[4e] Opposed Rolls in Combat?

[Dizlag]

I'm considering trying an experiment in my 4th edition game that's in the current Hackmaster Basic game I've been playing ... Opposed Rolls in Combat.

Specifically, having the defender of an attack make an opposed roll when their AC, Fortitude, Reflex, or Will defenses are attacked. So, instead of taking a "10+" to these defenses, they would roll a d20 instead. I want to try this for a couple of reasons, the main one being keeping everyone involved a bit more.

My question this forum is two-fold:

1) Will someone run the crunchy bits (statistics) on this to let me know if the attacker's effectiveness will go down or up and by how much?

2) Would you accept a whole night of gaming with this implemented as a test?

Thanks for your time everyone!

Dizlag

[Tequila Sunrise]

1) Will someone run the crunchy bits (statistics) on this to let me know if the attacker's effectiveness will go down or up and by how much?

Attacker effectiveness will go down by 2.5%. (The average roll on a d20 is 10.5, which is a 2.5% difference from 10.) The important thing that opposed rolls really change though, is they make everything more unpredictable and chaotic. And chaos favors the underdogs, which means the monsters. Depending on your group, this might be a good thing or a bad thing.

2) Would you accept a whole night of gaming with this implemented as a test?

I'd give it an honest go, but I personally find opposed rolling to be more trouble than it's worth, at least from the DM's side of the screen. But if you're the DM, I don't have to do all that math!

[Aron Times]

At one point during 4E development, combat worked like this (this was discussed in an pre-DDI Dragon issue). However, it drastically slowed down combat because it doubled the number of rolls in combat; for each attack roll, there was a defense roll.

In the end, the devs used the current combat system, where the defender takes 10 on his defenses and only the attacker has to roll.

[Tengu_temp]

It makes combat run slower and very slightly boosts defense in comparison to attack, but it's a perfectly reasonable mechanic apart from that. If you and your players prefer a lot of rolling, go for it.

A benefit of this mechanic is that various powers that let you reroll now affect your defense too. On the other hand, reaction powers that boost your defenses are less reliable now.

[TheEmerged]

At one point during 4E development, combat worked like this (this was discussed in an pre-DDI Dragon issue). However, it drastically slowed down combat because it doubled the number of rolls in combat; for each attack roll, there was a defense roll.

In the end, the devs used the current combat system, where the defender takes 10 on his defenses and only the attacker has to roll.

I've experimented with a few opposed-roll combat systems in the past, and the part I bolded in the above is the main reason I don't like them. The point Tequila raised about it causing wilder swings in the rolling is a valid point as well (although secondary to the slowdown for my tastes).

My favorite alternate attack roll system? Instead of having the monsters rolling to hit players, the player being attacked had to roll to be missed. Of course I say that as the GM

[Decoy Lockbox]

I'm currently running a campaign in which the players roll all the dice. Instead of attack rolls, monsters have static "attack values" or "spell DCs", which the players roll against. I converted the 4e "defense" scores back into the saves they were in 3rd edition (the phrase "make a will save" is just too nice to drop), and introduced a "defense roll", which can represent anything from dodging, parrying, shield blocks, or just plain luck (the defense roll bonus is just AC - 10). So instead of a fighter with reflex 25 being hit by a lich's ray (+13 to hit) 45% of the time, a fighter with +15 reflex can dodge a lich's ray (spell DC 25) 55% of the time -- same numbers, different feel.

The percentages are all the same, but it makes defense seem more active than just standing there.

I feel that using both attack and defense rolls would in fact slow down combat.

My favorite alternate attack roll system? Instead of having the monsters rolling to hit players, the player being attacked had to roll to be missed. Of course I say that as the GM

D'oh! Thats what I get for not reading that closely.

[Swiftblu]

The important thing that opposed rolls really change though, is they make everything more unpredictable and chaotic.

Not true. Rolling more dice (while keeping the average value roughly equal) actually reduces entropy. Consider the time-tested alternative to the d20 roll, 3d6. Both have an average value of 10.5, but the multiple dice of the 3d6 create a "bell-curve" distribution, meaning that values closer the average value are exponentially more likely to be rolled. On a 1d20 roll, every value has an equal chance of occuring. (actual 3d6 stats here)

Likewise, having both the attacker and defender roll is equivalent to having every player roll 1d20-1d20+10 instead of a normal d20, and again, more dice, less entropy. Forgive me if I don't calculate the chance for every value to appear, but there is only one dice combination out of 400 (.25% probability) that gives a maximum or minimum value (29, -9), but there are twenty combinations (.5%) that give the average value of 10.

