FE2 Observations and Musings

[Nicholas Bell 117]

I’ve been playing some heavily modified versions of Ojcar’s Armchair Aces for 1917.  Really appreciate the “heavy lifting” he did creating these.  Tweaking them is a breeze compared to starting with a 3W campaign.  Been working on attempting to decrease the casualty rates via a number of approaches – the hit boxes, other damage model settings, AI settings, gun accuracy and lethality.  As a base for starting, I’m using the Von S FM mods where available, and Peter’s to fill in any gaps that Von S hasn’t hit yet.

 

Probably the most important thing I “discovered” (realize this is not new to others, but it’s been since 2007 since I last played…) was that single missions and campaign missions played out entirely differently.   In campaign missions the AI pilots will at least on occasion break off from combat, while in single missions it’s a fight to the death.  So here I am doing test after test in single missions attempting to increase the probability of the AI breaking off, all for naught.  Jokes on me.

 

But the fact remains that the AI will break off the chase or bug out of a fight.   Unfortunately this doesn’t have anything to do with odds stacked against them or altitude disadvantage.  On the other hand I experienced an interesting situation while flying a Pup and being chased by an Albatros D3 for a good 5 minutes.  After I crossed the lines into friendly territory I turned on him and managed to fire off a few rounds in his general direction before running out of ammo.  I don’t think I hit him, but he obviously had had enough and flew off in the general direction of his lines.  OTOH I had a stubborn bugger follow me all the way back to base just last night.

 

I need to check out the AI data for the Brisfit, as I’ve noted that they will often break off, fly a ways off, then turn about and re-engage.  Lewis equipped Nieuports  also display this behavior which I suppose is related to the reload time assigned the weapon.  After they fire off all their reloads they will break off and fly home.

 

The Brisfits are nasty!  I understand why the Germans avoided them. They’re fast, maneuverable, and spit fire from both ends…

 

I’ve often wondered about the apparent ineffectiveness of AAA.   By using Debug and a lot of sight seeing after being shot down or while flying back to base I’ve noted a lot of 2 seaters along the front lines with light damage and no enemy aircraft about.   I was really surprised while chasing an Albatross and it was hit by a flak burst which blew off both right wings -  right in front of me.  And it was German flak aiming at me which did him in.  Unlucky bastard.

 

Playing a campaign in a single gun plane is a totally different experience than a 2 gun fighter.   Despite the higher rate of fire of a Vickers, it seems damned difficult (realistic?) to down any German plane.  Lucky if I get 1 kill in 10 missions…  While flying for the Germans the AI seems to do better with those single gun fighters against me… 

  

In my 1917 campaigns I’ve bumped up the experience of the Jastas and a few RFC squadrons (where appropriate like No 56) to 100 Experience and Morale.  I’ve noted that flying against these Jastas as opposed to flying with them is a different experience.  Against them, the Albatri AI pilots display what appears to be teamwork.  What they look like is a school of sharks!  I see them circling together, attacking together, and tag teaming me similar to wingman behavior in SF2.  What’s odd is I don’t see that while flying an Albatross.  Nor do I see similar behavior in the RFC units.   Now I know this has nothing to do with the code favoring the Germans, so it must be something to do with the AI data.  I also think the Albatros’s relative lack of maneuverability might have something to do with it.  The faster and more maneuverable allied aircraft have a tendency to really spread out with all their twisting and turning, and this may inhibit the AI cooperation.  More observation is necessary…

 

Need to look at the DH2 damage model.  On several occasions I’ve seen them on the ground going in circles, tipped up on one wheel.  The German fighters came down to strafe them, so they are not in DISABLED mode (AI ignores all aircraft in this state).  Some element of the DH2 which is tagged as HEAVY in the damage model needs to be changed to DISABLED or DESTROYED.

 

UFO landing.  Saw an Albatros  stop in mid air over an airfield and float pretty much straight down until it touched down, at which point it rolled forward a bit.

 

I saw an amazing display of AI piloting a week ago. A German 2 seater – I think it was an LVG – was attacked and did a slide-slip.  Cool, except it kept flying sideways and losing altitude.  Well, this pilot was so damn good he flew that plane sideways with the right wing pointing forward until he completed a 3 point landing.  Then he immediately proceeded to take off, this time with the propeller in front, LOL.

 

I didn’t know the RFC had Russian pilots.  I witnessed an SE-5a perform a taran attack on a German DFW.  Flew up to it and literally sawed off the DFWs tail, losing only his propeller in the process.  The German plane fell in 2 pieces, ending that crew’s existence.  The SE-5a pilot glided down to safe landing.

