I have a problem with the balance and what happens when the Allies have bad roll
I need help… The Hard AI just destroys me. Russia loses the start. Britain gets wiped out. I see no fun in this game at all. The Allies lose so much on the first roll that it is almost unfair.
Help me understand the game… It has to be more than just rolls of the dice. if it isn’t, then I want my money back…
Or I’m just terrible at rolling the dice…
Don’t think it is dice; 2 fig, 1 car, 1 bb vs 1 cru, 1 car, 1 fig is a total loss for Britain. The board seems to be set wrong. Germany and Japan can take out Britain’s navy without much worry. Like the original game where we came up with house rules of no first Russian attack, two hit battleship, bombers able to drop infantry. The game just seems, in my opinion, to be totally unbalanced unless the Allies get a good starting roll.
I had a chance to win vs the Hard AI; I took 11 away from Japan and got stupid. After that, the 7 bombers Japan built destroyed my navy and set the Allies back three turns. Game over because of my own stupidity. Since then, I have not even come close to taking the islands away. In fact, Britain loses 2 cru, 2 fig, 2 car, 2 trn, 2 inf and a sub to zero for Japan. That’s game over three times in a row…
I am guessing you’re talking about v5 in tripleA?
Perhaps you should start by risking less in your opening attacks? For example, as Russia just attack W. Russia with everything. Don’t bother with Ukraine or Belo or Baltic. If you go all in to W. Russia only there is a very good chance you will sweep the territory in the first round of combat, which decreases the chance that the German defender will put up so many hits.
Buy Artillery with the Russians, and don’t bother with Caucasus in the first round. Just trade it, and place all your units in Moscow.
With UK fly 2 fighters either to W. Russia or Arch depending on what G does with its cruiser e.g. if the cruiser Goes to sz 7 then you land them in Arch, if it’d stays in sz 5 you land them in W. Russia.
If you do that, you can take some of the dice out of the equation.
So for example, rather than running 2 attacks at 60% odds, just run a single attack at 90% odds. And make sure to give yourself fodder cover.
Don’t run attacks against sz37 with UK. That is taking a big gamble, and it sounds like gambling isn’t panning out for you lately instead be more conservative in the first round.
For example if you’re attacking sz61 with UK, don’t just leave it up to chance, instead of bringing only the fighter bring the cruiser or carrier too.
If you like I can post some save games for you, showing how the Allies can come back from fairly brutal dicings.
I think your issue might be managing the endgame, more than individual battles or the odds in any one of them. Taking either Axis capital is challenging because you have to set it up over multiple rounds and 1942.2 is a different beast than Revised or Classic. It’s much more challenging for Allies.
If it’s helpful I can show you my favorite openers for this board as Allies, with saves after the combat move phase for the first two rounds. I’ll check the archives when I get home.
It’s possible that where you are running into problems with Allies is how to use USA effectively, since it is the probably the hardest nation to play in 42.2. This is a little different than earlier boards where I think you could make the case that UK was the trickiest nation to manage.
Ps. Or even more expedient, if you’re willing to post a save game file from one of the matches where you lost to the Hard AI (either here or on the tripleA boards) I could look it over and give you some feedback on areas for improvement.
One thing I would really encourage you to do, is to buy more artillery with Russia. You will never back down the German advance by buying solid infantry in 42.2. As the Russians you need to at least threaten an effective “all in counter attack” or “counter strafe” against the Axis forces. You’ll never pull this off with an infantry wall, especially once Japan starts flying in fighter support. Basically what you want, (and this holds for human opponents as well as the computer) is an artillery piece to match every infantry unit. You can do something similar with armor, but its a lot more cost prohibitive. Otherwise what’s going to happen is, you’ll be able to defend but you won’t be able to go on offense, especially during the critical rounds when G tries to break your position on W. Russia/Caucasus. So don’t waste your artillery on pointless trades for low value territories, instead save them for the really critical battles of the mid-game and the end-game.
