Thursday, May 9, 2013

Creating and holding on to initiatives in GURPS combats.

Disclaimer: +Jason Packer has written down one of the point that I wanted to make. So I'll dodge the truly academy-style advice and try something more geared for GMs. I am missing the point a bit. You are warned.

Why did you leave your arm open like that?
Many good folks have toiled, racking their brains, to find an elegant way to pace GURPS combats. The obvious solutions is to encourage people to "Do nothing" or use evaluate actions. Do Nothing as in catching your breath, for example. The benefit from Evaluate actions rarely compensates for the lost turns that it exacts: why spoil a perfectly fine second of your life staring at your opponent if it isn't going to boost your overall "to hit" probabilities?

Other Melee-Academy Posts:

Traits of a good fighter (in a tiny nutshell)

Let's pretend that I get myself into a fight and go tactical. What I'd do first is to settle on an action, then posture and wait for an opening. If I am a really good fighter, I'd probably be more aware of the situation and capable to change my mind to catch an opening that is more suitable for a different course of action.

Here are some key concept: good fighters are good at:
  1. Picking a course of action that will occur fast.
  2. Seize the moment and switch tactics in a split second.
  3. Dictate the course of the combat by maintaining the initiative.

How well are these traits modeled in GURPS?

Picking a course of action that will occur fast

This is almost a given: GURPS mechanics is an equal opportunity system where everyone gets to make a decision on most turn. So in a way, it doesn't really matter which action a character takes, there will be a chance to give it a go. We can argue that a failed attack roll could be equated to a no attempt at all. This is particularly so because the defender will not roll for an active defense. Critical failures are exceptions, as they usually imply that something went wrong. So, in a way, selecting an combat action and option that maximizes the chance to force an active defense is one trait of a good fighter. It actually is a combined trait of a good player AND his/her character.

Seize the moment and switch tactics in a split second

Well, as far as split second decision goes, the RAW is taking you there. Players will necessarily take longer than a second to decide what to do with their character. There is no penalty for changing your mind from second to second. In a way, the RAW makes even the lousiest fighter have the same tactical flexibility as a grand master. I think that this trait should be purely a character trait rather than a player's: it is a matter of perception and training. I'll propose a (fun?) way to introduce some serendipity into combats shortly.

Dictate the course of a combat by maintaining initiative

The RAW guarantees all an action on most second (exceptions exist, especially when dealing with wounds). The concept of initiative is rather abstract, or at least as abstract as having the initiative in a game of Chess or Go. Playing such that you dictate to your opponent their course of action is the way to go, the mechanics for doing this is implicit. Playing blandly* is essentially throwing away this critical element of combats.

* bland as just trying to deduce hit points from an opponent without leveraging the environment and tactical dispositions.

What to do about it?

Solution 1: The trap of high probability to hit?

The average combat skills of PCs is so high that it is a better strategy to engage in a feverish flurry of blows (figure of speech, not the extra effort). Combat are dissatisfying unless they tend to go somewhere at a good pace. New players will thus want good odds to hit without fancy options: even the basic combat workflow is intimidating for new-to-GURPS players. Once that you start mastering combat, it is too late: all monsters and PCs are locked into high skill levels.   Maybe the solution to this problem is to de-escalate skills levels such that a sounder, more subtle approach to combat will be a better strategy. An other solution, as pointed out by +Jason Packer, is to make combat gritty and not shy away from factoring in tactical difficulties. I'd be curious to hear from people playing in low-powered campaigns to share whether combats are more technical, and also more fun. 

The first solution is useful only 1) if it is correct and 2) at design time for a campaign. What am I doing? I'm almost missing the point of this Melee academy... Let's try something else.

Solution 2: Shake the Evaluate action with a Combat opening option.

Every character has a guarantee that it will have one action on every turn, that evaluating will peak in three seconds, just like aiming. A player can assume that most usually  an active defense will not interfere with the next combat action. There is no such thing as an opportunity that opens and throws the combat flow into a different path because it made sense at the time. I'm not sure that we can address this without touching the RAW. I'd like to bring forward a small change that may do the trick.

The idea isn't new, +Douglas Cole  and +Sean Punch  have written about it before respectively here and here. I plan to experiment with a combat option for Evaluate : spotting a combat opening. It is all about Perception and training. I'm not sure whether this option should require a Per-based weapon skill check, or be automatic. The effect is a random combat opening: an transient opportunity that comes with a string attached, and must be used immediately. Since it's 2013, I decided to replace the random effect table with a grammar-based random effect generator.

