AI for MWiF - USSR

[npilgaard]

Yes, I fully agree that it is hard to read such long posts of dense text.
There is a lot (perhaps to much) of text and difficult to read - some of the posts were a part of an ongoing discussion at that point (I have included the posts here in their full length (almost - I have removed a little non-related stuff already), partly to allow for a full read, and partly because it was easier [:)]).

But actually I have already split them up - each post with list post no. (e.g. ' Bruce (67973): ') and then the post text in " " afterwards. So each " " indicates the text from a single post. But the one right above is especially long, and I have now split it into sections for easier reading.

Except for a single introducing line or two I have made no comments at all in the various posted texts (the original posts from the list were actually that long [:)]). So the content of the posts are all from the original poster.

Also, the posts were not necessarily taken from one thread - often they have been posted with no connection to each other, at different times. I have just brought them together, as they treated (roughly) similar subject. So they should not be seen as an ongoing coherent discussion, but rather just as independent, different aspects or tips on a subject.

Edit: Have now split the longer of the posts into sections for easer reading.

[npilgaard]

ORIGINAL: composer99
In particular, the CW will be very hampered by the effort to deliver the USSR the build points and whatnot early, since until the US is in the war and getting 50+ bps.

I'm not saying it's a bad idea to send a lot to the USSR (I almost said lend, but you won't be getting it back) - indeed, it's an excellent idea. But first make sure that Germany's going to go in and go in real hard. Building convoys is always good for CW, but they need to build a LOT more to pitch in big-time to USSR than they do if they're just keeping the lines to England open.

Yes, I agree. For the same reasons we often in our games see little LL from CW/US to USSR. Often it feels like to much trouble having to build, maintain and defend the convoy lines to USSR, and all that just to be able to give away BP! [:)] The CW / US players much prefer to spend those BP on their own units.
So I have never myself tried a game with a highly boosted USSR - I think that the original poster had quite some experience with it, though, and he advocated for it strongly, especially combined with the mobile armored defense vs. Barbarossa strategy. So definitely worth being included in the AI's options, imho.

[brian brian]

ORIGINAL: Froonp

ORIGINAL: dale1066

Thanks for that, thats pretty clear and how I'd expect it to be.

Only one question, I seem to remember, and I may well be wrong, reading somewhere that if you have no saved oil or oil resouces it is still possible to re-org 0.4 worth for free?

This is wrong.
You need to trace to available OIL to reorg, if using that option.
If you have OIL available, and you spend 0.4 OIL points to reorg only 2 FTR, then the total amount of OIL you must spend is 0 (0.4 rounded down to 0).
If you have no OIL available, you can't trace to any OIL in the first place, so you do not count at all.

I just learned something about the RaW Oil rule - if you have only one oil, you can only re-org units up to ten tenths or five fifths or what have you. You can't go to 1.4 oil unless you have more than one to trace to. Once you hit ten tenths the first oil is used up and you can't gain the 0.4 for free via rounding as you then have zero oil. Makes total sense but we had never thought about it and were always going to 1.4 when we had only one oil.

Perhaps the RaW oil rule needs it's own thread?

[brian brian]

The three key hexes to have for the Axis to defeat the Russians are Vladivostok, Bandar Shapur, and one of either Petsamo or Narvik (though taking Narvik gives ten CPs to the CW making their lines easier to build). It is good Allied play to give lots of aid to the USSR.

[composer99]

I would have said the key hexes for the Axis to defeat USSR are Leningrad, Rostov, Baku and Novosibirsk, myself.
 
Also, if I am not mistaken, Petsamo needs an HQ in the area or it is useless to the Axis. Likewise with Narvik (even if you can trace supply through Sweden).
 
That's an HQ you're not using either on the central front or to maintain supply in France or the Balkans.
 
As for AI stuff, the USSR AIO should always ask for as much lend-lease resources and build points as it can. For the USSR, there is no such thing as too much.
 
