Hey again everyone, welcome back. Today I'm going to continue with the topic of dice systems, and having learned my lesson last time, I'm going to jump right in to the first system of the day, Pathfinder 2e.
I figured that it would be good to cover pf2e right after dnd5e given their similarities. Both systems use the same core resolution system, so looking at them back-to-back should make it easier to see the minor differences between them. While Pf2e's dice resolution is much the same as 5e's, 1d20 + modifier, there are a couple of key differences that go a long way to differentiating the systems and making pathfinder feel unique.
A small change that has a big impact on the game flow is the way that pf2e implements critical successes and failures. In 5e, a natural 20 only has a special effect on an attack roll, in which case the attack becomes a critical hit, and death saving throws, where a character stands back up with 1hp. A natural 1, similarly, will only cause an attack to automatically miss, or a death save to count as two failures instead of one. In pf2e, all checks can have 4 different results: a critical success, a success, a failure, or a critical failure. Success and failure work as in d&d, but a critical success is achieved if a check's DC is exceeded by at least 10, and a critical failure happens when a roll falls short of the DC by at least 10.
Natural 20s and 1s have their own special effects that work with this success scale. A natural 20 on the die shifts a result “up” one category (from critical fail to fail, fail to success, or success to critical success), and a natural 1 shifts a roll “down” a category.
Most of the time this ends up working out the same as it would in 5e. A natural 20 on an attack roll is going to be a critical hit, and a 1 is going to be a miss. The place where this change really shines is when combined with the huge flat modifiers that pf2e characters can get to their rolls. A level 10 fighter can easily have a +20 to their attack roll, which means that against a low-level target that might have an AC of 20, that fighter is going to score a critical hit on over half of their attack rolls.
That's a pretty extreme example, but it demonstrates the overall point; a natural 20 isn't required for a critical success depending on the skill level of the character. This method allows character's that are particularly good in one area to critically succeed way more frequently. The critical success-critical failure scale also applies to almost everything, allowing critical successes on checks to craft something, improve an NPC's disposition towards a character, or any range of other things. I mentioned in my example the level 10 fighter having a +20 to their attack roll, which leads me into the other thing I wanted to talk about regarding pf2e, the proficiency bonuses.
Likely the biggest difference between pf2e and d&d 5e is how proficiency bonuses work. In 5e, proficiency is (mostly) binary, either a character has proficiency (or expertise) with a skill, weapon, armor, saving throw, etc. or they don't. The proficiency bonus of a level 1 character starts at a +2 and scales up to +6 by level 17. Pathfinder adds a bit of granularity, as well as potency, to proficiency. In pathfinder 2e, there are four levels of proficiency; trained, expert, master, and legendary. These levels of proficiency give a +2/4/6/8 respectively. Additionally, for anything a character has proficiency in, they get to add their character's level to their proficiency bonus for that trait. For example, a level 5 character with expert proficiency would get a +4-proficiency bonus from being an expert, and an additional +5 from being level 5, for a total proficiency bonus of +9.
Obviously, these numbers can start to get really big really fast. A level 10 character will have at least a +12 prof. bonus in anything they're at least trained in, and a level 20 character with legendary proficiency will be looking at a +28. None of this even takes into account ability score modifiers and magic items that can boost those numbers up even higher.
So, what's good about this difference, and what's bad? Well, a big upside to this method of proficiency is that a high-level character “feels” high level in their areas of expertise. A level 5 character with only the trained level of proficiency can have a total bonus of +11 to a skill, or +13 if they increase their proficiency to expert. This is a huge bonus compared to 5e, where expertise characters would only just be managing to keep up with the trained character, having a +11 at this level.
Now, I'm not trying to say that bigger numbers = better design. But, if we take the same example as last time, comparing a level 5 character to the baseline level 1 beginner in an area, the difference is pretty big. The 50% success rate DC for our level 5 character is 24, assuming they used a skill increase to bump up their proficiency to expert. This means our baseline +0 character can only even succeed on a natural 20, a 5% chance, by merit of the 20 bumping up their failure to a success. The chance of the level 1 character outperforming the level 5 in terms of roll total is only 7%.
