Pilot Training Alogrithm (Full Version)

All Forums >> [New Releases from Matrix Games] >> War in the Pacific: Admiral's Edition



Message


PaxMondo -> Pilot Training Alogrithm (7/18/2011 3:10:40 PM)

Wanted to have a bit of discussion on this alogrithm. The current alogrithm has the rate of skill learning inversely tied experience. In the game what this causes is for players to train low exp pilots in all the skills they want first and then focus on gaining their exp.

Not entirely inaccurate, but a bit counter-intuitive. You would think that higher exp pilots would be able to gain skills at least as fast as low exp pilots. In fact, you could make a good arguement that higher exp pilots would gain such skills quicker.

The gameplay issues are numerous. For example: You can't take your experienced pilots in naval search (PBY's) and use them in an ASW role (essentially ever as they gain ASW skill so slow). You have to wait until your new pilots train ASW and then gain exp.

My proposal, subject to discussion here and general approval by this august body, would be to modify the pilot traiing alogrithm slightly.

1. skill increase to be dependent only upon current skill level, not pilot exp. (change)
2. skill increase to be faster at lower skill levels. (no change)
3. skill cap to be 70. (no change)

In gameplay, this will slightly favor the IJ in the early game as it starts with more exp pilots, and will then benefit the allies in the later war as their pool of exp pilots expands. In other words, fairly minor impact overall. But will remove a bothersome (at least to me) dichotomy.

Thoughts?




Rob Brennan UK -> RE: Pilot Training Alogrithm (7/18/2011 3:24:49 PM)

I would agree with all 3 points . seconded.

One oddity I can understand from a game play perspective but not real life is the split between low bombing and regular bombing (ground esp but also naval upto a point) , you can have elite bomber crews who can 'put a bomb in a pickle barrel at 15k' but drop down to 5,999 feet and they lose the ability to hit anything at all until a multi-month training program has been attended. very odd.

As for favouring any side , im not going to go there as its a flame magnet [8|].




PaxMondo -> RE: Pilot Training Alogrithm (7/18/2011 3:33:09 PM)


quote:

ORIGINAL: Rob Brennan UK

As for favouring any side , im not going to go there as its a flame magnet [8|].

I know, and agree. But as I was putting forth the proposal, I felt it appropriate to cover all aspects of the proposal including this item. Hopefully, I stated it simply enough to not draw an overt number of flames ... we'll see.




LoBaron -> RE: Pilot Training Alogrithm (7/18/2011 4:13:31 PM)

Agree with the issue, but not neccesarily with the proposed solution.

In my opinion pilot training is too fast (yes, I know the current training speed is because people said it was too slow... [8|] ).

The problem there is that, as Pax Mondo already stated, training gain slows with exp gain. This renders initial replacement pilot skill next to useless.
The difference between training up a 10exp pilot compared to a 30exp pilot is neglectable, and so works against the historically finetuned replacement pilot
exp increase of the Allies and decrease of the Japanese.

Houserules against pilot training will not work because for this the feature is too integrated into the game.

The result is retort pilots, with a very similar skill level, and lots of thought, time and micromanagement needed to get the pilot on top in those skills the player believes neccesary.

I believe the solution could look like this:

1) Significantly reduce pilot training speed:

And I mean significantly. Make the gain of 10 points in skill a task of 40-50 days.
If someone objects against this proposal, I understand, but please read my other points before jumping conclusions.

I donīt believe for a second that pilots fresh out of combat training were facelevel to veterans. The get there by fighting.

2) Reduce the impact of exp on training speed and lower the training cap to around 50:

Skill is gained by surviving in a combat enviroment. I always considered a 70air/60def skill fighter pilot is the equivalent of
an experte pilot with a high number of combat missions flown.

Currently you can reap those pilots in high numbers for both sides without needing to fly a single combat mission.
With the current training system you get 70/60 in 2-3 months, these are pilots that were called "green" in WWII!

3) Increase the exp/skill gain for successful combat missions flown, maybe put a high random factor in here.

4) Declutter pilot training:

I do not have any figures here but I believe that the training settings are too closely related to mission flown.
There should be some general training options (3-4) with the rest coming from actually flying such missions.

If a pilot trains for becoming a fighter pilot, why do I have to separately train him for defensive maneuvering?
For naval attack, why do I have to train him on naval search and lowN?
For ASW, why do I have to train him on naval search and lowN or Naval bombing?
For ground attack, why do I have to double train depending on the altitude the pilot flies? Why is there not even a slight increase in naval attack skills?
For recon, why do I have to train the pilot on nav search?

I know that - for the above trainings - separate pilot courses are needed, true. Its different whether you have to identify ship types of take pics of ground formations.
And its different to attack at 10k compared to wavetop height or to hunt a sub compared to attacking a BB.

But this should be an incremental part of such a training, not a minutious decision by Admiral Nimitz.


