This is not the post that I expected to make this week. I simply ran out of time and could not finish either the article I had intended to post this week [about time travel] or the one for next week [the long-awaited followup to last year's Pillars of Architecture article], in time. Instead, I have had to reach into my list of possible future blog topics to extract one for which I had already done most of the hard work. Normal service will be resumed as soon as possible!
Most roleplaying games rely on the concept of rolling one or more dice to generate a random result. Die rolls imply extreme results; on a d20, those are ’1′ and ’20′, but every roll has (by definition) a minimum and a maximum. The question under consideration is whether or not those rolls should have an interpretation beyond the literal one of the result, ie a ’1′ being 1 removed from a ’2′ and a ’19′ being 1 removed from a ’20′.
Let’s start by looking at some of the ways that these extremes might be interpreted in a game context.
This is the most obvious interpretation of them all – it says there are no such things as critical hits or misses. At the GMs discretion, he might allow a minor enhancement of effect or severity of failure on a maximum/minimum result, but a result of three on 3d6 has, officially, no significance beyond being one worse than a result of four.
We use this approach in the The Adventurer’s Club Campaign, where the characters may be exceptional but they are still, essentially, human, for all the reasons listed in the ‘No’ arguement below.
Success/Failure that would otherwise not be permitted
This is the interpretation built into the standard d20 system. By stating that a “20 always succeeds” and a “1 always fails”, the system ensures that eventually characters can succeed at anything, no matter how ridiculously impossible it seems, and will eventually fail at any task (no matter how simple it appears to be) – provided that repeated rolls are either permitted or manditory.
A first-level dweeb will eventually hit the Invulnerably Exquisite Muckamuck, Most Skilled Swordsman In All The Universe – and the Muckamuck will eventually fail to snot that first-level dweeb between the eyes.
Of course, these possibilities do not rule out the likelyhood of consequences in the interrim – the Muchamuck will react to the dweeb’s feeble attempts (by striking back, or by rolling on the floor, laughing, or whatever else seems appropriate), and the dweeb is unlikely to survive long enough for his lucky number to come up.
In theory, the GM can apply these facts to his own advantage by employing mass attacks, knowing that about 5% of them will succeed, no matter how lop-sided a one-on-one contest would be (I say “in theory” because I remember attacking two PCs with 630 giant phasing fire ants, each needing an 18 or better to hit, but with the PCs unlikely to survive more than two or three such hits; rolling all 630 attacks; and missing with each and every one. The PCs were able to use area-effect spells to take the ants out before they got another shot at the brass ring. Nor is this the most extreme ‘improbable roll set’ that I have witnessed – I’ll tell the story of two Ians in Seventh Sea some other time).
This system has a couple of drawbacks. It provides a disincentive to improve chances of success beyond a certain margin, because you can never get better than a 95% chance of success (using a d20). It marginalises the effects of setting high difficulty numbers – because there is never less than a 5% chance of failure. It should, in theory, have an impact on the psychology of combatants.
Levels of Success/Failure beyond the normal
This is the interpretation that features in my superhero campaign, Zenith-3, and the spin-off Warcry campaign, and which used to feature in my AD&D campaigns. It is also the system utilised in Rolemaster. Instead of saying that there is always a chance of success or failure, this approach states that the results will be more effective IF the action succeeds.
Like the previous system, this has a couple of drawbacks. It demands that the GM be a little less casual about setting difficulty levels (or difficulty modifiers, or whatever the mechanism is), because they can have genuine effects on the outcome of the game. It encourages players to become rules lawyers, seeking to obtain every last bonus that they can wring from the system – which in turn mandates that the GM be equally pedantic. This, in turn, can slow the resolution of game actions down, especially in critical situations – exactly when the GM wants to keep the tension level high. And it can encourage players to to obsessively persue improvements in skills that they deem as system-critical rather than developing more rounded, broadly-capable characters.
Standard d20 also uses this system, by means of the critical hit multiplier, making some extreme results even more valuable – but there is no equivalent ‘critical miss’ threshold. The system is only fair because the opposition gets the same capability.
Once again, I can draw apon one of my existing campaigns as illustration – this is the interpretation that is utilised with non-combat rolls in my Fumanor campaigns (for reasons I can no longer recall, I was persuaded to keep the standard system for combat). If you roll a 20, you roll again and add twenty to the result – and keep going for as long as you keep rolling 20s. Equally, if you roll a 1, you subtract 20 from the result and (again) keep going for as long as you roll ’1′s.
eg1: roll 20. roll 20. roll 14. 20+20+14 = a result of 54.
eg2: roll 1. roll 1. roll 14. 14-20-20 = -26.