Just saying, this option makes combat less random, favoring the PCs in a game like 4e, wherein they are not usually the underdog. It'll takes twice as long to run your combats, though...

[Yakk]

The value of +1 and -1 modifiers to a d20 roll is reduced by a factor of 1.4 (well, the square root of 2) roughly.

Accuracy on 'tied' attackers and defenders changes by a small amount (depending on who gets ties). This effect is quite small.

As these are the modifiers that you gain as you level (roughly), the power curve of the game flattens. A level 3 goblins is more dangerous to a level 5 character, and a level 7 orc is less dangerous to a level 5 character, than prior to this. The effect gets most pronounced when you hit larger values.

This changes the XP curve, so you might want to think about revamping the XP system, as otherwise encounter budgets will be off.

It slows down play. One of the reasons it slows down play is that in 4e, the player rolling and the player recording information are two different players. If the DM casts a fireball at the players, the DM just says "15 damage ... Hit Alice, Bob, Charlie, missed Eve". The players can "out of band" record the damage they take, or the status effect, etc. If they are rolling, this turns into a negotiation for each player, who then has to record the damage after the negotiation. This slows down area attacks (as after each roll, they must negotiate), and even slows down single target attacks.

[Tequila Sunrise]

Not true. Rolling more dice (while keeping the average value roughly equal) actually reduces entropy. Consider the time-tested alternative to the d20 roll, 3d6. Both have an average value of 10.5, but the multiple dice of the 3d6 create a "bell-curve" distribution, meaning that values closer the average value are exponentially more likely to be rolled. On a 1d20 roll, every value has an equal chance of occuring. (actual 3d6 stats here)

Likewise, having both the attacker and defender roll is equivalent to having every player roll 1d20-1d20+10 instead of a normal d20, and again, more dice, less entropy. Forgive me if I don't calculate the chance for every value to appear, but there is only one dice combination out of 400 (.25% probability) that gives a maximum or minimum value (29, -9), but there are twenty combinations (.5%) that give the average value of 10.

Just saying, this option makes combat less random, favoring the PCs in a game like 4e, wherein they are not usually the underdog. It'll takes twice as long to run your combats, though...

When you're adding dice totals together, yes, randomness is reduced. But with opposed rolls, it works the other way. When everyone is rolling against each other, there are more opportunities for the underdogs [monsters] to get lucky.

The biggest reason, other than reducing die rolling, that I believe most games don't use opposed rolls is because DCs normalize combat/other events. Using DCs makes things less random, which allows the designer to better control combat/events statistically, which makes the game safer for the main characters who have the upper hand. I wish I were a mathematician so I could explain this better, but maybe someone else can lend a hand.

[eepop]

I'm currently running a campaign in which the players roll all the dice. Instead of attack rolls, monsters have static "attack values" or "spell DCs", which the players roll against. I converted the 4e "defense" scores back into the saves they were in 3rd edition (the phrase "make a will save" is just too nice to drop), and introduced a "defense roll"(bonus is just AC - 10).

I think I saw someone run the numbers for this when it was suggested in 3.5 Unearthed Arcana. If I recall, you want to subtract 11 to make the math come out right. Although the difference wasn't that extreme, and it is a bit easier to subtract 10 than to subtract 11.

[Tengu_temp]

Here are the chances of rolling various results on 1d20-1d20:

Spoiler

Show

-19 0,25%
-18 0,50%
-17 0,75%
-16 1,00%
-15 1,25%
-14 1,50%
-13 1,75%
-12 2,00%
-11 2,25%
-10 2,50%
-9 2,75%
-8 3,00%
-7 3,25%
-6 3,50%
-5 3,75%
-4 4,00%
-3 4,25%
-2 4,50%
-1 4,75%
0 5,00%
1 4,75%
2 4,50%
3 4,25%
4 4,00%
5 3,75%
6 3,50%
7 3,25%
8 3,00%
9 2,75%
10 2,50%
11 2,25%
12 2,00%
13 1,75%
14 1,50%
15 1,25%
16 1,00%
17 0,75%
18 0,50%
19 0,25%

Seeing that the chance of rolling 0 here is 5% and the chances for other results are lower, and that on a simple 1d20 roll the chance for any result is 5%, I'd say this method is indeed more random.

[Artanis]

I wish I were a mathematician so I could explain this better, but maybe someone else can lend a hand.

