I think all of these are good points, but I'm worried about focus. If the player has to watch the bow to determine when maximum accuracy is reached, it means taking focus away from the target at the moments before release. Maybe a solution could be something along the lines of the stealth eye, with different levels of brightness? The reticule starts out bright, then dims until maximum accuracy is reached? (This is in addition to the bow jittering, not instead of)

I'd say just give us some kind of Bow sway and perhaps the smaller circle, but that is it. That way it is still a challenge without it being distracting.

There is nothing wrong with an expanding/restricting reticule. It's used in just about every FPS including CS, COD, Battlefield etc. etc. The only problem is that the dispersion as indicated by the size of the circle is deceptive, in that it doesn't dynamically scale for targets at different distances. This, however, has not prevented games like CS from being mostly intuitive and enjoyable. Alternatively, since ES is a RPG, we COULD have a reticule that adjusts itself mathematically to targets depending on distance...the player would then always know what the actual hit area is per target. Might get a little confusing with multiple targets and different distances though.

I can't get my head around this claim. Why should it dynamically scale? The circle represents the sum of errors made by the archer. Since the circle is drawn around the reticule, these errors will cause further deviations from the aiming point as range increases - even when the circle stays the same, as indicated in my previous posts. In what way do we need a correction from this?

Edit: Marksmanship training with rifles is a good example in that regard. When practicing for 200m-ranges, units sometimes use 50m ranges instead. They just scale down the targets to match 50m instead of 200m. When viewed through the sights, the target appears to be as big as the one they'll fire at on the 200m-range. The results achieved are usually the same on the 50m-range as they are on the 200m-range. The errors that cause deviation on the 50m-range is multiplied by 4 times when they shoot at the 200m-range, but since the target is proportionally larger, the results end up being the same. If you viewed both targets through a dispersion circle, they would appear to be of equal sizes, even though one of them in reality is much larger.

I think my idea is more to put the player into a realistic aiming position with his bow, javelin or crossbow and have him aim himself as an archer would actualy do. But, to put the character skill in play, I propose to give 1/ a more or less good/stable aiming position (that is, a more or less correct one, more or less excentrated...) 2/ an arrow point/bow that will move/shake more or less around the direction the player is trying to aim. As for real missile weapons, the player has to wait the perfect time when the target is right in his trajectory.

With this, you can see the sequence of shooting as
1/ a first step to band the bow
2/ another one to assume the aiming position
3/ a third one of weapon stabilisation and actual aiming where the precision gradually increase, before disgrading again if the player waits too much
4/ the fourth step being the releasing of the arrow or a return to a resting position to start another cycle.

For each step, we can imagine the character stats and marksman skill influencing how fast he can be ready to shoot (SPD+marksman from step 1 to 3), how long he can sustain the aiming position (END+STR+Marksman for step 3 length) and, because of the mechanism previously exposed, how precise will be the shot (AG+Marksman for the stability and quality of the position).

Care to elaborate? At all?

The only part of your post I have a problem with is this, and it is possible I'm simply misunderstanding what you're saying here...but are you suggesting that the arrow will move about inside the circle, and that this movement will directly affect the striking point? That is, the player has to wait for the arrow to point directly at the target, and then release?

Haha well you're probably right. CS has misled me - on full auto past the 3rd bullet not a single shot lies inside the reticule.

Heh...I've never played CS or any FPS-games, so you might be right about the implementation in those games

But there shouldn't be any need for any dynamic scaling to compensate for range with this circle. I used to work in a tank platoon, and we estimated the accuracy of our tanks based on dispersion circles. After zeroing our tanks, we'd measure the distance between the impacts and calculate deviation in mils. Usually the deviation was about 0,5 mils for the most common ammo. This meant that the "dispersion circle" (abstract figure, didn't really exist) was 0,5 mils in diameter, ALWAYS. 0,5 mils at 3000 meters is 1,5 metres, so we could expect to hit targets sized 1,5 metres or larger at 3000m. At 2000 m, 0,5 mils = 1 metre. At 1000m, 0,5 mils = 0,5 metres.

So even though the size of the dispersion circle is constant (0,5 mils), the area it covers varies proportionately with range. It's the same principle that allows you to cover the sun with your thumb, even though the sun actually is somewhat larger than your thumb. With this dispersion circle, you will be able to hit targets at short ranges very reliably, but it will become harder and harder as range increases because the size of the target decreases relative to the circle.

