Unpacking Failed Attack Rolls in GURPS

There’s an interesting (drink!!*) thread over on the GURPS forums about what a “miss” means. It’s called Failed Attack Rolls, and there’s a concept in there that, especially coming from the source, makes one go “Hmmm” a lot.

Let’s start with two quotes, both originating from Sean Punch, AKA Dr Kromm, the GURPS Line Editor. Also, a note: I’m not looking to quote him to fight, or to agree or disagree. The thread made me think, perhaps even to reflect (I’ve been reading Steven Brust’s “The Phoenix Guards” and “Five Hundred Years After,” so if you detect a bit of Tazendra in my statement, you’re not wrong).

In any case: I reflect, perhaps I even wonder.

The Quotes

Here’s the original note by Sean:

Mostly this. You failed at your roll to capitalize on an opening and/or seize the initiative, so you stood there doing nothing but defending. You can fix this by increasing your aggression (All-Out Attack (Determined) for +4, at the cost of giving your enemy an opening), falling back on textbook attacks when there’s no opening (Telegraphic Attack for +4, at the price of attacking directly into your enemy’s strongest defense), or learning to fight better (improve your skill, at the cost of many hours spent in the dojo, gym, kwoon, or whatever).

Missed attack rolls aren’t blows that hit with insufficient force. Too many things in GURPS (Melee spells, Contact Agents, etc.) rely on a mere touch for that to be a good ruling. Blows that connect weakly are things like successful attack rolls met by unarmed parries that prevent all damage*, and successful attack rolls met by failed defense rolls where the ensuing damage roll fails to penetrate DR.

* It’s safe to assume that in an unarmed fight, not all punches and kicks stopped by unarmed parries are warded off or blocked. Most are the blows you see sport fighters landing by the dozen in a match. A skilled fighter rolls with (not Roll with Blow – the realistic version), turns from, or otherwise minimizes the damage of these; his efforts count as a GURPS parry. These cases do result in contact under the rules.

This quote is from the first post of the thread linked in the intro, but was from a different-but-similar discussion.

Some of the responses were predictable – and not because they’re insufficiently thought-out. They’re more or less the way I’ve played GURPS for years, so I’m naturally sympathetic to the viewpoint. So sympathetic, in fact, that it’s how I run my own games.

In summary, some of the various replies:

  • Missing a roll has consequences, such as ‘weapon unready’
  • Making an attack has implications, such as when you are below 0 HP
  • Nothing in the above takes away from ‘you did do something, and it was ineffective’
  • What the heck is wrong with ‘swing and a miss?’

Sean followed up with another comment:

The problem with “swing and miss” is that it’s hard to do.* Yes, a person can do it, but most swings don’t miss. It’s very weird to imagine people are swooshing weapons around every second and missing all the time. It’s much more like a real fight if they’re only seeing and/or capitalizing on openings every few seconds. Save swooshing for critical failures. It creates no inconsistency with the rules to assume this:

  • Attack, critical miss: Swooshed clumsily, with bad results for you.
  • Attack, miss: Simply didn’t act, for want of initiative or an opening.
  • Attack, hit, defended: Blow hit enemy’s shield or weapon (block or armed parry), connected lightly but ineffectually (unarmed parry), or was ducked or sidestepped (dodge).
  • Attack, hit, not defended: Blow arrived faster than enemy’s reaction, and hit squarely.
  • Attack, critical hit: Blow caught enemy flat-footed – they simply didn’t act.

* One of the most common criticisms of GURPS is in the vein of, “Gee, how can an ordinary person swing a baseball bat at someone’s head and miss? But that’s what happens with -5 for the default penalty and -5 for the head, even with the +4 for Telegraphic Attack!” The obvious answer is, “They didn’t swing.”

Now, it’s very clear from his words that he’s talking about melee combat. One can even perhaps take it farther, and read “swing” as “thrust or swing,” or as “any sort of melee attack.”

So, before we dive in (or take a swing, perhaps?) at thinking about the suggestions, I want to wave my own flag a moment.

The Articles

One of the comments Sean makes speak directly to things I’ve addressed in Pyramid articles, in two different ways. The key bit is:

“It’s much more like a real fight if they’re only seeing and/or capitalizing on openings every few seconds. Save swooshing for critical failures.”

