To hit calculation

[takkaria]

I've been fiddling around with my local copy of the game of late and wondering how to make things more fun and make mechanics more discoverable.

One thing I think the game is lacking at the moment is an obvious mechanic for the to-hit formula. Luckily, I was playing some second edition D&D the other day and the to-hit mechanic was pretty obvious there, so I'm pretty tempted to import THAC0 into Angband. It would require a big of a rejig, but I think it would make quite a nice fit, and make just what to-hit scores on weapons mean a little more discoverable.

I will try and get a patch up here in the next couple of days. In the meantime, discuss.

[Pete Mack]

I think the to_hit formula is pretty simple--no chance, if your to_hit is below 3/4MAC. Linearly increasing chance beyond that point, with a "natural" 5% chance of failure or success. (This is pretty much what I recall from when I played D&D way back in the dark ages.)

[PowerDiver]

The pure D&D mechanic means a progression of X * 5% values. That shortens the range over which your skill matters. To make it work, you need to shorten the range of skill. This is related to the discussions of device activation that unfortunately went nowhere.

Do you want the player to [fairly quickly] improve to 95% success, and thereafter make no improvement with additional plusses [weapon, stat, or level]? That's where the simpler methods lead. The current system allows unlimited progression. I also think there ought to be a way to convert hit rolls that are way above AC into criticals [rather than the current method I really dislike].

The only strange things currently are the 3/4 multiplier on AC and the 3 multiplier on to-hit. You could just rescale the values in race_info and class_info to use straight AC and convert +to_hit = +to_skill directly, and publish the exact skill number. Make all +to_hit transparent, and the player could compare skill to monster AC to see hit chance directly [could even be included in the monster description]. OK -- currently +4 AC on a monster requires +1 to_hit to offset so I guess all of the monster ACs would have to be rescaled as well, but that's still simple enough.

[Nick]

One thing I think the game is lacking at the moment is an obvious mechanic for the to-hit formula. Luckily, I was playing some second edition D&D the other day and the to-hit mechanic was pretty obvious there, so I'm pretty tempted to import THAC0 into Angband. It would require a big of a rejig, but I think it would make quite a nice fit, and make just what to-hit scores on weapons mean a little more discoverable.

I will try and get a patch up here in the next couple of days. In the meantime, discuss.

Too subtle, I think. Personally, I would have gone with "I'm moving to O combat".

[takkaria]

Too subtle, I think. Personally, I would have gone with "I'm moving to O combat".

Well, O combat with THAC0 would probably be a compromise solution.

PS. I also want cocaine in the black market.

[Nick]

PS. I also want cocaine in the black market.

Don't we all.

[PowerDiver]

PS. I also want cocaine in the black market.

Speed isn't enough for you?

[Rizwan]

Speed isn't enough for you?

Or mushrooms of hallucination?

[Atarlost]

There's no reason you have to do THAC0 with d20. You could use the same algorithm in 1% increments with simulated d100s.

[PowerDiver]

There's no reason you have to do THAC0 with d20. You could use the same algorithm in 1% increments with simulated d100s.

Makes no difference in the basic problem of a quick progression to 95% and no progress thereafter, which is guaranteed given the huge difference in possible AC values.

The only viable solution I can see is for Takkaria to allow a choice between current style and THAC0, by introducing a new birth option. Make that 2 new options, to allow for both d20 and percentile dice.

[AsimovIsaac]

sound intersting ....nice post tnx to share..

[PowerDiver]

Or mushrooms of hallucination?

Tragically, said mushrooms are never for sale at the market.

[Polyonymous]

Angband's to hit mechanism basically works as follows.

You always hit 5% and miss 5% of your attacks.

In the remaining 90% of the time, if your hit is greater than the target's AC (with all the scaling factors of *3 and *3/4 glossed over) and you double your hit, you will halve the miss chance. If the AC doubles, the miss chance doubles.

As Eddie has pointed out, in a lot of ways this is better than a linear system (hit chance is proportional to the difference between hitroll and armor). A linear system has a smaller sweet spot, making it harder to balance. In order to combat the inflated statistics deep in the dungeon (need to inflate so that more powerful characters do better against weaker ones) and keep the sweet spot from being irrelevant (always 5% or 95%), the scale of the linear system will have to chance as the stats inflate.

Once the scale starts changing, it quickly becomes basically what the system has now.

Currently outside the 5% hit and miss cases, a player has a [(hitroll - 3/4 * armor)/hitroll] chance of hitting.

THAC0 works as (hitroll - 1 * armor)/20 + some constant.

In order to prevent a THAC0-like system from quickly becoming an edge case (it has a fixed size sweet spot of 20 points between 0% and 100%, making it 1% increments doesn't change the fixed nature, either deep mages never hit or deep warriors always do) some sort of scaling will be needed to ensure the sweet spot stays relevant.

So it becones (hitroll - 1 * armor/20)/scaling + some constant

If we also realize that hitroll and armor scale as you go deeper I can rewrite that as (hitroll - 1 * armor)/20/(hitroll * some constant) since I can use hitroll as a proxy for the appropriate scaling factor. Instead of using 3/4 it uses 1 and /20 in various places. In the end it's the same functionality.

PS I greatly applaud your investigations into why things are as they are. Generally there's a really good reason for things. (like GW's fixing of object rarity * artifact rarity as a constant when creating a randart, it preserves the overall rarity of artifacts). Now if we just had all the design notes from the history of the game. If nothing else, this examination will help expose why angband does things the way it does (hit code is the same thing as an autoscaling THAC0, it takes care of the scaling without having to balance things, etc). Now if only Robert, Ben et al participated in the conversation.

PPS as you figure things out, add the reasons why to the code or to a separate design document that's part of the source archive so that a maintainer down the line can quickly see why things are what they are.

[ramela]

I'm curious to know why so many people prefer the THACO system over the base attack bonus system in D&D 3rd edition?

Baldur's Gate nostalgia?

Either way, optimally, the to-hit calculation should be simple. Preferably something that allows you to calculate the odds of a blow landing without resorting to paper or a calculator.

odds-of-hitting = to-hit-skill / (to-hit-skill + monster-AC) is pretty much the upper limit of complexity.

[Polyonymous]

THAC0 and d20 are basically the same system, the math is just explained a bit differently, in THAC0 you have a combat table where you look up the math, in d20 you do the subtraction yourself. As I recall THAC0 improved by one for each +hit, making it functionally equivalent to d20. It's a naming convention and an explanation paradigm, nothing more.