Edit: The image from my first post might illustrate this. There are 2 NPCs in this screenshot, one guard at medium range, and an orc at longer range. The blue ring is about the same size as the orc, but smaller than the guard because of the range difference. That means that the player will have a greater chance of hitting the guard than the orc.
http://img63.imageshack.us/i/marksmanexpert.jpg/

Sorry if this seems like lecturing

I like your idea, but I think there's a small flaw in the system, which can be corrected by having the center of the circle track the target you last pointed at with your cursor.

Basically the problem is that if you aren't as good with aiming, and if your character is an expert marksman, then the circle is going to work against you.

The arrow has a 100% chance of hitting within the circle, which means it has 0% chance of hitting outside the circle.

This means that if your (the player) aim is off the target, and you have a small circle because of expert marksmanship skill, then you have 0% chance of hitting the enemy, because the enemy is outside the circle. Now contrast this with the same situation but with a crappy marksmanship skill. Because the circle is now large, there's a greater possibility that the circle will overlap the enemy, even if your not aiming directly at him, and so you actually have a chance of hitting, your crappy marksmanship skill could potentially save the shot from your lack of cursor aim skill.

If we have the center of the circles track the enemy target, until your cursor crosses another enemy, or goes far outside the range of the target, the problem is fixed

naw, it was a good comparison. The issue i think is with how to effectively represent the status of the dispersion circle without having really intrusive recticle.

Well, I don't really see this as a flaw. What you're basically saying is that "if you don't aim at the enemy, you won't hit them", and that is correct. Yeah, the crappy marksman may get lucky, just as an incompetent gunner firing an AK from the hip might get lucky with a single shot, but the skilled marksman will still beat him by consistently hitting targets, not getting lucky.
I never found it that difficult to place the cursor accurately on a target, so I'm not sure I see the need for an "auto-tracking" feature...

I don't think it's the same situation, in your example you are only taking character skill into the equation, if the player misses, the crappy marksman character might get lucky, but the expert marksman character will never get lucky.

Is there ever a situation where an expert and a novice tries to do something, and the novice have a higher chance of a hit than the expert?

Don't get me wrong it's a small flaw, your system is still better than what the current is, and if I'm not mistaken, it's the system used in fallout 3.

Well, to me that seems like somethin of a false anology. The expert marksman doesn't need luck as long as he actually aims at the target - he will hit because of his high skill. Your comparison is based on a situation where the player for some reason doesn't place the reticule properly over the target. In those situations, he will also hit the spot he's aiming for, i.e. where the reticule is. If that isn't the intended target, the player should have placed the reticule over the proper target. This doesn't seem like an unreasonable demand; I never had any problems placing the reticule over targets in Oblivion. It seems like you're proposing an "auto-targeting"-feature.

I do classing archery myself for leizure only and like this idea a lot!. arrows tend to go upwards and away from the bow, then compensate by wobbling in the other direction, more then the other way around unless you snap your wrist inwards while releasing the arrow to limit the fletching hitting your thumb and bow shaft. I also love combining this with crossbows where untrained marksmen have a more accurate option in case it's needed. the skill in crossbows can affect both relead speed and accuracy, though to a lesser extent. would add variety to range combat too.

Edit: then again, my fletching is not made of feathers, but rubber. this wobble may have been lesser or not present with actial feathered arrows.

@Daydark
Adding to Andrimners post I understand some people might find it hard to put the reticule on a distant target (or put it at the right height above it to compensate for gravity or other factors), and that within this system these people will occasionaly get lucky and hit their target regardless of the poor player skill as long as they have low marksmen skill. I find it impossible to believe that if a person like this plays as an archer long enough to master the marskman skill they will still be unable to accuratly place the reticule the majority of the time. A novice may just get lucky and gain some marksman experience and maybe even some levels a few times but never enough to level up to the point where there is almost no dispersion solely by luck, he must have gained enough player skill along the way to aim the reticule at least reasonably well.

And to the person who commented earlier that size changing reticules work fine in fps (and they do i wont argue with that) may I remind you that TES is not an fps and arrows are not bullets, which are unaffected by gravity in almost every fps, and even when they are effected it is not nearly to the same extent as an arrow.