 

Was flying a Pfalz DIII escorting a couple of C-types on a deep recon mission.  Unusual mission for the Germans.  Things went south when we were intercepted by Camels.  My engine was knocked out at about 4000 feet altitude, so I did my best to glide back across the lines.  We were so deep I didn’t even get close.  Little did I know I had such excellent escape and evasion skills, for I turned up the next day at the squadron unharmed.  I’ve also had flight members forced down behind German lines make it back to the squadron, although usually wounded…

 

Murder.   I’ve been feeling guilty about an incident from a few days ago.  Flying an Albatros against DH-5s.  Now I’ve toned down the UFO-like FM of the DH-5 quite a bit, but it remains a maneuverable foe difficult to keep in one’s sights.  So here I am in a hard twisting fight with this DH-5, getting off short, high deflection bursts  when all of a sudden the pilot stops maneuvering and flies level.   We were not close to the ground.  Was he wounded?  I matched his speed, pulled up right on his tail and aimed for the pilots back.  He slumped forward out of sight and the plane dove straight into the ground.

 

 

 

[+Trotski 936]

Sounds like something right out of a Biggles novel !! ( I grew up reading all the Biggles books Jingoistic and very British , but pretty good yarns all the same ) Its true the Germans did have different tactics to the British and French and the Laffayette boys, fling a flight of SE5's or camels into a swarming circus of Albatri must have been quite daunting, hardly surprising the survival rates weren't that special ! It's maybe a blessing the DVII and DVIII didn't appear in large numbers earlier or the Entente may have struggled even more. Archie and Flaming Onions were not hugely accurate, they didn't have the gun prediction tech that the Germans in particular had in WWII , so the idea was to distract and annoy the enemy aircraft, rather than actually bring them down, of course when that did happen , schnapps all round for the lads !! I have a feeling that MG anti aircraft fire was more effective and dangerous than the Archie/Flaming Onions

 

The chap you murdered in the DH5 ( you filthy hun ! ) probably couldn't see you with all those wingy things behind and above him , so may be he thought you had swanned off, and was heading back to the mess for tiffin old chap , and then you murdered him you murderous swine !!! hahahaha

[+Geezer 3,569]

LOL!  Next, you'll accuse him of being a bounder and a cad.

[+Trotski 936]

an absolute rotter, most un-gentlemanly and certainly not cricket, couldn't he see a chap was orft for tiffin wot wot !!

Edited April 16, 2016 by trotski00

[+VonS 1,424]

Well this is a very stimulating post. It reminds me of some of the detailed discussions over on the Aerodrome website from a few years back. I usually tend to fly the single missions, of different kinds, intercept, offensive, army coop, lots of other types, and noticed a variety of AI behaviors if the enemy AI is set to "normal." My observations are that it seems that normal setting allows for more variety than the hard or easy settings for enemy AI, although this probably requires more experimenting with.

 

I was flying some more of the Middle Eastern theatre a few days ago and noticed that a pair of RFC-flown Morane Bullets did eventually break off their pursuit when we were at an altitude of about 1300 meters. I was in a Pfalz E.I. They simply turned around and went back to their base. I'm assuming that they ran out of ammo. and turned back. A stubborn Sopwith Schneider on the other hand still kept following me, while two other Schneiders that were also pursuing me finally turned around.

 

The persistent Schneider still kept following and closing in on me slowly, while I was out of ammo. myself, having shot down a couple of Schneiders earlier on and a Morane Bullet. Thinking what to do, and being about 100 meters above and 300 or 400 meters in front of him - and noticing that he was still firing short bursts at me, a couple of which were hitting my wings - I turned on him (which was crazy at that point since I had no ammo.) and dived.

 

He also went into a tight turn and dive, and we then circled about two or three times so that I got about 30 meters behind his tail - stayed there for about 10 seconds in that position and then ascended, doing a wide turn this time in the direction of my base again.

 

I checked over to see what the Schneider was doing - and it was still racing towards its base, not bothering to attack me or follow me again. It was an interesting engagement. It looked almost as if the AI was fooled or changed its level of aggression based on my unexpected move and decided to run, after I got on its six. I don't think that he was out of ammo., this fellow in the Schneider, but he just "decided" to leave (or run away) when I managed to get behind him. This was all experienced with enemy AI set to "normal."

 

Regarding the DH.2 FM, that's an interesting one - it is nimble with the ailerons, as historically noted of the type, but loses altitude easily in tight turns. I've found that turns are "safer" in it if the rudder is not used too much, and the ailerons gently. It gains speed best in a dive and has a top speed of close to 150 (148kph). There was some discussion of the DH.2 over on the ROF forums a couple of years ago I think - was it or was it not better than a Fokker Eindecker.

 

I've set up some "homemade" single missions for myself where I go one-on-one in a Fokker E.IIa with a DH.2 - usually it's not difficult to take the DH.2 down. In another mission where I go up against two DH.2s, in one Fokker E.IIa - I might get hit some if I'm not careful, but usually you can take out both of them if you stay above and use altitude to your advantage. The DH.2s can get away slowly if they decide to, but otherwise they don't seem as impressive as is often mentioned of the type. I did notice however that when flying a Fokker E.III, I have a more difficult time of it with biplane-types, often barely getting in a kill or being shot down quickly.