There is a very early prototype here: All that it does right now is to generate an effect with each page reload. If the idea isn't dumb, makes the playtest more awesome and generate a bit of collective brainstorming, I'd like to develop it further.

I may post later on the underlying model to generate these effects, but for now it may be just more fun to read some of the openings yourself. There are more than 200 permutations with tunable probabilities. Some openings are off because I didn't tune the grammar yet. If a PC declares this option, it either has to immediately take action (changing action) to take advantage or mitigate the opening, or convert to a RAW Evaluate action.  To avoid making this option too powerful, I've included conditions to counter-balance the benefits, and introduced the possibility of adverse effects. The adverse effect is there to shake things up a bit, and... well... there is no real game without a bit of gambling!


  1. Over on Jason Packer's blog I suggested that a program like yours could be leveraged to really make some hugely deep tables and options sing. Bonuses could vary with a total value of (say) the equivalent of +3 to hit, with an extra hex of movement counting as 2. So combos could be a +1 to hit the other guy or -1 to be hit, and either an extra hex of movement (maybe restricted to an advance or retreat) or a +1 to defend or a -1 to his defense.

    Pair that up with certain hit locations or attack styles (grappling vs. striking) vs. handedness (primary vs off-hand; if you have a 2-hand weapon you can kick instead).

    You could even have tables depending on what either the attacker or defender did in a given turn previously, though "just click the button" has great appeal.

    1. I am all for it, and Alex and I are thinking about it right now. A quick contest of perception or Oer-based skill could set the potency of the bonus/penalty. The sky is the limit... What is there is a thin straw man.

    2. A not too intrusive way to refine the opening would be to have a bunch of buttons with initial conditions. These tags would shoe the probability landscape of the opening in the shape like this

      [who] can do [what] with [bonus] if [ end state].

      For example.

    3. You could do it with both - there is a natural opening, something small, worth maybe no more than +2 to hit, -2 to be hit, +1 to defend, maybe a "free" retreat. That gets clicked every turn.

      Simultaneously, if your previous move was Evaluate, hit up a QC of Per-based combat skill, perhaps, and either have the same opening, but with more goodies, or something totally different, and potentially much more valuable.

      Other option (just brainstorming before I jump on my conference call): make a DX-based skill check each round, with margin of success or failure giving a general level of openness. The Per-based check would be to spot that opening.

    4. This is a good time for brainstorming and thinking about that. Why don't we open a doc and put ideas down? Laying the app to try things is simple.

      As long as in the end both GMs and players think it awesome and painless to use.

  2. Not sure I get the combat opening idea. As you say, it looks like it messes with the basics of combat to much. I like the optional Evaluate rules in Martial Arts (p100). I think it would be best used as you close to fight.

    I like the app, I wish the crit buttons worked.

    1. The critical buttons will be online within days.

      Combat openings is an experiment. I'll be testing a bit of it tonight with the palantir folks. But it is far from usable.

  3. The usual tactic in GURPS lethal combats is "hit him as often as I can with my primary weapon." None of the nifty stuff that was apparently pretty key in real fights as seen in old fight manuals ever shows up. You have one optimal tactic (say, Targeted Attack to the neck with my axe) and you use it over and over and over.

    The Combat Openings concept is to provide a game-mechanical way to incentivize alternate strategies and more dynamic fights outside of "he just rolled a crit and sliced his leg open."

  4. Returning to this after more thought.

    I'm wondering if it might be fun to generate for each fighter (and since the computer's doing it, it can be automatic or roll-intensive) a list of not just "openings," but also "closures" that represent areas that are better protected than usual. While triggering this off of your foe's evaluate seems a bit odd (I was totally balanced in my defense until my foe looked at me funny, now all of a sudden my legs get a +2 to defend, while my left arm is suddenly at no penalty to hit and -1 when I try and parry) there might be ways to do it that aren't quite so WTF-provoking.

    This is somewhat in response to some noodling I'm doing, and will likely write about soon, about something that makes for fast trips around a game table in terms of who does what, but isn't boring.

    We'll see.

    1. We're coding up the next iteration (actually I'm doing that right now). The emphasis is in creating events that do not cause WTF moments AND will achieve the desired effect to drive combat into interesting twists (without gobbing things down).

      In the next iteration, the size of the bonus, and potential blowback will depend on a contest of Per-based combat skills VS the defender's DX-based combat skill.

      There should be something to look at soon. I look forward playtesting these ideas... I suspect that there will be a good need for calibration.