Matrix Games Forums

After offering advice for almost 8 months on Burma in various AAR's [:'(], I have decided to take the plunge and start a game with njp72. This player currently has a well illistrated contest against dennishe.

The scenario is BigBabes 28C, with stacking limits. The home rules are simply no restricited HQ's crossing borders and no 4E naval attacks below 10K'. The latter is fixed in this particular executable, but we shall keep the home rule. njp72 understands my distain for homerules, especally in the middle of a game to compensate for some historical myth. Fix the 7 vs. 8 hex CV strike range game balancing option and fix the problem of airstrike detection at the target hex, and I might start thinking that the game has historical merits [;)] Otherwise it is a great wargame with very interesting challenges on both sides.

For a scenario #2 to have feasibly happened, I beleive one would have to imagine our previous policies with North Korea in recent history. I might conjecture that a policy of appeasement with unfulfilled promises of withdrawal from China along with standing down the Navy and Army. The Allies would have to not only ignore the IJ's build up, but 'donate' tons of oil, raw materials, and supplies (processed materials) for the IJ to have built such a formidable military force. In this scenario the United States continue investments in domestic spending while engaging in a diplomatic policy that Japan's dependence on trade with the West will deter war. With Japan spending over 30% of GNP on military spending for over 10 years these diplomatic policies have cumulated into building a "monster" Like Scenario #1 -- Japan cannot in the long run maintain this economy -- they must expand. Unlike scenario #1 -- they have stockpliled enough GNP to match the Allies industrial base through 1943 - mid 1944.

Despite this threat, the Allies remain in a "Hitler" first central focus. Winston and Roosevelt both concur; "to give the slant eyed bomb dodgers a chance" [8D]

My strategy .... survive "The Trama of '42" No IJN AV.

fix the problem of airstrike detection at the target hex

What are you referring to with this?

And Good Luck!

ORIGINAL: witpqs

fix the problem of airstrike detection at the target hex

What are you referring to with this?

And Good Luck!

If you notice, all aerial combat takes place over the target hex. Let us say I have a strike package leaving Imphal to hit Chang Mai. . The IJ have multiple airbases at Shwebo, Mandalay, Meikita ... full of fighters. A tatitical representation of operations would suggest that the detection should occur over the skies of Shwebo. However, in the combat report the detection will be at Chang Mai and none of the fighters at these bases will particpate in an interception. If fighters at Meikita and Taung Gvi were on LRCAP a percentage of them will appear at Chang Mai assuming the interception occurs at the target hex.

This algorithum works sort of great for pure CV-CV contests. It breaks down if LBA are involved. A strike package leaving CV's off of Baker Island headed toward Canton Is. presents a very interesting problem. The strategy of intercepting over Baker Island is not possible if the package is headed a full 8 hexes .. so . the optimum strategy is to escort strike packages 1 hex rather than LRCAP to an impossible distance.

Now a tactical circles of apollonius algorithum would calcualte various interception points per airbase operationally representing these points within hexes would represent the running battles of a long airstrike package.

The problem is that complex tactical rules are built into what amounts to as a strategic representation of air combat.

For me fundementally this precludes the notion of "tactical" home rules as a tool to somehow move the game toward an historical representation of WWII. There are fundemental design issues that preclude this, but .. these issues do not preclude engaging in a great game playing the game as the designers intended. As an example, I understand the designers intended to keep restricted units within borders but did not get the code done ...

Hopefully that is clear [8D]

Yup.

Because you could also have meant the feature whereby an attacking bomber force reads the conveniently placed placards identifying which grounds unit they are attacking! [:D]

ORIGINAL: witpqs

Yup.

Because you could also have meant the feature whereby an attacking bomber force reads the conveniently placed placards identifying which grounds unit they are attacking! [:D]

Well this is all "abstracted" .. I would assume that in a Hogan Hero's like manner partisans hold up candles at night configured in code to identify enemy units ...[:D]

This one does not bother me as much as the circles of apollonius problem ... "The Elf" engaged me in a conversation in GreyJoys thread .. just because one is a F-18 jock does not mean they understand the math involved with their trade or how to program these algorithums so they fit in a game ... [;)]

I´ll follow along! [:)] Is this your 3rd PBEM?

ORIGINAL: JocMeister

I´ll follow along! [:)] Is this your 3rd PBEM?