 

At any rate, FE2 provides a fun simulator-like experience where some of the myths and truths of these aircraft can be tested. That Lanoe Hawker survived for 5 or 10 minutes in his famous dogfight with Von Richthofen in such a crate then becomes even more impressive, considering that, at least in FE2, the DH.2 does not put an end to the "Fokker Scourge." Also significant is that the E.II variant of the Fokker Eindecker is the best of those types in FE2 - and may have been in real life too for the seasoned pilot. The E.II had slightly shorter wings than the E.III version, and was more maneuverable than the E.III, also a bit more maneuverable than the E.I version. Handling of the E.II was considered difficult, but flown well it is more deadly than the E.III, and also better than the E.I.

 

It's also pretty much impossible to do a modern "Immelmann" turn in the Fokker Eindeckers in FE2 (of any version of Eindecker), but old style Immelmanns, with a stalled vertical turn, sort of like a hammerhead, is no problem at all, and looping, if done carefully, is also possible in them - but be careful of the inverted flat spins. While they occur less in my tweaked FMs for the Fokker and Pfalz Eindeckers, they're still a real problem to get out of.

 

Happy flying,

Von S

[Nicholas Bell 117]

Good stuff, Von S.  Thanks for sharing your experiences.  I've got my head pretty well stuck in the Western Front '17-18.  But you've got me thinking....

 

Nick

The Murderous Cad

[+Trotski 936]

Yes, I think in real life, the lone gunman chasing the enemy aircraft home was both frowned upon, and unwise, but there are occasional incidents of it, by the lone wolf rebellious types. you turning in on the Schneider, is following the Bolcke dicta anyhow, always turn in to you attacker, he obviously though you have more ammo, and decided the discretion was the better part of valour !! the DH and FE pushers were a good stopgap, to try and combat the Eindekkers, and I have a special place in my heart for them, very dastardly and mutley !!

[+ojcar 157]

About the  AI settings:

 

-If you Choose Easy AI, the game change the missions to have more experienced pilots by your side than the enemy ones.

 

-If you choose Hard AI, your foes will be more experienced than your friends.

 

-If You choose Normal AI, the game will use the vanilla flies from the campaigns and you'll have more or less experienced pilots in both sides (and depending on what unit you are flying for)

 

I fly Normal AI. 

[+Stephen1918 1,474]

This is an interesting discussion.

 

I think in the campaigns, the AI pilots are more focused on their missions - if they're on recon, their goal is to get home with their info, so they're more likely to break off and head for home. Escorts are likely to stay with the planes they're escorting, defensive patrols would stay in their zone, etc. In single missions, the enemies you encounter are probably hunting and are likely to be more aggressive.

Edited April 16, 2016 by Stephen1918

[Nicholas Bell 117]

Interesting development.  I've cut the squadron morale levels in half and the results appear good in the limited testing I've done. In conjunction with eliminating the wing collision points I had a mission with results I've never had in any FE2 mission.  July 1918 and I lead 8 D VII on a cap.  I don't follow the waypoints, but head directly for the front line and then follow it north.  I spotted a couple of Salmsons heading northeast and order my flights to follow me and then engage, so we all started diving.  Then a flight of Spads appeared and 6 of my flight peeled off on their own to deal with that, while one of the 2nd flights wingmen decided to stick by my side.  I got distracted by what I thought were several Allied 2 seaters headed west (I use no HUD help at all) which turned out to be friendlies.  The Salmson's got away - they are pretty fast.  So my wingman and I turned about and climbed back into the fray which by that time included SE-5a's and Camels.  Typical huge furball, but loose and more dodging than firing (low morale level?).  

 

I did get involved with one very aggressive SE-5a and we went up and down and round and round for maybe 10 minutes.  They make me feel like I am standing still at times.  Any way, I finally got on his tail at mid-range and fired to get his attention.  While the log showed I didn't hit him, he climbed up and flew off to the west and eventually turned a bit evidently to hit a homeward bound waypoint (I kept padlock on him so I could watch his moves externally at distance).

 

By that time the whole dogfight had fizzled out.  So I RTB'ed.  End result - NO collisions, and NO losses to my flight.  And only 3 claims.  None by me, 2 by one chap and 1 by another.  Pretty impressive.  Mind you my gun accuracy ratings are 1/2 normal, I've upped the VolumeStructurePoint to 2000 and done a myriad of other tweaks.  But the most "historical" results I've seen yet in FE2.

 

I will test further, and perhaps cut the squadron morale levels further.  Sure would be good if any else could give this a try with their setup to see what the effect is.

 

Nick

[Nicholas Bell 117]

Bombers that are hit and immediately drop their bombs are given an RTB order, whether light or heavy damage.  Disabled status always results in RTB.

