MadTommy
10-09-08, 10:53
Written by [R-CON]Wolfe - reposted here for thse that did not see this... some very useful info in here.
The following is a re-post of something I wrote in the tester forums. I hope this helps your understanding of why it was changed the way it was, as well as some tips on how to make achieve better aim. It's a long read, but it explains a lot.
Oh and one extra note: Your accuracy is 10x better when scoped than non scoped. This is true for both iron sights and scoped weapons. If you fire your weapon un-scoped, your range is limited to about 10-20 meters maximum.
In regard to deviation:
Quote:Originally Posted by [R-CON]Wolfe
Weapon performance vs. Weapon performance in the hands of a soldier.
BF2 does not currently use weapon sway or body momentum which are two critical factors when determining weapon performance in the hands of a soldier. Without these factors, the next best thing is to simulate them by incorporation into the weapon deviation itself. This means that cones of fire will be wider and settle times will be longer.
This is also where wiki-wannabees and military advisers get their panties in a knot; they expect the coding to match the realism of the weapon while completely ignoring other factors that allow a realistic weapon to be used in unrealistic ways.
Some people want weapon performance to be represented by the coding, and weapon performance in the hands of a soldier to be represented by their mouse. The problem with that style of coding is that it allows players to aim/shoot as fast as they can move their mouse: turn your mouse 1/4 inch and your 180lb. soldier turns 360 degrees with little or no loss of accuracy. If you're trying to mod a super-twitch run'n gun, then this is the way to go and is exactly what we saw in PRv6. Sure, the weapon coding itself is realistic, but the weapon performance in the hands of a soldier extends into the realm of fictional super heroes. If that's what we want, then let's be honest about it and re-skin the uniforms to include capes.
All kidding aside though, deviation has to match soldier behavior, not weapon behavior. Otherwise, you'll never achieve any degree of realistic soldier combat.
Engagement distances.
The real-life M16 might have have an effective range of 500 to 800 meters, but that doesn't mean those numbers should be plugged directly into PR. Why? Because in-game, most engagements take place in relative close proximity to one another; usually 100 to 200 meters. If weapons had an effective range of 500 to 800 meters,then players can easily headshot at 200 meters.
Also, the maximum draw distance is only about 900 meters. This limits the sniper rifle to the same distance; 900 meters. If assault rifles had a range of up to 800 meters, that means the sniper rifle is only moderately better than a scoped assault rifle. This also opens the door for the assault rifle to engage and defeat a sniper at most ranges because despite less accuracy, the assault rifle can place more rounds down range, suppress the sniper's vision, and eventually score a hit/kill. This is a common scenario in PRv7 and it's a direct result of the assault rifle having too great a range for the dynamics of the game.
So, understanding the 900 meter limit, I decided to give assault rifles a 300 max range; basically allowing a 25% to 50% chance to hit. At 400 meters, you can suppress but anything beyond 400 meters is pure luck. This creates some breathing room for the marksman (effective up to 350-400) and the sniper (effective up to 950).
Effective ranges by weapon.
Based on the above, here are the ballpark effective ranges for each weapon type:
0-10 meters: Support Gunner (un-deployed version)
0-50 meters: Sub machine gun, Pistol
0-200 meters: Rifleman (Iron/reflex sight)
30-300 meters: Rifleman (scoped)
30-350 meters: Support Gunner (deployed version)
30-400 meters: Marksman
30-600 meters: LAT
50-950 meters: Sniper, HAT
In addition, every faction uses the same values for their weapon classes. In other words, the MEC assault rifle has exactly the same deviation coding as the Brit or US or China assault rifle. This is where the similarities end; their scope magnification, recoil, clip size, muzzle velocity, damage modifiers, and rates of fire are all customized to resemble their real-life counterparts. By doing this, it creates a relatively level playing field while still keeping the unique feel of each weapon.
EVEN IF, however, map sizes were increased, I would still argue against longer ranges if not only for the headshot effect. Again, if rifles are effective at longer range, then due to the BF2 engine limitations, that makes the same weapon SUPER accurate at shorter ranges and we're right back to headshots, twitching, and run'n gun.
Stances, settle times, and shift delays.