You don't really need to be a mathematician


Here's my attempt at an explanation:

Monsters are supposed to die. That's their job. If you make things more random, then...
*If you get lucky and deal a lot more damage than usual, it doesn't matter. All you've done is speed up what's going to happen anyways.
*If the monster gets lucky and deals a lot more damage than usual, you've got a dead PC, which isn't the usual result.

So long story short, if you add more randomness, you wind up with more dead PCs, but not more dead monsters.

[Decoy Lockbox]

I think I saw someone run the numbers for this when it was suggested in 3.5 Unearthed Arcana. If I recall, you want to subtract 11 to make the math come out right. Although the difference wasn't that extreme, and it is a bit easier to subtract 10 than to subtract 11.

All you have to do is subtract 10 from the player's armor class and defenses, then add 12 to the monster's attack bonus. Players evade attacks by meeting or exceeding the enemy attack value. For example:

(normal rules)
A fighter has 25 AC and is being attacked by an Orc who has attack bonus +11. The Orc needs a 14 to hit, which gives him a 35% chance to hit, and a 65% chance to miss.

(new rules)
A fighter has a defense bonus of +15, and is being attacked by an Orc who has an attack value of 23. The fighter needs to roll an 8 to evade the attack, giving him a 65% chance to do so. If he rolls a 7 or less (35% chance), he fails to evade the attack.

So as you can see, the math is identical, and the only difference is that the players roll the dice instead of the GM.

[Stegyre]

Seeing that the chance of rolling 0 here is 5% and the chances for other results are lower, and that on a simple 1d20 roll the chance for any result is 5%, I'd say this method is indeed more random.

What is it you mean by "more random"? I would use that term to describe a situation where the standard deviation was larger: a large standard deviation is "more random" than a small standard deviation.

Your calculations show the clustering we would expect from multiple dice versus a single die. With a single die, you have 20 possible results, and 80% of the results range from 3 to 18 -- 16 possible results, not surprisingly 80% of the total possible results, as it's linear.

With two dice, you have 40 possible results (so it is "more random" in that sense, at least), that 80% range is from -12 to 12. This is 25 of the 40 possible results, or 62.5% of the total possible results. That would be "less random," I would think: 80% of the results now fall within a narrower range of all possible results.

However, I wonder if that discussion doesn’t beg the real question: the numbers on the dice are not the real results. The real results are, as a first approximation (i.e., ignoring critical), only two: an attack hits or it misses. Someone want to make some reasonable assumptions about to-hit rolls and calculate those scenarios? (I could; I might; but I’m not a mathematician and therefore probably more prone to errors.)

[Master_Rahl22]

I wouldn't even need the math. Doubling the dice rolls and better chances of PCs ending up dead from lucky low level monsters means I would never come close to this.

[Stegyre]

I wouldn't even need the math. Doubling the dice rolls and better chances of PCs ending up dead from lucky low level monsters means I would never come close to this.

That's just the question though: does doubling the dice rolls increase PC losses? Just because someone (Artanis) says so, doesn't mean it's correct. As already pointed out, more dice rolling for the same event (however tedious that may make things) brings the event closer to the average. This means that (contrary to Artanis's assertion), PCs are less likely to get lucky and end their combats sooner, and monsters are less likely to get lucky and kill PCs.

The SRD makes note of this phenomenon when discussing substituting 3d6 for 1d20, recommending that you reduce an encounter's CR, because PCs are more likely to win when the bell curve is introduced.

[Jack Zander]

I've tried rolling for defenses before and I have to say, on my personal experience, it actually sped up combat. My group is always talking and not paying attention to anything until their turn comes up, and then they have to spend extra time figuring out what to do. When they were rolling dice all the time, even when it wasn't their turns, they kept into the game, paid closer attention, and things ran much faster. If both the DM and the player roll their dice at the same time, it doesn't actually take any extra time either.

Your millage may vary depending on the causes for slow play in your group.

[Behold_the_Void]

We use opposed rolls in the system we're making, I am rather fond of the system. It doesn't tend to bog combat down too much, and psychologically, it feels more active, like you're actually trying to defend yourself which is an important aspect in our system. In our rules ties go to the attacker, and all in all it works out pretty well.

[Dyllan]

Lets assume that the PCs need to roll an 8 to hit, and the bad guys need to roll a 12. That seems about average for the combats I've been in (admittedly, we're only level 5, so that may change as we reach higher level).

Using standard rules, the bad guys hit 45% of the time, and the PCs hit 65% of the time.