 

Fighters usually RTB when they have heavy damage, although they may defend themselves.  Going to experiment with changing component damage result states from disabled to heavy, the goal is to get more AI pilots to bug out and therefore reduce casualties.  However, heavy damage can result in odd behavior dependent on which components are hit and the AI may attempt to continue to fly but is not able to.  Disable status usually (but not always) allows for the AI to land deadstick.

 

As noted in my other thread, I've found a way to get more engine and fuel tank fires.  That has upset the balance I had with structural factors, VolumeStructurePoint, weapon and AI accuracy, and projectile weight.  Back to having too many planes downed in an engagement.  In one attempt to rectify this I lowered the weapon accuracy and accuracy to zero.  Interestingly, this had little effect on the AI's ability to hit or the number of aircraft shot down. Evidently bullets with no weight or direction still are deadly!    Seems only CannonFireAngle, GunnerAimOffset, and GunnerAccuracy have significant impacts on the AI.

 

 

 

[Do335 382]

 

Disable status usually (but not always) allows for the AI to land deadstick.

 

so much difference... in sf disabled usually means pilot bail and crash while landing!

About fires does FE has progressionchance and explosionchance etc, maybe tune them down some instead, so they may catch fire but not actually be destroyed that way?

[+VonS 1,424]

Hello Gents',

 

Thought I'd post here instead of starting a new post since these are musings on FE2 anyway. Has anyone been able to get the streamers going on the Morane Bullet, or the Le Prieur rockets on the Nieuport types? Also, what about the mystery/invisible Vickers machine gun on the late, "Vickers" Nieuport 16? I was tinkering with the Vickers Nieuport 16 today, thinking of "converting" it to a Siemens-Schuckert D.1, but the invisible machine gun is making things difficult. This is possibly so since the multiple Nieuport 16 variants pack available here was probably made for FE1 initially.

 

I did have good luck tweaking the basic "Nieuport 16" folder and data ini file for FE2 many months ago - that Nieu 16 has a single Lewis gun above the top wing, "baked" into the skin of the model I think, and works beautifully in the Mid east and Western theatres (assigned to the RFC and French). The late Vickers Nieu. 16 on the other hand is ghastly to tweak. I did manage to get the muzzle blast for the Vickers visible, but no gun.

 

Frustrated with the poor results, I tweaked the Nieu. 17bis data ini today instead, that I had previously tweaked anyway to remove the top Lewis gun, and it now behaves beautifully as a SSW D.1. The cowling should of course be opened at the bottom, a spinner added to the prop, and the top wing lowered by about 6 or 7 inches (in real world terms...don't know how much of a change that would require to the wing in game) - for it to look more convincing as an early Siemens (this is beyond my talents) - but for now I'm happy with this temporary solution. If somebody figures out how to make visible the Vickers for the Nieu. 16 - I'll try "converting" that one to a Siemens-Sh. D.1 instead since it already has the half-open engine cowling.

 

Here's a couple of pics of my current solution...will perhaps upload this for a Version 8.0 update. Siemens-Sh. D.1s were used on the Eastern front, not so much in the Western theatre where they were largely trainers.

 

Happy flying,

Von S

 

 

 

 

[Nicholas Bell 117]

so much difference... in sf disabled usually means pilot bail and crash while landing!

About fires does FE has progressionchance and explosionchance etc, maybe tune them down some instead, so they may catch fire but not actually be destroyed that way?

 

I have experimented a lot with the ProgressionChance and ExplosionChance.  The plane sit explodes if the ExplosionChance is set to zero.  Setting the ProgressionChance to a low value or zero results in the AI continuing to dogfight like normal - not good!  I gave the fuel tank a DAMAGE=DESTROYED tag, but that doesn't help, because the fuel tank can receive LIGHT damage and still burn.  With reasonable settings - FireCheckTime=6.0 seconds and ProgressionChance=8 percent the plane will continue to fight for many minutes before it finally explodes.  There does not seem to be a solution to get the AI to get an AI Action DESTROYED once they are on fire.  I may try to lower the Structural Rating of the fuel tank to get it to be destroyed faster, but that will likely just cause the plane to explode sooner.  i really would like the plane to burn until it crashes rather than exploding mid-air.

Here's a couple of pics of my current solution...will perhaps upload this for a Version 8.0 update. Siemens-Sh. D.1s were used on the Eastern front, not so much in the Western theatre where they were largely trainers.

 

Looking good.  Don't think lowering the wing is critical.  You would have to get someone to edit the LOD in Max or whatever program they are using.