For ranged weapon attacks, this is handled with my article “On Target” from Pyramid #3/77. I won’t beat around the bush here: much like Fantastic Dungeon Grappling from Hall of Judgment (and its 5e cousin, Dungeon Grappling) the rules there – treating Aim as a skill roll with an effect roll after, just like an attack – is simply better than the basic rules. It’s a simple system, using the same attack roll/effect roll paradigm as regular hits, the variable effects avoid the worst of “metronome syndrome,” and the trade-off of uncertainty and reward for using it is very high. The end result for ranged attacks is that you can spend a variable amount of time aiming before you decide that the delay/reward trade-off meets your needs, with the off chance of both critical misses (“my aim was crap, but I shot/loosed the projectile anyway”) and critical successes (“my stance and motion was perfect, my sight picture flawless, and I can shoot instantly that turn with my aim bonus!”) make it an exciting way to reward mechanically what folks do in the real world. So for ranged attacks, On Target is your answer to the “every few seconds” thing.

For melee attacks, the lull and flurry behavior is the direct objective of The Last Gasp, whose Action Point and recovery rules invoke additional complexity in order to achieve that same goal for melee combat. Games that have used it in actual play have reported outstanding effects; those that avoid it just don’t wish to add more book-keeping to their games even on a theoretical/consideration level.

Both approaches/perspectives can be right, depending on your table.

Point is, GURPS gets it right – very right – when it comes to the theoretical and practical pace of furious combat, as one can see from my article  Technical Longsword: GURPS gets it right.

My own perspective on the “what if you hesitated, not committed” question is that while it’s potentially a valid way to reframe the declaration and outcomes, for melee attacks it would need to have some gentle clarification in several places, and for ranged attacks, by and large, this is not how GURPS seems to expect things to work at all. The entire sections around Hitting the Wrong Target and Firing Through an Occupied Hex (both on p. B389) would need to be reworked in order to make sense with the hesitation rule.

I hasten to point out that this rework needn’t be that extensive:

  1. Check to see if you hit your intended target (this is RAW)
  2. If no, check to see if you hit the wrong target or pasted a guy in an occupied hex (also RAW)
  3. If also no, decide whether you wish to loose/fire anyway, or hesitate.

Why loose/fire anyway? There are optional rules that can be in play about laying down fire, having a miss cause a Fright Check, suppression fire, area fire, etc. that might have secondary effects even on a miss at the point target.

If you hesitate, I’d say that if you were aiming you probably lose it (there needs to be some consequence) if playing by RAW, or you invoke the maintaining your aim rules from On Target if you’re using those in order to keep it. If you were taking All-Out Attack, I’d maybe trade a larger bonus to hit (very much needed in ranged combat!) with a commitment to loose/shoot no matter what.

Or maybe go the other way: All-Out, since you’re giving up defenses, means that you have the OPTION to hesitate, while a regular attack means there’s no hesitation – you just shoot.

So it’s not a “alack and woe, the game is ruined because of this interpretation!” situation. But it definitely does need to be dealt with. It’s not crazy, but I think the assumption that “I declared an attack, so I swing!” is baked into the rules, and requires some flexibility to interpret with the “hesitation” or non-actionon in mind. So doable, but will require some deliberate thought, and doing that beforehand, and communicating that to your table, will be important for expectations-setting.

On Guard!

I think that if one were looking to turn a one-second GURPS combat turn into first a question of “do I see an opening worth exploiting/have I created an opening into which I can strike?” and only then a “did I whiff, get deflected/parried, or strike home?” question, and one that doesn’t result in table-flipping rage or game-killing boredom, one would have to design it in.

The most common situation that we’ll want to make sense is a near-duel, or at least lethal one-on-one combat. Two foes of nearly any relative skill level, but with similar training and equipment.

That is: if it can’t take care of two folks meeting for an unarmed fight, or two sword-and-board guys squaring off with each other, then when it gets into “man vs monster” or “many vs one” or “guy with shield and spear vs three octopus monsters carrying a total of twelve long knives” it probably won’t have a prayer.

Circle and Evaluate

One thing we see a lot when teaching weapon-and-shield combat is that if the students start off too close, you’re in the you’d better be responding right the heck now pace, which creates both a real and a false sense of vulnerability. When this happens, either both students get dead very quickly (double-weapon kills), or one instantly seizes a time-of-the-hand advantage, or they close to grappling distance.

