The Archive contains all manor of writings on harm. Countless civilizations from unnumbered worlds have decried evil, and sworn to do not harm But what is harm, you wonder? Fortunately, as with all questions, the Archive contains the answer.
Some fights are not worth fighting. Life has its frustrations and inconveniences. Most of the time, we endure this and move on. Minor nuisance are not worth wasting time on.
Then there are the other issues. The fights you cannot walk away from. The hills you will die on. Those times where you know you are right, and cannot back down, because it matters. Today is one of those days. It is time to talk about contested rolls.
Contested Rolls: That Which Should Not Be
Normally in D&D 5e, action resolution is achieved by rolling 1d20, adding a modifier, and comparing it to a target number, often called difficulty class or DC. The modifier is a combination of an ability modifier, a proficiency bonus, and some magic items. We use the same mechanic to make attacks, saves, and ability checks.
Mechanically, every time we make such a check there are 20 possible outcomes, some of which are successes and some of which are failures. If you have a +4 bonus, and the DC is 10, the possible results are all values from 5 (1+4) to 24 (20+4). Five values (5,6,7,8,9) are under 10, resulting in failures. The other 15 are successes.
As an aside, we could come up with a similar result by subtracting the modifier from the DC: 10-4 is 6, which is the number we need to roll to succeed. This was the basis of AD&D’s THAC0. Rather than using modifiers, there was a table showing the dice roll necessary to hit a creature with an AC of 0 (To Hit AC0) at different levels. In this edition, AC was descending rather than ascending, so an AC of 1 was 1 easier to hit, so you would need to roll your character’s THAC0-1 or higher to hit.
This approach uses a roll against a static target. However, there is another type of check the is used less frequently: the Contested Roll. One place where this comes up is when grappling. The attacker makes a Strength(Athletics) rolls, and the defender makes a Strength(Athletics) or Dexterity(Acrobatics) roll. Whoever rolls higher wins, with draws leaving the status quo unchanged.
The Problem
At a glance, this sounds fine. Both participants are actively involved, so should be in direct competition. However, there are some less obvious problems here.
Psychology
First, there is the potential to undermine good rolls. Rolling a natural 20 is exciting — that is why the game has critical hits! But with a contested roll, a natural 20 might not be enough, as the PC must also have a high enough modifier. More generally, it is not unlikely that they will experience the adrenaline rush of a fairly high roll, only to find out they have lost. The psychology of the dice game and the contested roll are in opposition, which is not the case the rest of the time. It is easier to get players excited about the game if the mechanics that feel like triumphs translate to real victories, so we should not disrupt this without good reason.
Statistics
Second, it does weird things to the odds of success. Normally, the PCs roll against a static DC. In most cases, DCs related to a monster are determined by its ability scores and proficiency — the exception is AC, which uses armour. With opposed rolls, the important value is the difference between the modifiers in use. A contested roll has 400 possible outcomes (20x20 possible roll combinations). If the initiator’s bonus is higher by 10, they will succeed at 345 attempts. If instead we used a passive value to set the DC (so 10+modifier), a bonus that was higher by 10 would guarantee success. At the other extreme, an initiator with a relative bonus of -10 would win 45 of the 400 contested rolls, but none of the static rolls.
This behavior is weird. If the initiator has a high modifier, contested rolls favor the defender. If they have a low modifier, they favor the initiator. This appears to discourage the use of mechanics leading to contested rolls: a PC who should be good at this will want to use the same ability score to make a static check so they are not disadvantaged; a PC who is not good at this will probably have better options that (how many wizards try to grapple trolls?).
And More
It is also sometimes unclear which side should have the advantage of the status quo. In a tie, does the sentry maintain the status quo (has not let anyone sneak in), or the thief (has not been caught)?
Finally, rolling more dice will take longer. This is not a huge issue, but if it is not gaining us anything we should avoid needlessly slowing down the game.
The Solution
The solution I have taken to using is passive values. The Player’s Handbook already suggests using both in certain cases, like hiding, so this is a minor deviation from rules as written. It does not significantly change the probabilities of outcomes — even with extreme relative modifiers, the effect described above is quite small. It is also extremely simple and quick. In these cases, the DC is 10+modifier.
As an extension of this approach, I very rarely use the passive skill values of my player’s PCs. When a contested roll comes up, they always make the roll. This means I need to do less, and they get to do more. Although this does leak some information, I do not mind. In fact, having players know they failed a perception check works well: they either have a moment of dread before being ambushed, or they know the world has more depth than they have explored.
The case could be made for using 8+modifier, as is used to determine saves. This gives whoever is rolling an advantage of +2. Given the infrequent use of shoving and grappling in most games, this bonus may be a good way to encourage it (especially as it may otherwise be a sub-optimal move). However, this is a more complex deviation from the rules, so I prefer the simpler solution.
Complexity
The underlying problem here is complexity. D&D is a very complex game, and that is part of the fun. However, we must resist the urge to make everything more complex for no reason. Contested rolls are not the only example.
Watery Depths
Consider drowning. Many TTRPGs have very confusing rules for drowning. The problem is that drowning can kill you, but does not do it in a way that looks like hitpoint damage. This break with reality upsets people, so we have an array of more complex systems instead.
However, this complexity does not gain anything. The sudden death mechanics of drowning are obscure enough not to scare PCs most of the time, and in the tense moments when they come up no one wants to stop the game to consult the rulebooks (for reference, the rules for drowning were in the third place I looked under Suffocation, not Unusual Environments or Underwater Combat). Doing things this way does not make the game better, but definitely makes it more complex.
Suppose instead we kept it simple. Drowning involves breathing in water. Breathing in poison deals poison damage, so we could do the same. To dive through a flooded corridor might require a Strength(Athletics) check. On a failure, the PC gains the Poisoned condition until their next short rest, and must make a Constitution Saving Throw or take 6d6 Poison damage. This is simple enough that I can do it on the fly. It is problematic enough to be taken seriously as a threat.
Time to Break the Rules (Again)
Complexity is not inherently bad. The issue is that complexity is also not inherently good. Complex things are harder, but not necessarily better. We should only use them when we want life to be hard.
Logistics
In my games, I use a homebrew encumbrance system. Players do not need to track every item that contributes weight, but they do need to track enough gear that most sessions have time dedicated to book-keeping. This is the intention. When encumbrance comes up — normally in town buying new gear — the game slows down as we consult rules. That is fine, as it is not disrupting a tense moment. What we gain is immersion, and an opportunity to make plans, as good preparation is rewarded.
My encumbrance system deviates from D&D core rules in several ways. I use different kit lists; I do not use the normal currency; carrying capacity depends on Strength Scores, not modifiers (and increases on odd levels instead). It breaks my rules about adding complexity, but is allowable because it improves the game.
As Luck Would Have It
Finally, I will concede that contested rolls have at least one legitimate use case. Sometimes, luck is a big part of a contest. In these cases, we may want the players to know that just rolling well might not be enough. Here, contested rolls are a way of raising the stakes. That is why I give such long explanations of things: it is important to know why we do or do not do things certain ways so we understand when it may be beneficial to do the opposite.
A Way Out of the Mess
Complexity is our friend and our enemy. Next time you come across a contested roll, or mechanic that does not fit neatly with the rest of the game, consider if the complexity is adding anything. If it is, then great — tell me in the comments! If not, why not replace it? It is likely no one else at the table actually knows how it works…