Jump to content
bies

Auto level autopilot is single engine airplanes which gives additional speed is being abused in multiplayer

Recommended Posts

So, my dear de-bunkers, I decided to do a quick test, so we would not need to have the same ping-pong for the next 8 pages, but could have something new to argue about. According to my own personal test, your reference test that de-bunks everything has been de-de-bunked. Results were pretty much in line with Dietrich's tests. 

 

Short description of test procedure:

Went to a quick mission, picked Bf109G2, full fuel, a single plane mission, altitude 4000 m, full throttle.

First test:  Find out optimal stab trim for level flight. Result -80%.

Second test: Start with -100% stab trim. After achieving top speed, change stab trim to -80%. After a couple of minutes disengage autolevel and fly manually.

Results: After 3 minutes with -100% stab trim: 495 km/h. Trim to -80%, two minutes later: 493 km/h. Disengage autolevel, two minutes later: 489 km/h.

Third test: Tested also flying manually from the beginning, same conditions as before. Could get to 486 km/h at max, but speed varied between 480-486 km/h.

 

Pictures of the second test below. Times when I changed trim and disengaged auto-level can be seen on cockpit clock.

2vtupo0.jpg334rx2t.jpg99jfnt.jpg

  • Thanks 1
  • Upvote 1

Share this post


Link to post
Share on other sites

So, around 10-15km/h maximum advantage doing the auto-level - that is 3m/s difference. Not enough to get out of the enemy's gun solution but very nice for cruising. Airplanes with full trims should gain less with auto-level on.

Share this post


Link to post
Share on other sites

Wow, what a thread! Lets see what happens now. Kemp Thaks for the test, and your patience and logic! :)

  • Upvote 1

Share this post


Link to post
Share on other sites

I'd like to see II./JG77_Kemp's test repeated by other people, in conditions where you can actually be sure the aircraft is flying level. I've tried it myself, on the Kuban map over the sea (so the HUD gives a meaningful altitude figure: it reads 'height over the ground' which isn't a lot of use over a non-flat surface). I found it difficult to maintain level accurately (in these conditions, the elevator and trim are very sensitive) but as far as I could tell, if you kept the height accurate, the speed matched the autolevel one. Any climb slowed it down, and any dive speeded it up. The only way I could slow by 10 km/h was by climbing.

Edited by AndyJWest

Share this post


Link to post
Share on other sites
4 hours ago, AndyJWest said:

I'd like to see II./JG77_Kemp's test repeated by other people, in conditions where you can actually be sure the aircraft is flying level. I've tried it myself, on the Kuban map over the sea (so the HUD gives a meaningful altitude figure: it reads 'height over the ground' which isn't a lot of use over a non-flat surface). I found it difficult to maintain level accurately (in these conditions, the elevator and trim are very sensitive) but as far as I could tell, if you kept the height accurate, the speed matched the autolevel one. Any climb slowed it down, and any dive speeded it up. The only way I could slow by 10 km/h was by climbing.

 

You cannot do this easily using full engine management because of the engine time limits. Kemp has to use the G-2 because it's maximum boost is limited to 1.3ata which allows 30 minutes full throttle. With  other 109s full throttle will put the MP into zone where the engine will start to fail before he can even stabilize a top speed never mind complete his tests. (Hence why this whole discussion is entirely theoretical).

 

He does not say what map he is on or what conditions - looks like Stalingrad summer, so I went with that: 12.00 as per the clock.  As you progress throughout the mission the radiators are on auto and gradually open during the test as the temperatures pick up. No way to tell Kemp's radiator settings: as he does not show the required data. And as you say, it is extremely hard to tell if the plane actually is level or climbing slightly while under manual control.   

 

What Zacharias' test on his pre-teen "G-6" demonstrated is that the auto-level button does not affect stab trim, and that stab trim affects speed with auto-level on. Kemp's pictures 1 and 2 merely confirm this: ( I take picture 1 to be auto-level at 100% stab, 2 as 80% stab, and 3 as manual flying at 80% stab. Although this is also a puzzle since he claims that 80% is optimum yet the speed in picture 2 is lower. But this may be incorrect, since the pictures are not properly labelled).  