[+VonS 1,424]

A quick update and some more musings. I'm happy so far with the results of the general aircraftaidata ini file that I've included with the 7.5 FM upload. Planes now fly well in formation without "waltzing" back and forth - the reduced value of 3.0 for formationroll seems to have been a good choice. I've also noticed that the gunneraimaccuracy setting is a very important one...and have been tinkering with possibilities. The default value was something like 20 or 30, but I had increased it to 40 in the aircraftaidata that's been uploaded here. Further testing revealed the following:

 

values of about 60 provide for challenging dogfights

 

values of about 80 provide for very dangerous fur-balls where one or two well placed shots can blow off your wing or kill your pilot, or give you an engine hit (and fire)

 

values above about 90, also tested at 100, sometimes give peculiar results where the ai hesitates as to what to do - to shoot or not to shoot (interesting stuff)

 

values of 40 are also more challenging than the default value for aimaccuracy, better than no alteration certainly - and a good choice for variety of dogfights encountered

 

I've set the aimaccuracy to 60 now and will upload the changes in a small 7.9 FM update - I've also included the additions for cannonfireangle and maxcannonrange for the different dogfight skill levels and results seem to be promising so far. The general aircraftaidata file is proving to be a good one since any values missing from the individual aircraft data inis are then "plugged" into the game by this file.

 

On an unrelated note, also changed now is the seating position of the pilots in the Morane Bullets (with their eyes at mid-level with the wind screen...they were sitting too low previously), and I've redone the Nieuport 17bis folder posted here with my earlier tweaks into a proper 17bis folder, a 17 C.1 folder (the common type of Nieuport 17), and the Siemens-Schuckert D.1 folder (pics are available above in this post, of that one). The three types now display their own quirks, with the 17bis a powerhouse more effective in many ways than the type 23, the C.1 being the delicate and slippery type it was known for, and the Schuckert D.1 being the "poorer" copy of the Nieuport, underpowered by 1917 and less maneuverable somewhat than the real 17. Wings can be broken when power-diving the C.1 and Schuckert more easily than on the 17bis. Will post these latest additions after readjusting the service dates for the different theatres.

 

Happy flying,

Von S

Edited May 12, 2016 by VonS

[Nicholas Bell 117]

Great stuff, Von S.   Thanks for sharing it.
 

I also separate AI data and work to only leaving aircraft specific AIDATA and DOGFIGHT* in each aircraft data file.
 

I cut the MaxCannonRange by at least half without impact positive or negative, other than the improvement of not seeing the AI pilots shot at impossible & ahistorical ranges (even for novices).  I would recommend increasing the MinCannonRange a bit for Ace and Veteran AI as the often the sub-50 meter values result in collisions between attacking fighters and 2 seaters– even with my eliminating the wing collision points.  I acknowledge this is ahistorical, as aces would fire from even less than 10 meters.
 

I’ve also increased the CannonBurstLength’s from a minimum length of 2 .0 seconds for an Ace in intervals up to 6.0 seconds for Novice.  These are more realistic than the impossibly short snap bursts the stock AI uses.  I’ve fired 7.62mm and .50 cal machineguns while in the service and was trained to fire 3 second bursts.  Less was really difficult and could result in a misfeed.  Longer was the tendency for most people but results in less accuracy, wasting ammunition.
 

A side benefit of this is to increase ammo usage with the “hope” of the AI running out sooner resulting in lower overall casualties.  Difficult for the Germans to run out of ammo with all that they carry,  but does have an impact on Allies with their Lewis guns.  Athough I don't have independent verification of this fact, but read in the WOFF forum at SimHQ that even Richtofen averaged 700 rounds fired per kill.  As an average this is believable as we don't read much about all his engagements where he did not score.  Sure would like to see this kind of average in FE2.
 

Tests on scenarios where I’ve edited the ammo load to 5% shows that the AI is given either a FORMATION or WAYPOINT command when they run out – not an RTB, unfortunately.   Even with FightWithoutAmmo=0, the AI will continue to dogfight especially in a large furball as they constantly go defensive because all the EA in their DefensiveAngle.    I’ve not tested it in small engagements, but theoretically it might be easier for the AI to disengage and follow their FORMATION or WAYPOINT orders.  In a large 18 vs 18 furball with 5% ammo loads no one was shot down, but the furball continued until I gave the rejoin order resulting in my AI pilots immediately following the Formation Order.  The enemy was given Waypoint orders and everyone went their way.

 

For this reason I suggest that players advance their waypoint to 8 immediately once engaged.  In this way any friendlies in the player’s flight will proceed to waypoint 8 if they are out of ammo, cannot see the player and not engaged.  This as opposed to wandering back into a dogfight only to get shot down.

 

Have you looked at the GunnerFireTime= values?  I “think” it is delay before pulling the trigger, or perhaps the interval between bursts, but have not been able to conclusively prove that.

[+VonS 1,424]

Hi Nick, thanks for the detailed comments as always. Here's what I have currently added to the aircraftaidata ini file (that's missing from the version 3 upload of that file).

