A Battle system: Archer vs Dragon » Forum Post by Climber

It seems that people have an aversion to the term "random" in the same way that many people have an aversion to the word "taxes," without recognizing the necessary nature of both. Having no dice rolls whatsoever will produce formula outcomes. I was once a very serious chess player, playing competitively in tournaments and studying strategy books. The reason why I turned away from chess is because it became a very predictable game of pattern recognition rather than creative analysis and broad strategy. If you are a fluent chess player playing another fluent chess player, the first half of the game is almost always the same and the end game follows a similar circumstance. What you are suggesting, in effect, is to make Elemental combat like chess.

I might be less averse to your battle system if your cielings weren't so low, which will invariably force the player to rush for bigger and better soldiers whom will act with impunity against soldiers who are a mere 1 or 2 technological generations behind them.

If I march up toward a dragon with my army, both sides will be able to determine very quickly whether the fight will be a decisive defeat or a decisive victory. If I have 5000 archers of similar calibur, I will know they have no chance to even touch a Dragon, so I will promptly and ingloriously withdraw without even trying to win. Fights involving a few large units will be even more predictable. This will result in an alarmingly few number of battles or, if the player/ AI isn't very wise, a lot of slaughters, with large units constantly blitzing toward the units that will deal them 0 damage and those very same low level units never, ever, ever attacking units that they are clearly and obviously incapable of damaging. Battles will end up like a Monte Python Colliseum, with units chasing each other in circles. I can guarentee that upon entering 3/4ths of the battles, I will know with certainty whether I will win or lose. If I know I will lose and stand to fight, I will be unable to acknowledge that I'm fighting to win, but instead that I am fighting to just inflict casualties.

If there is less certainty with whether my units will hit, it's much more likely I will stand to fight because I know that, even if it's a long shot, there is at least a small chance that I might win.

Now let me be clear--- I don't like randomness deciding the game on a macro level. Large random events or influences in the game I am against. But if you have many small random occurances, the player is forced to think creatively instead of mathematically. I can calculate in my head with a certain degree of confidence what I think my odds are of winning, and integrate that into my total strategic outlook.

Now, some people find it enjoyable to play a strictly mathematical game. I prefer playing games that challenge me to think creatively and mathematically.

Successfully updated karma reason!

Yeahhh this system kind of takes Demiansky's worries about the system brad briefly laid out and makes it a very, very large problem indeed. It exacerbates the 'damage ceiling' effect to the point where your some of your basic, poorly trained units won't even be able to affect some medium quality units, some of whom in turn won't be able to affect some high-quality units, and so forth. No # of low-end troops would be able to scratch even a single well-equipped, well-trained high-end 'buildable' unit. One Bear Paladin should not in any way be immune to 10,000 peasants.

Really, IMO the only types of units that these damage ceilings should apply to are when low-end units come face to face with the most powerful and rare creatures, like dragons. Not a single buildable unit should fall into this category.

And the fact that there isn't any randomness at all to this system isn't appealing, as Demiansky said. With zero randomness, combat is much less interesting.

Successfully updated karma reason!

Quoting pigeonpigeon,

reply 2

Yeahhh this system kind of takes Demiansky's worries about the system brad briefly laid out and makes it a very, very large problem indeed. It exacerbates the 'damage ceiling' effect to the point where your some of your basic, poorly trained units won't even be able to affect some medium quality units, some of whom in turn won't be able to affect some high-quality units, and so forth. No # of low-end troops would be able to scratch even a single well-equipped, well-trained high-end 'buildable' unit. One Bear Paladin should not in any way be immune to 10,000 peasants.

Well said. I've changed the forumla to 2.5%

Damage inflicted = DAM * [1 + (ATT – DEF) * 2.5%]

What it does is it broaden the damage ceiling. Now you cannot damage a high quality unit if your attack is 40 points lower than their DEF. If your ATT is within that 40 points, you just need to increase you Quantity of unit to deal more damage. Now your Bear Paladin can be killed by 10,000 peasants.

