Anvil Game Studios

Author Topic: For Devs | Urgent change | Musket accuracy  (Read 6929 times)

0 Members and 1 Guest are viewing this topic.

Offline Timo420

  • Ordinary Seaman
  • *
  • Posts: 25
  • 4. Ostpreußisches Infanterie-Regiment Nr. 5
  • Faction: Neutral
  • Nick: [Nr5]Kpt. Timo
Re: For Devs | Urgent change | Musket accuracy
« Reply #30 on: September 24, 2017, 04:42:57 pm »
I mean I only really have problems with bullet drop, accuracy is well bad but you have to take in account that in real life there where factors that made aiming alot harder, like the sounds, smoke and people dying next to you so I don't have a huge problem with the innaccuracy.

Offline TooL69

  • Surgeon's Mate
  • *
  • Posts: 105
Re: For Devs | Urgent change | Musket accuracy
« Reply #31 on: September 24, 2017, 06:22:30 pm »
And what is the speed of the bullets in the game?

------

“Benjamin Robins obtained muzzle velocities between 1425 fps (434 m/s) and 1700
fps (518 m/s) in 1742 with a ¾ inch (19.05mm) diameter ball and 45 inch (1.143 m)
long barrel. A century later Captain Alfred Mordecai studied gunpowder used for an
English musket and recorded an average muzzle velocity of 1561 fps (476 m/s) and
that 1477 fps (450 m/s) was adopted as the minimum velocity for proof of powder
when using 10 grams of powder, whilst 7.5g of powder achieved a velocity of 1550 fps
(472 m/s” (Roberts, 2008). After much research, it was concluded that data indicates
that the musket ball would have probably averaged, at the muzzle, about 1500
fps.(457 m/s).
This information is supported by Eyers (2006), in her Master’s research on the
‘Ballistics of matchlock muskets’. Eyers states that tests carried out in the 1980’s in
Austria by Krenn (1989), ((cited in Harding 1997), using small arms of the 16th, 17th
and the 18th centuries produced muzzle velocities between 450 and 500 metres per
second. This was obtained using flintlock muskets of 17 mm calibre with a powder
charge of 15 grams. Eyers concluded that 17th Century muskets, had velocities of
approximately 400-430 m/s and ranges of approximately 170-180 m when fired
horizontally.


A reference to Robinson's work:
http://spiff.rit.edu/classes/phys369/workshops/w10b/robins.html

As shown on picture 258.627....
The theory is correct but not in all the games. For example. in NW its much higher  and here is lower + there is gravity that is impacting the bullet speed and trajectory.

To add yes in the 17th Century the muskets had more grams of powder because then it didn't matter will the bullet actually hit the target... rather scare or to hit a big bunch of people in one place. It had higer shooting distance but the moving of a bullet was terrible. After working on barrels in Napoleonic and post Napoleonic era they orientated on bullet hiting the target for sure then rather gambling if it will hit anyone.

The muskets of the 30-year war were fairly light (4.5-5 kg) and had the same caliber as that of Brown Bess.
http://www.engerisser.de/Bewaffnung/weapons/Matchlockmusket.html

The charge of gunpowder at half the weight of the bullet can be explained by the need to pierce steel cuirasses of pikemen. In the 18-19th century, the charge of gunpowder was reduced to one-third the weight of the bullet. Because the probability of encountering an enemy in the armor has drastically decreased (but this is just my theory).

Note the muskets of the 17th century had a developed sight. According to Graz, matchlock musket has an accuracy of 30 MOA per 100 meters. This is very good.
Spoiler
[close]
« Last Edit: September 24, 2017, 06:32:00 pm by TooL69 »

