest I've ever come. Unfortunately, it still suffers from too little or too much deviation, and defies attempts to expand or modulate the results. It's difficult to say what the problems are. Partly, it depends too much upon information that isn't available for the whole world. Partly, it micromanages too many details past the point where they need to be managed - which is annoying when I want to add something (like tornadoes, for instance).
I've taken a step back and simplified. Mostly, by reducing the total amount of necessary data that I need to generate interesting tables. In fact, I'm a little stunned by how little data I do need - and how this enables me to get more results by producing more groups of rolls. A point I was trying to make with this post that got very little attention.
Next, I'm going to talk about what I've been doing the last two weeks, more or less getting on top of the problem and offering up its details now, at a time when I'm still looking for reasons why this new system fails. I can't see that yet - but that doesn't mean a failure hasn't already occurred.