Yep, except the fact that the player won't have a perfect visualisation of the direction. He will be positionned more or less correctly to assess it, but it won't be a very precise targeting system. The idea is that the character skills set a layer of unprecision and/or unstability between the player skill and the target. It works less on randomisation than the original circle idea, though the arrow movements will have a random dimension, but more as a penalty for the player who is confortably seating with his mouse in hand and possibly interpretating a character who has never wielded a bow of his life.

I like the circle idea, but my point was to try to adapt it without putting distracting and artificial colored circles in the UI. As some has pointed it already, I also prefere to feel in the skin of my character than in front of a balistic modelisator.

Well, they sort of are, but I won't go into semantics as to how gravity affects arrows. The gravity drop doesn't pose a problem at any rate, dispersion circles will still work just fine. Yes, I realise that I have said that the arrows will strike "within the circle", but that is merely a simplification. The dispersion circle represents deviation from the aiming point, which may be above the target. Remember that this is only an illustration; it only says that at your current skill level, the arrows will have a max deviation from your aiming point of, say, 60 mils. Obviously, the arrows will still drop below the reticule once released, and therefore below the centre of the circle, but the deviation from the aiming point still stays the same.

The "dispersion circle" as presented to the player stays the same; the "hit circle" around the target (a completely abstract creation only for the purpose of this post) will represent the actual impact, and it will be the exact same size as the dispersion circle. It's the actual dispersion (60 mils in the example above) that's important. The circle merely means that when you release the string, the arrow will have an exit direction somewhere within the circle, as opposed to dead centre.

Edit: Everything gets better with an example.

Your character has a marksman skill of 30. At this skill level, the circle is at 50 mils, meaning maximum deviation is at 25 mils in all directions.

You aim at a target some distance away. In order to compensate for arrow drop, you aim 10 mils above the target.

As you release the arrow, random numbers are drawn between 0 and 50 for deviation in X and Y planes. Those numbers are -15 X (horizontal) and +25 Y (vertical), so the arrow will deviate somewhat to the left and upwards. It starts out with an angle of 35 mils above the centre, because 10 mils superelevation + 25 mils deviation = 35 mils.

Assuming 10 mils superelevation was a correct estimation, the arrow will drop 10 mils as it travels towards the target, and strike 15 mils to the left of the centre of the target, and 25 mils above the centre of the target. Keep in mind that mils are angular, the amount of metres will change as range increases (One mil equals 1 metre at 1000 metres, and 0,5 metres at 500) - if the target was at 100 metres, this means you'll hit 1,5 metres to the left and 2,5 metres above the centre of the target, probably a miss. If the target was at 25 metres, the arrow will strike app. 0,3 metres to the left and app 0,6 metres above the centre - which could be a hit. (Obviously assuming that you estimated superelevation correctly)

These calculations might seem complex, but this is obviously not something the player will be troubled with. He will just be presented with an "estimated accuracy" based on max dispersion. (If we go for the visible circles solution)

This sounds a whole lot more complicated than it is In order to discuss gravity drop, maybe it's better think of the circle as an invisible funnel that surrounds the curved trajectory towards the target, expanding as range increases(at the mil/metre-ratio as above). The walls of the funnel symbolizes the deviation; that is - the arrows freedom to move. It will hit within the funnel. As I said, the dispersion circle is a simplification, but it does present all the information the player needs.

Well, the circle is merely an abstract idea to indicate dispersion. It needn't be included in the UI, it could just as easily be invisible. That's just a matter of preference.

About the suggestion, I'm still not sure if I fully understand it. My understanding now is that the arrow will move visibly, and somewhat randomly based on skill, and that it is possible to judge the trajectory based on this. To hit, wait until the arrow is at the "correct" point, and release. The higher the skill, the "easier" (smaller?) the movement is. If this is an incorrect understanding, I apologise.

The point is, the circle indicates a random devieance from the point the player is aiming at. That's a quite different idea from waiting until the sights are properly aligned and releasing. The point is that even though the character is positive of his aim, there will be some dispersion. This isn't caused by poor sighting, it's caused by other factors. (Such as improper release) Aligning the reticule over the target (aiming) is done by the player, but the release is done by the character. That's what affects deviation.