It is .. My first 2 games were scenario #1 .. the first game I AAR'ed ending mid 1943 with a Burma campaign that seized all of Southeast Asia, the occupation of the Marshalls and Gilberts, and the unraviling of the IJ's foray in Oz with NG safely in Allied hands... Second game ended Feb 1944 with a Burma Campaign seizing most of China using Commonwealth forces drving up from Southeast asia .... I did not AAR that game due to time and effort constraints although I posted occasional updates in repsonse to suggestions that Burma is impossible .... [8D]

This game I suspect will be much more challanging and different due to scenario #2 combined with BigBabes constraints ... my main objective is to avoid an IJ autovictory .. firm up a MLR .. then pick a strategy to get Allied victory before July 1945 ... or inflict IJ SCLS before that date [:'(] Pretty basic strategy I realize ..[8D]

Burma is a much tougher contest I think due to the looming possibility of a seaborne India operation.

I do not think the IJ can advance on more than axis out of seven possible axis effectively: Alaska? Mid-Pacfic Canton/Baker/Johnson/Christmas Is? Suva/New Hebrides? Oz? Burma/India/Ceylon? China? Russia?
njp72 has demonstrated keen aggession in the past ... successfully neutering China and destroying Russia in one game while keeping the USN at bay in the Gilberts .. We shall see how the game develops ...

ORIGINAL: Crackaces

ORIGINAL: witpqs

Yup.

Because you could also have meant the feature whereby an attacking bomber force reads the conveniently placed placards identifying which grounds unit they are attacking! [:D]

Well this is all "abstracted" .. I would assume that in a Hogan Hero's like manner partisans hold up candles at night configured in code to identify enemy units ...[:D]

This one does not bother me as much as the circles of apollonius problem ... "The Elf" engaged me in a conversation in GreyJoys thread .. just because one is a F-18 jock does not mean they understand the math involved with their trade or how to program these algorithums so they fit in a game ... [;)]

I'm sure he understands the math involved, which I think is the easy part of it. The harder parts are figuring out the algorithms for those interceptions and then programming them properly. Consider that CAP is not all at a point in the center of the circle (over the base). So there is a percentage of the time that they are closer or farther away. Or even equidistant though at a different location where they have used up some fuel already and can not intercept to the same distance from base as a result. And remember that fuel is not tracked for CAP that way (not for any aircraft AFAIK). Layer on top of that the likelihood of being tasked to intercept, plus the likelihood of actually intercepting, detection from each base, communication of detection from one base to another (including how good & how successful the comm. is) etc, etc. And when I say 'algorithms', I don't mean the formulae themselves, I mean the valid representative percentages to plug into the formulae. Sorting out the valid numbers to use as input the 'the' real issue. And of course programming it is beyond trivial, certainly 'out of scope' given the list of things they did (and more they tried to) accomplish with the release of AE. Because of the structure of the way that air combat is modeled, you are quite possibly/likely talking about needing to rip out a major structural beam here and there and replace it. Then when all that is done, test things and tune them to be 'right'. Including things that were not part of the original objective, but that are affected by some of the internal things that had to be changed (those structural beams that needed to be replaced to accomplish the objective).

I am not at all saying that it can't be done, or that it is not a desirable objective. I am saying that it is most likely a very major effort. I think that's the actual problem. How much improvement would it yield in comparison to the effort expenditure required, effort which is not even available to expend?

ORIGINAL: witpqs

ORIGINAL: Crackaces

ORIGINAL: witpqs

Yup.

Because you could also have meant the feature whereby an attacking bomber force reads the conveniently placed placards identifying which grounds unit they are attacking! [:D]

Well this is all "abstracted" .. I would assume that in a Hogan Hero's like manner partisans hold up candles at night configured in code to identify enemy units ...[:D]

This one does not bother me as much as the circles of apollonius problem ... "The Elf" engaged me in a conversation in GreyJoys thread .. just because one is a F-18 jock does not mean they understand the math involved with their trade or how to program these algorithums so they fit in a game ... [;)]