What would I love to have in the end?
Lower skill/exp pilots, with more broadly distributed skills, max exp of 50, the classic rookie pilot fresh out of combat training.
Then, depending on the main mission of the unit the pilots get assigned to, the exp and skills will increase further, but only as a result of participating on combat missions on the frontline.

This could help making pilot training a bit less important and micro than it currently is, and leave the 70-90 exp/skills to those pilots that historically
were only few and valued.


(said some months ago already, but something similar applies to some leadership values IMO. A leader with 86 aggression rating on a scale from 0-100 is a raging maniac, nothing else, but that leads us a bit off topic [;)] )





herwin -> RE: Pilot Training Alogrithm (7/18/2011 5:22:41 PM)

quote:

ORIGINAL: PaxMondo

Wanted to have a bit of discussion on this alogrithm. The current alogrithm has the rate of skill learning inversely tied experience. In the game what this causes is for players to train low exp pilots in all the skills they want first and then focus on gaining their exp.

Not entirely inaccurate, but a bit counter-intuitive. You would think that higher exp pilots would be able to gain skills at least as fast as low exp pilots. In fact, you could make a good arguement that higher exp pilots would gain such skills quicker.

The gameplay issues are numerous. For example: You can't take your experienced pilots in naval search (PBY's) and use them in an ASW role (essentially ever as they gain ASW skill so slow). You have to wait until your new pilots train ASW and then gain exp.

My proposal, subject to discussion here and general approval by this august body, would be to modify the pilot traiing alogrithm slightly.

1. skill increase to be dependent only upon current skill level, not pilot exp. (change)
2. skill increase to be faster at lower skill levels. (no change)
3. skill cap to be 70. (no change)

In gameplay, this will slightly favor the IJ in the early game as it starts with more exp pilots, and will then benefit the allies in the later war as their pool of exp pilots expands. In other words, fairly minor impact overall. But will remove a bothersome (at least to me) dichotomy.

Thoughts?



In reality, any numerical measure of human expertise grows logistically over about 4 years (10,000 hours) of full-time experience. That is, if it starts out at A at the lower limit and ends at B at the upper limit, it has the form B*A*(exp(2*years))/(B+A*(exp(2*years)-1)).

A nugget with about a year of training still has quite a way to go. Specialised skills grow the same way.




Insano -> RE: Pilot Training Alogrithm (7/18/2011 5:55:14 PM)

I believe the inverse relation to skill training is related to number of sorties flown and not the experience. In most cases these are the same thing but not in the case of starting a new game. Try it on some experienced pilots at game start before their sorties go from 0 to 1 and see if you agree with me.

In general I find it too difficult to get 85+ experience pilots and almost impossible to get 90+ experience pilots. I agree with the sentiment to keep the training levels down but it should be easier to increase through combat. Of course that brings up a problem with the game engine that it can't tell the difference between bombing a helpless ground unit far behind the lines and going in hot against a heavily defended target. The current system we have must be a kind of compromise about this. With that in mind I think any possible changes should be very, very small in order to not 'break' something else.




crsutton -> RE: Pilot Training Alogrithm (7/18/2011 6:15:18 PM)

Or, just get rid of this horrible "time suck" of a feature and have new pilots come into the game with the basic skills needed to have a chance of surviving and let them gain experience in combat. Saving me endless hours of clicking and allowing me a chance to rediscover my life again...[;)]




PaxMondo -> RE: Pilot Training Alogrithm (7/19/2011 4:17:21 AM)


quote:

ORIGINAL: LoBaron

Agree with the issue, but not neccesarily with the proposed solution.

In my opinion pilot training is too fast (yes, I know the current training speed is because people said it was too slow... [8|] ).

The problem there is that, as Pax Mondo already stated, training gain slows with exp gain. This renders initial replacement pilot skill next to useless.
The difference between training up a 10exp pilot compared to a 30exp pilot is neglectable, and so works against the historically finetuned replacement pilot
exp increase of the Allies and decrease of the Japanese.

Houserules against pilot training will not work because for this the feature is too integrated into the game.

The result is retort pilots, with a very similar skill level, and lots of thought, time and micromanagement needed to get the pilot on top in those skills the player believes neccesary.

I believe the solution could look like this:

1) Significantly reduce pilot training speed:

And I mean significantly. Make the gain of 10 points in skill a task of 40-50 days.
If someone objects against this proposal, I understand, but please read my other points before jumping conclusions.

I donīt believe for a second that pilots fresh out of combat training were facelevel to veterans. The get there by fighting.

2) Reduce the impact of exp on training speed and lower the training cap to around 50:

Skill is gained by surviving in a combat enviroment. I always considered a 70air/60def skill fighter pilot is the equivalent of
an experte pilot with a high number of combat missions flown.

