Accuracy graphs for ranged units

[serveurix]

8 hours ago, Grugnas said:

Honestly more crush instead of pierce damage would be really interesting, perhaps an alternative to ordinary sieges.

I don't agree. Slingers shouldn't be seen as low-grade siege unit. They shouldn't be able to take down buildings, or more precisely they should be as good as archers or skirms in that regard.

I don't think slingers and siege engine do the same kind of damage. 300 pebbles sent from 300 slings at the same time don't equal a catapult shot or a ram hit.

With the current mechanism we have buildings that are very resistant to hack and pierce, to resist "regular" units (spearmen, swordsmen, archers, skirms), and less resistant to crush, to allow the siege engines to do damage. That means that the slingers will always have an edge over skirms or archers in taking down buildings, which is not realistic. We could minimize that by giving buildings the same armor against crush than they have against hack and pierce, and give all the siege engines super high crush attacks. But at the same time we don't want siege weapons to be OP against units. We don't want rams to mow down infantry like they currently do. I think this is unsolvable if we stick to the current way we treat slingers.

Some thoughts on what we could do :

- We could treat the crush damage from the slingers as a different kind of crush damage than the one from the siege engines. Although that would be probably the most physically accurate solution, I don't like it because it requires to create a fourth kind of damage (I don't even know if the engine supports it) and overcomplicates the game, just for one unit.

- We could give slingers pierce damage instead of crush damage. This is more realistic because the damage they do will be comparable to archers and skirms, both against buildings and on the field. The problem is that it won't be easy to fine-tune to make it even more realistic (for example I expect pebbles to be much less efficient than arrows against cataphracts and any kind of heavy-armored units), unless we add hard counters.

- We could keep crush damage for slingers, and go for a hard counter system that would allow for easy fine-tuning of units, but would be less elegant and would require players to rely on memory a bit more.

 

I would like to add that making slingers do pierce damage would not make them "yet another kind of archers", like the archers do not (or are not supposed to) constitute "yet another kind of skirmishers". Slingers would be different from archers, on a general level, in terms of range, damage, firing speed, accuracy, and firing angle (although I don't know if the firing angle has an influence in the game, if you have any info on that let me know). (and on a lower level in terms of cost, armor and walking speed but that's just regular balancing, not what we're talking about)

[LordGood]

i think 'pebbles' underestimates the size of sling stones, or the weight of leaden bullets. If I'm not mistaken Xenophon recorded how Persian slingers would regularly hurl stones the size of a man's fist, and slingers were a staple of Celtic siege warfare on both sides. If the argument was to nerf their ability against buildings and siege sure, that can be done easily enough with multipliers.

also in that scenario, 300 Persian slingers would be throwing as much rock as a 6 talent caliber Roman ballista, so keep that in mind lol

also keep in mind that 6 talent roman ballistae dont exist and that we don't have Persian slingers

so more like 15kg of stone or lead at the small end of things. Still 32lbs. Thats about cannonball mass still

[Lion.Kanzen]

Why no add siege armor and siege crush like AoE3?

[serveurix]

2 hours ago, LordGood said:

also in that scenario, 300 Persian slingers would be throwing as much rock as a 6 talent caliber Roman ballista, so keep that in mind lol

also keep in mind that 6 talent roman ballistae dont exist and that we don't have Persian slingers

so more like 15kg of stone or lead at the small end of things. Still 32lbs. Thats about cannonball mass still

Agreed. The effect on a building wouldn't be the same though.

[Grugnas]

Slingers damage on a build is an abstraction to make the game more interesting, otherwise slingers should be faster in movement than skirmishers (pouch of rocks is less cumberson than a quiver of javelins and bigger shield), with higher range and accuracy which seems kinda imbalanced. Having history as main reference doesn't prevent to have some abstraction though.

Also, pierce damage damaging buildings is really unrealistic.

[fatherbushido]

(when we say history we mostly think of art, of strings, of names, of references...)

[elexis]

> Also, pierce damage damaging buildings is really unrealistic.

Didn't we come to that conclusion several times already, always increased the Pierce damage and are now at 99% and it's still too few? Anything speaking against making that 100%? (I guess the formula, but perhaps we can go to 99,99%)

[wowgetoffyourcellphone]

2 hours ago, elexis said:

> Also, pierce damage damaging buildings is really unrealistic.

Didn't we come to that conclusion several times already, always increased the Pierce damage and are now at 99% and it's still too few? Anything speaking against making that 100%? (I guess the formula, but perhaps we can go to 99,99%)

As far as realism goes, yeah slings and arrows doing damage against a barracks is pretty silly. But as a gameplay abstraction I don't mind arrows and lead bullets doing some damage. 99% pierce armor is fine with me. Still takes a darn long time for a bunch of slingers to take down a barracks at that level of pierce armor, as long as slingers don't have crush attack; significantly longer than it takes the same number of spearmen or swordsmen. Ima do some tests, brb. I could do it with maths, but fighting things out in Atlas is so much more fun!

