Tuesday, October 12, 2010

DBAS - Contemplating Moves

As part of my involvement with the Solo DBA development effort, I have a tendency to get on an idea of something to try, and then try to catalog all of the variations of that idea. The latest idea to get this treatment is available moves.

Problem

A fundamental problem for the solo gamer is to determine what move - from all of the possible moves - should be taken for the non-player, or programmed side. In a game like DBA, where you have a command and control mechanism that does not necessarily allow a player to move all of their units every turn, you end up with an additional twist, which is to evaluate the chosen "best" move for one unit against that of another unit and determine which should be taken in those cases where you are restricted and cannot move every unit.

So, to restate, the problem is two-fold:

  1. Determine, from the set of moves that an element or group could make, which is the best move to make, and
  2. Determine, from the set of best moves found in #1 above, which ones to make if not enough PIPs are available to do them all.

Solution (or Attempts at Solutions)

In previous versions of De Bellis Antiquitatis Solus (hereafter referred to as DBAS), my version of rules for the wargamer to use DBA for solo games, my approach has been as follows:

  1. Determine the "best" move for any given element or group of elements. (The "best" move is the one determined as having the highest score - see below.)
  2. Score that move, based on:
    • The Non-Player General's (NPG) current mood (called the Strategic Stance).
    • The category of the move (Combat, Defensive, or Approach).
    • Conditions resulting from the move, such as whether the following combat gives you a tactical advantage, retreats you out of a losing combat, moves you towards and objective, etc.
  3. Play out each move, from highest score to lowest, until the NPG runs out of PIPs for the turn.
This process worked well, and despite the sound of it, did not take a lot of time to calculate the scores of the various moves. (After awhile, you pretty much got the feel for which move was better than another.)

Where I did not like the system was that I tried to bite off too much; I was trying to provide a different scoring system for each Strategic Stance of the NPG (there were five). My last effort was to reduce the Strategic Stance Values to three - Cautious, Moderate, and Bold - and then provide a preference to the type of move (Combat, Defensive, or Approach) based on the Strategic Stance. A simply table shows how it works.

If the Strategic Stance is......the First Preference is......the Second Preference is......and the Third Preference is...
CautiousDefensiveApproachCombat
ModerateApproachDefensiveCombat
BoldCombatApproachDefensive

As you can see from the above, the NPG's mood determines the type of move favored. As I used a scoring system I simply add +4 to the score for the moves in the First Preference, +2 to the score for the Second Preference, and +0 for those in the Third Preference.

This is a good start, but it still requires you score out all possible moves for an element or group before you can figure out what gets the first PIP.

What About...?

So, one of the ideas is to rank the actual moves and consider them in order. Essentially this means cataloging the possible moves and giving them point values. If each move has a value, it probably needs a modifier based upon how good it is compared to other like moves.

For example, consider the basic Move to Contact, where one element moves into contact with another element. If Move A results in a combat of +5 versus +3 and Move B results in a combat of +3 versus +3, should Move A be valued higher than Move B because A has a better chance of winning the subsequent combat (all other things being equal)? In the past my answer was yes (and probably still is).

Now consider a Group Move to Contact (two elements, moving as a single group, move into front-edge contact resulting in no overlaps but rather initially two one-on-one combats). If the Group Move to Combat results in two +4 versus +3 combats does that move rank higher than a Move to Combat resulting in a single +5 versus +3 combat? As you might imagine, the resulting scoring system that takes all of this into account could probably get out-of-hand pretty quickly.

So, if you go down this path of cataloging moves and assigning scores to those moves, your next basic decision is, do you:

  1. Rank the moves in order.
  2. Compare each element/group that can make that move to determine which ones get PIPs first.
Or do you:

  1. Rank the moves in order in order to assign a basic score to that move.
  2. Modify the scores based upon additional factors.
  3. Execute the moves in score order.
The latter sounds right, on face value, but is much more difficult to pull off, I think. Next, I'll look at some simple moves and pose questions on how to go about tackling these issues.

1 comment:

  1. Interesting ideas. I wonder if you haven't already touched upon the solution earlier? In your single move to contact vs group move to contact example, could you not tie that in to the strategic stance? It seems to me that an aggressive stance should in principle attempt to do the most damage possible (ie, should make the group attack). An army adhering to the cautious stance might in principle prefer the single attack, as it is a surer bet.

    Some thoughts, anyway.

    Cheers,
    Aaron

    ReplyDelete