As you've noticed, all stances have been significantly increased over what they were in PRv7. This is a direct result of decreasing the effective range of the weapons which was explained above. The crouch and prone have the same modifier to help downplay prone spamming. The stand modifier is significantly higher to guard against jack-in-the-box shooting; the technique of crouching then rapidly standing/firing and crouching again. Unfortunately, BF2 does not offer any kind of deviation penalty to stance changes, so the next best thing was to noticeably decrease accuracy while standing versus crouching.
Settle times have also been significantly increased. To reach perfect accuracy, movement (walking/running) requires 2.5 seconds of settle time, and firing requires 1.0 seconds of settle time. Turning (mouse movement/tracking) requires a maximum 1.3 settle time but it's completely dependent upon how fast you turn. Following a moving target at 50 meters will not cause any turn deviation, but rapidly turning 180 degrees will. This is meant to simulate a soldier's momentum and helps against twitch shooters.
A shift delay is the wait time between trigger pulls. In PRv75, only the sniper rifle had a shift delay which prevented it from being fired faster than once every 1-2 seconds. No other weapon had a shift delay which meant that assault rifles could be single-shot fired as quickly as you can click your mouse (as much as 10 shots per second). This allowed for an exploit in the game which I won't get into here, but the bottom line is that shift delays were added to all assault rifles and pistols; basically anything that has a single-fire mode. In PRv8, weapons cannot be fired faster than is mechanically possible in real life; about 3-4 shots per second.
Putting it all together.
My ultimate goal of the new deviation model was three-fold:
1. Formulate a paper/scissors/rock hierarchy that balances weapon distance, versatility, and power.
2. Reward squad cohesion and teamwork through focused fire while punishing the run'n gun mouse-twitch Rambo.
3. Create realistic infantry combat engagements that last longer than 10 seconds and that use more than 3 bullets.
All in all, I think I succeeded at achieving those goals but it's not perfect, yet. There we so many fundamentals changed that trying to predict exactly how they will work in full-scale combat is impossible, so there's no doubt that tweaks will be necessary. But we're definitely on the right track. Combined with Chuc's excellent work on the new animations, PRv8 combat creates a more realistic tactical feel. After playing PRv8 CQB, PRv7 feels cheap and arcade-ish. It's a huge step forward.
I think people are having trouble with PRv8 because they're trying to play infantry like they did in PRv75 and that just isn't going to work. SLOW DOWN. Take your time. Don't be in such a hurry, your patience and aim will be rewarded.
In regard to settle times:
Every rifle in the game follows the same deviation settle times. To achieve maximum accuracy:
Quote:Rifles (includes both iron sights and scopes):
Per shot: 1 second
Mouse turn: 1.3 seconds (only if rapid 180deg turn)
Walking: 2.5 seconds
Marksman (same as rifleman)
Sniper:
Per shot: 8 seconds
Mouse turn: 1.6 seconds (only if rapid 180deg turn)
Walking: 8 seconds
Saw (undeployed):
Per shot: 1.6 seconds (allows 8 bullets before reaching max deviation)
Mouse turn: 1.9 seconds (only if rapid turn)
Walking: 3.3 seconds
Saw (deployed):
Per shot: 1.1 seconds (allows 12 bullets before reaching max deviation)
Mouse turn: 1.6 seconds (only if rapid turn)
Walking: 3.3 seconds
Grenadier launcher:
Per shot: none - dictated by reload speed
Mouse turn: 1.3 seconds (only if rapid turn)
Walking: 2.5 seconds
LAT:
Per shot: none - dictated by reload speed
Mouse turn: 1.3 seconds (only if rapid turn)
Walking: 2.1 seconds
HAT:
Per shot: none - dictated by reload speed
Mouse turn: 5.1 seconds (visual)
Walking: 5.1 seconds (visual)
Pistol:
Per shot: 0.2 seconds
Mouse turn: 0.6 seconds
Walking: 1.1 seconds
In regard to the sniper rifle:
Quote:Originally Posted by [R-CON]Wolfe
The primary concern was to emphasize the complete role of the sniper class, not glorify the sniper rifle itself. In other words, most players use the sniper rifle to kill everything in their field of view. By limiting each shot to 8 seconds, snipers would have to choose the priority of their targets.