Offline CrocHunter

  • Master's Mate
  • *
  • Posts: 30
  • CavGF Co-owner
  • Faction: Neutral
Re: For Devs | Urgent change | Musket accuracy
« Reply #32 on: September 24, 2017, 07:46:48 pm »
I think what Leon means by those 2 centimeters is how high is his point of aim above the target to hit at the distance of 60 meters. But while the crosshair is only 2 centimeters above the target from the perspective of the shooter, the point of aim is actualy several meters above the target which lead to a unrealistic drop from the point of aim to the target.
This picture with very rough measures shows how enourmous the drop is at the range of aprox 90 meters
Spoiler
[close]
There is some consistancy to the offset of the point of aim compared to the point of impact at set ranges and knowing these even with the random spread can lead to hits even at further distances. The issue may not be in the bullet's trajectory (as shown by Leon's image) as much as in the fact that the firearms at this stage of the developement are not at all sighted to any distance and that the projectile's point of origin is not aligned with the barrel but in a downard pointing angle, this would results in need to aim above the target even at close distances as is currently happening in the game. Changing the bullet's trajectory point of origin to zero with the aiming crosshair at a set distance could fix the currect issues.
Nr3 "Griff's Own" Husaren & Infanterie Regiment
Rittmeister Sernis Thausten

Offline Leon

  • Able Seaman
  • *
  • Posts: 52
  • Leon
  • Faction: French Empire
  • Nick: LEON
Re: For Devs | Urgent change | Musket accuracy
« Reply #33 on: September 24, 2017, 09:11:11 pm »
I think what Leon means by those 2 centimeters is how high is his point of aim above the target to hit at the distance of 60 meters. But while the crosshair is only 2 centimeters above the target from the perspective of the shooter, the point of aim is actualy several meters above the target which lead to a unrealistic drop from the point of aim to the target.
This picture with very rough measures shows how enourmous the drop is at the range of aprox 90 meters
Spoiler
[close]
There is some consistancy to the offset of the point of aim compared to the point of impact at set ranges and knowing these even with the random spread can lead to hits even at further distances. The issue may not be in the bullet's trajectory (as shown by Leon's image) as much as in the fact that the firearms at this stage of the developement are not at all sighted to any distance and that the projectile's point of origin is not aligned with the barrel but in a downard pointing angle, this would results in need to aim above the target even at close distances as is currently happening in the game. Changing the bullet's trajectory point of origin to zero with the aiming crosshair at a set distance could fix the currect issues.
Im sorry i didnt check pic good first time now i see what you ment to say
Yes just put your character into fpas mode and then mesure it
3rd peraon camer gives difrent perspective
« Last Edit: September 24, 2017, 09:13:35 pm by Leon »

Offline CrocHunter

  • Master's Mate
  • *
  • Posts: 30
  • CavGF Co-owner
  • Faction: Neutral
Re: For Devs | Urgent change | Musket accuracy
« Reply #34 on: September 24, 2017, 10:14:18 pm »
Did a both FPS and TPS test on the same place, I shot a little bit to the right so I could see where the shots are landing in grass.
Spoiler
[close]

Also did a small test for FPS accuracy and spread. I dont know the exact distance, but it was shot from the back of french A spawn on the desert map. I shot my whole ammo bag.
Spoiler
[close]

As I wrote in my first post, there seems to be some kind of projectile balistic trajectory with a random large horizontal, small vertical spread. The projectile path seems to be not paralel with the barrel of the musket but seems to be in a downard angle. A solution would for example to be to sight the muskets to a certain range, let say 70 meters (only for this example), so when you aim with the crosshair at something 70 meters away it would hit withing that crosshair with variation by spread. Easiest solution (only assuming, I dont know how exactly they did it in holdfast), would be to rotate the bullet point of origin trajectory upward so it would cross the crosshair at that set distance. This would do away with the unrealistic high aiming that is present.
« Last Edit: September 24, 2017, 10:39:05 pm by CrocHunter »
Nr3 "Griff's Own" Husaren & Infanterie Regiment
Rittmeister Sernis Thausten

Offline Feldmarschall_Ben

  • Petty Officer
  • *
  • Posts: 27
  • Gott schütze den Kaiser!
    • My Imperial Army forum
  • Faction: Neutral
  • Nick: FeldMrshl_Ben
Re: For Devs | Urgent change | Musket accuracy
« Reply #35 on: September 24, 2017, 10:43:43 pm »
I think it'll be beneficial for the dedicated devs of this game to hear what I and a lot of veteran players of musket games have to say. Since days of the original battlegrounds mod, I've seen this genre change and develop and so here is the most important aspect of the game that needs revision (yes, more than melee combat). I suggest reading the entirety of my post before commenting.