 

But what Kemp is trying to prove with his tests in picture 3 is that he cannot get the same top speed manually that he can get with auto-level, which is an entirely different issue. Unfortunately his screenshots do not show this - they merely show that he was not at the same top speed at the time the screenshot was taken, but we do not know why. The merest twitch while pressing the auto-level button can lose - or gain - 4kph.    

 

 

7 hours ago, Ehret said:

So, around 10-15km/h maximum advantage doing the auto-level - that is 3m/s difference. Not enough to get out of the enemy's gun solution but very nice for cruising. Airplanes with full trims should gain less with auto-level on.

 

No - at the same stab setting, which is pictures 2 and 3, the difference is 493-489 = 4 kph, not 10-15.    And 495-489 = 6 kph. About 10-15 kph my foot!

Edited by unreasonable

Share this post


Link to post
Share on other sites

Unreasonable, I know exactly what Kemp was trying to demonstrate. I was comparing the IAS I got with autolevel on and off. I wasn't comparing my airspeed measurements directly with Kemps (though they weren't significantly different from his autolevel on results). And I did the test under 'normal' settings, since this seemed the best way to ensure consistent results.  

 

 

 

 

Share this post


Link to post
Share on other sites

I just replied to your post since you mentioned the speed difference you got - the discussion below not specifically aimed at you, just my thoughts on his test post.  Should have replied to him, sry!

 

When I ran a test just now at what I though were Kemp's settings I also could not tell if I was faster or slower on manual, the speed difference being so small and sensitive.

 

Share this post


Link to post
Share on other sites
1 hour ago, unreasonable said:

No - at the same stab setting, which is pictures 2 and 3, the difference is 493-489 = 4 kph, not 10-15.    And 495-489 = 6 kph. About 10-15 kph my foot!

 

I was trying to read the best case from provided data... If it is just few km/h then it is a non-issue - even more so than it was not at 10-15. From my experience flying airplanes with full trims (P-40, P-39 and La-5/F) when coordinated the difference was imperceptible.

Share this post


Link to post
Share on other sites
19 minutes ago, Ehret said:

 

I was trying to read the best case from provided data... If it is just few km/h then it is a non-issue - even more so than it was not at 10-15. From my experience flying airplanes with full trims (P-40, P-39 and La-5/F) when coordinated the difference was imperceptible.

 

Fair enough - it seems like a non issue to me too. I would be much more worried about the timer issue in anything other than a G-2, rather than an extra (maybe) few kph.

 

(My foot much better now :) )

Share this post


Link to post
Share on other sites

Andy and unreasonable, thank you for your contributions and additional tests. I hope more people will test it, so we get more data about the matter.

While your tests did not provide numerical data or de-bunk the speed claims of OP, they provided valuable information otherwise. Namely these experiences:

7 hours ago, AndyJWest said:

I found it difficult to maintain level accurately

 

2 hours ago, unreasonable said:

it is extremely hard to tell if the plane actually is level

 

They seem to be more in line with OP's claim that "Using autopilot you are perfectly efficient beyound human capabilities" and de-de-bunk some of the de-bunkers claims like "computer isn’t doing anything that you can’t already do". And all that while you were specifically trying to fly level, in contrast to the mentioned advantages of being able to scan surroundings or figure out your whereabouts or fly through the clouds without artificial horizon.

Share this post


Link to post
Share on other sites

Since it is apparent at this point that Kemp is only interested in dragging out this discussion endlessly, and is incapable of actually producing any meaningful evidence to back up his claims (even his own screenshots seem to illustrate that the aircraft was climbing in his test), I have added him to my ignore list. I suggest that others do the same. I'm sure the developers will, since they don't modify code on the basis of 'exploits' that have been repeatedly shown to be insignificant.

Share this post


Link to post
Share on other sites

Now this starts to be a bit silly.

49 minutes ago, AndyJWest said:

Since it is apparent at this point that Kemp is only interested in dragging out this discussion endlessly

After 8 pages of ping-pong I actually did a test and presented my results, as requested, and that is the point that made it apparant to you that I only want to drag this discussion endlessely? Interesting counter move, even if not very logical in my mind.

 

49 minutes ago, AndyJWest said:

incapable of actually producing any meaningful evidence to back up his claims

Hmm ... you really find the de-bunkers "evidence" somehow more meaningful than my test results (which are in line with Dietrich's test results) and logical explanations of auto-level advantages, given in this thread?

