Goodmorning all

By the time you're rolling 10D2,  you really are wasting processing time. and it's not intuitive the curve you'll get as a result

One could fix that by adding M+ 10D2. to bring the probabilities into line..   but it's just easier to do

Level 1  = Gauss (11.5, 5)

Level 10 =  Gauss ( 15,2.5)   subject to the constrains reroll if outside of [min,max]. 

15 might be too low, or 2 too narrow . . . but it's really easy to see.  easier to vary (Gauss (10+lv/2, 5 - lev/4), and easier to balance.

The Communist Party of Elemental disapproves of 1DN model and joins the rally.

Now if we could only get the moderate left and the moderate right to join the political tides, I'm glad to see I have the extremists :- P

Dude. Gauss IS bell curve/ bell distribution.

Also, Not all Bells/Gauss are equal. We have at least 2 variables with Gauss (height of curve and standard deviation). I only hope we can make Gauss skewed right or skewed left as well (and skewed by how much).

Catholic fundamentalists throw their vote for "1dN is the work of satan!"

Seriously, we need gaussian! 1dN randomization is not enough to have interesting choices, it could work in a game like GC II, because in the end the statistics will favour the better fleet.

But in a game where tactical combat is present, you need the ability to weight your options. The only way I could see the current system working is by separating "to hit" and "damage" rolls. That way you can keep 1dN for attack (like d&d) because in the end is just a 0/1 situation (you either hit or miss, so nothing wrong with an omogeneous distribution of success rate). But totally random damage in the 0-80 rng?

Hell no!

At least open up those files for modding (if it is doable... don't know much about elemental files structure, yet)

edit:spelling

From https://forums.elementalgame.com/394530

I declare Victory is ours
" 

Battle Improvements

Damage Calculations “A lot of games use 1DN rolls,  they are quick, easy to implement, and WRONG.” (https://forums.elementalgame.com/392880)  

"

Go us

Yeah, what Tasunke said!

Let's keep it up... Is wishful thinking on my part to believe the majority of players would want this changed?

So, you are saying that we can put you in steel armor and I can club you as much as I want, and you will never get hurt?

Believe me, after the tenth time I have hit you on the head, you will start to feel it.  If I knock you down, you will feel it.

Can someone provide the percent chance of missing assuming a bell curve?   This didn't make sense to me until I started breaking down the odds for the current system.  I was surprised by the results.

Current system with an attack of 6, defense of 8.  I am also assuming the range is 0 - 6 and 0 - 8 respectively.

Damage  is the first row.

Second row is number of combinations that lead to that.  Last row is the chance of it occuring

    <1             1             2          3           4             5         6
    42             6             5          4           3            2          1
    66.67%    9.52%    7.94%    6.35%    4.76%    3.17%    1.59%

Can anyone provide the same for the bell curve or Guassain distributions.

To be honest, I'm really coming around to the system as is.  I've just accepted the fact that combat is dangerous.  Beef up your hitpoints, since you can't afford to skimp on them.

I voice my agreement that there are many ways better than 1DN, nearly any of them I would support.  Having Attack be different than Damage is useful for detailed battle and makes sense, but that would be icing on the cake. I hope SD will listen to this post.

This. So much this.

I've been having this same conversation with people for years. Not all of them are ignorant of statistics, but a lot of them are. It's a simple proposition - as you increase the central tendency of your random dice roll, you increase the probability that the higher trait will win, and the lower trait will lose. This is bad for game balance, because it accelerates the point at which the situation becomes unwinnable.

I've done a lot of calculation and experimentation to figure out a perfectly 'fair' dice mechanic - my criteria being 'if your skill is twice your opponents, you should win twice as often as he does.' The system that satisfies my requirements actually is a variation on the 1dN mechanic... it's something that would only work in a computerized environment, so I've never been able to do anything in tabletop.

The basic procedure: X = Attack, Y = Defense, Z = Minimum of Attack and Defense. Now generate 2 random numbers, one between 0 and X+Z and the other between 0 and Y+Z. If the first number is larger, attacker wins. Otherwise, the defender wins. Excluding ties, this generates a set of odds that exactly matches the ration X:Y.

I have a program written in Anydice that demonstrates the operation.

Use something like this, and a hit-miss system, and the stats are totally intuitive. If Attack is 10 and Defense is 20, then the attacker will miss 20 times for every 10 times they hit. Simple and easy to understand. Then you compare damage to hit points - if you do 1 damage and they have 10 hit points, it takes an average of 10 hits to defeat them. Again, simple and intuitive.

Also, I do agree with the idea that squads should not have a single attack value. A squad of 12 units, each with 6 attack, should be attacking 12 times for 6 damage, not once for 72 damage.

I dont mind a chance for missing, a chance for a parry, etc ...

HOWEVER, if it hits I'd rather have a reasonable result rather than using a giant mallet to hit for 1 damage. (and the sad thing, is that its possible regardless of the type of armor or damage absorption the player has ... they could be shirtless and still only get dealt 1 damage)

and I'm not even saying that 1 damage should be "impossible" ... but it should be far FAR more likely for (1 standard deviation from the mean) than a 1 or 0.

so ... Gauss (20, 4), or even just 10-30 damage. But honestly, I think its probably just the UNIFORM DISTRIBUTION which feels wrong.

Also, it would be nicer if each soldier is handled separately, rather than combining defense and combining attack. Just because 100 peasants are wearing leather armor, doesn't mean a dragon or a hero can't even make a scratch on em. A hero would probably be able to take down at least half of them, while a Dragon should be taking them out maybe 5 at a time. (more with fire-breath).

Also, I personally feel Run Speed vs Attack Speed is a big deal. The higher your run speed, the less points are required to move a square, and the higher your attack speed, the less points are required to make an attack.