It is more like a balance issue, instead of a issue of this concept

Quoting pigeonpigeon,

reply 2

the fact that there isn't any randomness at all to this system isn't appealing, as Demiansky said. With zero randomness, combat is much less interesting.

I disagree that Dice randomness in combat is more interesting. You cannot control a dice in anyway.

(An understandable & somewhat controllable) Randomness is there already without the need to roll your dice, because there are already many valuables in combats, like opponent casting a ATT buff spell, using the "dragon slaying arrow", and other strategy you don't expect.

The randomness of a dice roll is not interesting. With luck (but using the same skill playing), you win or loss a significant battle because of a really bad dice roll. Skills like scouting, using terrain effectively, building the right units/army and many strategic mechanism in the game will then become the highlight of any combat, instead of dice luck. I think it is cooler!

Successfully updated karma reason!

Tridus

Join Date 03/2008

+110

Quoting Demiansky,

reply 1

It seems that people have an aversion to the term "random" in the same way that many people have an aversion to the word "taxes," without recognizing the necessary nature of both. Having no dice rolls whatsoever will produce formula outcomes. I was once a very serious chess player, playing competitively in tournaments and studying strategy books. The reason why I turned away from chess is because it became a very predictable game of pattern recognition rather than creative analysis and broad strategy. If you are a fluent chess player playing another fluent chess player, the first half of the game is almost always the same and the end game follows a similar circumstance. What you are suggesting, in effect, is to make Elemental combat like chess.

There's randomness, and then there's *randomness*. A certain amount of randomness is one thing, and I'll agree that its not bad. Taken too far, you get outcomes that are completely unpredictable because the random factor is so large that it overwhelms everything else.

An overwhelming force controlled by a better player should not lose because the RNG did something odd. Once that happens, the game isn't about skill or tactics anymore.

But in a close fight? Sure, randomness is necessary and even makes things fun (it gets tense when you're about to die and have a 40% chance that the lightning bolt you're casting will kill the enemy this turn).

Successfully updated karma reason!

yea, I think anything using randomness should have a nice ... high bell curve, so that 60 times out of 100, the events will meet at 50% probability ... of either damage, attack, or defense range. This just means that out of the hundreds of permutations, 60% of them will be in the middle range (45-55% likelyhood) or so ... I think its hard to explain comparing percentages to other percentages, but its like 3 + 4 vs 4 + 3 ... they both arrive at the middle, but through different avenues.

I suppose what im saying is I want results to be weighted towards the center ... aka using various d6s instead of one epic d20.

I suppose one flaw to this system would be that there would be a minimum roll value of #of dice. I suppose with a minimum of 10% ... the # of dice used could be total value/10 ... or something. So an archer with attack of 4 woudl be a d4, but a creature of defense 40 would have 4 d10s for defence. A being of 22 defence woudl perhaps have 2 d10s and 1d2? not sure ... possibly two d 11s. Anyways, a proper formula would need setting up, but I have a feeling this is far down the road, and might be connected with the development of weapon stats/ weapon abilities.

I can see most units having fairly low attk/def (even if only the most basic have as low as 2-4), with a general max being around 10 or 20 perhaps.

+1

Successfully updated karma reason!

yea, I think anything using randomness should have a nice ... high bell curve, so that 60 times out of 100, the events will meet at 50% probability ... of either damage, attack, or defense range. This just means that out of the hundreds of permutations, 60% of them will be in the middle range (45-55% likelyhood) or so ... I think its hard to explain comparing percentages to other percentages, but its like 3 + 4 vs 4 + 3 ... they both arrive at the middle, but through different avenues.

To make it a little simpler, or perhaps just more conventional (mathematically), they could aim for the mean ATK/DEF rolls to be, say, 1/2 the max, with a standard deviation of 1/4 the max. That way, over large numbers, 66% of rolls should fall within the middle half of the possible range. Ultimately I'm not sure if it's make much of a difference, though.

I can see most units having fairly low attk/def (even if only the most basic have as low as 2-4), with a general max being around 10 or 20 perhaps.