This system has one big advantage over the standard system: it means that even if a dweeb attempting an almost-impossible task rolls a 20, he still has to get phenomenally lucky to succeed – it preserves the relative difficulty, while still giving a chance of achieving the impossible. Similarly, at the low end, the extremely skilled character has to roll phenomenally poorly to fail at a simple task.
It’s not perfect – in fact, it effectively replaces the drawbacks of the standard system with the drawbacks of the ‘success/failure beyond the normal’ alternative.
Cascading/Imploding Rolls Version 2
A variation on the concept is to add or subtract the next smaller die size on a critical result.
For example, on a ‘critical success’ skill check made on a d20, you would roll a d12 and add the result to the critical result; on a combat check where the effects are d6′s of damage, you would add 1d4 extra damage; if the effects are a d8, you would add a d6; and so on. On a ‘critical failure’, you would roll 1d12 and subtract it from the result, and if the character’s skill was still high enough to succeed, if the effect dice were d6s then you would subtract 1d4 from the result, and so on.
Cascading/Imploding Rolls Version 3
A further variation on the effects side would be to roll a number of extra dice (1 size smaller) equal to the number of maximum/minimum results on the effect rolls – so, for a critical hit, rolling 4d6 damage and getting two sixes amongst the results would add 2d4 to the damage total. On a critical failure that succeeds despite the low result, if the effect was rolled on 4d8, and the result roll included two ones, the total effect would be reduced by 2d6.
That can be quite significant in terms of impact: 4d8s, two of them coming up ones, give a range of results of 6-18 [average 11], while 2d6 have a result range of 2-12 [average 7] – so the total after subtracting one from the other ranges from is -6 to 16, with an average of 4. Compare that with the results of a standard 4d8 roll: 4-32, average 18. The normal minimum result is the typical result of the modified roll in the example.
This raises the question of how effect results of less than 0 are handled, which offers two further sub-variations for consideration.
This quite straightforward approach simply sets a minimum effect level of 0. If the effect total is less than zero, it is treated as having no effect.
But, given that we’re talking about combat effects, why can’t the attacker injure themselves instead of the target on a negative? Surely it’s possible to strike someone awkwardly and sprain a wrist, or stumble and twist an ankle?
And here’s yet another variation to consider: basing the number of additional dice on a critical success on the number of minimum results on individual dice, and vice-versa. That means that a critical failure tends to transform good effect results into mediocre-to-poor ones, and vice-versa. This waters down the extremes of impact on the effect totals by the system, and eliminates the risk of a ‘less than zero’ total.
Overall analysis of Variations 2, 3a, 3b, and 4
The variations offer a number of beneficial effects over the base version. The severity of impact that comes with a critical is moderated, because there is only a single ‘cascade level’ on success/failure rolls.
Both forms of Variation 3 and Variation 4 permit success or failure results to have an impact on effect results, potentially significantly, but to varying degrees. Because they apply a modifier to effect results based on the success/failure result, levels of effect are at least partly commensurate with that success/failure result instead of being completely independant. A great hit will have more effect than the average dictates, and a poor one will make a blow less effective than the average. These adjustments can be minor or can be extreme, depending on the variation chosen.
I have never seen these ideas used in a game system – but I like them so much that I am going to ask the players in my Fumanor campaign to try them out!
This is an option that I’ve never seen utilised, possibly because it can lead to one character getting a disproportionate share of the ‘air time’ in a game, but in theory it should work. Most game systems have some sort of limitation placed on what the character can do in a given slice of game time, especially in combat – D&D uses the “free” and “move” action types for the purposes. What this option suggests is that on a ‘critical success’, that action is downgraded in time requirement 1 step, while on a critical failure, it is upgraded one step.
In other words, in D&D: on a ’20′, a ‘move’ action would be considered a ‘free’ action, permitting the character to make a second move action in the combat round, or a ‘free’ action would be considered a trivial action.
On a ’1′, a ‘move’ action would be considered a ‘whole turn’ action, a ‘free’ action would be considered a ‘move’ action, and so on.
The biggest downfall of this system that I can see – not having actually tried it – is that it would encourage the DM to make players roll for everything.
Reroll minimum/maximum results
Another option that I haven’t actually seen employed, this will only work for combat and other situations where there is a seperate die roll for ‘level of effect’ to the roll for success or failure. On a critical success, you would get to reroll (once) any effect result die that came up the minimum, making it more likely that your success would actually have an impact on the outcome; on a critical failure, you would be forced to reroll (once) any effect dice showing a maximum result, making it less likely that your roll would achieve anything worthwhile.
