When most people think of improvisation in role-playing games, they generally think about inventing the story based on what idiot choices the players make. And that certainly IS a form of improvisation. And, honestly, you’re doing that s$&% all the time when you run games. Improvisation is nothing more than the behind-the-screen version of role-playing. When the players make some dumba$& decision based on their characters, that’s role-playing. When you – as the world – make decisions based on what the PCs do, that’s also role-playing. And that sort of “narrative improvisation,” the one that introduces or affects the story elements of the game is exactly what we’ll be discussing… some other time. Because right now, we’re talking about another kind of improvisation. A sort of improvisation that most GMs don’t think of as improvisation. One that they do ALL THE TIME without even realizing it. Mechanical improvisation.
Have you ever had a player ask “can I do the thing?” or “is it possible for me to somehow thing?” Have you ever forgotten a rule or had to calculate a target number for a die roll on the fly? Have you ever had to figure out what happens when a spellcaster tries to use a fire bolt spell underwater? THAT is improvisation too. Remember, improvisation is just “performance without plan” and the whole game is a performance.
If we want to get more technical – and we always do because technical leads to nice, long juicy articles that have been called “a waste of 5,000 words” by Bill “@DungeonBastard” Cavalier – if you want to get technical, we can divide call this sort of improvisation “improvisation of outcome.” Remember the basic formula:
- The GM describes a situation.
- The player thinks about the situation and their character and chooses an action.
- The GM determines the outcome of the action.
Improvisation of outcome occurs whenever the GM has to determine the outcome and discovers the rules have failed them in some way.
When Rules Fail
When you plonk down fifty goddamned dollars for a giant-a$& hardbound textbook of player rules and then another fifty f$&%ing dollars for a useless book of supposed GM advice, you’d think you were getting a complete game, right? You’d think the designers would have finished the damned thing. But, they haven’t. And that isn’t because they are greedy, heartless bastards who do half-a$&ed work. I mean, sure, sometimes that happens. But the problem is, to be complete, the rules would have to be infinity pages. If you want to get REALLY technical, they would need aleph-naught pages – because it would be a countable sort of infinity – but that doesn’t matter. Shout out to the two mathematician GMs who read this site.
Anyway, part of the reason the rules fail is because players can literally try anything they can imagine. And that means there are an infinite number of degrees of freedom in any given situation. So, occasionally, you’ll end up with a player doing something crazy that the designers never thought of.
But the rules can fail for other reasons. For example, the rules of most RPGs are hundreds of pages long and some of them are really nitpickingly detailed. And it’s easy to forget rules that don’t come up often. Even if you try to review the rules that are most likely to come up in a given game session, you might not foresee one of the idiot PCs ending up in the moat of the castle instead of walking across the perfectly serviceable drawbridge and now you need to resolve underwater combat with the sharkgators that you just invented to teach the player that sometimes the shortest distance between two points is the one that won’t f$&%ing kill them.
In those situations, if you give ANY f$&%s about flow or momentum or pacing, you won’t stop the game so you can review the rules in question. Instead, you’ll figure it the f$&% out yourself. Or you will make a ruling.
And then, there are cases where the rules do cover a situation and you actually know what they are, but the rules are kind of stupid. Stupid rules happen. Trust me. Read a few pages of the Sage Advice archive and you’ll see some of the stupid bulls$&% that is technically true of the D&D rules.
Stupid rules occur because every rule that ends up in the game is a choice. And design choices involve weighing all sorts of factors. Generally, the trade-off is balance vs. verisimilitude. That is, what makes for a well-designed, fair game vs. what makes for a consistent world that could exist in the human imagination. And sometimes you might disagree with the designers about what is more important. For example, D&D 4E swung very much on the GAME side of that argument. Lots of things happened in the rules that made literally no sense as part of a world that made any kind of sense. And D&D 3E swung very heavily on the consistency and verisimilitude side (well, far for D&D). Lots of tiny detailed little rules existed to serve the game world. Which is correct? Both. Neither. It comes down to what you enjoy in a game.
Other stupid rules exist because the designers were trying to avoid or circumvent problems and ended up creating new problems in the crossfire. For example, there’s a whole mess in D&D 5E right now about how an unarmed strike is a melee weapon attack but does not count as a weapon. And in 4E, there was a huge issue about how, because you were not your own ally, you could not use the Heal skill to give yourself mundane first aid. These things happen because closing loopholes in the rules to preserve game balance often creates NEW loopholes and the designers either have to decide which loopholes are more important OR they simply don’t see the loopholes.