 

49 minutes ago, AndyJWest said:

even his own screenshots seem to illustrate that the aircraft was climbing in his test

I was concentrating on keeping it level and ball in the middle and would say that I was able to do it relatively well. Maybe some people can do it better, some might be worse at it. Feel free to do your own tests to de-de-de-bunk my test results. If you are referring to the altitudes on the pictures, that is ground altitude. Feel free to test it yourself, just go to quick mission and fly on autolevel and see that the altitude value is changing all the time.

 

49 minutes ago, AndyJWest said:

I have added him to my ignore list

That is a nice way to disengage from discussion. I am sure it proves you were right about everything.

 

49 minutes ago, AndyJWest said:

I suggest that others do the same. I'm sure the developers will

Can you be sure that developers will put me on ignore list for running a simple test? Maybe you should help them understand the need for it:

On 21.5.2018 at 10:57 AM, AndyJWest said:

Make a concrete proposal, indicating exactly what change you are advocating,  presenting verifiable evidence for why the change is necessary, and a full description of what the side effects are going to be.

 

Edited by II./JG77_Kemp

Share this post


Link to post
Share on other sites
3 hours ago, II./JG77_Kemp said:

Now this starts to be a bit silly. After 8 pages of ping-pong I actually did a test and presented my results, as requested,  <snip>

 

True, and true.

 

However, there are three questions all tangled up in the debate about whether auto-level should be allowed at full throttle, the most important of which you are still refusing to address or acknowledge. 

 

1) Does auto level allow the pilot to fly level without carefully looking at his instruments, etc thus losing focus on other things?

2) Does auto-level allow the player to fly faster while level than manual flying while level?

 

For completeness:

 

2a) Does auto-level allow the player to fly faster level than while climbing manually at the same settings?

2b) Does auto-level force the player to fly more slowly than while diving at the same settings?

 

3) Is using auto-level advantageous in combat?

 

On 1) I think the answer is obviously yes. It is very hard to be sure you are flying exactly level manually, so if you are concerned about this of course you will be using attention that you cannot be using to look around and so on.

 

On 2) the tests so far are inconclusive since it is  so hard to tell if one is exactly level while not in auto-level. As previously pointed out, this can only be done over water and over time, so your screenshots demonstrate nothing.  In principle, if the answer is in fact yes, then that implies that engaging auto-level is not just turning on an auto-pilot affecting the control surfaces, it is actually changing the FM of the aircraft. This I find implausible given the developers' assurances that AI does not change FMs but merely flies them. So for me the burden of proof lies with those who claim that auto-level is actually faster. So the OP is at best unsupported by evidence, at worst simply wrong about additional speed.

 

2a and 2b are clearly "Yes" and "No" respectively. 

 

The OP claims that auto-level is an exploit. In my eyes, an exploit is something that gives an unfair advantage. We may agree as per question (1) that auto-level allows the player to fly level without losing SA, and (2) that if not really faster it is certainly no slower, but if flying level is in fact a disadvantage, as every fighter doctrine asserts, how is this an exploit?

 

The only situation I can think of where auto-level is really advantageous would be flying through cloud, but is this the real cause of the OP's complaint? Suppose that auto-level automatically disengaged when a cloud is entered. Would this calm the complainants? I very much doubt that.

 

I think the developers should leave it in. I recall at school we all complained about being made to have short hair cuts. On of the younger and more progressive teachers explained to some of us that if we had a short hair-cuts rule we would complain about that and test the boundaries with hair over the collar etc. If we were allowed to grow hair as long as we wanted (this was the early seventies :) ) we would then be forced to find something else to complain about and rebel against. So before you know it we would all be dope fiends like the boys at Westminster. 

 

Experience shows that MP always produces loads of complaints, so it may as well be about something as trivial as this.

 

  

 

Share this post


Link to post
Share on other sites

I personally think that auto level should be removed from Expert difficulty if plane does not really have it.

Share this post


Link to post
Share on other sites
24 minutes ago, Godspeed said:

I personally think that auto level should be removed from Expert difficulty if plane does not really have it.

So much for level bombing in the Peshka.

 

I personally think that if you cannot properly fly and operate an aircraft, autolevel must look like an exploit to you.

Share this post


Link to post
Share on other sites
9 hours ago, hrafnkolbrandr said:

If you're worried about dudes chasing you down in autolevel, or dudes running away in autolevel; why don't you just use autolevel?

