PDA

View Full Version : Use of Control Surfaces Should Slow AI



VF2_Tim
11-18-2007, 06:03 PM
There are still some outstanding issues affecting playability of the sim.

Most notable is how an ai plane being chased gyrates and rolls repeatedly while stomping on the rudder. Let's forget for a minute that these are actions no sane pilot would take at any altitude and certainly not under 300 meters. Instead, the most startling thing is how that twisting plane walks away from a clean and straight-flying pursuer that is at full throttle.

Use of control surfaces slows a plane down. Use of multiple control surfaces slows it down badly. Anybody pursuing a plane that is in a rolling rudder skid will catch that plane.

After years of tweaking and player feedback it is a bit disappointing that issues like this have not been addressed.

Here are some other issues that need attention IMHO:

1. AI does not black out. If I'm in a turning fight against an ai and I'm losing that turning fight then the ai is pulling more g's than I am. Therefore it is impossible for me to blackout but the ai does not. Not only do they not black out but they obviously do not experience a loss of performance as they approach what should be a blackout condition.

2. AI planes do not experience aileron lock or reduced performance at high diving speeds.

3. AI planes fly impossibly straight lines and exhibit unnatural acceleration that make them uncatchable when they decide to disengage.

4. AI plane's performance is not inhibited by being in a cloud.

5. AI needs to be much smarter about when and how they engage enemies. They can at times ignore an enemy directly in their sites and yet they will shoot over a player's shoulder at a plane that is on fire or belching black smoke and obviously going down.

6. AI planes don't disintegrate under extreme diving speeds.

7. Negative G cutout is way too sensitive. Even the slightest forward stick can kill an engine. I doubt this is modeled correctly. Even if it is realistic, it's a terrible characteristic for a flight sim.

8. AI planes don't experience over heating.

9. Trims are not consistent from one mission to the next. i can trim out my plane for a mission, then fly a second mission with the exact same plane and the trim needs to be readjusted on my stick.

10. Many of the planes wobble when the stick is neutral. Seems to me that a plane cutting through the sky would naturally try to hold a pretty straight line. In fact, if the stick were neutral, the air flow would force the control surfaces flat and the plane would fly a straigt and steady line. Instead we work to get on six and then have to struggle to settle the plane down for a shot.

11. Healthy bombers will follow a wounded formation leader that is in a death spiral.

12. This one can't be proven but it is my opinion that in this sim the closer a plane is to the ground the more prone it is to stall. If true, this is the exact opposite of reality since the air is much denser the closer the plane is to the ground. It is also my opinion that recovery from blackout takes much longer when a plane is pointed towards the ground. Sure, sure, someone might answer this one by saying that it just seems that way because the pilot is intent on not crashing in both of these situations. Maybe so. If it is all in my head, I apologize. If, however, it is true and someone thinks that's good programming, I hope they will reconsider.

13. Planes don't appear to bleed enough speed in a turn. This gives the advantage to the angles fighters. While this probably makes for more interesting playability it sort of short changes the boom and zoom aircraft.

14. Tail gunners seem a bit too accurate. I understand that they are there to keep the game honest but maybe dial them back a bit? I'm no historian but I'd wager there has been a much higher tail gunner kill perecentage in this sim compared to the reality of the war.

15. The multiple 'go-around' landings by an ai squad need to be reworked. Space them out and bring them in trail. That's how the humans do it.

I understand that we are in the latter stages with this sim and that reworks are getting pretty slim. It just seems a shame that the game may be retired with these blemishes after all the years of tweaking that have gone into it.

Probably too late for major reworks of this caliber but maybe they can be considered in the design of BOB or other future sims.

Sorry so negative. Sorry so long. This sim is head and shoulders above the rest but it's still not as good as it could be.

Please understand, I took the time to post this because I love this Sim.

Thanks for your time and attention.