Also, remember to always sink Axis transports when they are left undefended, the same thing with Axis bombers. This is the easiest way to catch the HardAi off guard, though it won’t work against human opponents.
Also, if the German subs hit in sz 7, remember to take the first casualty on the Russian sub (not the battleship!). TripleA will default to the battleship for the absorption, but this will increase the likelyhood that your BB gets sunk with fewer chances to return fire (since German fighter hits cannot be taken on the Russian sub, as no German DD is present.) That one is a rookie mistake that many people make.
There a number of other things I would suggest, but again its a lot easier if I have a savegame to look at to see where your games are diverging from the norm.
Attached below is a worst Case scenario for Allies. Notice that I elected to play it super conservative and only run a single battle with Russia, the battle in W. Russia. And wouldn’t you know it, those damned German’s hit 5 times! Just a nightmare of red dice. (I had to run this battle about a dozen times before I got a result this catastrophic hehe.) But with confidence in our superior brain power as humans vs the machine, lets give it a shot and see how things shake down, from this grueling start.
You can see here that by J1 the UK didn’t have much luck either. Although the Russian sub was able to steal a few hits away from the Germans in sz 7, every single Axis defender in all the other battles hit!
After such a morale reducing series of setbacks, one might be tempted to quit. But I’m going to give this a couple hours just to see whether Allied recovery is possible. I’ll report back with another save in a few
ps. here we are just a short time later, at the end of the third round. You can see that already, even starting from a really crappy position on the eastern front, Allies are starting to make progress. With a little more patience and just a couple Axis goofs, we might just be able to turn the tide back in our favor.
And now a round later….
Notice how the bad Allied luck continued with the US bomber being shot down on a risky AA run to close out the third round.
Can you see now the critical error in judgement that the HardAI has made? How USA is now positioned to cut off Japan’s access to all the rich islands in the south pacific, and how HardAi has left the Japanese transports practically defenseless (meaning that if I take the rich islands this round they cannot be retaken again for another round.) So this is the round where Allies start to turn the tide of the war, and as usual, it came from patiently pursuing a 5 year plan with USA.
If each one of the Allies, and USA especially, works at projecting power in one direction for 5 rounds, its usually enough to make the gains necessary to start turning the war in that theater. Then you can divert resources back in the other direction to try and prevent “center collapse” to Axis. But honestly, even if Moscow were to fall, a strong position for US vs Japan can still be enough to play into the deep endgame, so I am optimistic. Lets give it another couple rounds and see
pps. And here we are entering the mid-game the critical fifth and sixth rounds. If as Allies, your situation on the gamemap has not changed appreciably from the start after 6 rounds, then chances are you’re not pursuing a concentrated enough strategy.
Lets pause to look at UK’s move here in the fifth round. After the daring Carrier action by USA in the Pacific, its time for Churchill to start making long term decisions. It’s also time for the supreme Allied command to determine whether they want to pursue a full KJF and trade Moscow for Tokyo entering the deep endgame, or simly contain the Japanese and try to save Russia, redirect on Germany. I say go with your gut at this point.
Here is the situation in the 6th round as UK. Note that Moscow can still be held, UK has 44 ipcs and USA has 50 ipcs and has just entered the production game in the south Pacific, by purchasing factories there.
For me this would be a fun game to play out as Allies, with a fair amount of confidence that, although its not yet decided, Allies can probably carve out the edge going into the endgame.
Thanks… I haven’t been playing much as work has taken a large part of my life right now. Basically, I don’t keep up the same strategy and as the allies, tend to not think of how the countries can defend each other. I’ve been studying the save games you posted in the other thread and noticed that I get very distracted and don’t make the AI pay when it makes poor moves or don’t see a good landing spot for my fighters. Didn’t notice they can get to Archangel from the sea zone the cruiser usually is in. I also expect the AI to do things that I would do, like destroy the transport and carrier moved from India towards Africa by using the two fighters in a 3 move and then bringing the battleship and carrier from the East Indies for the fighters to land on with the remaining move.