So, in any given game, you might run into a situation where either there is no rule for something, or you don’t know the rule, or you do know the rule but it’s stupid.
The Rules Are Not the Most Important Tool In the Game
First and foremost, let’s get this out of the way, the rules are NOT the game. And they are NOT the most important tool in the GMs arsenal. And if you view them that way, you’re going to f$&% yourself and f$&% your players. Let me show you something.
This is a puzzle. It’s really simple. Take a pencil and, without lifting the pencil from the paper, draw exactly four straight lines that connect all nine dots you see below. Take a moment now to try it.
Did you get it? Lots of people have trouble with it. After you’ve given it a try or decided not to bother, click this link so you can see the solution.
Why do so many people have trouble seeing that? Because we tend to imagine constraints that aren’t there. It’s a human thing. Our brains are wired to constrain problems to make them easier to think about. In this case, people tend to imagine that their lines can’t go OUTSIDE the nine dots. They treat the outer row of dots like an invisible wall.
And we run into the Nine Dots Problem ALL THE TIME in RPGs. The minute you create a rule that handles situations A, B, and C, people will assume the omission of situation D means that D is not possible. I could go into a huge rant about how that presentation of the rules of RPGs creates that problem. But I won’t. That’s an article in itself.
THAT is why, when I teach people how to GM, I don’t start with “in D&D, you resolve an action by rolling a d20.” THAT is why I start with “a player says a thing and you decide whether it’s possible or impossible, then decide whether you need rules to figure it out, then you use the rules.” That is SUPER IMPORTANT. Because the first and last thing that should be involved in every GM decision is YOUR BRAIN.
Remember in those old game shows when Alex Sajak would ask a question and a player would give an ambiguous answer that was kind of similar to the answer written on the card. And Alex might say “close enough, it’s actually “Noether’s First Theorem.” Or he might say “can you be more specific?” Or he might say “I’m sorry, that’s not it.” But very occasionally, he would actually say “hey, I’m not sure if that’s good enough to count. Judges?”
THAT’S how the rules should work in an RPG. They should be like the judges on Who Wants to Spin the Wheel of Jeopardy. Monty makes most of the decisions. He only looks to the judges when he isn’t sure how to rule.
See, the D&D rules are really backa&$wards in this respect. Because they spell out all these rules and then give the GM permission to break them. What really should happen is that the rules should spell out how the GM’s job is to decide what’s possible or not and that the rules are there to help resolve outcomes. But, WotC and Paizo still aren’t rushing to cut me a check. So whatever.
Arming Yourself to Make a Call
Okay, so what DO you do when the rules fail? First of all, don’t panic. After all, this is precisely why you exist. You exist because the designers of RPGs know they can’t make infinite rules that feel good for all tables, so they give you the power to run the game.
First of all, every RPG has one or more core mechanics. Every RPG has a basic, underlying rule for resolving ALL actions. For example, in D&D 3.5 and Pathfinder and D&D 4E and D&D 5E, it’s this:
Roll 1d20 + Appropriate Ability Score Modifier + Other Modifiers vs. Static DC
Now, there’s a little more to it than that, sure. A few little extra rules to keep in mind. For example, the DC might be an Armor Class if it’s an attack meant to cause damage. And if it is an attack, you might add a Base Attack Bonus if it’s Pathfinder or D&D 3E. If the person taking the action is using a skill or tool they are proficient with, they add their Proficiency Bonus in 5E. In 4E, they add +4. In 3.5E and Pathfinder, they add skill ranks. In 4E, an attack can target one of four different defenses: AC, Fortitude, Reflex, and Will. And sometimes the action requires a saving throw, in which case, in 3E and Pathfinder, you add the Base Save and in 5E the saving throw might add the Proficiency Bonus.
There’s a few other things to know as well. For example, you need to know how to modify rolls. In 3E and Pathfinder and 4E, generally speaking, a favorable condition is grants a +2 bonus and a detrimental condition is worth a -2 penalty. You can go up to +4 or -4 for really serious stuff, but generally you don’t. In 5E, Advantage and Disadvantage mechanics take the place of most situational modifiers. DCs that aren’t coming from creature abilities generally run in multiples of 5. The 3.5E, 4E, Pathfinder scale generally says 10 is easy, 15 is moderate, 20 is hard, and so on. 5E starts at 5, then 10, 15, 20. DCs for saving throws are calculated in a few different ways. For example, in 3E and Pathfinder, a saving throw DC is generally 10 + Ability Modifier + Half the Level/Hit Dice of the creature making the attack. In 5E its 8 + Ability Modifier + Proficiency Modifier.