I don’t think this option is one that will ever be utilised in a real game, simply because a ‘critical failure’ is still a failure, and therefore has no effect dice attached to the result that has to be rolled. But, in theory, it would still work.
The downside is that it doesn’t make a huge degree of difference, and therefore reduces the impact of critical successes and failures – so much so that they simply become “better successer” and “worse failures”. They aren’t criticals any more. And I’m not entirely convinced, given that fact, that the results would justify the extra time consumed by the system.
Skill Improvement On A Critical
This was the approach taken by a friend of mine in his Traveller campaign back in the 80s and 90s (was it really that long ago?). The basic idea is that on a critical success, you are presumed to have learned something, and can increase the relevant skill; on a critical failure, you are also presumed to have learned something, and so can increase the relevant skill. Over time, you will get better at the skills that you actually use, and remain static in those you don’t.
A variation requires the character to make a second (successful) roll against the same skill after a critical success in order to gain the improvement, while it remains automatic on a critical failure. This weights the benefits according to the premise that you learn more from a failure than you do from a success – sometimes you succeed because you ‘get lucky’ and learn nothing because it was accidental.
A further variation requires the character to make a second (successful) roll against the same skill after a critical failure in order to gain a second improvement.
Limits can be placed on the number of skills that the character can upgrade in a given game session if they still seem to progress too fast.
The same GM also takes the attitude that a critical success or failure always produces especially spectacular consequences, without necessarily impacting the level of effect that results from the game mechanics.
Taking Longer To Succeed
I came up with a new take on the notion of critical failures for my Shards Of Divinity campaign which also replaces the ‘take 10′ and ‘take 20′ system in the standard d&d 3.5 rules. The assumption is that with many tasks, the character can and will keep trying until they either succeed (and know they have succeeded) or give up. Since eventual success is therefore inevitable in a situation in which a 20 always succeeds – it’s merely a question of how many rolls it takes – it seemed to be more efficient to combine all these rolls into a single one, and to use the result to determine how much more time than normal the task takes.
Once the system was in place, further reflection showed that even with skill checks that are normally ‘success or failure’, ‘all or nothing’, most rolls could be interpreted using this paradigm. Searching an area for traps or secret doors? Normally a ‘you find it or you don’t’ roll, there’s no reason not to say ‘it’s been 1/2/3/whatever minutes and you havn’t found anything yet’. Certainty and confidence are suddenly replaced with uncertainty and doubt. Even when something has been found, the characters can often be left unsure that this was all there was to find! Instead of racing through challenges, the players feel under far more pressure, and are generally a touch more cautious.
Further, by not announcing the DC for success, the players can make openly all the rolls that the GM would normally make in secret, so they feel even more in command of their characters. And the GM can inspire paranoia simply by asking for a check on a whim.
It works for Spot, too. Eventually, a character will succeed in spotting whatever is in front of their nose – but will it be in time? It only requires a single spot roll, at whatever DC the GM considers appropriate, and a determination of how long it takes the characters to become aware of the potential danger or object of interest. All the GM then has to do is keep track of what the characters do until that time is reached for someone. IF the item to be spotted is still there, AND if the PCs havn’t been attacked by it, the first to succeed will then spot the painfully obvious.
Here’s the table that I use:
+1 = +10 sec, +2=+30 sec, +3=+1 min, +4=+5 mins, +5=+10 mins, +6=+15 mins, +7=+20 mins, +8=+30 mins, +9=+45 mins, +10=+1 hour, +11=+1½ hrs, +12=+2 hrs, +13=+4 hrs, +14=+5 hrs, +15=+8 hrs, +16=+12 hrs, +17=+24 hrs, +18=+1 week, +19=+2 weeks, +20=+1 month
the “+1″ and so on is the difference needed to turn a failure into a success. For each failure above +20, double the number of months.
The table assumes a 1-round action – if the GM thinks the basic task will take longer, simply multiply the extra time accordingly.
Arguements for the ‘Yes’ case:
Okay, so there are a whole gamut of alternatives, and the fact that I use so many of them in my campaigns suggests that I’m predisposed towards the ‘yes’ case. So why?
One of the big reasons to say yes: spectacular achievements give a campaign a more epic, sweeping, and dramatic flavour. It’s going too far to suggest that this is the only difference between high and low fantasy, but criticals of some sort are an essential componant of the first. The four-colour antics of superheros practically demand a criticals system (which only makes it more surprising that the official Hero system doesn’t have one). Catastrophic failures and Cataclysmic successes come with the territory, so why shouldn’t they be inherantly built into the game rules?