I'm sure he understands the math involved, which I think is the easy part of it. The harder parts are figuring out the algorithms for those interceptions and then programming them properly. Consider that CAP is not all at a point in the center of the circle (over the base). So there is a percentage of the time that they are closer or farther away. Or even equidistant though at a different location where they have used up some fuel already and can not intercept to the same distance from base as a result. And remember that fuel is not tracked for CAP that way (not for any aircraft AFAIK). Layer on top of that the likelihood of being tasked to intercept, plus the likelihood of actually intercepting, detection from each base, communication of detection from one base to another (including how good & how successful the comm. is) etc, etc. And when I say 'algorithms', I don't mean the formulae themselves, I mean the valid representative percentages to plug into the formulae. Sorting out the valid numbers to use as input the 'the' real issue. And of course programming it is beyond trivial, certainly 'out of scope' given the list of things they did (and more they tried to) accomplish with the release of AE. Because of the structure of the way that air combat is modeled, you are quite possibly/likely talking about needing to rip out a major structural beam here and there and replace it. Then when all that is done, test things and tune them to be 'right'. Including things that were not part of the original objective, but that are affected by some of the internal things that had to be changed (those structural beams that needed to be replaced to accomplish the objective).

I am not at all saying that it can't be done, or that it is not a desirable objective. I am saying that it is most likely a very major effort. I think that's the actual problem. How much improvement would it yield in comparison to the effort expenditure required, effort which is not even available to expend?

Well the conversation started pretty arrogantly [8D] So you know from which I speak from .. I was a systems architect for Lucent Technologies ..solving software problems was a way of life I have never flown fighter jets, but I do understand math, systems thinking, and programing algorithums [:'(]

I think a minor step in the right direction would be very doable and worthwhile .. simply abstract intercept like the ol' Luftwaffe game .. first base crossed initatates detection/ intercept as per the current rules with current leaky CAP/LRCAP rules. That is the first base crossed becomes a "target" for intercept purposes

The ramifications of the current system means concentrating aircraft at potential targets and generating huge furballs ... which also taxes the current system .. . games like Greyjoy vs. Radier stopped because of these furballs ....If I know I at least get a crack with my early CAP .. I use them on defense rather than focus on sweep or whatever to get an offensive operation in ...

Anyway there are lots of constraints that keep this engine from being an historical similation and no amount of home rules is going to get players to get this game to act as a similator ..so .. I gleefully accept this is an excellent game!

[:D] I don't recall following that earlier conversation, but I guess it ticked you off.

Maybe this one is a -2 improvement! [;)]

ORIGINAL: witpqs

[:D] I don't recall following that earlier conversation, but I guess it ticked you off.

Maybe this one is a -2 improvement! [;)]

Ya not sure .. and many others are very very happy with detection at the target base .. so I am a lone voice in the wilderness ... I do think it would be a better game from teh standpoint of enabling a more diffuse defense rather than hording fighters at targets ..

njp72 I believe lives in the land of Oz .. so I suspect the small talk will get serious in about 3 - 4 hours [8D] We should know the results of day 1 and what platforms we have to work with ...and get a feel for turning turns around ...

Looking forward to this one!

Glad to see you back in the saddle. [8D]

ORIGINAL: obvert

Looking forward to this one!

Glad to see you back in the saddle. [8D]

Thanks Obvert! It should be a very interesting contest. I do not think njp72 is a complainer and I surely will not be giving up .. I do suspect he will make it a close AV contest before I can get the big Allied machine rolling ...[&o]

I quite agree - I've played against him - it will be an interesting match - no doubt

ORIGINAL: MAurelius

I quite agree - I've played against him - it will be an interesting match - no doubt

Thanks MAurelius ..

Turn is complete but I await the movie ..the only thing I know is that flak was effective .. [8D]

He is 17 hours ahead of me .. so his turn will arrive in my morning and my turn will arrive while he is sleeping .. ... a turn a day at most for sure ...

ORIGINAL: Crackaces
"The Elf" engaged me in a conversation in GreyJoys thread .. just because one is a F-18 jock does not mean they understand the math involved with their trade or how to program these algorithums so they fit in a game ...

I was a systems architect for Lucent Technologies ..solving software problems was a way of life I have never flown fighter jets, but I do understand math, systems thinking, and programing algorithums [:'(]

Actually is funny, bordering arrogant, to assume anybody who flies combat aircraft packed with complex software doesn´t, or does not do it eqally well as you do. You are talking about people working on difficult mathematical problems under high pressure as a routine training.

A lot of people here do understand mathematics and deal with complex software solutions, including Ian, and myself.

I work as IT architect for an international ISP/Telecommunications company. Means I create IT impact assessments, design and maintain high level system architectures, participate in creation of inter-application functional/technical designs, and do IT project management more or less as a sideshow.