I’m hoping to play my friend this weekend and try one of the all bomber buys just to see what happens. He is convinced, as I was, that the game is hopelessly unbalanced and the allies have no way to win. We played 1942.1 about 3 times and I never got to do much because as Germany I did the 6 tank push and got hit by some terrible dice early on. Never have played Japan and most of the time, just lose focus and want to try something other than the infantry push. As Russia, I will buy the 4 inf, 3 art the first round and try to hold West Russia with fighter support. USA gives me a lot of trouble because I see so many things that I want to do and end up doing none of them. I will forget to move pieces in the non-combat phase and then end up being a move behind.
I like the idea of 3 tanks per turn in India, single fighter and infantry for Britain as long as they can afford it. Get 6 or 9 tanks in India and with infantry/artillery it starts to become a major distraction for the Axis.
There’s usually a fair amount of drinking involved and my strategic thinking turns to when the next shot of whiskey is going to happen…
Anytime dude! I think its a cool map, and the HardAI is pretty entertaining as well. It does some great moves, but also has some kinks that can make the steam roll against the computer highly gratifying, when you just want to beat up on the machine. Also, hopefully some of these exploits will be looked into in future AI updates, so its always good to keep at it.
I think the main thing to keep in mind, especially with USA, is to pick a direction and gun for it, with steady purchases for like 4 rounds before you make a major move out. Sometimes moving out too soon is where people get hung up, and they trade units before they have enough force in the area to defend the gains they’ve made.
I also updated the post above with savegames through the 6th round UK. Notice how as a human I make use of destroyers as blocking units, whereas the HardAI doesn’t seem to get the concept quite so well. So here USA’s fleet is more or less protected from attack at a cost of just 16 ipcs sacrificed, so secure India, the South Pacific, and pretty much the whole rest of the game in the Pacific. A bargain basically
Axis advantage, hard liquor! heheh I know how it goes. I smoke like a chimney myself.
I loaded the save games up to where Britain is going to move. At this point, I see 6 fighters and 1 bomber for Japan that can pretty much wipe the existing British navy out unless it move or there is some fighter support with the carrier. I don’t see how a couple of fighters can get there so I don’t really have a good idea what to buy to prevent having to move the navy away. I guess I’ll get the next save game to figure it out.
Yeah, my one turn thinking did me in again. I saw where the fighters could land as the map was set, not as it would be after the Britain moves. Now, the fighters are closer so they can land back in Germany so somethings gotta give. Japan is in trouble and Germany keeps moving it’s tanks around in a vain attempt to position for the fatal attack on Moscow.
Interesting on the destroyer blocks and the factories. Not sure what you are going to do however it will be fairly easy to rebuild the carrier and fighters so the navy will be pretty much untouchable at that point. Losing 12 for Japan is close to terminal as it cannot support the factory with much and now must regroup to protect the main island. Plus you can bring up 7 inf/arm/art for pushing Japan out of Asia.
One thing the AI really should do is keep fighters supporting the carriers when available. Otherwise, it’s a poor excuse for a destroyer. I was close in one game against the AI like this however I didn’t position my navy correctly and everything got smoked by the Axis air force. I will check back in over the weekend as I will be out and am not going to load TripleA on my company laptop.
The destroyer block also forces Japan to move the navy somewhere else or it will get pounded by the pack of subs, destroyers, battleship and cruiser. You are also able to reinforce immediately and the IPC advantage makes it a poor choice for Japan to get into a trading war with USA.
USA ices it at the end of the 15th round!
final Allied Victory savegame below…
HardAI did indeed use the Japanese air in a daring assault on the Royal navy, but the gains USA made in the pacific were too much for Axis to handle. After the airstrike on the British fleet, allies adopted a plan to stall Japan while reinforcing “the center wedge.” Took a few more rounds, but ultimately Allied victory was inevitable. You can see also how UK is able to build up over a few rounds in the midgame, to a point where it can begin to threaten Germany again, even after losing its entire Atlantic fleet and all its transports out of UK.