It is important to note, however, that you don't always have to wait 8 seconds. Depending on the range of your target, you may only have to wait 4 seconds. Why? The sniper rifle's maximum range is about 900 meters. If you want to hit a target at that range, you'll have to wait for the cone of fire to reach its minimum point. But if your target is only 400 meters away, you will only have to wait half the time because it's half the distance. This actually creates the need for distance calculations, much like a real-life sniper would have to do but not in the same way.
Of course, the problem with an 8 second delay is that a real-life sniper doesn't have to to wait that long between shots. Once their position is set and calculated, it's very easy to place multiple shots on target with little or no delay between them. So why not allow this in PR? In this case, the concern was if the per-shot delay was something like 3 seconds at 900 meters, a sniper would only have to wait 1 second at 400 meters, giving a sniper the ability to wipe out an entire squad in 5 seconds with perfect accuracy. While that's a good thing for real-life snipers, it can ruin the gameplay in PR.
An alternate idea is to decrease the per-shot delay to about 4-5 seconds but increase the shift delay to 3 seconds (a "shift delay" determines how often a weapon can be fired). This would help resolve the unrealistically long per-shot delay while not allowing snipers to kill 1 enemy per second at shorter ranges because the shift delay would only allow the weapon to be fired once every 3 seconds.
In the end, successfully incorporating the sniper class into the mod is difficult. It's very much an "I" weapon in a "Team" game. How do you allow use of the weapon without giving the green light to lone-wolfs, skeet-shooters, or those who grab it just because they think "snipers are cool", all the while taking it out of the hands of someone who could put it to much better use. The 8 second delay was an attempt to make people think about their motivations for using the sniper rifle. Do you want to headshot everyone at 900 meters? Or do you want to intelligently choose your targets, and more importantly, provide intel to your team instead of squeezing off rounds.
All of these points were heavily debated during the coding process, and while the decision was made to use the 8 second delay, it doesn't mean that it's set in stone. If it doesn't work, it can be changed at any time. Just the same, I would encourage everyone to give it a chance. Change is always hard at first but it always seems to grow on ya. Just like the removal of the minimap and other fundamental changes, these ideas were hated a first but now we wonder why we didn't change it sooner.
The following is a re-post of something I wrote in the tester forums. I hope this helps your understanding of why it was changed the way it was, as well as some tips on how to make achieve better aim. It's a long read, but it explains a lot.
Oh and one extra note: Your accuracy is 10x better when scoped than non scoped. This is true for both iron sights and scoped weapons. If you fire your weapon un-scoped, your range is limited to about 10-20 meters maximum.
In regard to deviation:
Quote:Originally Posted by [R-CON]Wolfe
Weapon performance vs. Weapon performance in the hands of a soldier.
BF2 does not currently use weapon sway or body momentum which are two critical factors when determining weapon performance in the hands of a soldier. Without these factors, the next best thing is to simulate them by incorporation into the weapon deviation itself. This means that cones of fire will be wider and settle times will be longer.
This is also where wiki-wannabees and military advisers get their panties in a knot; they expect the coding to match the realism of the weapon while completely ignoring other factors that allow a realistic weapon to be used in unrealistic ways.
Some people want weapon performance to be represented by the coding, and weapon performance in the hands of a soldier to be represented by their mouse. The problem with that style of coding is that it allows players to aim/shoot as fast as they can move their mouse: turn your mouse 1/4 inch and your 180lb. soldier turns 360 degrees with little or no loss of accuracy. If you're trying to mod a super-twitch run'n gun, then this is the way to go and is exactly what we saw in PRv6. Sure, the weapon coding itself is realistic, but the weapon performance in the hands of a soldier extends into the realm of fictional super heroes. If that's what we want, then let's be honest about it and re-skin the uniforms to include capes.
All kidding aside though, deviation has to match soldier behavior, not weapon behavior. Otherwise, you'll never achieve any degree of realistic soldier combat.
Engagement distances.
The real-life M16 might have have an effective range of 500 to 800 meters, but that doesn't mean those numbers should be plugged directly into PR. Why? Because in-game, most engagements take place in relative close proximity to one another; usually 100 to 200 meters. If weapons had an effective range of 500 to 800 meters,then players can easily headshot at 200 meters.