All of these are valid outcomes! But the “circle and evaluate” issue that most folks (including me!) seem to acknowledge as the expected outcome requires a bit of setup space. Enough that you start, as we say in class, “out of measure,” which in GURPS lexicon means “I can’t reach you with a Step-and-Attack, All-Out Attack (Long), or Committed Attack (Extra Step).”

I actually don’t worry much about “but I can reach you with a Move-and-Attack!” distance, since that kind of attacks is still going to allow a quick response.

Still, if I’m in, as Sean notes, a “You failed at your roll to capitalize on an opening and/or seize the initiative, so you stood there doing nothing but defending” kind of mode, it’s because I’m looking for an opening. Or trying to make one.

That sounds to me like most of my actions are really something sitting between an Evaluate and a Feint. A motion designed to provoke a response.

Mechanically, that should look like one of these things:

  • A Quick Contest of weapon skill, much like a Feint, or
  • A penalized skill roll, with an effect determined by margin of success (and the penalty might be the difference between the other guy’s skill roll and 10; effectively an “average” defense)
  • A penalized skill roll, with an arbitrary effect roll (say 1d), perhaps boosted by +1 for every -2 you take to your skill roll

The quick contest method is pretty GURPSy, but it means an opposed roll for every action every turn. That way does not lie speed of play. The other two aren’t bad; the first is just “you have an average roll on the first test for the foe, and can simply write down a delta from 10 as a penalty (or a bonus for sub-10 skills). The last way bounds the effective bonus you can get at 1d. Another version of this would invoke something like the “training bonus” from Technical Grappling. As you go from +1, +2, +4, +7, +10 above your stat, you get +1 through +5 to your effect roll on a success.

Whatever, the point of the probe, evaluate, or maneuver is to see whether or not you create or perceive an opening, either right then, or on the next turn.

The flip side of this is that reactions to a “hey diddle diddle, straight up the middle” attack on the defensive end really should be higher in this case.

The Dance of Death

I will admit that I like the effect roll version, since that take the place of the normal attack/damage roll. And Technical Grappling established various uses of control points, including “defend better” and “reduce hit location penalties,” which translates fairly well into “exploit opening.” Building up enough feint/control/positioning advantage until you can make a strike (lower penalties to hit, or lower defenses) and then utterly deny a counter (an active defense, or perhaps defensive feint, so that the other roll simply can’t succeed).

That’s how we try and fight in viking training, and how it seems a lot of manual-based HEMA works – build up position and openings, and position so one can kill and fade without being open in return.

But that means that, as Sean notes, to force an opening would be an All-Out and/or Telegraphic Attack, and do do that all at once – in one turn – might be a rapid strike. Seeing an opening and making an attack in the same second, rather than maneuvering for the opening, would thus be covered under existing rules, and against a competent foe, would be very challenging. You wouldn’t get a defense roll, the foe would get a +4 to spot your opening (the +2 to defend mirrored) as well – because you’re leaving yourself open.

One Door Closes, Another Opens

That’d be another way to go, too. Where a regular attack lowers the threshold for the foe to spot an opening by a bunch, Committed still more, and All-Out even more. Defensive Attacks would open you less, but do less damage. That sort of thing.

That would be another reason to do the maneuver dance for position: to use those points to close off the line you open in your attack.

About Feints

I’ve ignored it, but honestly, making the Feint the go-to mechanic each turn instead of an attack would arrive at the same place with minimal rewriting. If the basic defensive target were more difficult – Dodge, Parry, and (especially, perhaps) Block were simply higher, which would demand a Feint or Setup Attack to score, that would probably achieve the same goal with less fuss.

Feints/Quick Contests can be pretty swingy, though, so one might instead treat a Feint or Footwork play as a Quick Contest of Parry or Dodge scores, which would flatten things a bit, keep margins of success lower.

What’s The Point?

Really, this meandering is just that. Wondering out loud how you’d take “attack spam,” which is more or less how a lot of fights work in many games, and GURPS is one of them, into something where more careful action/reaction is modeled abstractly, but modeled as “the first thing you do.”

Modeling it as control points, advantage, or centerline control – whatever – does a few things. It quantifies, it. It makes it similar in kind to the attack/damage buttons we have now. It means that no turn has to go by without someone doing something (and that can include “attacking” to fight more defensively, by spending your points to close out and bolster your own passive target number for someone to spot an opening on you).

That would allow a lot of footwork, and would have the side-effect of really punishing engagements on an infinite featureless plain. Punishing the players and the GM, because boring. But I’ve seen fighters and fights like that too, where one combatant has a really good eye for distance and openings, and like Iceman in Top Gun, just waits for the other guy to get frustrated, lose patience, and force something.