You could actually fit the core rules of D&D 3.5 or Pathfinder or 4E or 5E on ONE PAGE. Everything else in the rulebook is details, modifiers, and exceptions. And honestly, if you ignore all that other crap, you’ll probably be fine. The entire fifty pages in the skills chapter of the 3.5 PHB and Pathfinder Core Rulebook is actually just rules for setting DCs in specific situations. And the sick thing is they all fit into the very general scale of 5, 10, 15, 20, 25, 30. You can skip all the rigmarole of finding the “balancing on difficult surfaces” table in the rules, assess a DC based on the difficulty, and get your DC within 2 or 3 points.
Know your core mechanic. KNOW IT inside and out. Understand the scale for DCs and understand how die rolls are calculated. Understand the different types of rolls and the different ways to modify them. In general. That will let you resolve almost any action.
And remember, this whole “using the rules” thing is STEP TWO! Step one was you deciding whether the action was possible.
And now we come down to the real meat and potatoes of running a game using the rules. What’s possible? How do you decide? How do you decide whether an action CAN succeed?
Now, there are lots of GMs who will give you lots of pedantic stupid rules. They will say things like “always say ‘yes, and…’” or “say ‘no, but…’” or they will talk about the Rule of Cool where a thing is possible because it’s awesome or they will say “let the dice decide.” Those GMs are dumba$&es. Yes, even Bill “Dungeon Bastard” Cavalier. He’s f$&%ing wrong. And now I AM calling him out. But he called me out first.
Here’s the problem with ANY of those rules. They encourage mindlessness. They encourage GMs NOT TO THINK. And that is in-ex-f$&%ing-scusable. There is NO good rule for deciding what’s possible that can be summed up in a single sentence. Not a single f$&%ing one. And if you have one, you’re wrong. Because there aren’t any.
And THAT is why we discussed Sir Bearington last week and the twin bugbears of verisimilitude and tone. The Rule of Cool is a good rule for the Ninja Turtles, but it’s a terrible rule for Edward Carnby vs. Call of Cthulhu.
Deciding what is possible comes down to three questions. And we discussed two of them in the Sir Bearington article. The first is “does it make sense based on how this world works that such a thing SHOULD be possible” and “does it fit the way this world feels that such a thing IS possible?” Often, you’ll get the same answer to both questions. But sometimes you won’t. And when you don’t, you have to make a judgement call about which is more important: the way the world works OR the way the story feels. One of them has to give. And there is no right answer. There’s just YOUR answer.
But there is a THIRD question when it comes to mechanical improve. Apart from the way world works and the way the story feels, there is a tricky question we’ve never talked about before.
The Third Triforce: The Way the Game Plays
Remember how, in Legend of Zelda II: The Adventure of Link, suddenly, we got blindsided by the existence of a third Triforce (and also that Princess Zelda wasn’t the real Princess Zelda because the real Princess Zelda had been in a magical coma for like a hundred years – didn’t see that coming). Well, we’ve talked A LOT on this site about the twin concerns of “how the world works” (AKA consistence, verisimilitude, realism, etc.) and “how the story feels” (AKA tone, engagement, etc.). And those things come into writing adventures and running games. But, the thing is, there’s a third concern. And it is a concern that only comes into things when you start f$&%ing with the rules themselves. It’s the way the game plays. It’s the mechanics. It’s game balance.
Yeah, okay, we’ve talked about game balance a few times. Mainly in terms of fairness and balancing encounters and all of that crap. And I’ve always shied away from calling that “game balance.” Because it really isn’t. That stuff comes down to setting proper challenge. Game balance is how all the bits of the system come together as a whole to help you– the GM or adventure writer or whatever – figure out what outcomes you expect and how likely they are. It’s a subtle thing.
Here’s the deal. You know that if it’s 5E (for example) and you put four to five 6th-level PCs against a CR 6 creature, they should be able to defeat it and they are unlikely to actually end up with a PC or two. That whole thing works because, when the system was designed, the clever designers figured out how all the numbers and probabilities should work so that it works that way. That’s game balance.