Currently you can reap those pilots in high numbers for both sides without needing to fly a single combat mission.
With the current training system you get 70/60 in 2-3 months, these are pilots that were called "green" in WWII!

3) Increase the exp/skill gain for successful combat missions flown, maybe put a high random factor in here.

4) Declutter pilot training:

I do not have any figures here but I believe that the training settings are too closely related to mission flown.
There should be some general training options (3-4) with the rest coming from actually flying such missions.

If a pilot trains for becoming a fighter pilot, why do I have to separately train him for defensive maneuvering?
For naval attack, why do I have to train him on naval search and lowN?
For ASW, why do I have to train him on naval search and lowN or Naval bombing?
For ground attack, why do I have to double train depending on the altitude the pilot flies? Why is there not even a slight increase in naval attack skills?
For recon, why do I have to train the pilot on nav search?

I know that - for the above trainings - separate pilot courses are needed, true. Its different whether you have to identify ship types of take pics of ground formations.
And its different to attack at 10k compared to wavetop height or to hunt a sub compared to attacking a BB.

But this should be an incremental part of such a training, not a minutious decision by Admiral Nimitz.


What would I love to have in the end?
Lower skill/exp pilots, with more broadly distributed skills, max exp of 50, the classic rookie pilot fresh out of combat training.
Then, depending on the main mission of the unit the pilots get assigned to, the exp and skills will increase further, but only as a result of participating on combat missions on the frontline.

This could help making pilot training a bit less important and micro than it currently is, and leave the 70-90 exp/skills to those pilots that historically
were only few and valued.


(said some months ago already, but something similar applies to some leadership values IMO. A leader with 86 aggression rating on a scale from 0-100 is a raging maniac, nothing else, but that leads us a bit off topic [;)] )



LoBaron,

Not disagreeing with your suggestions here at all, but they are significant and would need testing to tune.

All I am suggesting is a minor tweak. Something that I feel would move us in the right direction. Baby steps. I think with enough support, we could introduce it to Michael for consideration and that might be within his mandate to implement.

The question is: Do you feel my suggestion is moving in the right direction? Is it worth testing in the beta build?




LoBaron -> RE: Pilot Training Alogrithm (7/19/2011 5:28:27 AM)

This or something similar could be worth a try.
The disadvantage I see by slowing training with skill gain only is you get to 70exp fast.
70 exp is high. Much too high for a combat trained pilot imo as it governs coordination and op losses.



TBH I donīt think much will happen in this direction in AE, and I tend to agree with crsutton although express it a bit different. [;)]

Current pilot training was a nice idea at first but is a major annoyance and work compared to gameplay benefit.

At the moment all it does is, as soon as two equally experienced players match each other, it creates high skill retort soldiers with identical exp
and skill levels, because what and how long you need to train is obvious by now.
This offsets the whole replacement pilot developement over the war and does in no way represent exceptional individuals who stood out from the masses
of mediocre pilots like it was in WWII, and this with a lot of time investement.

Basically it destroys more than you get from it.

I am sure there was a whole lot of work put into this part of the game, and its sad that it lost its benefits somewhere along the road, but as it is now it should be replaced
completely. No minor tweak will change its assimilatory effect on skills or the mass of micromanagement you need to get equal quality pilots.




Sredni -> RE: Pilot Training Alogrithm (7/19/2011 5:37:06 AM)

I'd like to see skill gains through use improved. Currently some pilots gain almost nothing from steady use, and others gain tons. Your navS and ASW pilots are never going to advance beyond the low 70's you train them to, fighter pilots will gain almost nothing in air and absolutely nothing in defense, but your ground bomber pilots through steady use will end up with grndB skill in the 90's.

Skills like recon or navB or airskill, navS, ASW, transport, ect ect you Have to train up to a high level because your pilot is never going to advance much beyond that once put to use.




herwin -> RE: Pilot Training Alogrithm (7/19/2011 5:52:20 AM)

Suppose your replacements are 10s, a green nugget is 30, and an expert after three years in the fleet is 90. (These figures are reasonable, by the way, for both experience and practiced skills.) The rate of increase should be about 20/year. Assuming 20 missions a month, there should be a 1 in 12 chance of a +1 in experience and/or an applicable skill per mission. Cap training at 50 for skills and experience.




LoBaron -> RE: Pilot Training Alogrithm (7/19/2011 6:13:14 AM)

The problem here is to discern gain from actual combat and from flying combat missions without enemy contact
with a very similar algorithm.
I agree with you that there that the skill gains are highly variable depending on mission type, but I have no
idea how to solve it in a way that does not impact other gameplay spects or simply delays the effect or increases
another effect in a negative way.

The reason why ground bombing has this exaggerated effect, is because every single pilot has a combat
mission with enemy contact every single time he flies an attack.
The same skill gain is subdued on ASW because an actual engagement is a rare event. Same for NavS missions
or naval atttack.

