## Tuesday, October 18, 2011

### Avoiding DTER

If you want to create a more elaborate table, the first thing you must dispense with is the flat, linear construction of "die total equals result."  You see the table everywhere:  roll a d6; 1 = bugbears, 2 = goblins, 3 = orcs ... and so on.  And then, when you have completed the die roll, and the encounter, you go back to the same die roll to produce another encounter.  And after you've played using a table like this for a long time, you get sick of tables and go looking for some other aspect of D&D in which to get interested.

Now I want the gentle reader to recognize that even a board game, the thing that most D&D players decry as the evil thing that D&D has superseded, is more complex that what the DTER (die-total-equals-result) methodology offers.  Let us consider, because it is familiar, the game of Monopoly.

Let's postulate the car, the thimble and the iron sitting upon the GO square.  Two dice are rolled for each.  The car gets a four and a five and moves to Connecticut Avenue and buys it.  The thimble gets a two and a three and moves to Reading Railroad and buys it.  The iron rolls two threes, landing on Oriental Avenue - buying it - and then rolls again, getting a five and a two and ending up on States Avenue - which the iron also buys.

When the next turn starts, the pieces do not move back to the beginning and start again.  No, they proceed from the point where they are at and new results occur.  Each turn is a stepping stone towards the next turn; each previous turn creates a formula that changes the position of the pieces for the rest of the game ... and in this formula a multiplicity of games results that are unpredictable and therefore consistently interesting to people.  Monopoly is still played as a game for this reason.

However, the DTER method does NOT provide this.  When you roll the die that determines that you throw orcs at the party, DTER tables do NOT then provide a stepping stone for the next roll.  The next roll occurs completely separate from the previous roll and for that reason the table is absolutely and completely useless for the construction of a sandbox world.

What would be the alternative?  Well, to describe that, allow me to propose another example, also from the familiar file.  Consider the game twister.

Here we have an extremely simple catalyst.  We have four fundamental body parts: your left and right hands, and your left and right feet.  And we have four colors: blue, yellow, red and green.  When the arrow is spun, determining this particular body part must be applied to that particular color, we have 16 possibilities.  This does not seem like very many ... but the game transcends the possibility in favor of something more important:  the building up of result upon result.  It isn't just that you put your left hand on the yellow circle, it is that then something will need to be done with your right hand, and then your right foot, and so on, while other people are also attempting to contort their bodies likewise.

If you are going to design tables for D&D you must think at least along the lines of Monopoly and Twister.  How does this result lead to the next result?  How does the combination of these results ultimately create new and unique sequences or troubles that are, at present, unexpected or unfamiliar?

In making a weather table, for instance, if I randomly produce a likelihood for high humidity; and I randomly produce a temperature; and I randomly produce the occurrence of a warm front or a cold front ... then how does the combination of these things produce OTHER results, such as fog, violent storms, high winds, a beautiful dry sunny day, and so on and so forth.  To simplify it, if I roll three d6, they give me 16 possible results, balanced on a bell curve.  But if I roll three d6, and the specific results of each of the three dice affect the result of the other two, then I produce 216 possible results.  To make that more clear, if I designate the first die to be a type of vessel, and the second die to be a given weather result, and the third die to be a tactic for avoiding the oncoming vessel, then the combination of all three becomes a necessary interpretation of how A affects B which is in turn affected by C ... and not just designated possibilities.  Do you see?  The results are logical possibilties, given the perameters of the model you construct.  If A lands at Connecticut, and B lands on Reading, and C lands on States, we have a model to guess at who is the most likely person to successfully obtain the Electric Company.

And if the next die 6 I add to this mix does NOT reset the game back at GO, I have 1,296 results to interpret.  And the next die produces 7,776 results that can be interpreted.

Finally, if we then understand that the individual constructions which produce the interpreted results can then be balanced to fit a given circumstance, so that 3d6 can be used to determine the type of vessel, and a number of different dice and results can be used to determine the weather, and player choice can determine the tactic, we have an untallied number of results ... with possibilities worthy of making the game forever interesting.

No, obviously, it isn't easy.  But the payoff is greater than crappy DTER tables can offer.

Eric said...

.... how is a system like this going to work without climate defined for each hex? I can see how correctly defined tables will generate random weather from climate, but climate is funky and local.

The NOAA does US forecasts quite precisely - and it will spit them out in easy-to-parse computer-readable formats like XML. If only they had a similar tool that would give you the average temperature and precipitation for each month of the year for a given location....

Alexis said...

I have a methodology for that, Eric. I use the average stats for a particular region and a particular time period as a base line from which the dice produce deviations ... which are in turn diminished by certain other random results of the die. The result is an ebb-and-flow that washes back and forth around an average ... and departing from that average becomes quite unlikely.

For example, at present the party is in Cuxhaven, in August. The average high for Cuxhaven in that period is 68 F, and the average low is 57 F. From my research the record for August is 86 F (high) and 40 F (low). The average swings within that range (as near as I can measure) about 99 times out of a hundred; so that means in 100 Augusts I'm breaking a record on one side or the other. I'm good with that.

If it becomes September, the base line falls, the rest of the pattern then begins to fall with it naturally without my needing to adjust; as the base temperature lowers, the random element steadily brings the new temperature down.

The same is true if the party moves from one climactic area to another ... as they move inland, for example, and onto another table of averages, the baseline steadily refigures the weather for the new area.

Since storms and the like happen as a result of random warm-and-cold fronts coming together, and the fronts are determined randomly by the more eccentric changes in the description above, and the humidity is determined by the average of the region also, I can see randomly what days a storm occurs. From the level of the eccentricity, the greater the shift from a cold to a warm front, the bigger the storm and therefore the stronger the wind. If the eccentricity is high enough, and it is a part of the world where it occurs, a tornado can result.