MUSKET INNACURACY AND UNWIELDINESS

The bedrock and bread and butter of a musket game, and a necessary component that needs to be done right and also feel satisfying. Muskets of the napoleonic era are no where near as innaccurate as protrayed in films and videogames of the era.

Trained soldiers could hit a human-sized target at 70-90m and volleys to inflict maximum damage were aimed to be conducted just under this range at 60m. Various soldier accounts of the era point to unwiedly accuracy becoming too big a burden at and beyond 150m.

In the 1814 To All Sportsmen, Colonel George Hanger wrote, “A soldier’s musket, if not exceedingly ill-bored, will strike a figure of a man at 80 yards; it may even at a hundred; but a soldier must be very unfortunate indeed who shall be wounded by a common musket at 150 yards, providing his antagonist aims at him; and as to firing at a man at 200 yards with a common musket, you may as well fire at the moon and have the same hope of hitting him."

The reason for closing the distance was not only to get within the musket's kill distance, but to strike fear through the visual and auditory threat a massed enemy force would present.

HOW DOES THIS RELATE TO HOLDFAST?

At the moment the preferred engagement distance and the distance you can be confident (key word: confident) in hitting an enemy (half the time) is around 8-10 meters. The bulletdrop mechanic needs to be redone. For every 15 meters you need to adjust upwards about 1 full reticle. Meaning at 20 meters you need to aiming as if you are lobbing a grenade by your enemy's feet. At 50 meters you need to be aiming at distant clouds for your shot to come back to earth and land near your target. I don't seek to offend, but this would not be the case if someone working on the game would have ever fired a musket of the era in question, ever seen someone fire a musket, read specifications and tactics relating to these weapons, or sought out first hand accounts on how these muskets functioned and were utilised. But that's okay, the drop can be easily fixed by straightening the trajectory and implementing slight bulletdrop beyond 60-90 meters.

The bulletspread of the muskets is another aspect of shooting that needs to be corrected. At around 10m distance the spread seems to be 1-2 meters. Double the distance, and this increases to 3-5 meters. Again, painfully unrealistic and unrepresentative of napoleonic era weapons. Muskets had around a 1/3 meter grouping at 50 meters meaning that a shot at an enemy 50 meters away nearly always delivered a hit. Beyond this distance, the grouping increases slightly and at 100m a soldier could still expect his shots to land within 1-3m of one another.

In other words, Holdfast has a real musket's spread at 100m come into effect at around 15 meters. You can thank blind luck beyond 10 meters if you land a shot. Before any claims are made by someone that they can land shots from 25-40m consistently, the answer is no. You cannot consistently shoot at that distance in Holdfast. I've topped the scoreboard various matches by adjusting the reticle for ridiculous bulletdrop, shooting into masses, and firing quickly. However, most of the kills beyond 20 meters have been on a soldier 2-3 meters to the left or right of my target and rarely directly where I was aiming as should it should be due to the inconsequential spread muskets have within 50 meters distance. Realistically, this spread of over 3 meters would happen at around 90-100 meters. This too can be easily fixed by tightening the shooting pattern significantly according to real musket spreads.

HOW DOES THIS CHANGE HOLDFAST'S GAMEPLAY?

There is enough argument from a realism standpoint to warrant a change to the shooting mechanics. At the moment, the game feels like a miniature, arcade FPS/TPS game set in the musket era because of the 10-20 meter engagement distance and the ineffectiveness of the muskets and rifles. The unit portrait and minimap do not help remedy this at all. However, having been an avid lover of community line battle and siege events for a long time, far before the warband NW mod, I must say that the engagement distance portrayed in this game is a real threat to the atmospheric and strategic depth that could and should be present in this game. Even from watching early event gameplay I noticed how the musket's effective range of around 10 meters causes lines to, after frustratingly trying to trade blows at 30 meters and beyond, close the distance to fire a determining volley and end what could've been an incredible back and forth mass volley engagement if only the muskets had realistic drop and spread. After playing the game for a day and a half straight, I can confirm my doubts.

Another point that must be taken into consideration is that cavalry will not be punished for blind charges at enemy infantry. Closing the gap will be incredibly cost-effective for a cavalry force seeing as the musket's spread will not be seen even as a mild deterrent.