If you say for example, the solution is to simply increase the exp/skill gain for flying combat missions without enemy contact,
you know what you would create? Elite fighter pilots who became such by flying circles on CAP over their airbase without
ever spotting an enemy plane.
Tough task...


Anyways, IMHO the player should be confronted with the results of training, not the training itself.

If though Michael does not want to touch this pandoras box I can understand only too well.




PaxMondo -> RE: Pilot Training Alogrithm (7/19/2011 9:36:59 AM)


quote:

ORIGINAL: LoBaron

Anyways, IMHO the player should be confronted with the results of training, not the training itself.

If though Michael does not want to touch this pandoras box I can understand only too well.

+1

I quite agree. Like you, I haven't really seen the solution yet.




herwin -> RE: Pilot Training Alogrithm (7/19/2011 3:41:38 PM)


quote:

ORIGINAL: LoBaron

The problem here is to discern gain from actual combat and from flying combat missions without enemy contact
with a very similar algorithm.
I agree with you that there that the skill gains are highly variable depending on mission type, but I have no
idea how to solve it in a way that does not impact other gameplay spects or simply delays the effect or increases
another effect in a negative way.

The reason why ground bombing has this exaggerated effect, is because every single pilot has a combat
mission with enemy contact every single time he flies an attack.
The same skill gain is subdued on ASW because an actual engagement is a rare event. Same for NavS missions
or naval atttack.

If you say for example, the solution is to simply increase the exp/skill gain for flying combat missions without enemy contact,
you know what you would create? Elite fighter pilots who became such by flying circles on CAP over their airbase without
ever spotting an enemy plane.
Tough task...


Anyways, IMHO the player should be confronted with the results of training, not the training itself.

If though Michael does not want to touch this pandoras box I can understand only too well.


OR studies showed WWII aircrew improved faster in training than in combat.




LoBaron -> RE: Pilot Training Alogrithm (7/19/2011 4:16:21 PM)


quote:

ORIGINAL: herwin


quote:

ORIGINAL: LoBaron

The problem here is to discern gain from actual combat and from flying combat missions without enemy contact
with a very similar algorithm.
I agree with you that there that the skill gains are highly variable depending on mission type, but I have no
idea how to solve it in a way that does not impact other gameplay spects or simply delays the effect or increases
another effect in a negative way.

The reason why ground bombing has this exaggerated effect, is because every single pilot has a combat
mission with enemy contact every single time he flies an attack.
The same skill gain is subdued on ASW because an actual engagement is a rare event. Same for NavS missions
or naval atttack.

If you say for example, the solution is to simply increase the exp/skill gain for flying combat missions without enemy contact,
you know what you would create? Elite fighter pilots who became such by flying circles on CAP over their airbase without
ever spotting an enemy plane.
Tough task...


Anyways, IMHO the player should be confronted with the results of training, not the training itself.

If though Michael does not want to touch this pandoras box I can understand only too well.


OR studies showed WWII aircrew improved faster in training than in combat.


I agree.

But in WitP AE exp and skill is not what common sense suggests. Many attributes you would adress to the experience of a combat veteran (advantage in spotting/identifying enemy planes,
keeping the tactical superiority depending on what you fight, in a confusing situation and under stress, instinctively knowing the best distance to pull the trigger, evading attacks, knowing the capabilities and
how to best counter specific enemy airframes,...) are distributed between the different skill levels (in this case def/air skill). Exp mostly affects coordination and landing a damaged plane, or a plane in
bad weather.

The game distinction between skill and exp is artificially chosen and does not always differentiate what you learn in a combat enviroment as compared to what you learn in training.

You can teach a pilot to act instinctively in a combat enviroment and to perform every task needed for an air battle in training. Still the usual trained pilot (except one or the other gifted) was in a position
of inferiority when engaging veteran pilots. Training makes the gap smaller but cannot eliminate it.

We only have air/def skill to represent this gap, as that exp does not cover the whole spectrum that makes a veteran pilot a veteran.


Btw: what I find interesting would be a random cap on skills per individual pilot as an additional attribute. This could reflect the different talents nicely. But as the rest of the discussion this
is probably only a topic for WitP 3 if there ever is one.




witpqs -> RE: Pilot Training Alogrithm (7/19/2011 6:41:38 PM)

I think pilot training now - in the Beta - is fine. I think the issue is the contribution that Skill/Experience makes to combat performance. It's just too high, reduce it and that would address your concerns vis a vis what could be achieved by what I think of as a 'fully trained but not combat experienced' pilot. It would simultaneously tone down the carnage wrought by highly skilled pilots to a more realistic level.




Cannonfodder -> RE: Pilot Training Alogrithm (7/19/2011 7:36:31 PM)


quote:

ORIGINAL: witpqs

It would simultaneously tone down the carnage wrought by highly skilled pilots to a more realistic level.