That sounds good to me, for buildable units anyway. A big enough range so that the rabble will be significantly weaker than the elite, but small enough so that the elite won't be godlike in the face of the rabble (that position should be retained for channelers and fantastical creatures ).

Successfully updated karma reason!

Quoting Tasunke,

reply 5

yea, I think anything using randomness should have a nice ... high bell curve, so that 60 times out of 100, the events will meet at 50% probability ... of either damage, attack, or defense range. This just means that out of the hundreds of permutations, 60% of them will be in the middle range (45-55% likelyhood) or so ... I think its hard to explain comparing percentages to other percentages, but its like 3 + 4 vs 4 + 3 ... they both arrive at the middle, but through different avenues.

I suppose what im saying is I want results to be weighted towards the center ... aka using various d6s instead of one epic d20.

I suppose one flaw to this system would be that there would be a minimum roll value of #of dice. I suppose with a minimum of 10% ... the # of dice used could be total value/10 ... or something. So an archer with attack of 4 woudl be a d4, but a creature of defense 40 would have 4 d10s for defence. A being of 22 defence woudl perhaps have 2 d10s and 1d2? not sure ... possibly two d 11s. Anyways, a proper formula would need setting up, but I have a feeling this is far down the road, and might be connected with the development of weapon stats/ weapon abilities.

I can see most units having fairly low attk/def (even if only the most basic have as low as 2-4), with a general max being around 10 or 20 perhaps.

I agree with this statement. However, I would like that bell curve to continue on indefinately rather than be capped. Also, it's very, very important that we mark the distinction between "randomness" and "probability." Randomness is when something completely unpredictable happens that you are umable to consider in decision making (a meteor crashing into the battlefield of its own valition and killing some of your units without you even knowing this was possible.) Probability can look like randomness, but it is uncertainty that you are aware of and can act upon accordingly (for instance, you might be aware that a meteor can fall into the battlefield with a 50 percent chance and kill X number of soldiers.) So in effect, probability is something the player can actually have control over with proper planning. Randomness I am against. Probability I am 100 percent for. What Climber is proposing is a system without probability.

Successfully updated karma reason!

Quoting Climber,

reply 3

(...)Well said. I've changed the forumla to 2.5%

Damage inflicted = DAM * [1 + (ATT – DEF) * 2.5%]

What it does is it broaden the damage ceiling. Now you cannot damage a high quality unit if your attack is 40 points lower than their DEF. If your ATT is within that 40 points, you just need to increase you Quantity of unit to deal more damage. Now your Bear Paladin can be killed by 10,000 peasants.
(...)

If I understand correctly, you sum the ATK and DMG of all units in the group and compare it against the other units DEF. And what if there are couple of units on the opposite side? Should we sum their DEF too? As you can imagine, this way a group of 100 peasants would have enormous advantage against a dragon, which would be silly. There is a problem however, if you don't sum all the DEF of the defenders, why do you sum all attack of the attackers? Assuming, that I overlooked some of your descriptions, you take ATK & DEF of each unit of attackers & defenders (let's say we can have only the same units in a squad), and then multiply it by the number of attackers, so that you know how many defenders to kill ( killForces(defenders, damage_dealt/defenders.size()) ). If that is true, than I am apologize.

What would I improve? Let me start from saying there are basically two parts of the equation: the DMG, and the ATK/DEF part.
Damage inflicted = DAM * [1 + (ATT – DEF) * 2.5%]

DAM - damage part (duh!).
[1 + (ATT – DEF) * 2.5%] - ATK/DEF part.

First, your whole function is linear. That is ok, but only for the ATK/DEF part. Why? The basic logic is such: I have slightly stronger fighters, I get slightly bigger advantage (damage). It's all simple, it's all linear. Ok, so why not to put it also into the damage part? For two reasons:

When 1000 peasants attack a dragon, it's impossible for all of them to strike at one time. There is simply not enough space. A few of them can fight easily, but after 20 peasants, it gets too crowdy near the dragon, so they can't freely attack. I think it's a logical, real-life based and empirical proof.
Real life examples doesn't seem to move devs (which should always strive for better gameplay, not better realism), so there is one advantage of diminishing damage (based linearly on numerical superiority) is that player is forced to invest in better units (those with higher ATK). I am stating that quality > quantity (to some point), if you didn't notice .