Everytime someone approaches me promising a "simple" "minor step" to change an irritating fundamental software property, I immediately suspect that he, or she, have either not understood the issue at hand, have no idea about the complex (side)effects of their proposed solution (usually the reason they call it "simple"), neglect the impact such a modification has on an overall balanced environment, underestimate the ammount of (regression) testing required on implementation, or some/all of the above.

In 99% of those situations my suspicion turns out correct.

I think a minor step in the right direction would be very doable and worthwhile .. simply abstract intercept like the ol' Luftwaffe game ..
first base crossed initatates detection/ intercept as per the current rules with current leaky CAP/LRCAP rules. That is the first base crossed becomes
a "target" for intercept purposes

The strike algorithm is point to point. The CAP/LRCAP intercept routines are not, they are more complex than that and abstract a lot.

Just an example: What you are referring to as "leaky CAP" is part of those intercept routines. If you just imagine it as leaky CAP, you miss a big part of the air model. Properly understood you can already create situations mimicking an interception along an ingress path.

Creating a virtual (first default? moving? from where is the CAP then pulled? and how is this decided? from the base detecting? from the target base? from all bases along ingress path? does it depend on fighters present with specific CAP and range settings? if yes how does this work with the current fatigue model? with recon/navS intercept? with CV battles and pickets? with multiple strikes on several targets in this environment?) target along the ingress route, to simulate radar detection and ingress intercept, impacts much more than you seem to assume.

To make it short: it would whack the air model in most situations, including those you want to adress with your solution, except if you implement several additional
major changes.

Since the code is unable to determine a location of the strike - except at base of origin and at target -, only a virtual direct ingress route could be simulated, with the current detection model of radar an inbound strike would always get detected early, with the current reaction algorithm to detected strikes you could draw a whole airforce onto a couple of incoming bombers. With all possible implications. And thats just one minor example.

It only takes seconds to imagine several technical issues, and exploits, possible by such a change of game mechanics, one worse than the other.

Yes, you can find ways around that. But it makes your proposal neither "simple" nor "minor". It would require a complete revamp of the air model, including the DB it is based on. Also I am pretty sure this is not allowed within the juridical limitations, constraining the AE devs to work with large parts of the original GG code.


Simply put, there are no programmers left to deal with changes on such a scale, there is no legal ground to make such changes an option, and there are no testers available to minimize the risk on implementation for a project of this scale.

If you really have been a systems architect you will see this immediately.

The ramifications of the current system means concentrating aircraft at potential targets and generating huge furballs ... which also taxes the current system .. . games like Greyjoy vs. Radier stopped because of these furballs ....If I know I at least get a crack with my early CAP .. I use them on defense rather than focus on sweep or whatever to get an offensive operation in ...

Seems you did not play the recent beta yet. Sounds like you should give it a try.

I haven´t yet myself, but my next PBEM will surely be latest beta, and I look forward to it. According to what I hear it resolve the issue experienced by rader and GJ pretty nicely.



Good luck with your PBEM. [:)]

ORIGINAL: Crackaces

It is .. My first 2 games were scenario #1 .. the first game I AAR'ed ending mid 1943 with a Burma campaign that seized all of Southeast Asia, the occupation of the Marshalls and Gilberts, and the unraviling of the IJ's foray in Oz with NG safely in Allied hands... Second game ended Feb 1944 with a Burma Campaign seizing most of China using Commonwealth forces drving up from Southeast asia .... I did not AAR that game due to time and effort constraints although I posted occasional updates in repsonse to suggestions that Burma is impossible .... [8D]

This game I suspect will be much more challanging and different due to scenario #2 combined with BigBabes constraints ... my main objective is to avoid an IJ autovictory .. firm up a MLR .. then pick a strategy to get Allied victory before July 1945 ... or inflict IJ SCLS before that date [:'(] Pretty basic strategy I realize ..[8D]

Burma is a much tougher contest I think due to the looming possibility of a seaborne India operation.

I do not think the IJ can advance on more than axis out of seven possible axis effectively: Alaska? Mid-Pacfic Canton/Baker/Johnson/Christmas Is? Suva/New Hebrides? Oz? Burma/India/Ceylon? China? Russia?
njp72 has demonstrated keen aggession in the past ... successfully neutering China and destroying Russia in one game while keeping the USN at bay in the Gilberts .. We shall see how the game develops ...

I havn´t played Scenario 2 but initially I think the extra troops can really make a huge difference. So I think its really important as the allies to get that MLR set up fast as hell and far back enough for it to have time to grow into a good defense in depth!