What Carnage? So you are saying experienced pilots should not chew up rookies?




witpqs -> RE: Pilot Training Alogrithm (7/19/2011 10:11:35 PM)

quote:

ORIGINAL: Cannonfodder

quote:

ORIGINAL: witpqs

It would simultaneously tone down the carnage wrought by highly skilled pilots to a more realistic level.


What Carnage? So you are saying experienced pilots should not chew up rookies?


There's often a fine line between being brief and to the point, and being unclear. I tried to be the former, and somewhat achieved the latter!

Air to air they should chew up rookies, no doubt about it. I am fairly certain that would not change because the difference between the skill sets and experience levels would still be very great.

I am mostly talking about other types of effects. For one example, by the time they reach a certain skill/experience level, even Japanese pilots (who have much less effective technology and doctrine/tactics at their disposal) get ASW results that are akin to what you would expect of 1980's NATO versus WWII era subs.

I've also seen plenty of comments about the results achieved by MB and HB against land troops, airfields, and so on. The same for attacks on ships.




LoBaron -> RE: Pilot Training Alogrithm (7/20/2011 6:54:44 AM)


quote:

ORIGINAL: witpqs

I think pilot training now - in the Beta - is fine.


Hey! Guys! Somebody said something about beta again. Do we want to switch over? Soon. I am getting the feeling we play a completely outdated game...
[sm=00000976.gif]



PS: witpqs, I think you hit the nail when saying that high skill makes certain aspects of the game uber.
And there are simply 2 ways to settle that:

- lowering the impact of high skill: this is difficult I think, since when skill is used, so about everywhere, the classic 0-100 distribution results in (un)successful dice rolls. You would have to change EVERYTHING.

- lowering the skill and the ability to gain high skill: I am more and more convinced that this could be the way to go: an artificial cap as additional attribute for each pilot, averaging around about 60.
Certain named pilots have no limit at all (the named pilots for example) while others had bad luck and get capped as early as 55. The value would be hidden but obvious as soon as a pilot reaches its
max potential as he does not increase through training anymore.
You still have your high skill pilots, you still can build elite fighter squads, but much less as compared to now. And youd average out on a skill distribution of around 60, which with the current way of
handling skill values is a good, but not exceptional pilot.




Cannonfodder -> RE: Pilot Training Alogrithm (7/20/2011 7:04:14 AM)


quote:

ORIGINAL: witpqs

quote:

ORIGINAL: Cannonfodder

quote:

ORIGINAL: witpqs

It would simultaneously tone down the carnage wrought by highly skilled pilots to a more realistic level.


What Carnage? So you are saying experienced pilots should not chew up rookies?


There's often a fine line between being brief and to the point, and being unclear. I tried to be the former, and somewhat achieved the latter!

Air to air they should chew up rookies, no doubt about it. I am fairly certain that would not change because the difference between the skill sets and experience levels would still be very great.

I am mostly talking about other types of effects. For one example, by the time they reach a certain skill/experience level, even Japanese pilots (who have much less effective technology and doctrine/tactics at their disposal) get ASW results that are akin to what you would expect of 1980's NATO versus WWII era subs.

I've also seen plenty of comments about the results achieved by MB and HB against land troops, airfields, and so on. The same for attacks on ships.



Some of those remarks you made seem to be related also to experience, but I am pretty sure that payload and ammount of bombers have a huge impact on land troops and airfields. Especially if they are undisrupted or barely disrupted by flak! That is my experience anyway.

Some ASW seems to be a bit to effective though, but only if you keep your subs in the same hex for too long!




oldman45 -> RE: Pilot Training Alogrithm (7/20/2011 5:36:46 PM)

Is there a chance we are blurring training and combat experience? I will admit I do not micromanage the pilots like most of you do. So I don't notice much of what you all are talking about. Here's my take on this.

We actually have too many categories that the combat algorithms use when it comes to the pilots. The fighter pilots skills should improve based on # of combat missions and number of skilled pilots in their unit. Their skills should be a few categories that only involves air to air combat. Then you would have your single engine naval bomber pilots with their skills limited to naval attacks. Army bomber pilots would have their skills limited to ground attacks when they join their squadrons, and that would be split to those that fly heavy bombers and those that fly medium and ground attack. Finally you have the multi-engine naval pilots who join their squadron armed with search/asw skills. You can have these guys flying training missions till the cows come home but their skills can only be honed by getting into combat. Using the army pilots as an example, the medium bomber pilots know how to fly their ships very well when they join their squadron. They can get their planes to the target, lets say its an airfield and they will put their bombs on that target. What is the minimum skill number for that to happen? Then the wing commander announces they will learn how to attack naval targets at low level. You take the squadron out of combat and they train for a few weeks with their skills starting at 0 and after a few weeks it reaches "x". Once they start flying missions then you will see the improvement.