Ultimately, one could probably make a system that did this. It might even be fun. Certainly it could model the sort of duel of maneuver that you can see even before blades clash. It would likely accommodate Setup Attacks (probably not Feints as written, though, which would be subsumed into the new Evaluate/Probe/Footwork/Positioning mechanic) pretty well too.

Defenses would need to be higher in this case, and playtesting would be required to figure out how much. Instead of 3+[figured quantity], that number might need to be more like 6 or more. That way, against someone with Weapon-16, you’d be looking at a base defense of 14 or more, even higher with a shield.

What would need to be done is to find the right mix. “Something has to happen every turn” gets too frantic, too fast, and every boxing match ends in death of a combatant after a few rounds, if that (a one-minute round is, of course 60 turns in GURPS). If it’s “crit fishing with feints” that’s not full of verisimilitude either.

It’s finding the right level of risk/reward. Where a few turns of footwork and feint are not just a good idea, but really fun. Resources are involved or decisions are made with tactical value. Fighting in teams can be the equivalent of “Everyone’s Invited” from Setup Attacks (Delayed Gratification, Pyr 3/52) where one can feint and draw off defenses so a friend can help. Grappling would be invoked to break stalemates and make footwork advantages or attack mistakes punishing.

GURPS in particular is amenable to this sort of fighting and tactical play, because in many cases, the death spiral is severe enough (and can be made more so if needed) that avoiding exchanges of wounds is already baked into play to a certain degree.

Parting Shot

It would be an interesting experiment to try out, and some of the suggestions above require fairly minimal violence to the rules as written. I do think, however, that while a reinterpretation of “miss” into hesitation isn’t an impossible reading of things, I also think that very few folks actually read it that way, and fewer play it that way.

But I’ve also written two different articles, one for melee attacks, and one for ranged attacks, that point squarely at the “slow down the attack spam” factor. Verisimilitude is only one reason. The other is that slowing it down actually makes the battlefield smaller. A friend has more than one second of move to be able to come aid a comrade. Folks could play “max defensive” to fend off multiple foes until they could fall into a fighting line. And of course, footwork and feinting and binds and all the things that make real weapon fighting interesting would come into play.

Lastly, some of the advantages and disadvantages of weapons might be able to be highlighted more. A one-handed spear might contribute very little to defense in close, or a shortsword be something folks actually want to take.

Hard to say. But against human foes, a simple rule or mechanic that makes footwork, evaluation, feints, and setups happen as emergent behavior would still be fun to play. I assert this with some level of empirical evidence, because I’ve enjoyed the hell of out of the simplified Fantastic Dungeon Grappling rules, as well as I won’t run a GURPS game that doesn’t do On Target for Aim – recognizing that The Dungeon Fantasy RPG isn’t really GURPS and making that rules switch isn’t quite an easy toggle if you’re trying to keep it to the Dungeon Fantasy RPG core with few changes.

One could even invoke morale rules, especially on NPCs. If you keep your “advantage-o-meter” pegged at maximum (and there would need to be a maximum), then after a bit of footwork and a few aborted attacks by an outclassed foe . . . they might realize it, have to make a Fright Check, and if they fail, either yield, run away, or do something desperate (and then get killed, likely).

Which again does not strike me as implausible.


*Some folks have verbal tics. This is mine. When I’m thinking about something in a podcast or discussion, I’ll respond to the point with “That’s interesting!” as I consider what was said and how to reply. My friend Christopher Rice calls me out on this all the time, and has, apparently turned it into a drinking game.

4 thoughts on “Unpacking Failed Attack Rolls in GURPS

  1. A missed attack roll could be other things too.

    * Attacker started to attack but didn’t follow through.
    * Attacker attacked but details of their & target’s positions/weapons meant it failed without requiring significant defensive effort.

    Part of the to-hit roll is that the target is a moving human combatant, or else there would be another +4 to hit for an unmoving target.

    Unless it was removed in 4e, another material effect of making an attack is that the attacker’s weapon is considered in the target’s hex for purposes of other figures who might want to target that weapon with an attack.