Also, the maximum draw distance is only about 900 meters. This limits the sniper rifle to the same distance; 900 meters. If assault rifles had a range of up to 800 meters, that means the sniper rifle is only moderately better than a scoped assault rifle. This also opens the door for the assault rifle to engage and defeat a sniper at most ranges because despite less accuracy, the assault rifle can place more rounds down range, suppress the sniper's vision, and eventually score a hit/kill. This is a common scenario in PRv7 and it's a direct result of the assault rifle having too great a range for the dynamics of the game.
So, understanding the 900 meter limit, I decided to give assault rifles a 300 max range; basically allowing a 25% to 50% chance to hit. At 400 meters, you can suppress but anything beyond 400 meters is pure luck. This creates some breathing room for the marksman (effective up to 350-400) and the sniper (effective up to 950).
Effective ranges by weapon.
Based on the above, here are the ballpark effective ranges for each weapon type:
0-10 meters: Support Gunner (un-deployed version)
0-50 meters: Sub machine gun, Pistol
0-200 meters: Rifleman (Iron/reflex sight)
30-300 meters: Rifleman (scoped)
30-350 meters: Support Gunner (deployed version)
30-400 meters: Marksman
30-600 meters: LAT
50-950 meters: Sniper, HAT
In addition, every faction uses the same values for their weapon classes. In other words, the MEC assault rifle has exactly the same deviation coding as the Brit or US or China assault rifle. This is where the similarities end; their scope magnification, recoil, clip size, muzzle velocity, damage modifiers, and rates of fire are all customized to resemble their real-life counterparts. By doing this, it creates a relatively level playing field while still keeping the unique feel of each weapon.
EVEN IF, however, map sizes were increased, I would still argue against longer ranges if not only for the headshot effect. Again, if rifles are effective at longer range, then due to the BF2 engine limitations, that makes the same weapon SUPER accurate at shorter ranges and we're right back to headshots, twitching, and run'n gun.
Stances, settle times, and shift delays.
As you've noticed, all stances have been significantly increased over what they were in PRv7. This is a direct result of decreasing the effective range of the weapons which was explained above. The crouch and prone have the same modifier to help downplay prone spamming. The stand modifier is significantly higher to guard against jack-in-the-box shooting; the technique of crouching then rapidly standing/firing and crouching again. Unfortunately, BF2 does not offer any kind of deviation penalty to stance changes, so the next best thing was to noticeably decrease accuracy while standing versus crouching.
Settle times have also been significantly increased. To reach perfect accuracy, movement (walking/running) requires 2.5 seconds of settle time, and firing requires 1.0 seconds of settle time. Turning (mouse movement/tracking) requires a maximum 1.3 settle time but it's completely dependent upon how fast you turn. Following a moving target at 50 meters will not cause any turn deviation, but rapidly turning 180 degrees will. This is meant to simulate a soldier's momentum and helps against twitch shooters.
A shift delay is the wait time between trigger pulls. In PRv75, only the sniper rifle had a shift delay which prevented it from being fired faster than once every 1-2 seconds. No other weapon had a shift delay which meant that assault rifles could be single-shot fired as quickly as you can click your mouse (as much as 10 shots per second). This allowed for an exploit in the game which I won't get into here, but the bottom line is that shift delays were added to all assault rifles and pistols; basically anything that has a single-fire mode. In PRv8, weapons cannot be fired faster than is mechanically possible in real life; about 3-4 shots per second.
Putting it all together.
My ultimate goal of the new deviation model was three-fold:
1. Formulate a paper/scissors/rock hierarchy that balances weapon distance, versatility, and power.
2. Reward squad cohesion and teamwork through focused fire while punishing the run'n gun mouse-twitch Rambo.
3. Create realistic infantry combat engagements that last longer than 10 seconds and that use more than 3 bullets.
All in all, I think I succeeded at achieving those goals but it's not perfect, yet. There we so many fundamentals changed that trying to predict exactly how they will work in full-scale combat is impossible, so there's no doubt that tweaks will be necessary. But we're definitely on the right track. Combined with Chuc's excellent work on the new animations, PRv8 combat creates a more realistic tactical feel. After playing PRv8 CQB, PRv7 feels cheap and arcade-ish. It's a huge step forward.