I think each of us has our own idea of what the base line should be when the pilots reach their squadrons. No matter how much they train that number should not really change much until they are tried in combat.

I think over the years we have created a monster with all the micro-managing that really should not be so intense.




CV 2 -> RE: Pilot Training Alogrithm (7/21/2011 12:35:09 AM)

Without knowing the exact effect of the numbers on game play (which the devs hold as a closely guarded secret), its hard to propose anything.




Alpha77 -> RE: Pilot Training Alogrithm (7/21/2011 5:49:05 PM)


quote:

ORIGINAL: Cannonfodder


quote:

ORIGINAL: witpqs

It would simultaneously tone down the carnage wrought by highly skilled pilots to a more realistic level.


What Carnage? So you are saying experienced pilots should not chew up rookies?


Just read a book with some figures... from USAAF and Luftwaffe means European TOW. But I guess also valid for Pacific. Conclusion of Author (for Luftwaffe) is that somehow the "Experten" shot down a major number of enemy planes. Means suggest Exp and skill is really THAT important....in real life. He also comes to the conlcusion that "rookies" often got shot down before they even could reach any level to be able to shoot down Allied planes at all (for 44+45). He also provides numbers and percentages but I donīt have the book here. It was written by an US pilot, he met with some aces from US,GB + Ger. I might quote some of the numbers if of interest here later (weekend). The book is called "Fighter Aces". I found some minor errors in it however, like he says the B17 has 10 20mm cannons (?) that could not prevent their demise however, but only when longrange escort got to the front in high numbers the air war in Europe changed in favor for the Allies. The conclusion is true, but the B17 has MGs not guns. Laso he confused the Me109F and E model at one occasion: That might not distract from the validity of his data though. But itīs only for fighters.




Nikademus -> RE: Pilot Training Alogrithm (7/21/2011 6:03:47 PM)

Yes.....Luftwaffe experten accounted for a great majority of kills. This does not mean however that rookies can't. Part of the issue with the Luftwaffe example is that often the combat formations were set up so that the experten got the kill shots while rookies or lesser exp'd pilots covered their backs. This helped up personal scores but didn't share the wealth too much. Key thing to remember is that WWII air combat isn't a tenth as bloody as it often results in this game (depending on the #'s engaging per air battle.) If it were, then everyone would be an ace. Think about it.....it "only" takes 5 kills to be labeled an ace....yet Aces were minorities. I believe it was in Bergerud that it was mentioned (and in other places) that most pilots never shot down even one plane.

Waxing poetic.....i always thought it was a mistake for G. Grigsby to try to portray indiv pilots in a game of this scope and size. He most likely did it because of his past experience with 12oclock high and BoB, but those games were air campaign specific and even then it was a database staticians nightmare to track all of em. (i personally never did.....i just looked at general exp levels of the squadrons) For a game covering the entire Pacific war...tracking each indiv pilot doesn't make sense. Game is already huge. Adding 12 different sub exp skill sets only exaserbated it IMHO. Not knocking the game or the deciders....just expressing an opinion. I know there are some who really like being able to micromanage pilot training. I think PacWar's system had a simple elequence to it that worked better in the long run as it generally took 6 months at least to train a trainee (exp 10) to fully trained (exp 50) which matches the USN ciriculum pretty closely. It was same for Japan but as in real life.....the Japan player usually didn't have the luxery to keep it's airgroups out of combat in "training" mode for the required time once the mid-late years of the war came around.

My experiences with UV, WitP and now WitPAE as an insider have really changed some of my past held notions on what makes a good wargame in an age of much more powerful computers. I am now of the view that nausiating detail levels are not necessarily either a good thing or a necessary thing. I see a similar issue with the new WitE which i'm hoping to try soon since i own it! :) Same issue there as here.....wheras past wargames sufficed to give a unit (say a brigade, or a division) a single exp rating with seperate exp values for leaders, WitE gives a seperate exp value for each component WITHIN the unit.....so a typical Panzer div could have 15 different sub components, each with it's own exp value.

was that really necessary for a game of that scope? When i see that clutter on the screen i groan. Maybe i'm just getting old.....when i was a teen i had all day litterally to parse through this stuff and play the game. As an adult......time is much more constrained. And then there's the simple question again.....does having this level of detail really make for a better game? Does a Pz div with 20 diff. exp levels fight differently in an operational wargame vs. one with a single exp value? Judging by the threads i browse....i'd say not. If anything the added detail levels create more loopholes and issues and exploits.

i'm starting to really like the idea of simple is better for games of this scale. Leave the minute detail to tactical wargames.