I congratulate you for sticking with me, hopefully the developers and community can come together on this issue and see that, for realism and gameplay purposes, this needs to be labeled as high priority on the early access to-do list right with a melee rehaul.

A strong shooting system must come first and foremost in this type of game. Good shooting and melee systems will propel this game forward while badones will hold it back even if new factions, weapons, and cavalry are introduced. Thankfully, it's easy to fix what's in place because of the base game the developers worked hard to create.

+1, especially the part about the gameplay.

Definately on your side on regards of the muskets accuracy!
Valid points which strenghten the gameplay in my opinion. Especially regarding linebattle events this would cause the game to be even more dynamic, a feature I´d really appreciate to avoid campers. I would definately like to see your suggestions come true.

The Imperial Army defends the honor of his Imperial Majesty, the Kaiser of Austria.

Offline Charles Caldwell

  • Regiment Leaders
  • Master's Mate
  • *
  • Posts: 203
  • Retired Founder of the SLRN [Naval Action]
  • Faction: British Empire
  • Nick: [1st L KGL]
Re: For Devs | Urgent change | Musket accuracy
« Reply #36 on: September 24, 2017, 11:34:14 pm »
I love the fact that its not just, point, click and kill. The random element of the ball is not only realistic but adds to gameplay imo. As mentioned, if we tinker too much with the accuracy then games will end quicker and we'll also end up with 125 snipers camping. Its going to be difficult for the Devs to balance, but currently its fun not knowing if this shot is going to be a kill or miss.


Offline Leon

  • Able Seaman
  • *
  • Posts: 52
  • Leon
  • Faction: French Empire
  • Nick: LEON
Re: For Devs | Urgent change | Musket accuracy
« Reply #37 on: September 24, 2017, 11:54:19 pm »
Did a both FPS and TPS test on the same place, I shot a little bit to the right so I could see where the shots are landing in grass.
Spoiler
[close]

Also did a small test for FPS accuracy and spread. I dont know the exact distance, but it was shot from the back of french A spawn on the desert map. I shot my whole ammo bag.
Spoiler
[close]

As I wrote in my first post, there seems to be some kind of projectile balistic trajectory with a random large horizontal, small vertical spread. The projectile path seems to be not paralel with the barrel of the musket but seems to be in a downard angle. A solution would for example to be to sight the muskets to a certain range, let say 70 meters (only for this example), so when you aim with the crosshair at something 70 meters away it would hit withing that crosshair with variation by spread. Easiest solution (only assuming, I dont know how exactly they did it in holdfast), would be to rotate the bullet point of origin trajectory upward so it would cross the crosshair at that set distance. This would do away with the unrealistic high aiming that is present.
Yes you are right i tested it iv shoot and saw that the project. Dosent start from the barrel. It starts about few cm to the right of the barrel.
Anyways this i think should be fixed so bullet comes straight from barell but other things i dont minde. For me bulletdrop is somehow natural.
Was pleasent time to discuse this things with you gents!

Offline CptBiym

  • Ordinary Seaman
  • *
  • Posts: 14
Re: For Devs | Urgent change | Musket accuracy
« Reply #38 on: September 25, 2017, 12:22:55 am »
Interesting discussion Leon and Croc, but more than that, helpful for me and the devs to understand exactly how the accuracy is busted. I apologize for any misunderstanding. I have not been too active here but am regularly posting my thoughts on the steam discussions as there are many users that simply don't understand the impact this has on the overall quality of the game.

I'm glad that a lot of new and veteran players can see the reasoning of my argument and why it was put forth with such fervor.

Charles Caldwell, how is it possible that you are a lieutenant in a rifles company and say the ballistics are currently realistic? You can say that you subjectively enjoy the current musket mechanics, but don't mention realism in the same breath. Please read my original post that way you and your fellow clanmates at least understand the ranges at which muskets, and in extension rifles, operated. Striking a balance between the current ridiculous spread at 20m and the realistic musket spread of 1-3 meters and bullet drop of 60cm at 100m will not have everyone become snipers. Compromising and having a real musket's 100m spread and drop come into effect at around 50-60m (rather than currently at 15-20m) is one way to solve this issue and make combat in public servers and linebattle events closer to reality and far more rich and rewarding.

