View Full Version : Rolling AC, any comments?

2007-01-22, 10:32 PM
As explained in the DMG sidebar (sorry, will edit for page), in our group we are used to roll AC to have opposed checks instead of a fixed number. This is done by substracting 10 to your AC and rolling 1d20 + the remaining modifier (example: you have a static AC of 17; then you have a mod AC of 7 and you roll 1d20+7 for AC).

Now, do you think this has any downside? Other than having more rolls to do. I mean, with the other alternative mechanics, the DMG offers advice to whether it could unbalance the game towards the monsters or the party. But here the sidebar lacks that bottom line.

My first reaction is to think it makes things more varied, so it favors situations where the party will be able to hit that monster instead of just being frustrated. But also means the opposite, I know.

Viscount Einstrauss
2007-01-22, 10:42 PM
Imagine a PC tank faces a group of goblins.

Now imagine the PC tank rolls a natural 1 while the goblins roll natural 20's.

Despite his entire class role and the innate weakness of goblins, I think the tank is about to be destroyed because he was effectively ruined just now.

Likewise, imagine that in reverse. Now the goblins are so easy that they might as well stab themselves instead of wasting the tank's time.

2007-01-22, 10:42 PM
It makes things varied, but ultimately it's identical to the normal AC rules. the 1d20 roll averages 10.5, meaning most of the time your AC is more or less identical. It is slightly harder for the attacker to hit the defender (note the .5), but in general all it means is that there is more randomness.

Randomness by definition isn't going to have a solid effect. It'll make some encounters much easier due to a string of bad luck, and some much harder since there's lots of lucky rolls. But in the end it works out the same.

If the system favors anyone it favors the defender, but both defenders are more or less in identical situations.

2007-01-22, 10:44 PM
Well you've already mentioned one downside, the extra rolls will take time. The other issue is simply adding randomness...randomness hurts the PCs over time. The NPCs / monsters are expected to be fodder so a string of bad rolls isn't that big a deal, but a string of bad rolls by one or more PCs could lead to a TPK.

2007-01-22, 10:57 PM
I would personally like a bonus based on class, sort of like BAB, but that's just me. I like wizards to be worse at defending themselves than warriors, even assuming the same gear.

2007-01-22, 11:14 PM
@ Wehrkind - That can be found here: Defense Bonus (http://www.d20srd.org/srd/variant/adventuring/defenseBonus.htm).

2007-01-22, 11:14 PM
Sure. All in all it doesn't seem like such a great idea, especially if statistically it's no big change or could even hurt PCs.

2007-01-22, 11:16 PM
Raum nailed this one. Nuff said.

2007-01-22, 11:25 PM
Well, smack my ass and call me Judy.


Viscount Einstrauss
2007-01-22, 11:27 PM
Hey there, Judy.

2007-01-22, 11:27 PM
Nice alternative but not exactly the same. No randomness there.

2007-01-22, 11:31 PM
I was talking about where he mentioned that a) it takes even more time to do andb) in the end, it ends up really hurting the PC's more because randomness always hits the guy with the most dice rolls.

2007-01-22, 11:40 PM
You want a comparison between:
1d20+mod vs 1d20+mod
1d20+mod vs 10+mod

Mods are roughly 1.5 times as effective in the fixed-target one as they are in the contested-roll one.

Ie, a +10 BAB under fixed target is about as good as a +15 BAB under random target.

The same can be said of AC: +10 AC under fixed target is about as good as +15 AC under random target.

That is the largest effect. AC and to-hit mods matter a bit less.

(the ratio is actually about sqrt(2), if you are curious. The value of a plus mod is roughly purportional to the SD of the random number generator used...)

Next, the tails are longer under d20 vs d20. Extreme differences of AC matter more.

Third, some interesting things happen to the mathematics. Calculating ideal power attacks isn't as trivial (but can still be done via gut feeling), high-threat range weapons are a touch better (less chance of their threat range being clipped).

Lastly, are you using the "1 is a -10" and "20 is a 30" rule, or "1 is an auto-fail" and "20 is an auto-success" rule?

With both people rolling, if you want 5% of swings to be auto, you don't need 1 is auto fail anymore. An auto-success by defence (natural 20) happens almost 5% of the time. If you also had auto-fail attacks, auto-miss happens almost 10% of the time -- twice as often as the RAW.


That useful?

2007-01-22, 11:41 PM
I've played two settings in a single campaign with these rules. It really just made things too cumbersome. I like that it mixes things up, because sometimes in battle you just get a lucky blow. However, this resulted in too many die rolls.

I don't know that my opinion really stands here, either. We were using variant armor rules (armor as AC and DR), and the DM was using 3d6 while we used a d20 (also upset the 'balance') for attack, skill, etc.

Overall, I like it in theory, but practically it was a mess.

2007-01-22, 11:56 PM
Thanks for the analisys, Yakk. It does seem less worthy the more I know about it. :)

2007-01-23, 12:30 AM
/shrug, it does make the game more gritty.

Heck, you can add in parry/block rules using it pretty easily. :)

The problem is, melee combat in D&D is pretty grindish. Making it take longer isn't always a good idea.

2007-01-23, 12:41 PM
An alternative is to have a minimum for this roll. That would reduce the randomness somewhat. If you choose a minimum of 10, you should limit the number of opposed AC checks allowed.