Said that, the function I would put into the C++ code (I hope you will understand it) as follows:

Code: c++
// Inline function calculating damage done to a defending squad.
uint32 damageInflicted(Squad& attacker, Squad& defender)
{
  uint8 atk = attacker.getAttack();
  uint8 def = defender.getDefense();
  uint16 squad_size = attacker.size();
  // In case of extremely low attack (in comparison to defense), the negative result
   // will be simply changed into 0.
  if (atk <= 40 + def)
    return 0;
  else
    return 5 * log(squad_size) + 1.0) * (1 + (atk - def) * 0.025);
}

I know the function isn't ideal, but it nicely reflects the actual power of a squad. Of course, we could implement a solution in which a bigger foe (in terms of size, or number of units present in the squad), could be attacked by a bigger number of attackers, but this is meddling with details, which is not important now.

Red.13 out...

Successfully updated karma reason!

If I understand correctly, you sum the ATK and DMG of all units in the group and compare it against the other units DEF. And what if there are couple of units on the opposite side? Should we sum their DEF too?

Since when was he summing over all attackers? I was under the impression that would be the formula for each constituent of the attacking force... Well I suppose it depends how combat is worked out.

If it's entirely group/squad based, then a single formula taking into account the numbers of units, # of participating units (whether explicit or abstracted like in your code) would probably be ideal.

But if the actual damage inflicted is done more like in TW (this is still quite feasible even within the context of continuous turns), then you don't need to account for # of units participating because it's all handled. The units close enough to strike the enemy, without friendlies or other obstacles blocking them, engage and the rest don't. As some fall, others take their place. In this scenario, each unit would have his own damage calculation done vs. a single unit within the opposing group.

Successfully updated karma reason!

Luck is fun. There should be some element of randomness but yet not too random so as to make strategies pointless.
Having 10 000 swordsmen swarming and slaying dragons is also fun. I don't exactly see that as being a problem or fun-killer.

I prefer something like
Damage inflicted = (ATT + 3D6) – (DEF)

This means that certain units that have very low attack will have absolutely no chance of damaging a high defense target.

in addition, there should be different defense values against different types of attack: melee defense, ranged defense and magic defense. A full plate steel armor should be of little defense against a magical lightning bolt.

For units that consists of squads of men, the number of men would determine the number of attacks per round rather than increasing or decreasing attack value. This means if 1 peasant cannot damage a dragon, 1 million will not scratch the dragon as well.

Successfully updated karma reason!

Quoting moondoggiee,

reply 10

Damage inflicted = (ATT + 3D6) – (DEF)

We aren't limited by dice in computers. If you want a 15 point spread with a minimum value of three you can just say
(ATT+3+random(0,14)) or something similar we'll understand.

Successfully updated karma reason!

Some minor update

% Ranged damage reduced = COVER % + SPEED/5
% Chance of completely evading non-elemental magic effect = RESIST %

For SOME elemental spells/effect, regardless of its spell ATT strength the victim has
% Chance of evading respective elemental magic effect = Fire/Ice/Wind DEF %

==

My intention is allow the lowest tier buildable unit can damage most, but not all, highest tier buildable unit, if there have overwhelming number.   If the ATT & DEF difference is 39, the lower tier unit needs about 40 times more in quantity to achieve their default DAMamge.    You need 40X in numbers to win, at the barest minimum.

My intention is not all highest tier buildable unit can be damaged at all by lowest buildable tier units, so gamer will have satisfaction that they can build very powerful units.

However, Highest tier buildable units (40-60) should not automatically be able to damage Highest tier Monsters (80-120).   Magic items, spells, and high level hero/Sovereign bonus is needed to fight these godly badass units (100-120) when the damage ceiling is now 40 ATT & DEF difference.

