Conceptualising Deviation
Posted: 2008-09-04 11:30
This thread is not a rant/***** thread about deviation. I want to discuss the theory and logic behind deviation of all sorts in this thread, so that it can later be linked as the reference point for all subsequent threads questioning why deviation needs to exist.
***
Before we continue, it must be understood that it is not possible to code weapons sway or swing into the BF2 engine.
Therefore, the possibilities to reflect inaccuracy are as follows:
Random Deviation.
Random deviation with a crosshair type thing that shows your relative accuracy (gets bigger when you're inaccurate, like the HAT) only while scoped.
Perfect accuracy, with a scope animation that moves too much to be used for perfect shooting (until you hit a target, and then back every shot up to the same place).
Why aren't soldiers perfectly accurate?
On a battlefield, numerous factors implicit in using a rifle, in the battle field and related to the condition of the individual soldier negate the ability to fire exactly where intended. They are discussed here.
Firstly
Fundamental differences between actually shooting a rifle, and aiming in a videogame;
The aiming of a rifle isn't as easy as moving a mouse over an icon. Firstly, a rifle has inertia - it is a relatively large object which requires extreme control to point in an accurate way. Secondly, the control mentioned in the point before must be generated by balancing the rifle at two different ends with different arms - this is a far less stable arrangement than a mouse already rested on a flat, smooth surface has. Thirdly, as the rifle is dependant on the movement of the entire body (which cannot itself be stabalised), and not just the wrist, involuntary movements (breathing, muscle shakes (accentuated by combat stress and adrenaline), and even heartbeat) detract from the stability of the rifle, meaning that a soldier may not be able to aim the rifle consistantly where he wants to. Fourthly and finally, some points which are easy to aim at as a PR avatar are actually physiologically impossible or difficult to aim at in real life (high/low angle shots being two good examples, but firing flat from just behind the crest of a hill is another).
This point may be practically refuted by pointing out that Tanks, APCs Light Vehicles and Aircraft have no deviation whatsoever, despite being at the mercy of many of the same effects, such as inertia, variations in controlability in different weather conditions and so on.
This point may secondarily be marginally refuted by pointing out that a soldier is more able to understand when he is able to shoot accurately, and will often forgo shooting until he knows he does have a perfect shot, something the game does not model at the moment (discussed in the third section).
Secondly
Differences between the Project Reality gameworld, and a real battlefield;
On a real battlefield numerous factors negate an individual soldier's ability to fire effectively. Disorientation from explosions, fog of war (no minimaps, remember?), unstable firing surfaces, realistic ballistics (bullets actually don't fly straight, or even on a parabolic arc, while different ammunition behaves in very different ways in differing conditions that the soldier may not have dope charts for), sights may (rarely) be incorrectly zeroed, guns may jam (not firing at all, on average reducing the number of shots fired on target) and finally visual distortions caused by heat are all factors on the battlefield that limit the ability of a soldier to shoot exactly where he intends. This list is by no means complete.
Thirdly
The state of the individual soldier on the battlefield;
This is the only piece of logic behind deviation which is contestable. I will explain it, and then explain the counter-arguement.
When a soldier is on the battlefield, his ability to make rational, intelligent decisions is limited by the high levels of stress inherent to the environment. Additionally, as there are no "respawns" his willingness to move out of cover to aim a perfect shot is likely to be much lower than it is in a video game. Because of these factos, a soldier on the battlefield isn't as likely to care about whether his shots are hitting the target or not, so long as the enemy is being suppressed, he is doing his job, and he is not under unnecessary danger. This is why the number of bullets fired to total kills remains so low - in a combat, individual soldiers are not concerned about their kill count, or KDR, normally speaking they simply want to survive, and the best way of doing that is to shoot at anything and everything so it can't shoot back.
The problem with this argument is that by modelling the individual soldier's tactical predilictions, the player's role in the game is usurped. The entire point of being the player is that he retains his bodily functions, and you take over his consciousness, using him as your avatar. In many situations the right tactical choice's effectiveness is inhibited by the avatar's excessive deviation caused by this modelling - a soldier who has managed to flank an enemy and is under no stress to take a shot will be able to make his shots count more than one who is under fire, but both have the same deviation models.
Summary on the realism of deviation.