Good day

Offline CrocHunter

  • Master's Mate
  • *
  • Posts: 30
  • CavGF Co-owner
  • Faction: Neutral
Re: For Devs | Urgent change | Musket accuracy
« Reply #39 on: September 25, 2017, 02:39:15 am »
So I got myself a server for further testing and to see the trajectory as shown by the game's built-in trajectory render and this is what I found.

I repeated the shots from my first and second post as that is a nice spot to shoot from.
First some notes, the point of origin for the bullet's trajectory is the crosshair, this is why on this following image, the point of origin is a little bit above, to the right and a bit behind the character. The angle at which I took the screenshot make it look more behind the character than it actualy is.
Spoiler
[close]
As the image shows, my theories from previous two post are proven wrong. The trajectories and points I have indicated on the image may be a little bit off but I think they speak for themselves.

As I noted previously, the bullet's trajectory originate in the crosshair, given that the crosshair is in a different spot, depending on the player's gameplay perspective a different trajectory will be made even when shooting from the same spot with aiming at the same target. Please note, that the trajectories goes differently due to the spread, but the point of origin of the shots is visibly different.
Spoiler
[close]

Given this realisation, I went to test a potentional exploit and found it to be existing. In the following picture, you will see the shooting player being hidden behind a wall, with the musket clearly pointing into the wall, yet still being able to shoot at targets behind the wall.
Spoiler
[close]

As previously noted, given that the trajectory's point of origin is different between FPS and TPS, a first person player will be at a disadvantage. Please note, that in the following screenshot, the shooter is standing in the same place and aiming at the same spot as in the TPS picture. The shot got stopped by the wall.
Spoiler
[close]

As a conclusion these observations were made quickly, so there may be some margins of error when taking shots or when taking screenshots as it is not easy to old the right mouse button, the freelook key and pressing F12 to make screenshots.
Nr3 "Griff's Own" Husaren & Infanterie Regiment
Rittmeister Sernis Thausten

Offline Leon

  • Able Seaman
  • *
  • Posts: 52
  • Leon
  • Faction: French Empire
  • Nick: LEON
Re: For Devs | Urgent change | Musket accuracy
« Reply #40 on: September 25, 2017, 06:31:03 am »
Damn i lost a track from all this long quotes:D
Yes as you can see you can shoot over the wall  while aiming trough it. That should be changed and also by meaning that make bullet fall a little bit smaller. Not a bunch but a little bit. + as i said trajectory of bullet from barell.

Hint: use regular infantry becuse  they have "normal" settings for shooting.
Light and rifles have if i am correct slightly better accuracy.

And you could of just asked i would give you pass to my server wich i creates for tests and reg:D
Best 10 euros ever spended

Offline Lecourbe

  • Regiment Leaders
  • Master's Mate
  • *
  • Posts: 491
  • Par le fer quand le feu manque !
    • 61e Régiment
  • Faction: French Empire
  • Nick: [61e] Lecourbe
Re: For Devs | Urgent change | Musket accuracy
« Reply #41 on: September 25, 2017, 07:28:45 am »

Offline Charles Caldwell

  • Regiment Leaders
  • Master's Mate
  • *
  • Posts: 203
  • Retired Founder of the SLRN [Naval Action]
  • Faction: British Empire
  • Nick: [1st L KGL]
Re: For Devs | Urgent change | Musket accuracy
« Reply #42 on: September 25, 2017, 09:27:22 am »
Clearly the mechanics of the muskets and rifles in this game needs looking at, however the overall accuracy of the weapon perhaps should not be overly tweaked...

Hows this for a sobering thought:

Quote
W.Müller author of Elements of the Science of War (1811) recorded the following results:

In a trial of well trained & competent men, fired at targets representing a line of 'Cavalry' (Not troops - so a greater target) at 100 yards.

The hit ratio at this range was a surprising 53% success rate, just over half the shots hit their target.

At 300 yards, it fell to 23%.....disappointed huh!

It was commented that 'Ordinary' soldiers fared far worse in the trials.