Game balance is also built around the idea that each participant in the game, each player, should contribute roughly equally to the outcome of the game. That is to say, the 6th-level fighter and the 6th-level rogue and the 6th-level cleric and the 6th-level wizard should be on roughly even footing in terms of swinging the outcome of a particular event. Why? Well, you might think it’s so that no delicate little special flower players get their feelings hurt because some PCs are better at things than others. But it isn’t really that. I mean, sure, that’s kind of important if you don’t blow it out of proportion (a story for another time), but the real reason is so that ANY party of 6th-level characters with ANY mix of races and classes and abilities works roughly the same. Otherwise, the GM or adventure writer or whatever would have to design encounters based on the exact makeup of the party.
I’m not here to argue how well or poorly D&D or Pathfinder or any other game actually does that. And if you try to do that in my comment section, I will delete the s$&% out of it. Because that s$&% just leads to edition flame wars.
So, let’s say you have a player and she asks about the possibility of stabbing a cyclops in the eye to deal damage and blind it. This actually happened at my game last week. So you say “sure, make an attack roll.” And she succeeds at hitting the creature in its armor class. And now you deal damage to the cyclops AND give it the ‘blinded’ condition until it receives magical healing or regeneration. Blinded, the cyclops now suffers disadvantage on all attacks and all attacks against it gain advantage. Pretty cool, huh?
But now what? Because the next time the party fights anything, they will try to gouge out eyes. Because gouging out eyes is better than just attacking. And they are able to trivialize fights against anything that has between one and five eyes and needs those to see. Suddenly, you’re unable to plan the outcomes of battles effectively and thereby find the sweet spot of “challenging but not frustrating” that represents a good gameplay experience. You’ve also inadvertently made piercing weapons much more valuable than slashing or bludgeoning weapons. And since most ranged attacks are piercing attacks. You’ve also created a situation in which the players will favor shooting the eyes out of most foes before they even get close enough.
All Improv is a Precedent
And that brings us around to this rule: every piece of improvisation, be it improvisation of outcome (mechanical) or improvisation of story elements (narrative) sets a precedent. Once you bring something into the world or establish a rule about how the world works or what’s in the world. Players will always assume things work in a consistent way. And I’ve already explained why things really NEED to work in a consistent way. Remember, randomly changing the rules ruins the very concept of role-playing.
The moment you decide “sure, you can stab something in the eye the same way you make any other attack and you get the blinded condition for free on top of your damage,” you’re making a rule about your game forever and ever and ever. And that means every but of mechanical improv can have far reaching consequences.
Now, that isn’t to say you can never go back on a decision or undo it or change it again. You can. But the more often you do s$&% like that, the more you frustrate your players. Each time you have to make a change to something the players expect, you’re wandering closer to that “act at random because you will never know the rules of the universe” point. And it doesn’t take much to push players into that.
Resolving the Dilemma
So, you have a situation where you don’t know a rule or there isn’t a rule for something or the rule is stupid or you can’t remember the rule. And now you have to decide what happens. It might that a player is attempting an unusual or weird action. It might be that two elements of the game are interacting weirdly. It might be that two elements of the world are interacting weirdly. And you’ve got to decide how to handle it. And you’ve got to decide how to handle it quick. Because it is stopping your game. Right now. Excitement is draining from your game while you sit there and ponder how these stupid rules should work. Your players are getting bored. No. Don’t go read the rulebook. That’s even more boring. Make a f$&%ing decision.
The decision comes down to three questions: “how would you expect it to work if this were a real world with real people” (verisimilitude), “how would you expect it work if this were a movie without seeming out of place” (tone), and “how would this work if this were a game that can be broken if things swing too unfairly one way or the other” (balance).
What makes this difficult is that things can be all over the map. Rarely is anything OBVIOUSLY fine or OBVIOUSLY terrible. You rarely get things that are unrealistic, atonal, and unbalanced. Sometimes, you get lucky and get something that is realistic, tonal, and balanced. But more often than not, you get stuff that is kind of floating somewhere in the middle.
The cyclops eye thing is a perfect example. Realistically, as far as MOST PEOPLE’S understanding of melee combat, the cyclops eye thing is fine. Completely acceptable. Tonally, it fits with most fantasy adventure games. The trouble is, from a balance perspective, it can screw everything up.
And balance is hardest part. It’s pretty easy, once you’re cognizant of them, to assess verisimilitude and tone. They are gut checks. Smell checks. “Does this SEEM unrealistic?” “Does this FEEL out of place?” If you feel weird about either of those questions, you probably have your answer. But balance is way harder. Because it’s a game design thing. And it requires you to foresee a lot of crap. But generally, it comes down to looking out for three big warning signs.