In the meantime, the CW and US AIOs should, based on their assessment of Germany's overall strategy, decide early on (probably no later than S/O 1940) how much lend-lease they are going to send to USSR and how it will get delivered. They should coordinate with USSR so it knows where to rail its factories in event of Barb.
 
They could probably have four levels of lending: High, Moderate, Low, and None, with None to appear only if Germany did a Sealion or late in the war (mid '44 on) when USSR doesn't need it. Low would be 1-5 resources and/or bp; Moderate would be 6-12 resources and/or bp; High would be 13+ resources and/or bp.
 
These numbers can be tweaked. The CW and US need to decide what levels they will be lending at and when, and build out convoys, plan out escorts, etc. accordingly.

[Shannon V. OKeets]

ORIGINAL: composer99

I would have said the key hexes for the Axis to defeat USSR are Leningrad, Rostov, Baku and Novosibirsk, myself.

Also, if I am not mistaken, Petsamo needs an HQ in the area or it is useless to the Axis. Likewise with Narvik (even if you can trace supply through Sweden).

That's an HQ you're not using either on the central front or to maintain supply in France or the Balkans.

As for AI stuff, the USSR AIO should always ask for as much lend-lease resources and build points as it can. For the USSR, there is no such thing as too much.

In the meantime, the CW and US AIOs should, based on their assessment of Germany's overall strategy, decide early on (probably no later than S/O 1940) how much lend-lease they are going to send to USSR and how it will get delivered. They should coordinate with USSR so it knows where to rail its factories in event of Barb.

They could probably have four levels of lending: High, Moderate, Low, and None, with None to appear only if Germany did a Sealion or late in the war (mid '44 on) when USSR doesn't need it. Low would be 1-5 resources and/or bp; Moderate would be 6-12 resources and/or bp; High would be 13+ resources and/or bp.

These numbers can be tweaked. The CW and US need to decide what levels they will be lending at and when, and build out convoys, plan out escorts, etc. accordingly.

Now that you have said it - obvious.[;)] Before then, I hadn't realy thought it through.

[npilgaard]

Re: USSR stuffing the border.
Stumbled on the below post in the wifdiscussion forum - it presents a very structured decision process with assessments of the success of various USSR / German strategies (he has explained the details behind his assessments in other posts - I can see if I can locate some of them if you like).
Maybe this formalized structure can be of interest regarding the AI. Also, the continual reevaluation of garrison ratio and calculations needed for this should be well suited to the AI.

Regards
Nikolaj

----------------

(...)

Let

AOV = sum(Value(OS)*P(OS|A))

Where
AOV = Average outcome value
OS = Outcome State, ie the situations on the board when a given
outcome happens
A = The action/choice/strategy that I make
Value(OS) = the "value of the board" given the OS. Gernally, this is
the same as the chance of winning given the OS
P(OS|A) = the probabiliy of OS happening, given Action A.

Whenever I make an action or choice in the game, I try to select the
action or strategy A that maximizes AOV, the average outcome value.
Very often, this implies another action, than the action whose most
probable P(OS|A) gives the highest Value(OS).

The 2d10 analogy is chosen for simplicity. If you make a +8 assault,
and roll average (11), you will loose 2 units and the defender 1. The
change to AOV can then be expressed as -1 unit (ignoring the negative
vaue of flipping), and the attack seems bad. However, if calculating
the full AOV, attacking means that the value will increase by about
1/2 unit (+ the negative value of flipping times probability + the
value of taking the hex * probability + the value of the defender
shattering * probability).

Applying this to the stuffing example we must evaluate the following
probabilities and values:

P(Stuffing successful | Stuffing attempted)
P(Russia pull back | Stuffing attempted)
P(Germany pull back | Stuffing attempted)
P(Stuff broken | Stuffing attempted)
Value(Stuffing successful)
Value(Russia pull back)
Value(Germany pull back)
Value(Stuff broken)

The alternative strategy (not trying to stuff) is used as a null
hypothesis (NH). For simplicity, it makes sense to estimate the
values above relative to the NH.