After that section, only the most die hard of anti-deviationists would argue that there should be no deviation at all in PR. There are obviously iron clad reasons why a real life soldier can't use a rifle the same way we use a mouse, and that that should be reflected in the game. After this, the question becomes more one of how much deviation there should be, and in what situations. At the moment, there is something of a community schism about how much deviation is realistic, and how this amount compares to how much there is in game. Few argued that CQB was realistic in the .7 family, and likewise, few argue that sniping is realistic in either .7 or .8. These points have their reasons that are covered in the next section, but standard infantry weapons are still open to debate. Before you form an opinion on the matter, consider the following thing that may have been responsible for not hitting your target:
BF2 Hitreg: The BF2 engine is probably the worst modern FPS engine in terms of handling whether someone shot someone else or not, which some might think would be important to a shooting game. The critical factor here is how the engine places models with relation to lag. Basically, the game assumes you have a static ping of 100, meaning that if your ping is anything lower, objects will appear in front of where their hitboxes are by the amount your ping is less than 100 in milliseconds, and if you are on a higher ping, they they will be in front of the models you see. There are console commands you can do to minimise this effect, and a perfectly legit program that helps you set them called Bf2HitFixer. Some might argue that BF2's **** hitreg alone negates the need for random deviation, as real life soldiers always hit what their bullets physically land in, however this is an invalid arguement, because this effect usually only happens when a target is moving, or has very recently moved, a stationary target at 500 meters would be fair game for a rifleman without deviation (which is unrealistic) - no dice Johnny.
Reasons that do not relate to marksmanship in Reality
The Style of Gameplay Desired by the Devs
This is probably the primary reason deviation is as significant as it is.
A common criticism of the 0.6, no deviation style of play was that the player with better reflexes could expect to win a vast majority of engagements, in spite of tactical decisions made by either side. The counterpoint to this is that ultimately some soldiers are going to have better reflexes, that will give them an edge in battle - moreover, by diminishing the role of reflexes, you are merely emphasising the "metagame" aspect discussed above - that the player's role is being diminished in order to accomodate a more "tactical" style of play. The pro-deviationists counter again by saying that reflexes are still important to the game, while the former group points out that tactics were still important before. The debate continues with many unsubstantiatable claims from both sides which do not resolve the issues.
Secondly, the dev team has consistantly talked about the style of firefighting they want to be "epic" and "intense". Certainly in real life, static gunfights tend to last for a lot longer than they do in video games, usually for hours, with only low double digits dieing on each side unless fire support assets are used. Some of the reasons for this are discussed above, in the state of mind of the soldier section.
To this end, the Dev team also introduced a suppression system where a soldier being shot at, even in cover, would be unable to return fire, and would basically be forced to hide (the system isn't perfect, but it works well, despite there being workarounds). There are a few things that could be said to counter this focus on suppression (which so far very few have tried), there are either that suppression is "too much" (soldiers aren't suppressed as severely as they are in PR, and are able to return fire more accurately (which is itself rebutted by the fact that they are also more afraid of losing their life, and are thus less likely to take the risk of dieing just to get a "kill"), and secondly, that suppression doesn't neccesarily play as large a tactic in all war as the Dev team has made out. This second arguement assumes that different armies conduct war in different ways, and so for example while a US centric tactical doctrine might focus on always pinning an enemy and then flanking him, an Australian infantry doctrine might revolve more around remaining in cover and concealment, and waiting for the enemy to enter range so riflemen can engage accurate fire, using machine gun assets to break up concentrated enemy attacks. In this second arguement, by making infantry rifles less accurate even while totally concealed and stationary (or moreso while tracking a target), you actually diminish the effectiveness of all tactical doctrines that do not conform to the Dev's vision of how combat "should" be.
(Disclaimer, this was a poor example. If someone can provide a better example of another tactical doctrine that would fit here, please do.)
What is not open to contest is that when there is no deviation (0.6), a run and gun style of play is encouraged, as there is no reason to adopt a hard-point defensive stance to wait for others to enter your killzones. When this occurs, battles do not happen anything like they do in real life, and this raises a more philosophical design question - should Project Reality impose unrealistic restrictions on the individual soldier (sights that don't represent where the gun is really pointing) in order to force the overall shape of the battle to be realistic, or vice versa?
(At present, other factors make the battle's shape significantly unrealistic, such as rally points, which are the in game representation of reinforcements, which logically should have to be set at a boundry of a map but presently may be set anywhere subject to the risk of being overrun, or Forward Outposts, which represent a defended position which has a constant presence at it, but are largely unable to defend themselves without squads dedicated to doing so, and even then, only to a limited degree because of the largely ineffectual nature of the defenses that can be erected there. However these are side issues, largely unrelated to the issue of deviation, and continuing with them runs the risk of tu quoque fallacy, such as the Vehicle Deviation rebuttal above).
***
(Additionally, it is worth pointing out that many of the original design goals set out by the PR team have been reversed - criticisms that rifles were too inaccurate in vBF2, that they didn't do enough damage and that aircraft were unkillable are again being bandied, only this time in the PR forums. However gameplay remains remarkably different to vBF2.)
Please discuss.
Please try not to talk about your personal opinions, but rather talk in hard facts. (So not "I think deviation is too much because I can't hit things in XXX....." but instead "There is too much random deviation from the prone position, because in PR you are unable to hit a standing target at 100 meters, with five timed shots, which a real soldier can do.")