Using the double roll rules, assuming we subtract 10 from the combined result (there has been talk of 11, but we'll work with 10), the bad guys hit on a combined roll of 2, which (according to the Tengu-temp's chart) hits 43.25% of the time.

The PCs need a result of -2 to hit, so they would hit 61.75% of the time. That's a 3.25% swing against hitting for the PCs, and a 1.75% swing against hitting for the bad guys.

If we increased the PCs stats by 2 across the board (so they needed only a 6 to hit, and the bad guys needed at 14), then the PCs hit 70% of the time (versus 75% in the current system), and the bad guys hit 35% of the time - just as in the current system.

To really tweak it out, lets say the PCs hit on a 2, and the bad guys need a 18 to hit. In the current system, the PCs would hit 95% of the time, and the bad guys would hit 15% of the time. In the dual-roll system, the PCs hit only 83.5% of the time, and the bad guys hit 21.5% of the time.

I think what this shows is that, the further you get from an even fight, the more it favors the underdog. This is likely because of the outlying numbers results of 0 through -9 and results of 21 through 29 - these make attacks that could not hit except on a 20 now possible, and attacks that could not miss except on a 1 now possible. That favors the underdog who needs those rolls. It doesn't just affect those rolls, however... as you approach those extremes, the outlying results become a bigger percentage of the total number of rolls that hit (or miss if we're looking at the PC attacks).

So yes, this system favors the underdog, not because it's more chaotic, but simply because it allows for a wider variety of numbers, thus marginalizing the flat bonuses (ie, high AC and high attack bonus) - in much the same way you would if you rolled a D30 and used 15 as your base AC.

[Artanis]

That's just the question though: does doubling the dice rolls increase PC losses? Just because someone (Artanis) says so, doesn't mean it's correct. As already pointed out, more dice rolling for the same event (however tedious that may make things) brings the event closer to the average. This means that (contrary to Artanis's assertion), PCs are less likely to get lucky and end their combats sooner, and monsters are less likely to get lucky and kill PCs.

I probably should've said that I meant randomness did that, but it's too late now


Edit: Something I just thought of that will mess with the math no matter whether or not you subtract 10 or 11 - natural 1s.

Normally, while a natural 1 always misses, it won't matter in most cases because a roll that low anyways. That is, a natural 1 misses anyways - and thus the automiss doesn't matter - unless your attack bonus is one point below the target defense or better (e.g. if the defense is 10, your attack bonus is +9). If your attack bonus is that high, then the enemy isn't a threat anyways, so nobody cares.

However, if a natural 1 attack roll still automisses on the opposed roll, then that rule will come into play in many, many more cases.

[Tequila Sunrise]

I've tried rolling for defenses before and I have to say, on my personal experience, it actually sped up combat. My group is always talking and not paying attention to anything until their turn comes up, and then they have to spend extra time figuring out what to do. When they were rolling dice all the time, even when it wasn't their turns, they kept into the game, paid closer attention, and things ran much faster. If both the DM and the player roll their dice at the same time, it doesn't actually take any extra time either.

Heh, that's an interesting way to keep everyone focussed. I'll keep that in mind the next time my players' attention wanders.

[Dizlag]

Hi everyone!

Thanks for all the replies to the original post. It really helped a lot and we actually tried something new out this past weekend. Instead of opposed rolls, the player's rolled their defenses, so basically the "10+" on defenses was replaced with a d20 roll and the monsters got the "+10" to their attacks which set the target number.

We added a special rule for rolling a "20" or a "perfect defense". The defender got a basic attack as an immediate reaction. We also added a special rule for rolling a "1" or "fumbling the defense", the monster got a critical hit.

Then it got us thinking about the player's attack rolls, specifically rolling a "1" or "fumbling the attack". We gave the monsters a basic attack as an immediate reaction in that case.

It seemed to work out great! Although there were just three of us players, seemed to go faster and it took a bit off the DMs mind having us make all the rolls. It put a little more of our own destiny in our hands (and dice!) and overall worked out pretty good.

Again, thanks for the hints, tips, and responses!

Dizlag

[Yakk]

*nod*, that particular inversion should work perfectly.

The perfect defence/fumble is also pretty good. It is slightly against the players, but only in the sense of variance (and it makes solos and brutes more dangerous).

Note that monsters that critical on a 19-20 should cause you to fumble on a 1-2. ;-)

[Dizlag]

Yakk,

Thanks for the reminder about the 19-20 critical range for monsters. I'll make a note of it and send it to the other DMs.

Dizlag