(This is of course a quite coarse grain with regards to outcomes. For
more exact understanding, one may need to add more results, such
as "Germany break the garrison J/F", "Germany breaks the garrison
M/J", etc., but that is not necessary to show the main point.)

Now, my personal assessment, is as follows:

P(Stuffing successful | Stuffing attempted) = Small
P(Russia pull back | Stuffing attempted) = Medium
P(Germany pull back | Stuffing attempted) = Large
P(Stuff broken | Stuffing attempted) = Very small
Value(Stuffing successful) = Large positive
Value(Russia pull back) = Small positive (at least Super Balbo is
avoided)
Value(Germany pull back) = Medium positive
Value(Stuff broken) = Very large negative

Total AOV = Small to medium positive (for Russia), so a rational
Russian should always go for it, imo.

Now for Germany, the choice is to attempt a Barbarossa. Here the null
hypothesis is to follow a close the med strat from the beginning:
P(Stuffing successful | Barb41 attempted) = Small
P(Russia pull back | Barb41 attempted) = Medium
P(Germany pull back | Barb41 attempted) = Large
P(Stuff broken | Barb41 attempted) = Very small
Value(Stuffing successful) = Large negative
Value(Russia pull back) = Zero or maybe small positive (Since the
stuffing anyway made Germany unable to build the best barb41 units)
Value(Germany pull back) = Small to medium negative
Value(Stuff broken) = Large positive

Total AOV = small to medium negative. On average, the
close_the_med_41 strat is significantly better for Germany.

This is why I claim that a rational Russia should always prepare to
stuff, and a rational Germany should always do a close_the_med_41 if
assuming that Russia is rational, when using the RAW stuffing system.

All of course, according to my estimates about the P's and Values. In
particular, the relative probabilities of the the different P's
depend on how much information each player has. If Russia has little
information, the P(Russia pull back) gets smaller compared to P(Stuff
broken) and if Germany has less information, the P(Germany pull back)
gets smaller compared to P(stuff successful).

This is why it is so important for both sides to have maximum
information and to deny that information to the opponent.

(...)

Cheers
Hakon

[Shannon V. OKeets]

I would be interested in the other posts if you can find them. Hakon has contributed a lot to the AIO discussion for MWIF.

What he is doing here is an Operations Research analysis using Basian (sp?) analysis. It most often is shown using a decision tree with the 'leaves' of the tree being final outcomes and the 'branches' decision points/alternatives. The value of each leaf contributes the value of the branch, weighted by the probability of the leaf outcome occuring. Then each subbranch can be added to the value of the 'higher' (or 'lower" depending on how you draw the tree) branches. This lets you evalutate each alternative at each decision point.

I completely agree with his use of low, medium, and high, rather than numbers. Though the analysis may be quite rigorous, it is never going to encompass the full range of possibilities in WIF. That's because of the random unit draws, the weather, and the initiative rolls, just to name a few of the more important highly dynamic variables. Given that perfect precision is impossible, using precise numbers is silly/arrogant/delusional. Far better is to use the fuzzy evaluations low, medium, high. I prefer the scale: Excellent, Very Good, Good, Fair, Poor, and Awful. This igives a slightly broader range of outcomes and imposes a desireability bias in the labels - which in this case I want to have, since the analysis is for a competition (us versus them).

I disagree with his recommendation for always choosing the alternative with the better outcome. Instead, I want the AIO to choose the clearly superior choices (Excellent, Very Good), and avoid the inferior ones (Poor, Awful). For those in between (Good, Fair), the AIO will use a random number to decide, where the probability of choosing the better outcome is weighted by how much better that outcome is. My purpose for doing that is I do not want the AIO to be too predicatble. I see predictability as a major weakness in head-to-head competition - poker comes to mind as an excellent example of why predictability is bad. Blackjack, of course, is a counter-example, but that is because the house is forced to take hits/not take hits. In WIF, there is nothing constraining your opponent from taking advantage of his foreknowledge when you are too predictable.