I think this way of handling proficiency is a lot more satisfying in terms of progression and having characters “feel” competent in their areas of expertise, but of course it has its own drawbacks. Probably the one consequence of this method that I'm not a huge fan of is the way that it makes certain skills almost required to be at least trained in at higher levels. Without at least the trained level of proficiency, a character entering a high-level scenario has essentially no chance of succeeding. An easy example is stealth. Suppose a level 10 party has a single member with no training in stealth. In that situation, any group stealth (outside of exploration activities) essentially must exclude that character, as their bonus to stealth will be so low that they'll likely only succeed on a natural 20. This isn't necessarily a huge issue given that characters gain at least one skill increase every 2 levels, so it isn't difficult to become trained in those skills where every member needs to make their own check, but that ends up feeling like a tax more than a voluntary investment.
The way that pf2e handles bonuses to rolls is also an interesting point of differentiation to talk about. D&D 5e does allow for bonuses to rolls outside of advantage/disadvantage, but those effects lack a general ruleset to manage them. This is what allows specific rolls in 5e to have some pretty insane bonuses in a system with typically low static modifiers.
Pathfinder 2e fully supports numerical modifiers to rolls, but puts limits on how many and which bonuses can apply to a rolls. Modifiers to rolls in pf2e are for the most part categorized as circumstance, item, and status. The big limiter on modifiers is that a character can only benefit from a single bonus and a single penalty from each category. This means that if a character has 10 circumstance bonuses and 10 circumstance penalties to a roll, they only apply the single largest bonus and penalty to the roll.
This does a lot to cut down on bonus stacking and tracking, but there can be situations where certain lower value bonuses or penalties will persist longer than higher value ones, meaning they still need to be tracked, just not necessarily applied to a roll.
Overall, I'm personally much more a fan of the way pf2e's dice system is implemented, but it does definitely add a bit of complexity compared to 5e's, which is generally reflected in the other parts of the systems as well.
So far, the dice systems that we've explored have been pretty simple on the whole. Pathfinder has a lot of good content and options that make the system as a whole a bit more complex, but the dice system itself is kept on the simplistic side of things. So, to give an example of a dice system of the more complex side, I'm going to talk about Shadowrun 5e; a system that's a bit notorious for being crunchy.
Having played quite a bit of Shadowrun 5th edition, I can say that its reputation for being complicated is entirely deserved. Shadowrun one of the most complex rulesets of the RPGs I've played, while also unfortunately being plagued by a lack of proper editing and organization in the sourcebooks. Despite that, Shadowrun is one of my favorite systems, and a big contributor to that is the core dice system.
In Shadowrun 5e, dice rolls are comprised of a pool of d6s, the number of which is based on the character's abilities and augmentations. The typical formula for determining a character's die pool for a roll is to add the relevant attribute and skill rating for the roll, then apply bonuses and penalties depending on gear, magic, or the situation. Each d6 that lands on a 5 or a 6 is considered a “success”, and the outcome of the test is based on the total number of successes rolled. Standard tests have a “threshold”, which is the number of successes required to pass the test. In other cases, such as when attacking another character, the test is “opposed”, meaning that both characters involved make their relevant roll, and the character with more successes achieves their goal.
I want to try something a bit different for the start of this analysis, considering this is the first time we have two different base resolution systems to compare. So, let's compare Shadowrun's method to the d20 + modifier method in a vacuum. With d20 + modifier, the distribution always retains the same shape, regardless of the value of the modifier, the only things that change are the minimum and maximum values. The success based d6 die pool, on the other hand, quickly turns from a flat distribution at 1 die towards a right-skewed distribution as more d6s are added to the pool. The distribution also becomes increasingly thin-tailed as the pool of dice gets larger, meaning that a higher percentage of rolls will fall closer to the mean number of successes, which is 1/3 the dice pool. You can also see that while the maximum number of successes changes with every die added, the minimum always remains at 0 successes. The probability of not getting any successes on a roll does get progressively smaller as the number of dice increases, but the chance is always greater than zero.