 

This is the most important point to me. If everyone can abuse the same "cheat" then is it really a cheat? People earlier in the thread already said that this is really only a "problem" in long chases. Those are hardly the entirety of air combat. If someone is using this "cheat" in those cases, just make an exception and use it yourself too. Problem solved. Why all the fuss to lower a plane's speed in auto-level by some arbitrary amount just so people can catch up or run away without auto-level? I'd bet the majority of times a plane's getting away or catching up it has very little to do with auto-level anyway. 

Edited by obit

Share this post


Link to post
Share on other sites

Unless you are approaching critical altitude (unlikely given the way multiplayer combat usually goes), in a 'long chase' of an aircraft with similar performance that stays level, you are probably better off climbing slightly initially, rather than staying at the same altitude as the plane you are chasing. You'll get a little behind to start with, obviously,  but then benefit from increased TAS and start to catch up again. And then have the height advantage as you catch up.

Share this post


Link to post
Share on other sites
5 hours ago, ZachariasX said:

I personally think that if you cannot properly fly and operate an aircraft, autolevel must look like an exploit to you.

 

Bashing other poster's flying skills, based on nothing, is an interesting approach in an argument, especially it has nothing to do with logic.

 

Let's try to have a little logical excercise.

Let's imagine we have people that could navigate without GPS and people that could not navigate without GPS. Which group is more likely to fight for keeping GPS?

Now, let's imagine we have people that could fly their planes efficiently manually and people that could not fly their planes efficiently manually. Which group is more likely to fight for super-efficient autolevel in a competitive MP environment?

Share this post


Link to post
Share on other sites
22 hours ago, II./JG77_Kemp said:

So, my dear de-bunkers, I decided to do a quick test, so we would not need to have the same ping-pong for the next 8 pages, but could have something new to argue about. According to my own personal test, your reference test that de-bunks everything has been de-de-bunked. Results were pretty much in line with Dietrich's tests. 

 

Short description of test procedure:

Went to a quick mission, picked Bf109G2, full fuel, a single plane mission, altitude 4000 m, full throttle.

First test:  Find out optimal stab trim for level flight. Result -80%.

Second test: Start with -100% stab trim. After achieving top speed, change stab trim to -80%. After a couple of minutes disengage autolevel and fly manually.

Results: After 3 minutes with -100% stab trim: 495 km/h. Trim to -80%, two minutes later: 493 km/h. Disengage autolevel, two minutes later: 489 km/h.

Third test: Tested also flying manually from the beginning, same conditions as before. Could get to 486 km/h at max, but speed varied between 480-486 km/h.

 

Pictures of the second test below. Times when I changed trim and disengaged auto-level can be seen on cockpit clock.

2vtupo0.jpg334rx2t.jpg99jfnt.jpg


My understanding from this is that the difference in max speed between autolevel max and full-on manual maximum is 9 km/h (495 km/h -486 km/h). If one uses autolevel to reach max speed, and then switches to manual, 2 minutes later the speed has dropped by 4 km/h. So, in a 109, the difference between someone who uses autolevel and their pursuer is 4-15 km/h. I'm using the max value provided of 486 km/h, since this was a number you were able to achieve and, in a focused pursuit, I'm assuming any pilot will be focusing all their energy on achieving best speed. The worst possible value would be 15 km/h. The best is 4 km/h. The middle value is 9km/h.

Unfortunately the test does not include radiator settings. If the radiators are on automatic, that could be confounding the data. EG, since the manual test was done after the autolevel tests, the engine is likely at a higher temperature, and therefore will have more open radiators, causing increased drag. I'm assuming all three tests were done in the same mission. If they were done in a separate mission that might eliminate the issue, if this is the case with this data then that may remove this as a confounding factor. Radiator changes may account for why the best speed stab trim doesn't get better speed than the 100% setting as tested.

What we need to see is something from the manual test to ensure that the plane is level at the time, and the radiators being set manually for consistency.

Compare this with other tests that users have done, we have several tests showing minimal to no difference, and some showing small differences. Some of those tests have been done with 109s, others with planes with 3 axis trim. The 3 axis trim planes show no difference, or close to none, when the plane is properly trimmed. The 109, due to lacking aileron and rudder trim, cannot be set that way, and so must be manually corrected for. This manual correction can introduce excess drag. This may put planes like 109s, yak-1b and I-16 without rudder trim at a disadvantage. 