[npilgaard]

Sounds good regarding the AI.

ORIGINAL: Shannon V. OKeets
I would be interested in the other posts if you can find them. Hakon has contributed a lot to the AIO discussion for MWIF.

Lol - hadn't noticed - then he can probably explain his assessments better than I can
I will see what I can dig out, though.

[hakon]

It's been a while since I've been writing much on the strategy threads on this forum. I had a rather long discussion with Patrice on the stuffing topic about a year ago, I think, where most of my argument was presented. Basically it goes:

- Russia should always stuff (reason: little to loose, very much to gain)
- If Russia stuffs, Germany should not try a 41 barb. (Reason: little to gain, more to loose)

Which means that all games will be close the med games.

While I agree that this logic may not apply to the AIO, since it doesnt have the same ability as a human has to learn from mistakes, create counter strategies on the fly, etc, I feel it does apply to human players. (Which is my prime concern.)

What I am hoping for, eventually, are rules changes, that:
- Allows Germany to go through with the historical actions, regardless what the USSR does, that is that they get to attack Russia in 1941 even if they finish Tirpitz, wage a battle of the atlantic, even takes some losses in France or in a Battle of Britain, and even allows them to have some forces in Africa while doing this.
- Prevents a knock out of Russia if Germany goes for the Super Balbo strategy (as it's been called, one of my posts on this strategy has been copied into the German strategy thread), ie the strategy of building a large air force of more than 10 (maybe even 20) aircraft prior to Barbarossa..

Secondary concerns, include:
- Reduce the effectiveness of the Super Balbo and Super Alex strategies in general.
- Create more friction beteen nations allied to each other, for instance by altering the victory point system into a system with separate victory conditions for each country. (The Britannia board game has a good system for this, for instance.)

Since all of the above really require changes to RAW (imo), I understand that it is currently out of scope for MWIF, which is why the discussion has been going on in the wifdiscussion forum. What I hope, is that MWIF, when released, will lead to the game being played more and more competitively, which should then lead to imbalances like this being fixed in patches, which may or may not be led (or followed) by corresponding fixes to RAW.

I expect one of two to be the case:
- MWIF implements a good stuffing strategy for Russia, which is quite often (or always) used by the AIO. This should frustrate German players enough for them to demand a nerf of that strategy.
- MWIF doesn't, in which case, super Balbo will dominate Russia so badly that people will also complain.

Of course, this should also show up more in multiplayer games.

Cheers
Hakon

[IKerensky]

My USSR Strategy:
 
  Hoard all the oil on the map.
  Stack STR and a CP in the Caspian to DOW persia, STR the MIL and walk a CAV to Teheran ( harder in CWiF but possible ). Could need a HQ.
 
  Have the STR stack back to europe and while Germany and France fight off ask for Bessarabia, pray the German say no. STR Bomb Ploesti out of existence. Axis are game over.
  Keep the STR stack until barbarossa and STR bomb ploesti ( at night if necessary ) until it is rumbled.
 
  By STR I mean every russian bombers with 3-6 STR bombing and 10+ range. Even if you dont use your FTR at set up for that anyway you are going to build them better planes.
 
  Depending on the Oil rule version played Germany and especially Italy are entirely dependant on Ploesti to work. You can even start thinking about going to Irak from Russia with a paradrop then a destruction of the Oil field.
 
  If you play with factory destruction then a good strike on Ploesti ( not too hard to achieve with the amount of STR the russian can muster : easily several attacks with HQ reorg at top column ) can just win the war, and Berlin and a lot of other area could be reached by the russians bombers. If the CW do the same on the other side the Germans will soon be in deep poopoo.
 