 

novice

 

CannonFireAngle=8.0

MaxCannonRange=300

MinCannonRange=50

CannonBurstLength=7

 

green

 

CannonFireAngle=6.0

MaxCannonRange=300

MinCannonRange=50

CannonBurstLength=7

 

regular

 

CannonFireAngle=5.0

MaxCannonRange=200

MinCannonRange=30

CannonBurstLength=5

 

veteran

 

CannonFireAngle=3.0

MaxCannonRange=200

MinCannonRange=20

CannonBurstLength=3

 

ace

 

CannonFireAngle=2.0

MaxCannonRange=150

MinCannonRange=20

CannonBurstLength=2

 

The cannonburstlength info. I've just added today (thanks to your advice) and will see what happens with it - ideally I'd like to see the novice pilots more quickly getting rid of their ammo and disengaging from a dogfight (that would be more historically realistic).

 

Speaking of ammo usage/load, this is a topic that I've been digging into somewhat, also on the forums at TheAerodrome website - and I've tried in many cases to give historical approximations for ammo load in the data inis for the planes (although there are several files I'd like to revisit regarding this). For example, the Germans well into 1916 used 200 or 400 clip belts in their single Spandaus (on such types as the Eindeckers). By the second half of 1917 from what I could research, total load that could be carried went up to about 1100 clips, this is 550 clips per Spandau on the two-gun planes like the later Fokkers, also the Albatros. Loads were usually kept a bit lower to save weight and not hit performance of the aircraft too much. Typical loads of the later war period would then be something like 450 or 500 per machine-gun, for the Germans. I frequently see a load of "380" given to the Spandaus in many of the data inis and have changed this in several cases, but will more consistently revisit that data somewhere down the road.

 

Entente powers were using 200 and 250 clip belts well into 1916, maybe also 1917, on guns like the Vickers, changing over somewhere in late '17 or early' 18 to the typical 400 clip belts that are seen in several of the data inis for Entente aircraft. The Lewis guns would have the smaller 47-bullet drum until about late '15/ early '16, and would standardize on the 97-bullet drum by the summer of '16 approximately. This is all interesting since it means that some of the Nieuports from '17, also earlier examples of the Camel, should have about 250 bullets per gun (this means a total of no more than 400 or 500 rounds for earlier Camels, even though they have two guns....there may be several other examples to draw from, especially on underpowered aircraft of both sides in the war, with shorter belts being chosen to save weight).

 

Overall, this is another excellent (and historical) way of cutting the ammo loads down in the game and making the dogfights more realistic - shorter gun belts coupled with longer gun bursts - I see this as an excellent solution (particularly to have novice pilots disengage more quickly).

 

I've increased the GunnerFireChance from the default of 70 or 80, to 90 (in the current upload of the aircraftaidata ini file) - this change from what I could see in previous testing makes the gunner fire more frequently, or may also impact on his determination to fire at targets - I've left that setting at 90 since I think the more frequent firing is once again more realistic for novices.

 

GunnerFireTime I haven't tested yet - and left at its default value - I think that setting is either interval between bursts or delay from the gunner fixing targets on the enemy and deciding actually to shoot. It seems that that setting is linked to GunnerFireChance somehow - will test in a later iteration of the aircraftaidata file since I think there may be something important in those settings.

 

The suggestion about advancing to waypoint 8 makes a lot of sense. I've also found that in certain tight situations it's best, after a dogfight, especially after running out of ammo myself and being outnumbered in a fur ball - to signal my flight to regroup, and we would then do a wide ascending turn, at full speed, out of the area and on to another waypoint. Usually I would give the order to regroup while already ascending and turning out of the area - the flight would soon follow - the higher up I was at the time, the more quickly they would get out of the area too, with stubborn enemies following but usually not being able to climb so quickly (works best to give the order to regroup when you are already at least 500 meters above the rest of your flight, since it then pushes the remaining dogfight below into the "vertical"...and is safer than giving the regroup command from the same altitude, since many from your flight then leave their six open to lucky shots).

 

Happy flying,

Von S

Edited May 13, 2016 by VonS

[Nicholas Bell 117]

 

Here's what I have currently added to the aircraftaidata ini file (that's missing from the version 3 upload of that file).

 

Looks good!

 

Something else to consider is the improvement in the Allied synchronization gear.  The data is there, but not used – in the form of Vickers Mk 1, Mk 1_1 and Mk 2 which have realistic ROFs correlating with the improvements in the gear.  For whatever reason 3W stock aircraft all use the slowest firing Mk 1s.

 

I use JSME to swap out data.ini’s of aircraft as required to reflect the gear used at the time.  In 1916 I use the Mk 1.  In 1917 the new Allied aircraft (SE5a, Camel, DH4 & 5, SPAD 13, etc) get the Mk 1_1.  Everyone gets the Mk 2 in 1918.   Simplified, I acknowledge, as aircraft were manufactured with newer gear or upgraded in the field, but not a lot of specific information is available.  One may want to give Sopwith Pups and Tripes the Mk 1_1 in their waning days of late summer 1917.  Not sure about when the Spad 7 got upgraded and were used until the end of the war (which was something I didn’t know until recently).

