Comprehensive Stats Chart (4/27/11)

Pantalaim0n · 29-04-2011, 10:36 PM | **Post: #21**

(29-04-2011 09:28 PM)rukqoa napisał(a): Naw. IIRC, it's a new mechanism introduced in BF2142 for the LMGs (which have decreased deviation if you go full auto in 2142) to prevent them from getting absolute perfect accuracy when zoomed in.

minDevModZoom means the absolute minimum deviation a gun can have when zoomed. A devmod usually implies a multiplier, but a min/max in front of anything is always a limit.

I don't have empirical evidence for it, but I don't have evidence for a lot of other things either.
Good read: http://forums.gamespy.com/battlefield_2_...620938/p1/

You shouldn't trust outdated threads. I've done enough tests in bf2 and 2142 over the past few years to tell you that deviation is calculated by

(mindev*mindevmodzoom+firedev+speeddev+turndev+miscdev)*devmodzoom*stancemod

Even the official documentation states "Be aware that MinDev is affected by both DevModZoom and MinDevModZoom when the weapon is in zoomed mode. Simple Math: MinDev * DevModZoom * MinDevModZoom * Stance = Accuracy". You should be well aware of that as you quoted that one yourself a while back.

If a machine gun was not to get perfect accuracy you'd just put a max (or min, the numbers being negative values) on firedev. The 2142 EU MG has mindev 1.18, firedev capped at -1.05 so the dev doesn't get too low. Even better, the 2142 machine guns have their mindevmodzoom defaulted at 1 because otherwise it'd mess with the formula.

I'm pretty confused actually what'd make you think that the term mindevmodzoom refers to some kind of absolute limit. There hasn't been any limit before and the term doesn't sound much like it either.

rukqoa · 29-04-2011, 10:54 PM | **Post: #22**

(29-04-2011 10:36 PM)Pantalaim0n napisał(a):
(29-04-2011 09:28 PM)rukqoa napisał(a): Naw. IIRC, it's a new mechanism introduced in BF2142 for the LMGs (which have decreased deviation if you go full auto in 2142) to prevent them from getting absolute perfect accuracy when zoomed in.

minDevModZoom means the absolute minimum deviation a gun can have when zoomed. A devmod usually implies a multiplier, but a min/max in front of anything is always a limit.

I don't have empirical evidence for it, but I don't have evidence for a lot of other things either.
Good read: http://forums.gamespy.com/battlefield_2_...620938/p1/

You shouldn't trust outdated threads. I've done enough tests in bf2 and 2142 over the past few years to tell you that deviation is calculated by

(mindev*mindevmodzoom+firedev+speeddev+turndev+miscdev)*devmodzoom*stancemod

Even the official documentation states "Be aware that MinDev is affected by both DevModZoom and MinDevModZoom when the weapon is in zoomed mode. Simple Math: MinDev * DevModZoom * MinDevModZoom * Stance = Accuracy". You should be well aware of that as you quoted that one yourself a while back.

If a machine gun was not to get perfect accuracy you'd just put a max (or min, the numbers being negative values) on firedev. The 2142 EU MG has mindev 1.18, firedev capped at -1.05 so the dev doesn't get too low. Even better, the 2142 machine guns have their mindevmodzoom defaulted at 1 because otherwise it'd mess with the formula.

I'm pretty confused actually what'd make you think that the term mindevmodzoom refers to some kind of absolute limit. There hasn't been any limit before and the term doesn't sound much like it either.

Alright you seem to be right. I do recall using that formula in some calculations. I forgot why I have a different perception of how accuracy works now. I'll include it in the next release, and thanks for your contribution! Smile

I think what got me confused was that you can set other "max" commands in the files, like fireDevMax...

Pantalaim0n · 29-04-2011, 11:06 PM | **Post: #23**

Ah I also recall the reason for the implementation of mindevmodzoom. It was about sniper rifles. In bf2 the mindev was set to 5 (i.e. really high) while the other devs weren't much different from ordinary weapons. To compensate for the high deviation, devmodzoom had to be really low (0.005/0.01). However as you can see from the formula this devmodzoom sort of caught any deviation, so snipers could fire on the move while scoped without much impairment. To get rid of that they introduced mindevmodzoom because it catches mindev only. It basically forced snipers to stop moving to get an accurate shot. Smile

rukqoa · 29-04-2011, 11:10 PM | **Post: #24**

(29-04-2011 11:06 PM)Pantalaim0n napisał(a): Ah I also recall the reason for the implementation of mindevmodzoom. It was about sniper rifles. In bf2 the mindev was set to 5 (i.e. really high) while the other devs weren't much different from ordinary weapons. To compensate for the high deviation, devmodzoom had to be really low (0.005/0.01). However as you can see from the formula this devmodzoom sort of caught any deviation, so snipers could fire on the move while scoped without much impairment. To get rid of that they introduced mindevmodzoom because it catches mindev only. It basically forced snipers to stop moving to get an accurate shot.