  Also sail your subs out of Vladivostok and declare on Japan at a big chance of turn ending impulse, try to split them at theirs CP lines with USA and cut it. If you succeed you seriously upped the ante for your side to win the war with early US intervention. Also you can bet that next time Japan will burn a decent amount of Oil each turn just to keep his CP lines safe, wich is even better.

[wosung]

ORIGINAL: hakon


- Russia should always stuff (reason: little to loose, very much to gain)
- If Russia stuffs, Germany should not try a 41 barb. (Reason: little to gain, more to loose)

Which means that all games will be close the med games.

That's true from what I read in the wifdiscussion forum http://games.groups.yahoo.com/group/wifdiscussion/

Close the Med seems to be the dream scenario, conceptualized by an Australian game company for an essential anglo-american group of customers: To play/or beat Rommel.
BTW: Nothing against ADG, nothing against anglo-american customers. Without them both: no MWIF.
But a more variable setting would be nice.

And for the Russo-German stuffing: What exactly is this game mechanism aiming at?

ORIGINAL: hakon
What I am hoping for, eventually, are rules changes, that:
- Allows Germany to go through with the historical actions, regardless what the USSR does, that is that they get to attack Russia in 1941 even if they finish Tirpitz, wage a battle of the atlantic, even takes some losses in France or in a Battle of Britain, and even allows them to have some forces in Africa while doing this.
- Prevents a knock out of Russia if Germany goes for the Super Balbo strategy (as it's been called, one of my posts on this strategy has been copied into the German strategy thread), ie the strategy of building a large air force of more than 10 (maybe even 20) aircraft prior to Barbarossa..

Secondary concerns, include:
- Reduce the effectiveness of the Super Balbo and Super Alex strategies in general.
- Create more friction beteen nations allied to each other, for instance by altering the victory point system into a system with separate victory conditions for each country. (The Britannia board game has a good system for this, for instance.)

So do you think the balance issue is a problem of action limits and of balancing the role playing element (naval oriented CW vs. land oriented Germany)?

I like the "more friction" part for MWIF Product 2. It would be nice to have a interface for inter allied bargain: lend lease of a/c for raw materiel. Oil for invading France in 43 instead of a med strat. Esp. for playing against the AIO this would be fine.


Regards

[Froonp]

ORIGINAL: wosung

ORIGINAL: hakon
- Russia should always stuff (reason: little to loose, very much to gain)
- If Russia stuffs, Germany should not try a 41 barb. (Reason: little to gain, more to loose)

Which means that all games will be close the med games.

That's true from what I read in the wifdiscussion forum http://games.groups.yahoo.com/group/wifdiscussion/

Close the Med seems to be the dream scenario, conceptualized by an Australian game company for an essential anglo-american group of customers: To play/or beat Rommel.
BTW: Nothing against ADG, nothing against anglo-american customers. Without them both: no MWIF.
But a more variable setting would be nice.

This is a huge major concern for me too.
"Close the Med" looks too much like the ideal scenario, which is wrong IMO for a WWII based wargame.

[hakon]

ORIGINAL: wosung

So do you think the balance issue is a problem of action limits and of balancing the role playing element (naval oriented CW vs. land oriented Germany)?

Balancing can be done in a number of ways. Simply making Russia stronger would balance the super Balbo, but could cause the Axis to avoid battle as long as possible. Making russia stronger AND making the surprise turn more effective (for instance by requireing some russians to be at the border at the time of the dow), could have the opposite effect, that is to make a Barbarossa the only way for Germany to survive, since those extra forces could make Russia too strong if not hurt sufficiently during the surprise turn.

But doing something about the logistics system or the action limit system could work too.

In raw, the only factor that limits Germany's speed of advance in 1939 is the need to rebase aircraft to the front. Of course, with unlimited italian air forces, that problem disappears, and a Russian strategy of gradually pulling back become far less effective. In real life, of course, the German land forces was just as limited by overextended supply trains as the air force was, but for game purposes, limiting the air force is enough.