VW-IceFire
11-18-2007, 08:40 PM
Good post Tim. To the point but without the usual arrogance and bravado we see in posts like these.

I agree with most of what you've said...the AI do some wonky things and I'm starting to see many of these things. Some of them we do know about from the developers...particularly regarding clouds as the AI disregards them altogether. Dive speeds I'm not sure about but I do know that some planes roll rates at high speed when controlled by the AI is unreal to say the least. Oddly...some planes have no problem while others do.

I do disagree on negative G cut out. It may be a bit sensitive but it was a realistic issue for some planes and not others. This will undoubtedly be back in Storm of War as this was an issue for British fighters during the Battle of Britain and it wasn't until the invention of "Mrs Schillings Orifice" (not a joke) did the RAF fighters have any chance in a negative G push over. German fighters utilized fuel injection and didn't have the same level of problem.

I don't understand the issue regarding trim either. It is something I constantly change...all the time. As speed and altitude vary it needs to be changed to keep the plane flying straight.

The rest are definitely issues in one manner or another. Some of them can be solved with technology upgrades but I doubt we'll see them in this game. I hope Storm of War takes it to the next level and approaches some of these issues.

Lurch1962
11-19-2007, 04:29 PM
9. Trims are not consistent from one mission to the next. i can trim out my plane for a mission, then fly a second mission with the exact same plane and the trim needs to be readjusted on my stick.

10. Many of the planes wobble when the stick is neutral. Seems to me that a plane cutting through the sky would naturally try to hold a pretty straight line. In fact, if the stick were neutral, the air flow would force the control surfaces flat and the plane would fly a straigt and steady line. Instead we work to get on six and then have to struggle to settle the plane down for a shot.

===============================

9) I've found trim to be always consistent when the same plane/loadout are flown from mission to mission. I'm sure you realize that trim is a variable that depends upon airspeed, altitude, fuel quantity, and bomb load, among other factors. If you're starting a mission from takeoff, or even in the air, the trim that applies at the start will almost certainly differ from that which applied to your eventual cruise speed, etc.

10) Not sure what you mean by "wobble". But it seems that this point goes toward trim (or the lack of it in some axes). In the heat of combat one often changes speed considerably, and if trim hasn't been adjusted to suit, it can cause difficulty in holding a steady tracking shot.

--Lurch--

SeaFireLIV
11-23-2007, 06:50 PM
Some of what you say is right and some of what you say is either `I think` or wrong.

Points I noticed...

1.`Ai don`t black out`.

This is true, but Oleg has assured us that AI never makes a manouever that would black it out and this has made sense to me for a long while. Of course, it would be nice if AI could make drastic moves and black itself out too, but I guess this was Oleg`s compromise.

"7. Negative G cutout is way too sensitive. Even the slightest forward stick can kill an engine. I doubt this is modeled correctly. Even if it is realistic, it's a terrible characteristic for a flight sim."

How can it be a terrible characteristic of a flight sim if it`s realistic? It`s a simulater. When I first tried this in the I16 it was a pain, sure, but I learned how to avoid this (go inverted and dive) and now I still love the little plane. I fly with it a lot offline and on and get my fair share of kills with it now. It also taught me something about WWII early planes I didn`t know without having to read a boring history book and for that I am grateful. Give it time and it`s easy. It certainly should not be change to suit what you `think`.

"10. Many of the planes wobble when the stick is neutral. Seems to me that a plane cutting through the sky would naturally try to hold a pretty straight line. In fact, if the stick were neutral, the air flow would force the control surfaces flat and the plane would fly a straigt and steady line. Instead we work to get on six and then have to struggle to settle the plane down for a shot. "

I`m not an aeronautical engineer, but I know it`s not that simple. The air up there is not just a flat smooth surface, it undulates with layers that ebb and flow resulting in some form of turbulence no matter how small. You are in vintage WWII aircraft that produce twisting torque in most aircraft that you must trim to help it fly straight. these are just a couple of facters to consider.