==
On Randomness…   Let me say:

If there is an important battle & both sides are completely evenly matched (including equally skilled gamers managing the battle).   The final outcome will seemingly boil down to Randomness.   They cannot be sure if they win/lose because of skill, or sheer randomness/luck.

Assume the same battle is contested instead with my battle system devoid of any randomness.   No matter how evenly skilled both players is, both will learn something from this battle. It is because they know randomness is not involved.   The end result is completely dependent on their skill/action (e.g. making an unnecessary move, or having slightly not appropriate weapon/unit design before the battle happens, really minor maneuver mistakes etc, etc).   I know I will be a happier gamer in this 2nd case.

Removing randomness in tactical combat, is a suggestion to improve the game. Randomness appears in other parts of the game is a must.

=
I believe the debate on whether “10000 swordsmen should kill a non-breathing Dragon/Drake” or not, depends on the fact that the battle should lasts 12-24 continuous turns (or whatever finite timeframe) at most. You can say that 1 out of these 10000000 swordsmen must have a chance to kill it because of a lucky strike. However, time is finite (say the battle last for 1 hour or 2 day in game time), you can easily assume that the lucky strike of low tier unit does not happen within the timeframe of the combat.

Therefore if there are huge ATT & DEF difference e.g. >40, yes, some highest end monster cannot be killed by lower tier units. (But this is a fantasy game, there should be some magical means to bridge a gap of (0-20 at most), in a case by case basis)

If you make a battle system that make Lucky Strike happens somewhat often (say 0.01% chance to critical instant kill per sword swing), are you willing to wait a long long time for such lucky strike by press the "END turn"? Or if you make it happen so often (2% chance), you will then be too happy sending hordes & hordes of lowerly swordsmen to the Dragon. What’s the fun with that, compared to the fact that you somehow need to find & equip enough rare “Arrow of Dragon slaying”?

Successfully updated karma reason!

Frogboy

Join Date 03/2001

+2394

I think some of you may be over-thinking the battle mechanics necessary to effect a satisfactory outcome.

Let's take a dragon who has say a 20 attack and 10 defense.

Its minimum defensive roll is going to be 10% so it will always get a 1.

Now, if you have a really basic archer its attack and defense might be 1 attack (bow and arrow) with just a shirt (0 defense) with 1 HP.

His arrow won't pierce the armor of the dragon and thus no matter how many of them you have, they will die.

You would need a much better type of archer (weapon wise). Maybe an archer equipped with a crossbow or a longbone that does 2 attack and 1 defense. You get enough of them and that dragon will go down.

Successfully updated karma reason!

Quoting Frogboy,

reply 13

I think some of you may be over-thinking the battle mechanics necessary to effect a satisfactory outcome.

Let's take a dragon who has say a 20 attack and 10 defense.

Its minimum defensive roll is going to be 10% so it will always get a 1.

Now, if you have a really basic archer its attack and defense might be 1 attack (bow and arrow) with just a shirt (0 defense) with 1 HP.

His arrow won't pierce the armor of the dragon and thus no matter how many of them you have, they will die.

You would need a much better type of archer (weapon wise). Maybe an archer equipped with a crossbow or a longbone that does 2 attack and 1 defense. You get enough of them and that dragon will go down.

Yeah, I have to go with Frogboy on this one and I have to defend his method. If you start adding a massive host of elements to consider, battles with anything over 100 soldiers will get very confusing, and putting all of that detail to use will be virtually impossible.

Successfully updated karma reason!

I believe the debate on whether “10000 swordsmen should kill a non-breathing Dragon/Drake” or not, depends on the fact that the battle should lasts 12-24 continuous turns (or whatever finite timeframe) at most. You can say that 1 out of these 10000000 swordsmen must have a chance to kill it because of a lucky strike. However, time is finite (say the battle last for 1 hour or 2 day in game time), you can easily assume that the lucky strike of low tier unit does not happen within the timeframe of the combat.