To keep WIF's simplicity, removing the ability for Italy to take air while Germany takes land, would solve the problem. For instance, impulse type could be per side per map (letting Russia be counted as a speparate side for this purpose), so that if Germany took land in eastern Europe, Italy had to take land there too. This would have the added benefits of also fixing super Alex, and also giving Germany a chance to sail her subs while performing Barbarossa.

Another solution, that lends itself better to a computer game than to a board game, would be to create a more realistic logistics system. First of all, it coudl be made necessary to repair rail lines in the Soviet union. This coudl be made by placing an HQ or ENG on a hex, which would repair the rail in the hex at the end of the impulse, and only if the hex already were next to a hex that had already been repaired. This would limit the long term advance of the German army in Russia to one hex per impulse.

Even more detailed supply systems could be imagined, such as he one used by the otherwise relatively simple Matrix game, World at War, where supply is produced, and has to be moved to the front using, and limited by, existing infrastructure.

[brian brian]

I like the impulse-per-side-per-map idea a lot. Let oil limit unit activities, not the current system that encourages playing to the rules system too much.

RaW does have individual country victory conditions though. It's just that WiF takes so long to play, there is a lot of social pressure not to play the game to win as an individual (and thus not to stuff...c'mon man, don't be so boring, or let's just get rid of that rule for this one game....) For a lot of people a game of WiF is as much an experience as a contest. I think the computer game, with less invested in all those pieces and trips to the game room and everything, will help bring out the competition within each side a little bit more. But it might show more of the doubters just how powerful (errr, maybe 'rigid' is a better term) the stuff strategy is.

And the WiF logistics rules are quite a giveaway to the players, to keep playability high. If Russian rail would need to be converted, the true amount of shipping required to support D-Day should be modeled as well, in addition to how many tankers the Axis can get a hold of in the Med and several other large logistics questions that are swept under the playability rug.


For close-the-Med games, I've always thought Spain is a little weak. On September 1st, 1939, they had the most combat-experienced army in Europe. Germany would overwhelm their front line eventually, but the partisan warfare behind them would be vicious and effective. I also think Stukas are far too effective in mountain hexes. With DoD1 and WiF5, if anyone was dumb enough to DoW Spain during the Civil War, they would fight against both the Nationalist and Republican units simultaneously. This seems to be a more plausible simulation of what a 1940 Spanish campaign would be like....hordes of experienced infantry ready to take on the invader.

[npilgaard]

ORIGINAL: hakon
Balancing can be done in a number of ways.

One option would be to play without WiF Edit: not without WiF [:)] but without PiF - that would drastically reduce Italy's ability to build a to-large air force.
In our group we have played without PiF the previous game and it worked out very well (everyone was very pleased with the result). We are about to start a new game, and we will play without PiF again.

But iirc it is not possible to disable the PiF-option in MWiF (?).
I don't know the data/code structure, but (without the proper knowledge) it don't seem to difficult to just remove a number of units from the available aircraft units, and thus allowing the PiF-option to be turned off. It will have a major impact on AI, though - maybe that is the reason not to do it

[npilgaard]

To elaborate a bit (from a post in wifdiscussion-forum):

At first glance it may sound unsatisfying to leave out so many units - after all, most experienced WiF-players like the added complexity and flexibility that extra units provide. Also, they allow to focus more
on specific types of units; and one could argue that it leads to imbalance to play with e.g. SiF but not PiF.