Alpha77 -> RE: Pilot Training Alogrithm (7/21/2011 6:33:54 PM)

Tend to to agree with last post.  Tracking every pilot is really over the top (same with the division Nick posts about in WITE - in WIR this was not the case, I guess the system of WIR was enough for me).... but as long we can leave it alone somehow itīs ok. And every player can choose if he wants to use it or not. Only problem might be, if the game getīs more expensive and unnessary cluttered (interface etc) with the upping of the details which a certain portion of players do not want, but need to pay for it anyway. But still when you can leave it alone I am ok with it.

I guess eg. the Division in WITE also give an average rating of all the subunits and you donīt need to look at any single of them ? The same in AE you also get the av. exp level etc. for the whole unit.

I adopted a simpler system:

Train fighters at air, then train general trainin

Bombers, ground, then general

Same for other types. General often adds some unwanted or not needed skills but I take this as compensation that I use a simpler system. If you train only general for eg. fighters air will also improve faster I noticed. So it seems general also looks at the plane type anyway to a certain level. Probably if your enemy is not too deadly (AI?) or use some rules in PBM you can get decent pilots only using general training (? this is more a theory from me not proven!)





witpqs -> RE: Pilot Training Alogrithm (7/21/2011 7:34:01 PM)

quote:

ORIGINAL: Alpha77

quote:

ORIGINAL: Cannonfodder

quote:

ORIGINAL: witpqs

It would simultaneously tone down the carnage wrought by highly skilled pilots to a more realistic level.


What Carnage? So you are saying experienced pilots should not chew up rookies?


Just read a book with some figures... from USAAF and Luftwaffe means European TOW. But I guess also valid for Pacific. Conclusion of Author (for Luftwaffe) is that somehow the "Experten" shot down a major number of enemy planes. Means suggest Exp and skill is really THAT important....in real life. He also comes to the conlcusion that "rookies" often got shot down before they even could reach any level to be able to shoot down Allied planes at all (for 44+45). He also provides numbers and percentages but I donīt have the book here. It was written by an US pilot, he met with some aces from US,GB + Ger. I might quote some of the numbers if of interest here later (weekend). The book is called "Fighter Aces". I found some minor errors in it however, like he says the B17 has 10 20mm cannons (?) that could not prevent their demise however, but only when longrange escort got to the front in high numbers the air war in Europe changed in favor for the Allies. The conclusion is true, but the B17 has MGs not guns. Laso he confused the Me109F and E model at one occasion: That might not distract from the validity of his data though. But itīs only for fighters.


My original comment is about the whole ball of wax, not specifically about air to air combat.




Alpha77 -> RE: Pilot Training Alogrithm (7/21/2011 7:55:27 PM)

Yeah, but I donīt have enough sources to the other balls of wax to add worthwhile comments, so I posted about air to air [:D] I could imagine in other missions it might be the same, but guess you meant this paragraph:

"I am mostly talking about other types of effects. For one example, by the time they reach a certain skill/experience level, even Japanese pilots (who have much less effective technology and doctrine/tactics at their disposal) get ASW results that are akin to what you would expect of 1980's NATO versus WWII era subs. "

Well one needed to research Japanese ASW to give a better answer. In my limited knowledge the Japs were always behind in ASW they got better later, but never as good as USA/Britain. I have not seen these deadly ASW effects of the Jap personally (well planes sometimes hit, but often the sub is not even damaged, rarely sunk). I noticed hoewever the boost the Jap getīs since 43 for surface asw. They are now much better in detecting allied subs and make succesful DC attacks. Except Jap merchant gunners, they were/are always good (in game) [;)]

But to the time Japan got better in ASW, it was already too late, too much ships sunk, too much good pilots lost etc.




witpqs -> RE: Pilot Training Alogrithm (7/21/2011 9:05:34 PM)

quote:

ORIGINAL: Alpha77

Yeah, but I donīt have enough sources to the other balls of wax to add worthwhile comments, so I posted about air to air [:D] I could imagine in other missions it might be the same, but guess you meant this paragraph:

"I am mostly talking about other types of effects. For one example, by the time they reach a certain skill/experience level, even Japanese pilots (who have much less effective technology and doctrine/tactics at their disposal) get ASW results that are akin to what you would expect of 1980's NATO versus WWII era subs. "

Well one needed to research Japanese ASW to give a better answer. In my limited knowledge the Japs were always behind in ASW they got better later, but never as good as USA/Britain. I have not seen these deadly ASW effects of the Jap personally (well planes sometimes hit, but often the sub is not even damaged, rarely sunk). I noticed hoewever the boost the Jap getīs since 43 for surface asw. They are now much better in detecting allied subs and make succesful DC attacks. Except Jap merchant gunners, they were/are always good (in game) [;)]

But to the time Japan got better in ASW, it was already too late, too much ships sunk, too much good pilots lost etc.



[:D] OK! It seemed like several people thought I was only referring to air to air. I think the elite vs rookie carnage would be little changed because the differences between their skills would still be quite great.