Therefore if there are huge ATT & DEF difference e.g. >40, yes, some highest end monster cannot be killed by lower tier units. (But this is a fantasy game, there should be some magical means to bridge a gap of (0-20 at most), in a case by case basis)

If you make a battle system that make Lucky Strike happens somewhat often (say 0.01% chance to critical instant kill per sword swing), are you willing to wait a long long time for such lucky strike by press the "END turn"? Or if you make it happen so often (2% chance), you will then be too happy sending hordes & hordes of lowerly swordsmen to the Dragon. What’s the fun with that, compared to the fact that you somehow need to find & equip enough rare “Arrow of Dragon slaying”?

A good illustration, and it would definately be a problem if that lucky strike is too low or too high.

Successfully updated karma reason!

Quoting Frogboy,

reply 13

I think some of you may be over-thinking the battle mechanics necessary to effect a satisfactory outcome.

Let's take a dragon who has say a 20 attack and 10 defense.

Its minimum defensive roll is going to be 10% so it will always get a 1.

Now, if you have a really basic archer its attack and defense might be 1 attack (bow and arrow) with just a shirt (0 defense) with 1 HP.

His arrow won't pierce the armor of the dragon and thus no matter how many of them you have, they will die.

You would need a much better type of archer (weapon wise). Maybe an archer equipped with a crossbow or a longbone that does 2 attack and 1 defense. You get enough of them and that dragon will go down.

your first post about the dragon really set them off

Successfully updated karma reason!

So it's just binary system when it comes to ATK/DEF? If I have ATK >= DEF, I can inflict damage = DMG * number of units? How will higher values of ATK increase the damage? Will it be linear?

Successfully updated karma reason!

Well personally i don't care what algorithm that is used.

All i need is a UI that tells me before i attack what is the most likely outcome of the melee attack (by giving me the centre of the bell shape)

I don't want to waste my time sending my peasants to attack the dragon only to realize it is impossible for me to scratch it.

I don't want to be forced to have to mentally calculate the odds in my head by comparing my units attack power with the enemy's defense etc.

Successfully updated karma reason!

Tridus

Join Date 03/2008

+110

Quoting red1939,

reply 17

So it's just binary system when it comes to ATK/DEF? If I have ATK >= DEF, I can inflict damage = DMG * number of units? How will higher values of ATK increase the damage? Will it be linear?

If it works in any way like GalCiv 2 (speculation, but since it was mentioned that both stats have rolls its entirely possible) then it works like this. Say we have 2 units:

Archer, ATK 3, DEF 1, HP 10

Guardsman: ATK 2, DEF 2, HP 10

When the Archer attacks the Guardsman, both units roll between the floor (currently 10% of their value, so 0 in this case) and their relevant score. So the Archer rolls from 0-3, and the Guardsman's defense from 0-2.

Compare the two. Attack Roll - Defense Roll = damage done (no damage is done if its 0 or negative). So if the Archer and Guardsman both roll 2, no damage is done. If its 3 and 0, 3 damage is done. The same would get done when the Guardsman attacks the Archer.

Now, lets say the Guardsman is now fighting an elite Deathguard Commander, ATK 9, DEF 12, HP 20. In this case the Guardsman's attack still rolls 0-2, but the Deathguard has a "floor" for its defense roll of 1. So it rolls 1-12. At most the Guardsman will only do 1 point of damage. The counterattack however rolls 0-9, so could be anything from no damage to a near fatal strike.

As I said, this is speculation based on how GC 2 worked.

Successfully updated karma reason!

Quoting Frogboy,

reply 13

I think some of you may be over-thinking the battle mechanics necessary to effect a satisfactory outcome.