Clearly the main problem here for HardAI, was its decision to back the IJN away from the home island and sail towards Panama. No human player would ever do this, but for the HardAI this proves its achilles heel. Once the rich islands are captured, containing Japan and Germany becomes just a matter of time.
Here USA cracks Tokyo in double hit over two rounds. While UK positions to threaten Berlin and prevent any hope of Axis recovery. A human would likely have given up hours ago heheh, but HardAI is a trooper
Under the “game” tab at the top of the screen, you can click “show history” to see the play by play for every round/phase. Click the little “+” symbol to the left of the game round. This will expand the individual game round and show all the phases in that round for each nation, purchase, combat, non com etc. There you can get a sense of what I was aiming for, how I got smoked a few times by HardAI, and how the plan was adapted at various points in response to what HardAI was doing.
in v5 HardAI Germany plays fairly well into the endgame, but Japan runs into some major problems once you disrupt their transports and start squeezing the Manchuria factory. Next time I’ll show a more KGF approach from a similarly weak Russian starting position. If I get a few hours free.
Have fun in the game man, and best of luck!
Thanks for the info; I’ve been loading the save games and then looking at the history round by round and then the individual steps by each power. I like the car, sub, des build up by USA for a couple turns and then go hunting for income. The islands have always seemed like a good source for increased income for USA. I was looking at the buys and thought that a bb, car, des first round would be interesting followed by 3 sub, 2 des, 2 inf second round and then trn, inf and other support units on the third round and then moving out on round 4. Who knows, I may download the source code, build it and check out the logs to see why the IJN Hard AI is getting stuck by South America. It seems that the path finding is having trouble with the canal being closed and the intent is to get to the Atlantic to support Germany. I would also like to see the AI take a few more chances when playing with dice. Pathfinding with navy, transports and air seem to be the weak points of the AI as you have noticed. Unlike the fixed board of chess, the movement possibilities are very complex as well as unit interactions. I don’t have any experience with game AI however some sort of database of opening moves and preferences would seem like a valuable addition. Nothing hardcoded in the game code for specific maps, just some rules in the map database to give the AI a starting point.
Hoping to get a live game and will do my best to keep focus for 5 turns as USA on the Pacific. Positioning of the Russian stack is important and making sure it can support WR, CAU and Moscow to trade territories or punish a bad move is one of the things I have trouble with. Also, need to be aware of when UK and USA can reinforce to prevent a German tank blitz.
Thanks for the info. I want to find some new ideas because the friend that I’ve played AAA with is convinced the game is unplayable and I’ve never even played a game with the new map other than the AI.
After all, the worst that can happen is I make a bad move or don’t see something and get to play again once I can convince him the game is worth playing.
I posted a bunch of savegames in the AI development thread, showing a couple different approaches with some ideas on how to get the edge.
Let us know how the game goes against your buddy. We’ll be pulling for you dude
@craykirk - Glad the new Hard AI is giving you a challenge. I don’t read these forums much so if you have specific feedback on the Hard AI then I’d recommend posting over at the TripleA dev forums: triplea.sourceforge.net/mywiki/Forum
The naval priorities need some work still as the AI will often end up getting its ships stuck in various places and doesn’t always do a good job of defending its transports. I’m currently working on the air AI in order to minimize the number of times the AI leaves planes in undefended locations.
If you or anyone is interested in actually assisting with the AI development just let me know and I can give you an overview of how to get started.
If you or anyone is interested in actually assisting with the AI development just let me know and I can give you an overview of how to get started.
Welcome to the forum, redrum, and thanks a lot for your efforts to improve the AI in TripleA.
Can you please be a bit more specific on “assisting with the AI development”.
Do you need Java programming skills or do you need just testers?