As far as ASW goes, if the IJ player systematically trains the heck out of LBA pilots and has plenty of LBA on ASW... Well, as Allies I was NOT keeping my subs in one hex, I was using patrol zones, etc. etc., and when my opponent's pilots reached a certain level of skill/experience it got very bad. Eventually I lost 10 subs in one month and had to pull everything back. If your opponent doesn't train for ASW that might not happen.

But everybody trains for naval attack, ground attack, etc. so the effects will be seen there by most people.

To be clear, I think a change in this area is very unlikely. I do think that addressing it through changing training would make matters worse because the early war pilots would never have equals, and they did IRL. I'm posting really to help avoid what I see as the mistake of changing training that has gotten quite good in the Beta.

BTW, I also never wanted the detail of individual pilots in WITP. When AE came along with increased detail and management, I also didn't want that. But - I realize that most people did/do, so I go with it and make suggestions aimed at making it more manageable for players and (hopefully) more realistic. [8D]




wpurdom -> RE: Pilot Training Alogrithm (7/21/2011 11:48:13 PM)

First, I agree with those who say the whole individual pilot model is likely a mistake.

Second, training should be slower.

Third - the pilot training model , including individual control and fast accumlation of skills, is a big part of of making the IJ player unnaturally competitive from 8/42-8/43. (The other factor is downplaying the huge advantage in mechanics (maintainence)the Allies and esp. the US had - its not how many planes you produce, it's how many you can keep operational on dirt strips in the jungle that counts). If you abstract the training model, keeping up the IJ skill level will have less plausibility, unless it is assume in an option choice at the beginning "better trained IJ pilots." So it's going to be very difficult to change without ruining the acceptability of the game as a game.

But if we're really going to model the path of individual pilots, I really wonder about this gradual progression of training in fighter pilots. For other pilots I can understand the gradual progression of skills, but everything I've read about fighter pilots suggest to me that prior to the "Top Gun" schools fighter pilots came naturally in three categories which bcame exposed in the first three combats or so (at least if they had the level of training for basic survivial of pre-war airforces or the USAAF, USN, or USMC) - naturals, competants and targets. It seems to me, that for the naturals, their skills peaked in the first several combats and then stayed at a high level until combat fatigue set in. These folk did the great majority of kills. Then there were competent pilots who would get the occasional kill from using buddy tactics or going after bombers. Then there were the many who would never get a kill, would try to get home alive, and in the meanwhile would perform a very useful role in providing protective coloration for those who did the work.

I'm not dismissing totally what herwin states about training being more effective than missions. It's almost certainly the case with the non-fighter pilots. Even with the fighter pilots, acquiring the basic skills to survive your first several combats may not occur best in the combat situation. But at least before the Vietnam era, I think the only real way to see who had the skills to be a killer was to send him into combat.




ChezDaJez -> RE: Pilot Training Alogrithm (7/22/2011 11:55:47 PM)


quote:

ORIGINAL: herwin

Suppose your replacements are 10s, a green nugget is 30, and an expert after three years in the fleet is 90. (These figures are reasonable, by the way, for both experience and practiced skills.) The rate of increase should be about 20/year. Assuming 20 missions a month, there should be a 1 in 12 chance of a +1 in experience and/or an applicable skill per mission. Cap training at 50 for skills and experience.



I would have to disagree, Herwin. I think you are basing your studies on peacetime training. By the end of one year of flying actual combat, a pilot will have gained all the experience he is going to, probably sooner depending on how much combat he saw. Any experience gains afterwards would be minimal. The only variable thereafter is experience gained from transitioning to new aircraft and encountering new enemy aircraft.

In wartime the learning curve is much, much shorter. You either learn fast or you die. A pilot will gain more experience on that first combat mission than he will have in all the previous combat training evolutions put together, assuming he actually engaged the enemy and survived. And he will be a seasoned pilot after 5-10 combat missions of engaging the enemy. He has learned what parts of his training work and what doesn't. And he will have the benefit of the veteran pilots around him for advice. After about 10 missions, he will have developed a set of skills that will allow him to 1) survive and 2) inflict damage on the enemy.

Consider this, the 8th Air Force in England rotated individuals out of combat after 25 missions. Those missions were typically compiled within 4-6 months assuming 1-2 missions per week. They were well seasoned veterans that I would consider to have an average set of WitP style experience in the 70's.

Personally I think that experience for the average pilot in the game should max out somewhere in the 70s with only a select few getting into the 80's and 90's. I think a better way to determine experience would be to have on map training max out at about 40 and then have experience gains based on the number of missions flown and kills, possibly 1 point of experience for every mission up to a max of 70 and 2 point for each kill with no max. This would be a more linear approach than using an inverse ratio but would be easier to implement.

Chez




Page: [1] 2   next >   >>

Valid CSS!




Forum Software © ASPPlayground.NET Advanced Edition 2.4.5 ANSI
0.0546875