  2. A few things stand out to me from this discussion that seem to be missing from GURPS.

    The first is a more cinematic view of combat. If you actually watch a fight, there’s lots and lots of false starts or extremely ineffectual blows, or moments where someone is going to attack and then begins to lose their balance and has to correct, or aborts from an attack to a defense, and this is often with professional fighters, who are much better than you are. Thus, this view that “fighters rarely miss” is based more on the choreography of a cinematic battle, where each character successfully executes their intended move, whether that’s to land an attack, or deflect said attack, and it becomes a battle of strategy and wits, where skill and prowess is a measure of your ability to control a fight, rather than how well you can execute your moves.

    I think this is a fine way of wanting to see a fight. A given campaign is more likely to be influenced by action films and kung fu flicks than by real MMA matches or HEMA.

    The second is the circling and evaluating, as you describe. Not only does this happen, but there’s a contest there: the one who circles and evaluates better tends to win the fight. He reacts when he should, when the other is momentarily distracted and gets that initial edge, which he leverages into a big edge and then wins the fight, often remarkably quickly. A lot of good defenses aren’t in the form of dealing with a single attack, but in breaking the other’s initiative. A good fighter “gets inside the OODA loop of the other fighter,” and another good fighter knows how to break the OODA loop of the attacker, to regain initiative.

    I very rarely see RPGs that handle these models of combat well, and when they do, they tend to be so abstract that the GM isn’t sure how to describe them. GURPS, while “wrong,” has the distinct advantage of describing physical actions in combat; I don’t think I would remove that. But I definitely find myself leaning towards some way to insert tactics, gaining initiative and trying to keep it/break it, etc into the fights of GURPS, whether they’re sword duels or starfighter dogfights.

    I’ve found myself increasingly drawn to your idea of control points. I might talk to you about that again soon, if you have time.

    1. I hope it came across as such in the article, but I also disagree with “fighters rarely miss.” I’ve found that at least in unarmed combat, folks miss all the time. Rank amateurs just whiff; more experienced fighters get blocked. With weapons, the biggest issue I’ve run across is not a miss with longer weapons but the double-kill, where both take wounds. We spend so much time hovering at the max reach of our weapons that many misses are blows that come up short. That may well be the most common type, actually, and is measured in the scale of sub-1-yard that GURPS calls a hex. One might even call it “A Matter of Inches.”

      Happy to chat control points, too!

  3. This was a damned great post, in my humble opinion.

    A thought. If I’m reading you right, the one of the most promising lines of thought here would have you do the following:
    1) You take something like an Evaluate maneuver to try to find an opening, maneuver for position, or whatever;
    2) Then you make a (roll-low) skill roll to see if you managed to find one.
    3) If you succeed, then you make a (roll-high) effect roll roll to assess how significant the opening was.

    I like that a lot, since it makes Evaluating feel much more dramatic and significant by giving it a lot of mechanical interest. It also raises the possibility of making interesting character design choices to capitalize on one or another stage of this process. For example, you might invest in being really good at the initial skill roll to evaluate (whether because you’ve bought it up as a technique, or perhaps – just spitballing – because you’ve got a perk that allows you to float the relevant skill from DX to IQ or Per or something for this purpose). Or you might be very good at the *effect* roll. Could one perhaps base the Effect roll on an existing stat, just as normal combat damage is based on ST? Unsure. But perhaps would be good if some fighters were better at it than others?

    In any case, it raises the possibility of some fun builds, allowing us to differentiate mechanically between fighters who hang back and feel their opponent out before committing, and other fighters who charge right in instead – and so on.

    One thing to note – I’m not sure if it matters or not – is that the opponent gets no “defense” to *either* roll. Just to state the obvious, the analogy here is with the existing roll-to-hit, then roll-for-damage sub-system. But in that sub-system, the defender gets a chance to defend at both stages – first an Active Defense against the hit, then DR and such against the damage. Is it making things crazy complicated to wonder if there might be equivalents here? I suppose the reason I ask is that if the defender doesn’t get a chance to defend differently against these different phases of the attack, then I start to wonder whether it’s really worth differentiating between them at all.

    That is, I start to wonder: what’s the difference between the skill roll to see how well you Evaluated, and the effect roll to see… how well you Evaluated? And then I begin to think that it might be better just to go back to the simpler system of making a single skill roll (perhaps opposed). But then I dither, because it’s nice to have some randomness in there – it seems like that’s one of the nice aspects of the “On Target” rules for ranged fire.

    This is now a long ramble, but I really just wanted to say that I liked this post a lot. Interesting!

Leave a Reply

Your email address will not be published. Required fields are marked *