 

Anyway, the Mk 2 firing at 750 rpm will “assist” in quicker ammunition depletion by the Allies – and also make them somewhat more deadly.  I cannot say for sure the extra 200 rpm when firing has increased German losses, but then I am using beefed up damage models across the board.  The increase Allied firepower may be too deadly for “fun” (TK’s guiding design theme) using stock aircraft data ini’s.

 

It’s worth noting that decreasing the ammunition load (as you mention was done) will have an impact in the simulation, as aircraft gross weight is calculated adding fuel, pilot and ammunition weights to the EmptyMass.  You can watch the gross weight drop in debug mode as fuel and ammo is used up.  Less weight improves flight performance.

 

Fuel tank capacity is something that needs to put on the list to cross check, especially in 3^rd party models.   Although I realize the information is difficult to track down, I’ve seen some MaxFuelAmount values that look suspect, and which could well be impacting flight performance.

[+VonS 1,424]

This is all very fascinating information, hadn't thought about the synchronization improvements too much but it makes sense that the rate of fire would increase by the second half of 1917 approximately, this will help further with ammo depletion and make FE2 more realistic. Have now tested the general aircraftaidata ini file further, with one-on-one dogfights with a DH.2 and Fokker E.IIa (comparable opponents) - and noticed that aces now shoot only when very near, sometimes being too conservative to shoot unless very close. Will readjust the maxcannonrange values a bit for more realism - perhaps something like 600 or 800 for the novices, 400 for the regular pilots, and 300 and 250 for veterans and aces - this will probably result in more variety and quicker ammo depletion (150 and 200 seem a bit too short from what I've tested, although I will admit that historically aces would open up only when right on your tail...although René Fonck is rumored to have been accurate even when shooting from a distance). Will post back with further results.

 

On an unrelated note, it dawned on me that it's more realistic to have separate aircraft folders for Belgian Nieuport 16 and 17s, and have them co-exist in the Western theatre with French Nieuports, also to have a separate folder for an RFC Nieuport 17 than only one folder where you have to switch nations in the game's menu. This will provide for more historical realism. Also, I will add a Belgian Hanriot to the Western theatre but designate it as VERY_RARE so that it shows up less than other aircraft types in that theatre. The Italian theatre has had a Hanriot for a long time already.

 

This will all be included in a 7.9 maintenance update. I will leave machine gun rate of fire, also gun belt lengths, for further testing, to be incorporated in an 8.0 update that will feature another 7 or 8 important aircraft for the Western theatre, as listed below:

 

Beardmore or French-built Nieuport 12

BE.12

Hannover CL.II

Hannover CL.III

Albatros C.III

Avro 504C

Farman F40

Vickers FB5

 

I may leave these aircraft for a 9.0 update and focus on revising gun belt lengths and rate of fire advancements in an 8.0 update (I think this may be more beneficial, to increase realism further).

 

Happy flying,

Von S

[+Geezer 3,569]

The code may not be up to this, but your quest for greater realism might include ammo feed problems?  Ever wonder why the flexible cloth ammo belts, which worked fine when used by the infantry, were replaced by more complex and more expensive metal links in aircraft?

 

Answer: the cloth belts would flex too much under high-G manuevering, resulting in a mis-feed and jam the guns.  Also, some ammo feed systems were poorly designed at first, which all but guaranteed a mis-feed.  French aircraft were especially bad in this regard - early Nieuports and SPADs used circular ammo drums, rather than ammo boxes.  The drum system in all Nieuport 11s and early batches of Nieuport 17s required the ammo to travel as much as three feet and also bend 90 degrees before feeding into the gun!  Later versions of the Nieuport and SPAD dropped the faulty drum design and used the more conventional ammo box located under the gun.  Shot below shows drum system in SPAD 7.

 

Smooth, reliable ammo feed has been a problem since WW1.  The B-52 tail turrets I worked on also had ammo feed problems.  Despite being propelled by powered starwheels all along a 10-foot length of flexible chute, ammo belts would usually kink and jam before all the ammo had been fired out.  Operational units considered 70% fire-out to be "acceptable."

Edited May 14, 2016 by Geezer

[+VonS 1,424]

This is great stuff Geezer, thanks for posting. I remember reading somewhere that the Germans continued using cloth belts well into 1918 but had ironed out some problems with the system, while the Entente was running away from the cloth belts because of all the problems. Maybe the Germans were trying to save weight on the planes by continuing with cloth instead of going with steel belts? An interesting theory that I'll think about for the 8.0 update of the FMs...maybe something along the lines of greater frequency of gun jams, and longer wait times to unjam would approximate these problems in FE2, especially with earlier planes. More jams with cloth fed ammo, fewer jams with the steel belts...but also to have shorter jams with steel fed ammo later in the war. I'll think about this in terms of smaller ammo loads and also the version of gun being used on the plane types (Vickers 1, 1.1, 2) since the rate of fire becomes faster later on....this will make it more exciting overall, with greater realism. Coupled with arbitrary engine failure, it looks very promising.