If we take the test at face value, at worst case scenario we can see a 15 km/h advantage in top speed to an autoleveling. And even within the same test, perhaps only a 4km/h top speed. 
Now the question: does that difference translate to any actual, fundamental difference in-game when fleeing/pursuing in a straight line?

Lets subtract 15km/h from the top speed of a G2 and compare it to it's best performing likely foes. All info is from tech specifications provided by devs as the max performance in game.

Sea Level:
109G2: 530 km/h - 15 = 515 km/h
Yak-1 S. 69: 514 km/h
Yak-1b: 530 km/h

Result: Yak-1 has no advantage, Yak-1b has advantage (15 km/h).

2000m:
109G2: 577 km/h - 562 km/h
Yak-1 S.69: 549 km/h
Yak-1b: 567 km/h

Result: yak-1 has no advantage, Yak-1b has advantage (5 km/h)

Once you're higher than 2000m the advantage increasingly goes to the 109s.

So what we see from this is the following:

If the above tests are accurate, and we take the worst case scenarios, we have the best possible opponent for the 109G2 being able to catch or flee from a 109G2 at sea level, having a possible very slight advantage at 2000m, and probably no advantage beyond that.

So a guy in a 109G2 who tries to escape a Yak-1b will get caught at sea level, possibly get caught at 2000 m, and everything else he will toodle away with impunity. And this is at combat settings. With emergency power a 109G2 could pull away for a while and gain separation, or catch his enemy. A vanilla yak-1 gets caught or left in the dust even with a 15 km/h boost.

Assuming these tests are correct (and we have some caveats, as we haven't seen radiator settings or screenshots of the manual flight, which could account for much of the discrepancy), and taking the absolute worst case of the results presented, against the best-performing likely opponent, we have a minimal impact on gameplay. If we take the best case scenario (4 km/h), the difference evaporates. 

Clearly, anyone who tries to autolevel away from a 109G2 is taking their life in their hands, since the G2 will catch them on combat power unless they're a 1b. On emergency power the 109 will probably catch them anyway.

The 109, meanwhile, can get away from most of its opponents at any altitude except very, very low, even with a significant speed disadvantage from not autoleveling. Only the 1b can catch you, and that's at sea level.

So we can say that, even in the worst case, autolevel doesn't change the dynamics of the game much, if at all. The Yak-1 still can't catch you, the Yak-1b might catch you down low, and the La-5 series 8 always could catch you down low. Flying in a straight line for prolonged periods is nigh-on suicidal anyway, so anyone doing this for advantage has a short life expectancy. 
 

Share this post


Link to post
Share on other sites
1 hour ago, RedKestrel said:

...
If the above tests are accurate...

 

They aren't. See my post in response.  If you repeat the test over the sea on the Kuban map, where you have an accurate altitude measurement in the HUD to work from, you will find that that keeping altitude accurately is difficult, but that if you do, the difference between autolevel on and autolevel off speeds is negligible. The only way you can reproduce Kemp's result is by climbing. Try it for yourself.

 

 

Share this post


Link to post
Share on other sites
Just now, AndyJWest said:

 

They aren't. See my post in response.  If you repeat the test over the sea on the Kuban map, where you have an accurate altitude measurement in the HUD to work from, you will find that that keeping altitude accurately is difficult, but that if you do, the difference between autolevel on and autolevel off speeds is negligible. The only way you can reproduce Kemp's result is by climbing. Try it for yourself.

 

 

I don’t have the Kuban map so unfortunately I can’t.

 

either way, it doesn’t make a difference. The small advantage his tests “proved” doesn’t change the game enough to be exploited.

Share this post


Link to post
Share on other sites
2 hours ago, Trooper117 said:

I bet if I come back here in 10 years time this effin bollox will still be going... 

 

I am beginning to think you actually may be right.

This is unbelievable...

 

 

  • Like 1

Share this post


Link to post
Share on other sites
19 hours ago, ZachariasX said:

So much for level bombing in the Peshka.

 

Good point some exceptions must be.

What i mean is single engine fighters can use this for advantage and if people want realism it should be removed.

 

Edited by Godspeed

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×