Hmm.. Can't they just increase speedDev so you can't fire accurately while moving?

Pantalaim0n · 30-04-2011, 12:10 AM | **Post: #25**

(29-04-2011 11:10 PM)rukqoa napisał(a): Hmm.. Can't they just increase speedDev so you can't fire accurately while moving?

Of course, in fact because of the high mindev there wouldn't be any noticeable difference really if speeddev was 100 times greater. I'm fairly sure they just hadn't thought of it. It does some good nevertheless because it allows for greater fine-tuning of weapons which are meant to be usable unzoomed AND zoomed. Basically when unzoomed you want a weapon to be mediocre when standing still and worse on the move. When zoomed however you want it to be highly accurate but still get a huge punishment for moving. The MG3 for example has 2.3 mindev and a 2.0 speeddev, so the share of mindev is actually greater than the speeddev one. When zoomed however mindev goes to 0.05 of its value so speeddev has a much greater impact on the final accuracy. All these things can be applied to firedev as well of course. The ridiculous mindevmodzooom on the MG3 makes it pretty obvious that the gun is supposed to be fired in single-shot only. Let's check four things, the first two shots of a burst when unzoomed, and the first two when zoomed (assuming a crouched stance).

Shot one unzoomed: dev=2.3*1.05=2.415
Shot two unzoomed: dev=(2.3+0.25)*1.05=2.55*1.05=2.6775

The second shot barely has any effect on the accuracy.

Shot one zoomed: deviation=(2.3*0.05)*0.6*1.05=0.115*0.63=0.07245
Shot two zoomed: deviation=(2.3*0.05+0.25)*0.6*1.05=0.365*0.63=0.22995

The second shot in the burst has over three times the deviation of the first. And actual weapon efficiency rather goes with deviation^-2 so it's even worse.

rukqoa · 30-04-2011, 09:12 PM | **Post: #26**

(30-04-2011 12:10 AM)Pantalaim0n napisał(a): It does some good nevertheless because it allows for greater fine-tuning of weapons which are meant to be usable unzoomed AND zoomed. Basically when unzoomed you want a weapon to be mediocre when standing still and worse on the move. When zoomed however you want it to be highly accurate but still get a huge punishment for moving. The MG3 for example has 2.3 mindev and a 2.0 speeddev, so the share of mindev is actually greater than the speeddev one. When zoomed however mindev goes to 0.05 of its value so speeddev has a much greater impact on the final accuracy. All these things can be applied to firedev as well of course. The ridiculous mindevmodzooom on the MG3 makes it pretty obvious that the gun is supposed to be fired in single-shot only. Let's check four things, the first two shots of a burst when unzoomed, and the first two when zoomed (assuming a crouched stance).

Shot one unzoomed: dev=2.3*1.05=2.415
Shot two unzoomed: dev=(2.3+0.25)*1.05=2.55*1.05=2.6775

The second shot barely has any effect on the accuracy.

Shot one zoomed: deviation=(2.3*0.05)*0.6*1.05=0.115*0.63=0.07245
Shot two zoomed: deviation=(2.3*0.05+0.25)*0.6*1.05=0.365*0.63=0.22995

The second shot in the burst has over three times the deviation of the first. And actual weapon efficiency rather goes with deviation^-2 so it's even worse.

But deviation at <1 is usually unnoticeable at medium range. So I'd say the MG3 is pretty good if you burst. Even if you don't burst, there's a max fire deviation, and I wouldn't say it's very inaccurate overall.

Pantalaim0n · 01-05-2011, 12:07 PM | **Post: #27**

I'm more into long range battles simply because the weapon allows for it. It's usually best to keep the enemy as far away as possible because neither assault rifles nor submachine guns are as accurate as the MG3. And snipers have to get a lucky headshot to do any harm as anything else will be healed within moments and/or the sniper will be shot within moments by the medic.

On medium range (let's say 40m) it's probably a matter of preference. I daresay single-shot will be just as deadly while not wasting ammo but tapping the mouse button all the time can be tiresome.