***
Before we continue, it must be understood that it is not possible to code weapons sway or swing into the BF2 engine.
Therefore, the possibilities to reflect inaccuracy are as follows:
Random Deviation.
Random deviation with a crosshair type thing that shows your relative accuracy (gets bigger when you're inaccurate, like the HAT) only while scoped.
Perfect accuracy, with a scope animation that moves too much to be used for perfect shooting (until you hit a target, and then back every shot up to the same place).
Why aren't soldiers perfectly accurate?
On a battlefield, numerous factors implicit in using a rifle, in the battle field and related to the condition of the individual soldier negate the ability to fire exactly where intended. They are discussed here.
Firstly
Fundamental differences between actually shooting a rifle, and aiming in a videogame;
The aiming of a rifle isn't as easy as moving a mouse over an icon. Firstly, a rifle has inertia - it is a relatively large object which requires extreme control to point in an accurate way. Secondly, the control mentioned in the point before must be generated by balancing the rifle at two different ends with different arms - this is a far less stable arrangement than a mouse already rested on a flat, smooth surface has. Thirdly, as the rifle is dependant on the movement of the entire body (which cannot itself be stabalised), and not just the wrist, involuntary movements (breathing, muscle shakes (accentuated by combat stress and adrenaline), and even heartbeat) detract from the stability of the rifle, meaning that a soldier may not be able to aim the rifle consistantly where he wants to. Fourthly and finally, some points which are easy to aim at as a PR avatar are actually physiologically impossible or difficult to aim at in real life (high/low angle shots being two good examples, but firing flat from just behind the crest of a hill is another).
This point may be practically refuted by pointing out that Tanks, APCs Light Vehicles and Aircraft have no deviation whatsoever, despite being at the mercy of many of the same effects, such as inertia, variations in controlability in different weather conditions and so on.
This point may secondarily be marginally refuted by pointing out that a soldier is more able to understand when he is able to shoot accurately, and will often forgo shooting until he knows he does have a perfect shot, something the game does not model at the moment (discussed in the third section).
Secondly
Differences between the Project Reality gameworld, and a real battlefield;
On a real battlefield numerous factors negate an individual soldier's ability to fire effectively. Disorientation from explosions, fog of war (no minimaps, remember?), unstable firing surfaces, realistic ballistics (bullets actually don't fly straight, or even on a parabolic arc, while different ammunition behaves in very different ways in differing conditions that the soldier may not have dope charts for), sights may (rarely) be incorrectly zeroed, guns may jam (not firing at all, on average reducing the number of shots fired on target) and finally visual distortions caused by heat are all factors on the battlefield that limit the ability of a soldier to shoot exactly where he intends. This list is by no means complete.
Thirdly
The state of the individual soldier on the battlefield;
This is the only piece of logic behind deviation which is contestable. I will explain it, and then explain the counter-arguement.
When a soldier is on the battlefield, his ability to make rational, intelligent decisions is limited by the high levels of stress inherent to the environment. Additionally, as there are no "respawns" his willingness to move out of cover to aim a perfect shot is likely to be much lower than it is in a video game. Because of these factos, a soldier on the battlefield isn't as likely to care about whether his shots are hitting the target or not, so long as the enemy is being suppressed, he is doing his job, and he is not under unnecessary danger. This is why the number of bullets fired to total kills remains so low - in a combat, individual soldiers are not concerned about their kill count, or KDR, normally speaking they simply want to survive, and the best way of doing that is to shoot at anything and everything so it can't shoot back.
The problem with this argument is that by modelling the individual soldier's tactical predilictions, the player's role in the game is usurped. The entire point of being the player is that he retains his bodily functions, and you take over his consciousness, using him as your avatar. In many situations the right tactical choice's effectiveness is inhibited by the avatar's excessive deviation caused by this modelling - a soldier who has managed to flank an enemy and is under no stress to take a shot will be able to make his shots count more than one who is under fire, but both have the same deviation models.
Summary on the realism of deviation.