Now that we know what the resulting distributions look like for the two methods, as well as how they change, we can try to translate those things into what they mean for characters using those systems. To put it in a way that's a bit easier to understand intuitively, let's say that we're measuring how far a character can jump. For a character in a system using 1d20 + modifier, they will always have a set range in how good their best result is in comparison with their worst result. Let's assume each “point” on the result equates to 1 foot of jump distance, so the difference between a character's minimum and maximum distance on any given jump is always going to be 19 feet. When a character gets better at something, like jumping, their best result and worst result both become better, but the range of results remain the same, and it's always entirely random which result they get within the range of their worst to their best. With our analogy, a character will be able to increase the minimum and maximum distance they can jump, but their best and worst jump distance will always be 19 feet apart, making their possible jump distance 1-20 feet, 11-30 feet, 41-60 feet, and so on.
For the d6 dice pool system, let's use Shadowrun's method and say each success on the roll equals 3 feet of jump distance, but a character can't jump farther than 4.5x their agility attribute in feet. With this system, the range of top end results changes based on how “good” a character is at something. A pretty average person (3agi, 3athletics) is going to get around 6 feet of distance on average, and up to 13.5 feet as their maximum. The better they are, the higher their best result becomes, but their worst result remains just as bad no matter how good they become at something. Which is to say, a character can increase their maximum jump distance, but their minimum distance will always be 0 feet. What does change in regard to their minimum jump distance is the chance of getting that result. A lower-ability character (6 dice total) has an 8.8% chance to completely fail their jump, a decently skilled character (9 dice) has a 2.6% chance, and a “professional” character (12 dice) has a .8% chance. As their dice pool increases, our jumper starts to get more consistent at getting acceptable jump distances, while also increasing how far they are jumping on average. They will still have a pretty abysmal jump every once in a while, but those become rarer the better the character gets at jumping.
When we look at the methods like this, I really favor the way the dice pool system represents things, as I think it does a lot better job giving a “human” representation of things. After all, when we look at our analogy, no matter how good of a jumper you are, there's always a chance you lose your footing right before you jump and fall on your face before the jump even starts. However, a better jumper will be able to much more consistently make jumps that are right around their mean jump distance. I think the one flaw I really notice with the dice pool system in this regard is the top end. Even though the very highest results rapidly become incredibly unlikely, there's still a possibility of getting a result of three times the average, which doesn't conform to the way that people usually perform at a task.
We should also look at things from a smoothness of experience perspective. Any time a variable number of dice is involved in a core resolution system it's going to slow things down, especially for people not familiar with it. Whereas the 1d20 system allows you to grab a die then figure out what your bonuses are, with a dice pool you need to figure out your dice pool first, then start grabbing dice until you have enough. Having a success-based system does help with some of the time cost involved with a dice pool system, as once the dice are rolled it's pretty easy to count up the number of dice that are successes, but it's still going to be slower than a single die method.
I do think that this slowdown generally gets to be much less of an issue when players get used to the system, especially during longer campaigns when players end up having a good idea of what their dice pools are, or at least know where to look on their character sheet. It's also possible to remove the slowdown almost entirely by making use of newer tools like virtual tabletops. Roll20 is the most well known TTRPG virtual tabletop and has support for a ton of systems, but there are other great options like FoundryVTT. These can be used to play games entirely online, or as an aid for in-person play.
I think the analysis we've done ended up covering most of the things I wanted to touch on for the strengths and weaknesses of Shadowrun 5e's dice system. To finish up, I want to mention something that isn't part of the dice system itself, but I think contributes to it in a pretty significant way, edge. Edge is most like inspiration in d&d or hero points in pf2e, but much stronger. Unlike inspiration and hero points, edge is its own attribute that can be raised during character creation or with experience points. Characters start with a number of edge “points” equal to their rating in the attribute, can spend them at will, and regain a point of edge after a good night's rest. There are multiple ways to use edge, with the most common being to “Push the Limit”. When edge is spent this way, the character gets to add a number of dice to their roll equal to their edge attribute rating. They also get the benefit of “exploding” 6s on their roll, meaning that each die that lands on a 6 is counted as a success, and then rerolled, counting any successes on the rerolls as additional successes on the roll.
I'm personally a big fan of edge, as it is both really fun to actually use, and it gives characters a sort of ace in the hole, something that they can save for a bad situation and spend to have a big influence on the way a situation plays out.
Anyways, that's all I had for this time. I think the more system-agnostic comparison of d20 and dice pool d6 in this post was a bit more of what I had in mind originally for this series, so for the next post I'm going to focus on the roll&keep system from 7th Sea and Legend of the 5 Rings.
Until next time.
-JGDev