I haven't figured out hurricanes yet, but since the party is in Europe, that isn't a problem.

This is more or less what I've been working on all week, and I'm almost ready to launch it. The biggest change is that I've been working on devising the weather for four daily periods, not just for a whole particular day; this creates complexity, but I am going to like it when I can tell not just that it rains, but how long and during what parts of the day.

It is all in an excel file, and it isn't even very complicated. It just works on the principle that yesterday's weather matters today.

Oddbit said...

This inspires me to consider an encounter version of such a thing with a slightly different method.

Consider the possible encounters as a stack of cards. Rolling onto that particular option causes the top card drawn. The cards are in a specific order.

Orc card stack
1. Orc scouts
2. Orc war party
3. Orc Elite war party
4. Retreating orcs

As the party defeats each in turn they face the developing opponents. They are spotted. They are attacked in force. The opponent tries to crush them, then the orcs retreat from the obviously superior foe.

Each encounter type being a simple story in itself.

Jeremy Morgan said...

Color me VERY intrigued. Please tell me you're going to post this here when you're done. I would love to see this in action.

Eric said...

As a Pacific Northwest resident, I humbly suggest that you'll want more than four daily weather periods; 12 seems about right.

Alexis said...

No good. I've been to the Pacific Northwest, where if you have to change from your polyester-mix sweater to your wool sweater it's considered a change in the weather.

Come try the continent, where not only the barometer goes up and down, but the temperature too.

Eric said...

I'd still argue for a tighter breakdown: 4 per day feels like the 3E designers going "eh, this breakdown of the Beaufort Scale into 7 categories is good enough." Yes, until you get out on the sea.... plus, the idea of all storms lasting a minimum of six hours is a little odd. (This is, of course, purely my opinion, and I don't know how much duplicated effort would be involved in a change.)

Alexis said...

You're jumping to conclusions, Eric.

1) Rain may not last the entire period. I am merely saying that there is a chance four times a day for rain, not necessarily that each time it occurs it will last a given period.

2) You presume I am saying four six-hour periods. No. The day is naturally divided into two periods, the high temperature and the low temperature; the other two periods are merely the half-way points in temperature. This makes sense to me.

Fuck man. You really burn me sometimes. Last week you sent me this utter piece of trash that gave crappy rules for weather changing once per day, and now you sit there on your ivory fucking throne without doing a fucking lick of work and give me shit because four periods a day aren't enough for you.

YOU want twelve fucking periods a day? YOU build the format.

Eric said...

What you've got going on here is really impressive. A good set of weather effects is hard enough, never mind random generation. I think I understand why a lot of commenters have suggested using real-world recorded weather conditions, despite the fact that these are going to be spottiest in the wilderness areas where they will have the most interesting effects.

On point 2, does the table also spit out what times of day each temperature/pressure extreme occurs?

To cover the final point, I said those random weather rules were TERRIBLE. Some of the weather EFFECTS weren't bad- the wind specifically- but that random weather table was an absolute joke, and I asked you to disregard it when I first mentioned it. When you brought that table up, I said it was stupid.

I'm trying to provide concrete, constructive feedback; it's harder here, since you're laying out the system in generalities, rather than laying out trade tables that I can proofread.

I'm a believer in saying one's first impressions out loud, in situations like this- as long as they're qualified with statements like "just my opinion" or "my initial thought is...." You get a lot of bad ideas this way- common sense is neither common nor sensible- but in my experience some good ideas do slip through that would get ground to dust in a more measured discussion.

I can understand how this can be annoying sometimes. "Wow! I can't make anything that cool, but I can tell YOU what you're doing wrong with just a glance!" I wish I had started off asking why you chose just four daily divisions- it would have been a lot more productive.

Alexis said...

Eric, I'm glad you didn't take it personally. I like to bite people and have them bite back.

I really didn't get anything from that 3e document that I didn't have already from somewhere else.

Sorry, for the times of day I'm using the crutch of daybreak and mid-afternoon for low and high, since those are the most common lows and highs of the day. I have to keep things reasonably simple somewhere. I'll fill in the gaps and trends as a DM.

Eric, the last point. Since no one, anywhere, ever, has even considered building a climate system for an RPG based on dividing up the day at all, sometimes your constructive feedback sounds like, well, this: it is as though I've proposed for the first time in history that rye and coke can be mixed together, and your issue is that I used C&C on my first try.

Eric said...

"no one, anywhere, ever, has even considered building a climate system for an RPG based on dividing up the day at all"

A little hyperbolic... but that system from ENWorld's still a gun-shaped prop, not a gun, whether or not it has "realistic moving trigger action."

Aaand I'm really torturing the metaphor, so I'll stop.

.... I guess what I was trying to aim for in bringing up that 3E ruleset was "what is the required level of granularity for weather effects?" I'd thought, on first analysis, that the wind-speed effects at least were approaching the bottom end of "detailed enough to be useful." I still think they would be, in a strictly land-bound campaign.... but I've been convinced by this discussion that they aren't fine-grained enough for seafaring.

I guess I'm just wondering how you see the weather-effect table working out- how granular? Temperature minus windchill plus humidity-based misery index equals the functional temperature, between -70 and 140* F. 21 divisions? 42 divisions? Add a "50-80 degrees, function normally" section to the middle, or keep all divisions equally sized?

* 140F seems like a decent top end for your campaign; no improbable adventures in magma-pool filled caverns.... although it might be worth extending the tables for a big fight in a foundry....