After that section, only the most die hard of anti-deviationists would argue that there should be no deviation at all in PR. There are obviously iron clad reasons why a real life soldier can't use a rifle the same way we use a mouse, and that that should be reflected in the game. After this, the question becomes more one of how much deviation there should be, and in what situations. At the moment, there is something of a community schism about how much deviation is realistic, and how this amount compares to how much there is in game. Few argued that CQB was realistic in the .7 family, and likewise, few argue that sniping is realistic in either .7 or .8. These points have their reasons that are covered in the next section, but standard infantry weapons are still open to debate. Before you form an opinion on the matter, consider the following thing that may have been responsible for not hitting your target:
BF2 Hitreg: The BF2 engine is probably the worst modern FPS engine in terms of handling whether someone shot someone else or not, which some might think would be important to a shooting game. The critical factor here is how the engine places models with relation to lag. Basically, the game assumes you have a static ping of 100, meaning that if your ping is anything lower, objects will appear in front of where their hitboxes are by the amount your ping is less than 100 in milliseconds, and if you are on a higher ping, they they will be in front of the models you see. There are console commands you can do to minimise this effect, and a perfectly legit program that helps you set them called Bf2HitFixer. Some might argue that BF2's **** hitreg alone negates the need for random deviation, as real life soldiers always hit what their bullets physically land in, however this is an invalid arguement, because this effect usually only happens when a target is moving, or has very recently moved, a stationary target at 500 meters would be fair game for a rifleman without deviation (which is unrealistic) - no dice Johnny.
Reasons that do not relate to marksmanship in Reality
The Style of Gameplay Desired by the Devs
This is probably the primary reason deviation is as significant as it is.
A common criticism of the 0.6, no deviation style of play was that the player with better reflexes could expect to win a vast majority of engagements, in spite of tactical decisions made by either side. The counterpoint to this is that ultimately some soldiers are going to have better reflexes, that will give them an edge in battle - moreover, by diminishing the role of reflexes, you are merely emphasising the "metagame" aspect discussed above - that the player's role is being diminished in order to accomodate a more "tactical" style of play. The pro-deviationists counter again by saying that reflexes are still important to the game, while the former group points out that tactics were still important before. The debate continues with many unsubstantiatable claims from both sides which do not resolve the issues.
Secondly, the dev team has consistantly talked about the style of firefighting they want to be "epic" and "intense". Certainly in real life, static gunfights tend to last for a lot longer than they do in video games, usually for hours, with only low double digits dieing on each side unless fire support assets are used. Some of the reasons for this are discussed above, in the state of mind of the soldier section.
To this end, the Dev team also introduced a suppression system where a soldier being shot at, even in cover, would be unable to return fire, and would basically be forced to hide (the system isn't perfect, but it works well, despite there being workarounds). There are a few things that could be said to counter this focus on suppression (which so far very few have tried), there are either that suppression is "too much" (soldiers aren't suppressed as severely as they are in PR, and are able to return fire more accurately (which is itself rebutted by the fact that they are also more afraid of losing their life, and are thus less likely to take the risk of dieing just to get a "kill"), and secondly, that suppression doesn't neccesarily play as large a tactic in all war as the Dev team has made out. This second arguement assumes that different armies conduct war in different ways, and so for example while a US centric tactical doctrine might focus on always pinning an enemy and then flanking him, an Australian infantry doctrine might revolve more around remaining in cover and concealment, and waiting for the enemy to enter range so riflemen can engage accurate fire, using machine gun assets to break up concentrated enemy attacks. In this second arguement, by making infantry rifles less accurate even while totally concealed and stationary (or moreso while tracking a target), you actually diminish the effectiveness of all tactical doctrines that do not conform to the Dev's vision of how combat "should" be.
(Disclaimer, this was a poor example. If someone can provide a better example of another tactical doctrine that would fit here, please do.)
What is not open to contest is that when there is no deviation (0.6), a run and gun style of play is encouraged, as there is no reason to adopt a hard-point defensive stance to wait for others to enter your killzones. When this occurs, battles do not happen anything like they do in real life, and this raises a more philosophical design question - should Project Reality impose unrealistic restrictions on the individual soldier (sights that don't represent where the gun is really pointing) in order to force the overall shape of the battle to be realistic, or vice versa?
(At present, other factors make the battle's shape significantly unrealistic, such as rally points, which are the in game representation of reinforcements, which logically should have to be set at a boundry of a map but presently may be set anywhere subject to the risk of being overrun, or Forward Outposts, which represent a defended position which has a constant presence at it, but are largely unable to defend themselves without squads dedicated to doing so, and even then, only to a limited degree because of the largely ineffectual nature of the defenses that can be erected there. However these are side issues, largely unrelated to the issue of deviation, and continuing with them runs the risk of tu quoque fallacy, such as the Vehicle Deviation rebuttal above).
***
(Additionally, it is worth pointing out that many of the original design goals set out by the PR team have been reversed - criticisms that rifles were too inaccurate in vBF2, that they didn't do enough damage and that aircraft were unkillable are again being bandied, only this time in the PR forums. However gameplay remains remarkably different to vBF2.)
Please discuss.
Please try not to talk about your personal opinions, but rather talk in hard facts. (So not "I think deviation is too much because I can't hit things in XXX....." but instead "There is too much random deviation from the prone position, because in PR you are unable to hit a standing target at 100 meters, with five timed shots, which a real soldier can do.")
)