Those were the main points i disagree with though there are less important others. Your other points about wounded bombers and AI not being affected at some dive speed are relatively true. Not seeing in cloud is true as well.

M_Gunz
11-23-2007, 08:35 PM
Originally posted by VF2_Tim:
1. AI does not black out. If I'm in a turning fight against an ai and I'm losing that turning fight then the ai is pulling more g's than I am.

That's all there is to it huh?

Don_X
11-24-2007, 02:34 AM
15. The multiple 'go-around' landings by an ai squad need to be reworked.
Don't you mean "crashes" ? http://forums.ubi.com/images/smilies/88.gif

VF2_Tim
11-25-2007, 07:45 PM
That's all there is to it huh?

Please explain.

VF2_Tim
11-25-2007, 08:07 PM
How can it be a terrible characteristic of a flight sim if it`s realistic? It`s a simulater. When I first tried this in the I16 it was a pain, sure, but I learned how to avoid this (go inverted and dive) and now I still love the little plane. I fly with it a lot offline and on and get my fair share of kills with it now. It also taught me something about WWII early planes I didn`t know without having to read a boring history book and for that I am grateful. Give it time and it`s easy. It certainly should not be change to suit what you `think`.

I am not disputing that negative cut out was an actual characteristic of several wwii aircraft. I question how this is modeled into the sim. If, for instance, i'm on a bandit's six and need less than a degree negative pitch to put him in the crosshairs then i should be able to get that without stalling an engine with no chance of a restart. This negative pitch may indeed cause a loss of fuel flow to the carb and therefore a cut out. But the electrical system is still working, airflow is keeping the prop turning and the turning crank is keeping the fuel pump going. (or so i assume) So when I use my tiny amount of neg pitch and get bandit back in my reticle i neutralize the stick, fuel flow returns and the feathered prop has all the momentum it needs to fire the cylinders and bring my engine back to life. Right? (i'll rely on the engineers to set me straight cuz i'm just assuming here) ok, i accept that i should lose some speed and rpms. But i just don't think it should kill an engine with no chance of restart.

Now a prolonged negative g maneuver would flood the cylinders and possibly cause an unrecoverable engine outage. but a 1/4 second nudge forward on the stick?

It just doesn't seem right to me.

VW-IceFire
11-25-2007, 09:08 PM
Depends on how much negative pitch actually...what your probably doing is trying to follow a bandit that just went negative G on you and you pitch down to follow...except that is enough negative G to starve the engine of fuel. If it was a prolonged problem rather than an instant one it wouldn't have likely warranted such a drastic response by the RAF during the Battle of Britain.

Although Wikipedia is by no means a fantastic source it is a good summary:

http://en.wikipedia.org/wiki/Miss_Shilling's_orifice

Good place to start...better places are well written history textbooks on the Battle of Britain or air combat in general. I'm not going to say with 100% certainty that the game does it completely right but my guess is that, like everything else in the sim, it does it more right than wrong.

Another thing to keep in mind that the sim does a terrible job at is that negative G over a prolonged period of time (several seconds) for ANY WWII era aircraft is bad because the oil will run out of the engine. For instance the P-38's flight manual says that flying inverted or maneuvering in the negative G for longer than 7 seconds will potentially damage the engines and that after 7 seconds the potential will increase until the engines are out of oil and they seize. This is a common mention in most manuals I've seen...even the fuel injected German planes aren't immune.

Point being that negative G cut out must have been an extremely quick and instant problem that needed solving while several (greater than 2 seconds in my mind) seconds long possibility of negative G cut out is so long that its then in the range of talking about loss of other things in the engine like oil and coolant and so forth. So by all reasoning it should be something that happens so quick to illicit such a response from pilots.

Case in point is Spitfire and Hurricane pilots who ignored the warnings, nosed over to follow a 109, and found their engines stopping. I cannot remember if, when reading about this, some of those pilots even crashed their planes or had to bail out.