I remember playing old "target practice" games where a crosshair would zigzag across a bulls-eye, and you had to click at the right moment when the cursor is right on top of the centre of the target to hit. I always got pretty annoyed because that's nowhere near how it's done in real life. Assuming you have a decent stance, anyone can line up the sights properly and pull the trigger when the sights are properly aligned with the target once the basics are learned. The challenge is reducing the dispersion as you pull the trigger. Squeezing the trigger, holding your breath, proper grip, etc. That's what separates a good marksman from a poor one. Proper aim, and then minimum dispersion.

Also, I'm not partial to an idea where the player has to wait for an animation (as the arrow movements would be) before releasing. I think the pause could be frustrating.

You understood my point correctly. I think I also catch you argument. As I describe my option there is no more uncertainity, just a delay... Though both ideas could combine. Aiming with a bow is not as aiming with a firearm. There is much more muscular tension in the process. Moreover, even with firearms, aiming is much about waiting the perfect time. With big targets or close ones, you may not care too much for a slight deviation, but if you aim a very precise point in the long range, you can get impressed by the hardness to keep your arm perfectly stable. Now, after the player has aimed, there can still be room for approximative projectile release.

I think that we should keep in mind that making marksmanship a challenging task should go with making it a more efficient one. Oblivion had a ridiculous damaging system. You needed dozens arrows to kill any NPC (the same was true for close combat). With the systems we propose, the player will have rarely more than 2-3 shot before close combat. There should be a serious chance that a hiting arrow heavily hurt a normal creature. Marksmanship should also require the opportunity to engage long range combat, which mean being sometime able to spot your target a few hundred meters away, not 20-50m only as it was most of the time in Oblivion. Terrain and buildings should also provide secure places to shoot without being immediately charged.

I don't see how it's a false anology, like you say my comparison is a situation where the player for some reason doesn't place the reticule properly over the target, or really not for some reason, we know the reason, bad player aim. In those situations he wont hit the spot he's aiming for, because where his reticule is when he releases the arrow isn't where he intended it to be. If it isn't the intended target, how can he place the reticule over the proper target, when he was already trying for that? The starting assumption is that he has already failed to do so, how does that not seem like an unreasonable demand? You're literally saying the solution to the problem isn't getting into the problem, which is a universal solution to every problem there ever existed, but is not a valid one because you don't need a solution before you have a problem.

I am proposing an auto targeting feature, well limited really, but an auto targeting feature nonetheless, because it will mean that there's never a situation where a crappy marksman character will have greater odds than an expert. It's not because it was a major problem in Oblivion, because Oblivion NPC's don't try to dodge, an incoming arrow, they at most try to strafe a little. But they do move, and you do miss even when you really shouldn't given the supposed skill of the character you're playing. This is also a problem that the npcs wont be affected by, because they will have 100% auto aim, archers in OB will nail you from any distance provided you maintain your direction and speed.




Of course if the target is standing still, there can't be a problem, but if the target is moving across uneven terrain, and your moving too, and you have to adjust for arrow flight time, and gravity. This will undoubtedly result in misses that would not have been a problem for an expert marksman. It's not a big deal, as you say yourself the player most likely develop some form of skill enough to reasonably place the reticule well, but the flaw, although minor, lies in the fact that this will not give a true portrayal of an expert marksman, given that an expert marksman doesn't just aim reasonably well.

Ok I concede your point Daydark and I see how the movement and drop could confuse some players and result in the situation where a master might have a lower hit % than a novice because they have trouble aiming and never get lucky. I wouldn't mind if Beth added in a auto-targeting feature as long as it was an option that could be turned off as I find that this feature in other games makes them much less fun and rewarding for me.

It seems very silly to make changes to the archery system based on the assumption that players won't be able to aim properly. Archers with a high skill level should be rewarded with the *ability* to hit targets with greater accuracy (by means of a smaller dispersion circle) -- provided of course that they are able to keep their reticule on target.

We're not saying change the whole system just add an option for aim assist if you want it I wouldn't be against that even thoguh I'd make sure to turn it off if you dont like it then it doesn't affect you just helps those people who arn't able to correctly factor in arrow drop and movement inorder to "keep their reticule on target".

Well In FO:NV you had to have some skill in strength and guns to handle a sniper. So I know Skyrim will have something like this.