 

Happy flying,

Von S

[+Geezer 3,569]

Another possibility is ejecting empty cartridges.  Both the Spandau and Vickers were versions of the Maxim gun - all of them ejected spent cartridges down through the bottom of the gun.  Most aircraft collected the empties in boxes located beneath the gun, but some aircraft - mostly British - funneled the empties down a chute to exit the aircraft.  See upper right corner of Camel shot for ejection chute.

 

This is easy to accomplish in FE1/FE2 as a lot of the original SF code is still operational.  Example below is from my FE2 Gladiator.

 

 

[Gun01]
SystemType=FIXED_GUN
//GunTypeName=303CAL_VICKERS_MK1
GunTypeName=303CAL_Browning
InputName=FIRE_PRIMARY_GUN
MuzzlePosition=-1.881,0.334,-0.599
LightPosition=-0.5437,2.1606,0.0605
AimAngles=0.4,0.0,0.0
MaxAmmo=400
EjectShells=TRUE
EjectPosition=-1.973,-0.695,-0.70
EjectVelocity= 0.0,0.0,-2.0
//MinExtentPosition= -0.44, 0.05, -1.006
//MaxExtentPosition= -0.39, 0.15, -0.196

[+VonS 1,424]

Thanks Geezer for this info. I'll look into the Vickers on the Nieports too, can't remember now if those ejected cartridges as well, or collected them in a box...that's another great point that I will attempt to address in a version 8.0 update. I'm thinking that the five factors to address, for more historical realism, are:

 

(1) ejection or storage of empty cartridges

(2) length of cartridge belts (depending on period in the war used)

(3) gun jams and speed of unjamming (to simulate cloth or steel belts, also length of cartridge belts, shorter being better in terms of less jamming)

(4) rate of fire of synchronized guns (also of unsynchronized ones too, I remember reading that the Hotchkiss had an average rate of fire of something like 700 rounds/min but this was usually detuned to about 400 or 450 to minimize jamming)

(5) arbitrary chance of engine failure and/or fire (I'm thinking along the lines of fuel temp., overheating, things like that, not so much oil pressure loss since tampering with the oil settings under "Engine" in the data inis might lead to permanently poor engine performance during a flight, instead of the chance of problems arising...there is a +/- setting in the ReLoad program used with Red Baron 3D that helps regulate "air flow" to rotary engines, but such dynamic tampering might not be possible to that great an extent in FE2...)

 

Happy flying,

Von S

[+Geezer 3,569]

(4) rate of fire of synchronized guns (also of unsynchronized ones too, I remember reading that the Hotchkiss had an average rate of fire of something like 700 rounds/min but this was usually detuned to about 400 or 450 to minimize jamming)

 

Most of the aircraft gun systems used in WW1 were actually interrupters, not synchronizers.  The guns would shoot independently, but would be interrupted by a propeller blade - usually there was no synchronization of gun speed to engine/propeller speed.  The rate of fire was inhibited by the engine/propeller speed, which was slower than the gun's theoretical rate of fire.

 

Not sure if the code can simulate all of this complexity.  OTOH, you could have some fun.  Have a recording of the pilot cussing a blue streak while beating on his cocking levers with a hammer! 

[+VonS 1,424]

Most of the aircraft gun systems used in WW1 were actually interrupters, not synchronizers.  The guns would shoot independently, but would be interrupted by a propeller blade - usually there was no synchronization of gun speed to engine/propeller speed.  The rate of fire was inhibited by the engine/propeller speed, which was slower than the gun's theoretical rate of fire.

 

Not sure if the code can simulate all of this complexity.  OTOH, you could have some fun.  Have a recording of the pilot cussing a blue streak while beating on his cocking levers with a hammer! 

 

This is all fascinating info., I hope that I can incorporate some of it into further FM tweaks. Can't remember the exact post now on TheAerodrome, but a few years ago there was a big debate over there regarding average rate of fire for guns firing through propeller blades, considering that most props would rev exactly as much as the engine, with the exception of some of the sub-types of Hispano-Suiza that had reduction gears and the four-bladed propeller, later replaced with fixed gears once again. This of course is probably too much for the FE2 code to handle, to have faster rate of fire in a dive with engines revved, and to have slower rate of fire with engines cutting out or not giving full power - but I will see if some kind of "educated estimates" can be brought into the data files, to account for best and worst rate of fire.

 

Von S

Edited May 15, 2016 by VonS