 

 

EDIT:

In Delenda Est, 30 slingers of mixed rank, representing a likely force of slingers raised by a player, it took them 4 minutes to take down a barracks! lol The barracks had 96% pierce armor. This is okay to me. lol By stats alone, it would take them 12 minutes to take down a fortress, but in reality they never would because they would be killed by the arrows shot by the fortress, as the fortress outranges them, rightly so.

Edited October 29, 2017 by wowgetoffyourcellphone

[Nescio]

2 hours ago, elexis said:

> Also, pierce damage damaging buildings is really unrealistic.

Didn't we come to that conclusion several times already, always increased the Pierce damage and are now at 99% and it's still too few? Anything speaking against making that 100%? (I guess the formula, but perhaps we can go to 99,99%)

Units will always inflict at least 1 damage, right? If so, it doesn't really matter whether the armour is 90% or 99% or 99.99%, nor what the attacker's pierce damage is; what counts is the number of attackers and their attack rate. E.g. twenty archers need (2000/1.0)/20=100 seconds to level down a barracks.

[LordGood]

3 minutes ago, Nescio said:

Units will always inflict at least 1 damage, right?

noooooo thats how it's done in AoE but not here

[Nescio]

1 minute ago, LordGood said:

noooooo thats how it's done in AoE but not here

No? So health is not an integer? Interesting. What about other values, e.g. resources?

[LordGood]

You'll have to ask one of the programmers about that, I just know from my Ponies ascendant armor experiments. Takes a long time to put knights down with archers lol

[wowgetoffyourcellphone]

13 minutes ago, Nescio said:

Units will always inflict at least 1 damage, right? If so, it doesn't really matter whether the armour is 90% or 99% or 99.99%, nor what the attacker's pierce damage is; what counts is the number of attackers and their attack rate. E.g. twenty archers need (2000/1.0)/20=100 seconds to level down a barracks.

 

9 minutes ago, LordGood said:

noooooo thats how it's done in AoE but not here

 

7 minutes ago, Nescio said:

No? So health is not an integer? Interesting. What about other values, e.g. resources?

 

3 minutes ago, LordGood said:

You'll have to ask one of the programmers about that, I just know from my Ponies ascendant armor experiments. Takes a long time to put knights down with archers lol

LordGood is right. 

[wraitii]

Health is a decimal number in the code exactly so that we can handle fractional damage and avoid the "1 HP damage minimum" from AoE 2. On the GUI it's shown as an integer though.

 

I believe resources are the same.

[wowgetoffyourcellphone]

3 minutes ago, wraitii said:

On the GUI it's shown as an integer though

Thank god. lol

[elexis]

Except in some places. Also implying some issues like units walking around with 0HP (also carrying 0 resources) or hellenic buildings with 10% bonus having 3301 HP. Not really urgent luckily.

[fatherbushido]

49 minutes ago, Nescio said:

No? So health is not an integer? Interesting. What about other values, e.g. resources?

health is decimal

resources are integers (not need to cut the unit in which we count it, else we had to count it in another unit, if someone think what I say didn't mean anything, he is right, that's why we have to have integer resources)

[elexis]

(at leat I saw a resource-carrying tooltip stating 0/10 wood being transported. Edit: some time ago before a22)

[wraitii]

I think the number carried by units is a decimal, but the actual number of resources is supposed to be an integer - all costs are integers. But JS doesn't really have a hard difference there.

[Stan`]

Actually I think in newer versions of JavaScript you can have typed arrays. And it would seem that integer typed arrays are faster to iterate through.

Edited October 29, 2017 by stanislas69

[fatherbushido]

Ah yes resources have different meanings.

And there is missing sim rounding in some places (after modifications). (Missing, I don't say if it's right or bad)

 

 

[Palaiogos]

Is health a floating point value? How do you get it to not due weird things, as floating point values get messed up a lot of times.

[leper]

When anyone talks about something decimal in the simulation, which for reasons of being shorter and people being lazy and everyone knowing better (at least I hope the last one is true), they mean fixed point. That's the reason why we have cross-platform (or actually same-platform in some cases) multiplayer and a reproducible simulation which is quite important.

As for float getting messed up, well only if you decide that you like NaNs, apart from that the remaining issues are some platform specific functions, actually used internal precision for floating point operations, and rounding modes, most of which one can work around, which is partially done by JS itself and implementations adhering to it, or us taking some care to fix the other parts (the first paragraph is about the C++ part of the simulation, though for what we use floats in the JS simulation they should be considered equivalent to fixed point).

So yes, internally (simulation that is, though I think the rounding is only done in the GUI, as opposed to the simulation side that provides those values to it) health is a floating point value.

[tuk0z]

On 04/08/2017 at 1:33 AM, temple said:

Here's a graph of the differences in average damage per second between a21 and a22 for the basic ranged units. Their accuracy depends on the target's footprint

Really nice visualization work temple! Sorry if this misses any serious maths, here's just another video illustration of the former cav skirmishers' « accuracy » in a21 - At 2:00 precisely - Valir'hant :o]

 

Edited November 20, 2017 by tuk0z