I think this shows that those players expecting to always hit their targets with perhaps a revise hit system SHOULD be disappointed.
« Last Edit: September 25, 2017, 10:52:00 am by Charles Caldwell »


Offline CrocHunter

  • Master's Mate
  • *
  • Posts: 30
  • CavGF Co-owner
  • Faction: Neutral
Re: For Devs | Urgent change | Musket accuracy
« Reply #43 on: September 25, 2017, 10:02:24 am »
Thank Leon for the offer but I hosted the server localy from my computer, make testing a bit easier.

I did further testing and used other classes to see how it behaves and I am shocked about what I found.
I set up my position next to the cannon and used the split between the mountains in distance and the skybox as my aiming zone, I used a tree in the distance to target horizontaly.
Please note that on this following picture I moved a bit to the left after shooting to check the trajectory, which hit at the distance of 304 meters.
Spoiler
[close]
Observing the displayed stats in the top right corner I noticed several variations in values.
For the musket, the values are:
Muzzle velocity: 339.5, 298.5, 284.6, 264.5, 323.6, 319.7, 352.2, 328.5, 319.7, 327.7
Gravity:28.5,65.6,71,32,42.5,74.1,46.6,60.5,27.2,58.7
Impact distance:304.7,178.7,171.1,237.5,258.1,191.8,246.4,196.7,286.9,196.2

For the rifle, the values are:
Muzzle velocity: 126.4,126.6,112.8,128.3,126.8,110.1,126.6,116,127.2,113.3
Gravity:10.9,10.8,10.5,10.2,10.9,11,10.6,10.9,10.3,10.2
Impact distance:176.4,189.3,168.9,191.9,177.1,154,176.8,173.9,177.8,169.9

You can see from this data sample:
That "muzzle velocity" on the musket varied between 264.5 - 352.2 while on the rifle it varied between 110.1 - 128.3.
That "gravity" value on the musket varies a lot, with values between 28.5 - 71 while "gravity" value for the rifle was between 10.2 - 11.
That "impact distance" for the musket varied between 171.1 - 304.7 while on the rifle it varied between 154 - 189.3

From this data sample I can only conclude that given the overall consistency of the rifle's values a rifleman will have a predictable shooting pattern while his rigle actualy underperform in range compared to the musket which on the other hand have values all over the place and while it can shoot further, the random factor make shooting the musket a pain.

These trajectories are a sad joke and this innacuracy cannot be reasonably argued. I have seen some player here and on steam forums arguing that innacuracy is good as it prevent the game from turning into a "camping match" and put emphasis on melee which according to some was so prevalent in that period (an opinion based on games, movies and not actual sources). While a certain deviation from reality is needed to make the game playable and given this game dont atempts to be ArmA: Napoleonic Wars having values that are not exactly realistic is ok. But what is presented in this game at this moment cant even be called "arcadey", its simply wrong.
This level of innacuracy is bad regardless of what your historical perception is, if or when cavalry is implemented the inability of infantry to defend itself from cavalry apart of depending on bayonets would lead to cavalry dominating the battlefield. Even in the NW DLC for Warband where precision is higher and melee responsive infantry is easily killed by cavalry.
« Last Edit: September 25, 2017, 10:11:38 am by CrocHunter »
Nr3 "Griff's Own" Husaren & Infanterie Regiment
Rittmeister Sernis Thausten

Offline Charles Caldwell

  • Regiment Leaders
  • Master's Mate
  • *
  • Posts: 203
  • Retired Founder of the SLRN [Naval Action]
  • Faction: British Empire
  • Nick: [1st L KGL]
Re: For Devs | Urgent change | Musket accuracy
« Reply #44 on: September 25, 2017, 11:11:51 am »
Look I'm sure the ballistics in Alpha will be constantly be tweaked and for the better, and we as players will adapt with every change... but I dont appreciate the venom, bordering anger at which to 'Urgent Change advocates' put your cases forward. Chill guys!!!

Currently the game is more than playable, and the level at which you guys have dissected the ballistics is beyond the will or want of most players. However Im sure the Devs will look at your findings and take all views into consideration. Currently for me I hit most things I want (with a great deal of practice) and with that enjoyable random element as previously stated (whether Realism or subjective). Would I like to be more accurate hell yes, but not so much as to corrupt the gameplay too much. I hope you can see beyond stats and figures to appreciate this view. ;)