First, imagine what would happen if the same situation arose three more times in the very session you are running. Would you feel good giving the same answer three times? If you aren’t sure, if you feel like that would make you uncomfortable, there’s probably a balance issue.
Second, depending on the ruling, try to imagine how often you – as a player – would take advantage of that ruling. Would you ever have a reason not to? Would you try to use it again and again and again? Alternatively, if you were playing to win against the players, how often would YOU (as the monsters or villains or whatever) take advantage of it?
Third, how often do you think this situation will happen in your game going forward? Is it something that will happen multiple times a session? If so, it’s something you really do need to worry about. If it’s something that will happen rarely, it’s probably not an issue.
So, you work through the three issues: verisimilitude, tone, and balance. And you’re probably going to get a mix of okay and not okay. And it’s all going to be vague and nebulous. Sorry. That’s how it is. So how do you resolve the dilemma. Well, it might help to understand which of the three concerns are the most important.
Prioritizing Verisimilitude, Tone, and Balance
Okay. Let’s get down to it. Which one is the most important: verisimilitude, tone, or balance? Which is going to break your game? The answer is: they are all the most important. And THAT’S the problem. Any one of the three has the potential to wreck the game for you or for some or all of your players. Sucks, doesn’t it.
How you prioritize these things comes down to why you and your players are playing to begin with. If your group is mainly about winning challenges, balance and verisimilitude beat tone every time. If your group is playing as a collaborative storytelling effort, tone and verisimilitude are important. Some groups like to play wacky and loose. Tone can take a flying f&$%. Some groups are just playing an abstract dungeon crawler board game by the book. Tone and balance are important, but they don’t care about the fiction of the world. That’s what the rules are for.
What makes this even harder is that it’s not always a matter of priority. Sometimes, an action or rule will be so weird or wacky that doesn’t matter that tone isn’t a big deal for you. You just know that the action will bend even your light-hearted tone to the breaking point. You might not care much about balance, but you know that handing a single PC godlike power to change the world or a million gold pieces or something will just break the game apart. I have two different groups right now. One of which are experienced gamers who play to win. If I break the balance of the game even a little, they will exploit the hell out of it. My other group is a relatively new group of people who barely know the rules and like to get into character and explore the world and be creative. I can lighten up on the balance in that group and focus on tone. See?
And THAT is why all the rules break down. All the general advice of ‘always say yes’ or ‘say no, but’ or ‘rule of cool’ all assumes the same priorities for the same table every time. They are “one size fits all” rules. And THAT is why I rail against that crap. I’d much rather teach you how to think through the situation and arrive at answer than say something like “do whatever makes the game interesting” or “if it’s cool, it’s good” or “never say no.” And THAT is why I waste thousands of words.
Sorry. If you want a better answer than “think it through and make your best guess,” I can’t help you. But you don’t want my help there. Because THAT is where GMing style lives. That is what makes one GM different from another. And that is why different tables do things differently and feel differently.
Compromise: When Everyone Leaves Unhappy (Not Really)
So, you have to make a call. The rules failed and you have to fix it. And you’re in a weird pickle because the call you want to make doesn’t quite pass the three criteria. But its close. What can you do? Well, you can compromise. You can figure out what’s not sitting right and tweak it. And there’s lots of ways to handle that.
For example, if the rule involves a player taking a weird action and something about it concerns you, modify things. Generally speaking, you can offer the player an alternative. All you have to do is drill down to the outcome the player is looking for and find a way to give them that without the offending bit. The barbarian wants to run up the back of the halfling, leap off, and stab the giant flying bat in midair? Well, that’s silly. The halfling will just get trampled. But the street is crowded with carts and stuff. All the barbarian has to do is make a running leap off a cart to an awning and then to the giant bat. Done and done.
If there’s a balance issue, the best way to handle any balance issue to add a cost, consequence, or penalty. Sure, the player can make an amazing flying tackle and pull the giant bat out of midair, but both the PC and the bat will take falling damage. In this case, it often helps to try to find similar rules and see how they handle things.
In the end, the key is to make a call you’ll feel good about being a permanent part of your game. You want to allow freedom, maintain tone, maintain verisimilitude, and not break the game. If you can find a way to compromise to bring those things mostly in line, it’s probably fine.
Okay, we’re going to end with three examples. These are things that have actually come up online or in my home game or whatever. We’re going to assume the rules have failed. Either we don’t know the rule or there is no rule or the rule is stupid. It doesn’t matter why the rules failed or if there is a correct answer in the rules. I’m going to present the situation and then I’m going to talk through MY answer based on what I think is important. And while I am always right and therefore my answer will be correct, don’t be dismayed if our answers don’t match. Because, remember, you don’t need my permission to run your game any wrong way you.