I am of the firm belief that given two players of somewhat equal skill a concentrated push for China is impossible to stop. I think you can safely assume China will be on his agenda as it doesn´t require a massive effort and still is an almost given win. I also think you can expect an additional push on one of the other locations you mentioned. Probably not russia though? An early landing in India perhaps?

My second game is a DBB one. AA really makes a difference I can tell you that! In that game I pay attention to AA units and bring them with me. Something I never bothered to do in the vanilla game.

So far I love DBB and SL. Doubt I will ever play a vanilla game again!

Will be really interesting to follow this! Best of luck! [:)]

ORIGINAL: Crackaces

ORIGINAL: MAurelius

I quite agree - I've played against him - it will be an interesting match - no doubt

Thanks MAurelius ..

Turn is complete but I await the movie ..the only thing I know is that flak was effective .. [8D]

He is 17 hours ahead of me .. so his turn will arrive in my morning and my turn will arrive while he is sleeping .. ... a turn a day at most for sure ...

that's unfortunate.... with my time difference to Australia - from Europe - we were often able to play 6-7 turns even... (when nothing was going on...)

oh well - at least the both of you will have more time to make sure you don't rush the turn

Everytime someone approaches me promising a "simple" "minor step" to change an irritating fundamental software property, I immediately suspect that he, or she, have either not understood the issue at hand, have no idea about the complex (side)effects of their proposed solution (usually the reason they call it "simple"), neglect the impact such a modification has on an overall balanced environment, underestimate the ammount of (regression) testing required on implementation, or some/all of the above.

In 99% of those situations my suspicion turns out correct.

That is soooo true .. 1/P(failure) = 1/P(a) + 1/P(b) +1/ P(c) .... thus replacing the guts of this model would mean hundreds of thousands of cycles to ensure some reasonable expectation that the thing won't crash or fail ..

My contention is if you are going to design a "tactical" air game .. one would think that the first game engine design would begin with the thought of circles of apollonius? Such is my thought

My thought on the Elf being arrogant is that he could have just said what you said [but nicer [;)] We thought about it .. decided given scope, resources and time it could not be done reasonably. I have come to that conclusion. However, I still contend that this is a big bugaboo in the game .. the current engine desgin abstracts too much for a tactical feel .. such is my opinion ...

Since the code is unable to determine a location of the strike - except at base of origin and at target -, only a virtual direct ingress route could be simulated, with the current detection model of radar an inbound strike would always get detected early, with the current reaction algorithm to detected strikes you could draw a whole airforce onto a couple of incoming bombers. With all possible implications. And thats just one minor example.

It only takes seconds to imagine several technical issues, and exploits, possible by such a change of game mechanics, one worse than the other.

Yes, you can find ways around that. But it makes your proposal neither "simple" nor "minor". It would require a complete revamp of the air model, including the DB it is based on. Also I am pretty sure this is not allowed within the juridical limitations, constraining the AE devs to work with large parts of the original GG code.

Simply put, there are no programmers left to deal with changes on such a scale, there is no legal ground to make such changes an option, and there are no testers available to minimize the risk on implementation for a project of this scale.

Ahhh I did not know this .. since ship TF seem to intercept I thought that the ship intercept alogorithum wold extend into this situation .. clearly not ..

If you really have been a systems architect you will see this immediately.

I respected you until this comment ..

ORIGINAL: Crackaces

If you really have been a systems architect you will see this immediately.

I respected you until this comment ..

Why?

All I am saying is: if your job was what it sounds to be (programmer, IT expert, architect, platform designer, network
engineer, those designations usually get thrown around much too often to hint competence where there is none), at least after
my explanation it should be obvious to you why your suggestion is not an as easy and simple solution as you initially thought.

Nothing personal.

Ahhh I did not know this .. since ship TF seem to intercept I thought that the ship intercept alogorithum wold extend
into this situation .. clearly not ..

You are not so far off as you might think. Good comparision, but there are some obvious fundamental differences (based on what we already
discussed). One of the most prominent is: Aircraft have no route, don´t move hex-by-hex, and got no location, except their base of origin and
possibly their target base. Everything else is abstraction and probability, governed by a considerable number of variables. Some are player
governed, some not.

However, I still contend that this is a big bugaboo in the game .. the current engine desgin abstracts too much for a
tactical feel .. such is my opinion ...

Absolutely. You are for sure not the only one praying to the strategy gods for WitP III, with all the bells & whistles.