Frogboy, may I disagree?   The system I’ve proposed is very simple, intuitive and versatile; not overthinking, imho.   It can be summed up by:
1)    Damage inflicted is proportional to difference btw the attacker’s ATT and defender’s DEF
2)    Different attack method (e.g. hand-to-hand, archery, magic) has its own ATT value
3)    Magical/Ranged attacks check against its relevant DEF value

NTJedi and I believes more attributes than ATT/DEF/STRENGTH/SPEED are needed, to allow quick and standardized method to distinguish between units; and improve moddibility

I will like very much to like to hear from your perspective the reason that the proposed battle system produce a satisfactory outcome.
Before you start, I understand the 2 of the satisfactory outcomes from your posts.
A. Elegant solution to abstract how 1 high STRENGTH unit can hand-to-hand 10000 units by asymmetric turn (& relative STRENGTH), without much lost in the suspension of disbelief
B. Elegant solution to disallow huge quantity of low ATT unit damaging rare high DEF unit (i.e. prevent 500 non-magical swordsmen killing a dragon)

However, I think the proposed battle system, linking STRENGTH to attack volley, may not exactly called easy to understand for some people.

Successfully updated karma reason!

Quoting Frogboy,

reply 13

His arrow won't pierce the armor of the dragon and thus no matter how many of them you have, they will die.

I like this system. I think it makes sense, both in real terms and in game terms, to have powerful units that are unstoppable by weak units.

For one it prevents the classic Civilization "tank killed by spearman" situation that we all hate. Sure it is just a game but that really seems to jolt the realism and piss off the players. We just feel that it isn't right nice, in reality, there is no way a spearman can do anything against an M1A2. The spear would literaly be hard pressed to scratch the paint.

However it also provides a useful game mechanic in that it means you can't just use swarm tactics. You can't totally neglect your technology and go for raw numbers. Sure raw numbers are useful to an extent, but if an enemy gets units that are far enough ahead, well then too bad, all the numbers in the world won't do anything.

Successfully updated karma reason!

Im not sure using the spearman/tank analogy is wise here, although I do agree that wooden, stone, and copper weapons would do jack ... although I suppose it will both be based upon the material used and the quality of the weapon.

Basic Obsidian might not do much ... but Expert-Crafted Obsidian is probably one of the sharpest Arrow-heads you could make.

Successfully updated karma reason!

Now, if you have a really basic archer its attack and defense might be 1 attack (bow and arrow) with just a shirt (0 defense) with 1 HP.

His arrow won't pierce the armor of the dragon and thus no matter how many of them you have, they will die.

You would need a much better type of archer (weapon wise). Maybe an archer equipped with a crossbow or a longbone that does 2 attack and 1 defense. You get enough of them and that dragon will go down.

I cannot help but think of the archer that took down Smog in the Hobbit. Though I guess he had a magical family hairloom arrow or something.

Successfully updated karma reason!

I cannot help but think of the archer that took down Smog in the Hobbit. Though I guess he had a magical family hairloom arrow or something.

He was an excellent archer, had a magical family heirloom, and was quite literally told by a little birdie of Smaug's weak spot. And in the grand scheme of dragons in Middle Earth, Smaug was small-fry. I cannot imagine Tolkien having had any of the more powerful dragons being slain even by someone like Bard... It took one of the greatest warriors in the history of his world, and a magical sword capable of killing even a Vala (after all, Morgoth is prophesied to die by it) to kill Glaurung and he wasn't even the greatest of dragons (and much of his power lied in overpowering people's minds - he wreaked far more havoc through mind-tricks than by brute force).

Then there was Ancalagon the Black... He was so huge that he blotted out the sky, and his assault forced the host of the Valar to draw back... It took Earendil, father of Elrond (and the only human ever to set foot in Valinor) in his magical flying elven ship, 'armed' with a silmaril a full day to defeat him.

Basically, all the 'great' dragons of Middle Earth were slain by some of that world's greatest and most powerful heroes. And none of the lesser dragons even were slain by hordes of peasantry; even in the case of Smaug it was the efforts of a single, very lucky and very skilled individual archer with a magic arrow.

Successfully updated karma reason!

so Even Smaug, a lesser Dragon, was slain by a hero. With no help from peasantry (other than not immediately dying)

Successfully updated karma reason!