Situation one is based on an actual series of exchanges between fans and Jeremy Crawford and Mike Mearls which ultimately resulted in a mess of errata that further complicated the issue. But the actual inciting question was pretty simple. You’ve got a monk who multiclasses into rogue. Monks can make unarmed attacks using their Dexterity instead of their Strength and the damage is 1d4 at low levels and very gradually scales up to 1d10 by the very high levels. Rogues can sneak attack, once per turn, if they use a finesse weapon (one that uses Dexterity instead of Strength) assuming they have advantage against an enemy or an ally engaged with the enemy. The question was: can a monk/rogue make a sneak attack using an unarmed strike.
Got an answer? Good. Here’s mine. Yes, the monk/rogue can sneak attack with her unarmed attack. Tonally, there is nothing wrong with this idea in standard D&D. We’re already allowing the monk to punch through armor as effectively as a dagger or mace. And there is a magic involved in the form of spiritual ki energy. If we’re allowing monks at all, this doesn’t seem weird. Realistically, it’s a little odd, but, again, monks are odd to begin with. So, it fits the level of verisimilitude in the game as much as anything else does. Balance wise, most rogue weapons and most finesse weapons do as much or more damage than the martial arts strike to begin with. And while you can argue that the monk gets more attacks than the rogue to balance out the rogue’s extra damage from sneak attack, the limit of one sneak attack per turn doesn’t seem crazy. Especially when you consider the monk could use any other monk weapon and do the same or more damage than an unarmed strike and unambiguously qualify for the sneak attack. So, I’m fine with this one.
Situation two is an older one. Three giant octopi are lurking in a pool. The wizard decides to launch a fireball into the pool to blast the three octopi. Does the fireball spell work underwater? Would it matter if the wizard were underwater when casting it?
Think about it. Done. Okay, here’s my feeling. From a balance issue, there is no problem. A fireball is a fireball. It’s an area effect spell meant to blast enemies. Doesn’t matter if its underwater or in space or wherever. It’s meant to be used. But where I have a big problem is in verisimilitude. The idea of a fireball penetrating water and exploding bugs me. It wrecks my suspension of disbelief. As much as I know its magic and there’s no reason for it not to work, it still rubs me wrong. I don’t like it. And I could ignore that, but tonally, when you get into a fantasy world where physics is powered by things like the interaction of four elements and you have creatures that are vulnerable to specific elements based on their affinities with other elements, it’s also bizarre. After all, water beats fire. We know that. And tonally, my games run more to the direction of “real world by with magic” than “magical reality,” so I expect fire and water to still work like fire and water. So, no. No underwater fireballs. It’s weird.
Situation three. Can you run up and stab a cyclops in the eye to blind it?
Okay. Got this? Here’s mine. I already explained that I feel that in terms of tone, it fits just fine. In terms of verisimilitude, I’d forgive anyone who thinks it’s okay, but it’s a border case with me. Video games and RPGs tend to overestimate the ease of “called shots” in active combat situations and D&D makes combat appear much more static and turn-based than it is. A cyclops has a big eye, yes, but it is not THAT big compared to its body size. It is not located in the center of the mass. And it is pretty high off the ground compared to the average human. Combine that with the fact that both combatants are actively fighting one another, deflecting, defending, jockeying for position, and therefore most hits that land are matters of opportunity, which is why most blows are aimed for the center of mass. The thing is, though, that’s a personal hangup. D&D combat is closer to mythology than real warfare. So, while it strains credulity, it doesn’t break it. And it fits tonally. But then we get to the balance issue. And the balance issue is where we have a problem. For all of the reasons I outlined above. And part of me wants to disallow the action. The problem is, tonally, it’s cool. I like the idea. And I like the idea of creative tactics. I want to allow this sort of s$&%. I just can’t make it easy. So, I decide on a compromise. First of all, the player needs to tell me how they are getting high enough to hit the eye (if they don’t have a reach weapon) and that will probably involve a skill check. Second of all, the attack will have Disadvantage (which is effectively a -4 in D&D). But, if it hits, it will blind the creature. Third of all, I make it a point that this is only a possibility because the eye of a cyclops is a large target. Against an orc or other creature of similar size, it’s probably beyond the realm of possible.
That’s just how I’d do it.
But I’m right.