I think people are having trouble with PRv8 because they're trying to play infantry like they did in PRv75 and that just isn't going to work. SLOW DOWN. Take your time. Don't be in such a hurry, your patience and aim will be rewarded.
In regard to settle times:
Every rifle in the game follows the same deviation settle times. To achieve maximum accuracy:
Quote:Rifles (includes both iron sights and scopes):
Per shot: 1 second
Mouse turn: 1.3 seconds (only if rapid 180deg turn)
Walking: 2.5 seconds
Marksman (same as rifleman)
Sniper:
Per shot: 8 seconds
Mouse turn: 1.6 seconds (only if rapid 180deg turn)
Walking: 8 seconds
Saw (undeployed):
Per shot: 1.6 seconds (allows 8 bullets before reaching max deviation)
Mouse turn: 1.9 seconds (only if rapid turn)
Walking: 3.3 seconds
Saw (deployed):
Per shot: 1.1 seconds (allows 12 bullets before reaching max deviation)
Mouse turn: 1.6 seconds (only if rapid turn)
Walking: 3.3 seconds
Grenadier launcher:
Per shot: none - dictated by reload speed
Mouse turn: 1.3 seconds (only if rapid turn)
Walking: 2.5 seconds
LAT:
Per shot: none - dictated by reload speed
Mouse turn: 1.3 seconds (only if rapid turn)
Walking: 2.1 seconds
HAT:
Per shot: none - dictated by reload speed
Mouse turn: 5.1 seconds (visual)
Walking: 5.1 seconds (visual)
Pistol:
Per shot: 0.2 seconds
Mouse turn: 0.6 seconds
Walking: 1.1 seconds
In regard to the sniper rifle:
Quote:Originally Posted by [R-CON]Wolfe
The primary concern was to emphasize the complete role of the sniper class, not glorify the sniper rifle itself. In other words, most players use the sniper rifle to kill everything in their field of view. By limiting each shot to 8 seconds, snipers would have to choose the priority of their targets.
It is important to note, however, that you don't always have to wait 8 seconds. Depending on the range of your target, you may only have to wait 4 seconds. Why? The sniper rifle's maximum range is about 900 meters. If you want to hit a target at that range, you'll have to wait for the cone of fire to reach its minimum point. But if your target is only 400 meters away, you will only have to wait half the time because it's half the distance. This actually creates the need for distance calculations, much like a real-life sniper would have to do but not in the same way.
Of course, the problem with an 8 second delay is that a real-life sniper doesn't have to to wait that long between shots. Once their position is set and calculated, it's very easy to place multiple shots on target with little or no delay between them. So why not allow this in PR? In this case, the concern was if the per-shot delay was something like 3 seconds at 900 meters, a sniper would only have to wait 1 second at 400 meters, giving a sniper the ability to wipe out an entire squad in 5 seconds with perfect accuracy. While that's a good thing for real-life snipers, it can ruin the gameplay in PR.
An alternate idea is to decrease the per-shot delay to about 4-5 seconds but increase the shift delay to 3 seconds (a "shift delay" determines how often a weapon can be fired). This would help resolve the unrealistically long per-shot delay while not allowing snipers to kill 1 enemy per second at shorter ranges because the shift delay would only allow the weapon to be fired once every 3 seconds.
In the end, successfully incorporating the sniper class into the mod is difficult. It's very much an "I" weapon in a "Team" game. How do you allow use of the weapon without giving the green light to lone-wolfs, skeet-shooters, or those who grab it just because they think "snipers are cool", all the while taking it out of the hands of someone who could put it to much better use. The 8 second delay was an attempt to make people think about their motivations for using the sniper rifle. Do you want to headshot everyone at 900 meters? Or do you want to intelligently choose your targets, and more importantly, provide intel to your team instead of squeezing off rounds.
All of these points were heavily debated during the coding process, and while the decision was made to use the 8 second delay, it doesn't mean that it's set in stone. If it doesn't work, it can be changed at any time. Just the same, I would encourage everyone to give it a chance. Change is always hard at first but it always seems to grow on ya. Just like the removal of the minimap and other fundamental changes, these ideas were hated a first but now we wonder why we didn't change it sooner.