Nevertheless, it turned out very well - one of the best changes we have made to our games for a long time.
A number of advantages:
- limits the Italian/CW all-air strategies
- avoids the great FTR-race - without PiF there are only so many FTRs, and most major powers build them out reasonably quickly
- less aircraft available to be built means more BP for other things - we now see many units being built that we otherwise rarely see in our games - also, the poorer aircraft are now fielded (and often not scrapped) since even a poor aircraft means one more aircraft on table; when having played many games, (almost) any kind of change is
refreshing
- with less aircraft available it is a much harder choice where to deploy and use them - and more player choices makes the game more interesting and fun, imho
- air combat becomes more exciting - it now means much more to lose one or two good aircrafts
- the 'pool all axis ac in the 1 or 2 box in the Med and hope for a successful search roll'-tactic is less of a problem. Italy can still cause trouble with NAV/FTR in the Med, though, so not unbalancing imho
- and similar: avoid the huge bunch of japanese lba guarding 2-3 sea areas in the late war
- thus CVs become more important - even an occasional Italian CV or two
- air-to-air combats tend to be fought between smaller forces, which imho is a good thing, since the air-to-air combat system handles battles between large amounts of of fighters poorly - unlike the land combat system, which is based on relative strength (i.e. ratio), the air-to-air combat system is based on absolute strength (i.e.
difference), meaning that with lots of fighters, but same number on each side and roughly even strength, a single loss or two can mean that the odds become so bad, that the combat is pretty much lost (e.g. huge naval air battles near Japan in the end-game, where 10+ fighters (or much more if playing with multiple CVPs on each CV) participate on each side just a couple of losses can lead to a +4/-4 situation or worse, meaning that chances of success have decreased dramatically - with fewer CVPs the usual 'build lots and lots of cheap CVPs and
take losses somewhat lightly since there are plenty to replace with' is not seen in the same magnitude - it now takes 4 turns before a good CVP returns
- usually, in '44 and '45 the allied impulses can take a very long time to complete due to super-combined and lots of aircraft - to some extend this is avoided, meaning a better pace of the late game

All in all, this is of course a matter of taste, but we are very happy with playing without PiF (and pilots and CVP - using only the CVP from SiF) for a change (and as mentioned, we continue doing so, at least atm), and regarding play balance it seems to me to work out well. Of course it requires rethinking some of the usual strategies, but that is very refreshing when the games otherwise can be somewhat repetitive.

Just a few thoughts about whether to PiF or not to PiF

[IKerensky]

Use oil rules...

[hakon]

Although it may help balance out things a bit, I see two problems with playing without PiF:

1) It is one step towards playing with a fixed OOB. I really like the ability that a full WiF gives in terms of letting the player play with various "what if" scenarios. Historically, Hitler WAS criticised for building too few fighters (and too many bombers).

2) Without PiF, advance building of aircraft (unless also banned) becomes too much of a dominating feature. Paying 2 extra build points for an aircraft with 1-2 extra air-to-air factors is easily worth it.

In general, I prefer to fix the errors in the actual simulation over introducing arbitrary constraints like this. And in my opinion, the real problem is that having 2 countries, 1 that does a land and one that does an air, is too strong compared to just having one country.

[Sewerlobster]

ORIGINAL: hakon

Although it may help balance out things a bit, I see two problems with playing without PiF:

1) It is one step towards playing with a fixed OOB. I really like the ability that a full WiF gives in terms of letting the player play with various "what if" scenarios. Historically, Hitler WAS criticised for building too few fighters (and too many bombers).

2) Without PiF, advance building of aircraft (unless also banned) becomes too much of a dominating feature. Paying 2 extra build points for an aircraft with 1-2 extra air-to-air factors is easily worth it.

In general, I prefer to fix the errors in the actual simulation over introducing arbitrary constraints like this. And in my opinion, the real problem is that having 2 countries, 1 that does a land and one that does an air, is too strong compared to just having one country.

I'd have to say that I am leaning the other way on this. PiF has always been, for me at least, an unneccesary complication to the game. By abstracting the pilot training,i.e. eliminating the pilot bookkeeping, WiF returns to being a wargame. Don't get me wrong there are times I'll play w/pilots but it always feels incongruous with the scale of the game.

This is one facet I'd just as soon see eliminated if it gets the game out sooner.