Nothing is impossible, no-one is invulnerable
And yet, at the same time, criticals can be viewed as “the great equalizer”. This too, is a staple ingredient of epic campaigns and stories, whether it’s Smaug being layed low by a single well-placed arrow or Spider-man taking down Juggernaut after a truly epic battle in Amazing Spider-man Issue #230.
Conveys a sense of the fantastic being possible
While it’s possible to add colour through narrative alone, there is a deeper effect when the potential for achieving the fantastic is built right into the game rules, a sense of the fantastic being the reality of the characters. This can sometimes be hard for GMs to wrap their heads around, because it means that every plot, and every encounter, carries an inherant risk that it will escape the laboritory and run amuck.
Because the potential for extraordinary results is built into the rules, there is an added weight of verisimilute that is almost impossible to attach to the fantastic any other way – at least without sucking all the life and vibrancy from the scene with tiresome justifications! A critical hits system tells the players, and hence their characters, “the incredible is reality – deal with it.”
Permits a more relaxed attitude
Because there is always a chance, referees have to stress a lot less about whether this DC is too high, or that task is too difficult, or that AC is too extreme, and so on. The same thing is also true of the players – they also relax a bit, and take greater risks. Sometimes that can get out of hand, leading to players taking foolhardy risks – which is generally a signal to amp up the opposition a little, or spend more time preparing them and their tactics.
Arguements for the ‘No’ case:
Equally, there are some compelling reasons to consider the ‘no’ case. And, in many cases, these are the flip sides of the ‘yes’ arguements already presented.
Perhaps the most obvious justification for not having a system for criticals is the added realism that results. Adding colour and making the fantastic inherantly more plausible is all well and good, but you should never lose sight of the fact that all games are an imperfect suimulation, already replete with compromises for the sake of playability. The grittier you want the campaign to feel, the more plausible you want it to be, the more you have to control the element of the fantastic; added realism is a clear advantage that can be achieved by junking the notion of a criticals subsystem.
Results are always merited
Then, too, if there are no wild-card successes built into the system, it forces characters to earn every success. If that means taking the extra time to stack conditions in your favour, or to evade the most significant negatives to your chances, that’s what it means. Results are always merited according to the expertise of the character making the attempt if there are no criticals, and that has knock-on effects that emphasise the tactical aspects of the game, which in turn can further add to the realism of the campaign.
Forces GMs to keep tighter control of challenge difficulties
It’s easy for GMs to get lazy, adopting a ‘near enough is good enough’ attitude, when the potential for criticals can always get them out of trouble. When there are no escape clauses, the GM has to look at the actual chances of success and failure far more critically. “+1 to the difficulty” can become significantly important, making it impossible for character A to hit target B. The downside of increased realism is that the GM has to work that much harder at getting the challenge levels exactly right. That also spins off into a need for greater control of rewards and treasure, as “+1″ becomes more significant on that side of the ledger as well.
Better for maintaining tension
Greater uncertainty means greater tension. The players will never be sure that they aren’t in over their heads, that the enemy they are about to confront doesn’t have something up his sleeve. This can make players a little overcautious, especially when it comes to making plans for their characters, but if not taken to extremes, this provides yet another layer of interaction between the characters. Start by insisting that all such planning be done in character!
It’s clear that in this debate, there is no clear winner! Both arguemnts present advantages and potential drawbacks in equal measure. Overall, both options have more advantages than liabilities, which is a good thing; it means that there is ample scope for benefits to compensate for quirks and penalties that are associated with the different systems of interpretation for criticals.
If no general conclusion can be reached, then the next obvious thing is to consider the question seperately for each campaign. What is the tone of the campaign to be? Where is it lacking, and what could use shoring up – the fantasty or the realism? The gung-ho or the cautious planning? And further, with their varying subtexts and implications, if you do choose to have extreme rolls make a difference, which is the best system for this particular campaign?
My personal preferances vary amongst my different campaigns. The choices have been made to compliment the unique flavour that I wanted to endow to each campaign, exactly as I reccomended in the five-part first Lesson From The West Wing. There is no One Right Answer, and no One Wrong Answer. Each choice has consequences, and those can add to, or detract from, the flavour of the campaign you happen to be running.
So instead of blindly following the rules in this respect, think carefully of which approach most enhances your campaign, and don’t be afraid to mix-and-match as necessary. One system for skill checks and another for attack rolls? Fine. One system for critical hits in combat and a different one for critical misses? No problem. A different system in the Palace Of Dreams to the one in use everywhere else in the world? Why not?
Have you given any thought as to the critical success/failure choice you use, its suitability to and impact on your campaign, its pro’s and con’s? Do you know of any other interpretations that I’ve missed? Then I’d like to hear from you! The broader the palette of choices, the greater the chance we all have of finding the perfect fit!