The stop gap measure for the Merlins was rushed into service and then later improved and incorporated into standard production. So it must have been somewhat dire for them to do that.

Finally cutting out the engine will not cause it to die in the game. A Hurricane Mark I will start right up again. The I-153 and I-16 will not right away because they have no starter (these planes were started on the ground by a starter unit) but if you continue to nose over after the engine has stopped and gain about 200kph or more and then start the engine it will usually get going. If it doesn't keep pumping it because it will eventually. Nothings a total loss unless your really close to the ground.

Good luck! http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif

M_Gunz
11-25-2007, 10:56 PM
Originally posted by VF2_Tim:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">

That's all there is to it huh?

Please explain. </div></BLOCKQUOTE>

Starting with relative speeds and moving into what kind of pursuit you are flying, I have read
the same argument in a few sims where the poster disregards differences in path and speed.
Your statement as I quoted it is just such a blanket statement as those in the past.
When the pursued slows down through bleed more than the pursuer he will pull inside and may
even appear to loop over and be on the tail of the pursuer in seconds when all he did was
to get inside and then relax the stick a bit yet that has fooled many players but then it
has been the death of more than a few pilots IRL.

Do you know that the AI only has controls to the planes that fly the same way AI or human?
Do you know that AI doesn't overheat because they control engine as we can to not overheat?
That AI uses same trim and stick we do only with inhuman skill?
That AI does does not black out but can ride the edge by numbers?

Make some tracks, ntrk if you can and run them through playback in slow motion if need be.
Switch your plane to combat AP and leave it there, make a track of that.

AI does see in the dark and through clouds. They shoot with the precision that they fly with
BUT they don't have the best reaction times especially from 200+m, they 'think' poorly and
they are often predictable as in every other combat flight sim.

M_Gunz
11-25-2007, 11:02 PM
Originally posted by VW-IceFire:
For instance the P-38's flight manual says that flying inverted or maneuvering in the negative G for longer than 7 seconds will potentially damage the engines and that after 7 seconds the potential will increase until the engines are out of oil and they seize.

IIRC (been years) Lockheed training film at Zeno's Warbirds says more than 10 seconds causes damage.

Where the sump is and what kind of oiling system controls this, it is not a universal problem.

Letting the prop drive the engine is more of a universal problem due to where the oil enters
the bearings.

Don_X
11-26-2007, 08:02 AM
Originally posted by VF2_Tim:
I question how this is modeled into the sim. If, for instance, i'm on a bandit's six and need less than a degree negative pitch to put him in the crosshairs then i should be able to get that without stalling an engine with no chance of a restart. This negative pitch may indeed cause a loss of fuel flow to the carb and therefore a cut out. But the electrical system is still working, airflow is keeping the prop turning and the turning crank is keeping the fuel pump going. (or so i assume) So when I use my tiny amount of neg pitch and get bandit back in my reticle i neutralize the stick, fuel flow returns and the feathered prop has all the momentum it needs to fire the cylinders and bring my engine back to life. Right? (i'll rely on the engineers to set me straight cuz i'm just assuming here) ok, i accept that i should lose some speed and rpms. But i just don't think it should kill an engine with no chance of restart.

Now a prolonged negative g maneuver would flood the cylinders and possibly cause an unrecoverable engine outage. but a 1/4 second nudge forward on the stick?

It just doesn't seem right to me.

If you close the throttle your nose will dip anyway + you can then apply negative G without the engine cutting out.

BrotherJayne
12-22-2007, 01:21 PM
Originally posted by M_Gunz:
Do you know that the AI only has controls to the planes that fly the same way AI or human?
Do you know that AI doesn't overheat because they control engine as we can to not overheat?
That AI uses same trim and stick we do only with inhuman skill?


Really? From what everyone told me when i started this same kind of thread about 6 months ago, the AI has a simplified flight model, which it in turn flys perfectly.