I will open a new thread about this, as this important topic may get lost here…
It would be great if you could comment on this here:
I’m a software developer by employment, so I could probably look at a few things. One thing I thought about was creating some unique personality modules by map, packaging them in the zip file with the map (same format as a jar) and using Spring to inject the code so it would be used by the engine. This way, any map specific AI could be kept with that map. IJN could have some logic to prevent the route finding from getting stuck dealing with the canal (it appears this is what happens) and even paying attention to the islands when threatened.
Or even some standard openings in the map xml similar to chess programs that have standard openings.
@craykirk - I’ll add more info around the Hard AI development in the thread P@nther created so not to derail this thread. To address your specific point, I plan to do something similar to what you described though most likely just have an AI XML file with things such as openings, territory values, unit values, etc. Essentially a way for map makers to give specific guidance on their map for the AI. I have it on the list of things TODO in the main AI development thread here (- Add per map XML AI configuration): tripleadev.1671093.n2.nabble.com/AI-Development-Discussion-tp7585227.html
Boo… Game fell through as he wasn’t feeling well. I would like to get good enough that I can beat the AI however I don’t want to just mimic the save games that I’ve downloaded. Map awareness and unit counters are one thing I really need to understand better.
It’s just a lot more fun to play vs people than the computer. I just want to get good enough to get my friends interested in the game instead of thinking the way I did. They tell me they’ve tried everything and the Axis are overpowered.
I understand how long software development can take, putting in changes, making sure nothing else is broken, etc. As soon as you start assigning new priorities to different units and territories, it can upset the balance of how the land AI is currently working.
The main difference between v5 and Revised is the amount of time it takes for Allies to do anything meaningful. With UK you’re looking at either fighters or tanks for a while. Basically it takes three rounds minimum before you can move out safely.
Often times I like to buy 3 tanks in India with UK and save 13 ipcs for the second round. This gives you a chance to see what Japan does and where Germany puts their air. In old boards G based air out of France, but in v5 NW Europe or Skandinavia is better, because it allows you to cover sz7 with fighters. If G buys a bomber in the first round, there is basically no way good way to build naval out of UK in the first two rounds (even if your Canadian destroyer survives and you manage to sweep all the uboats.) So I usually plan for a third round entry into the game for the Royal NavyAtlantic. This gives you just enough time to be in sz 13 with your Australian carrier deck. Unlike Classic or Revised, where you could shuck into UK or Africa from Eastern Canada v5 really requires 2 fleets to properly manage an Atlantic crossing, one in sz 13 and one to go somewhere else.
In Classic positioning an Allied fleet by Brazil was fairly pointless and pretty much a losing proposition over the North. In v5 Brazil staging or launching to sz23 is a somewhat stronger play, but it still requires a build up, because G has a lot of air to set up their defense coverage vs fleets.
India is very hard to hold when Japan is Burma pushing, but if you can get 3 tanks out of it to support Caucasus in the first round I find this gives UK some teeth early on. The alternative of buying armor out of UK just isn’t as effective as in Revised. It takes twice as long as earlier boards, and there aren’t all that many places for the tanks to go. Africa takes a round longer to push across (even with the blitz) and Karelia and the North is much easier for Germany to cover with air (since there are no shucks up into Archangel and the transports provide no fodder cover when parked up there.) By contrast, India with the starting factory will let you get 3 into Russia quick style, and maybe more if the Japanese player is being cautious and conservative with their air. I posted a game last night showing how you can use India tanks for a KGF set up in the AI dev thread. Trying to use them against Japan is inviting disaster. Better to use them against G. The only question is how long to hold India before bailing. A human will force you out faster than the AI. Basically 3 tanks is all you really count on, anything beyond that is dependent on Japan’s moves.
USA is slow on the uptake, in Classic and Revised you could set up for a round2 move out. In v5 if you move out in the second round, you usually just get stalled. Round 3 is much safer since this will let you get the Battleship ready to launch, and you can have your pacific Carrier by Brazil with UK ships to support it.
The main reason your friends are probably frustrated is the Russian set up. There just really aren’t many safe ways to destroyer German units on attack without totally screwing your second round position. And we’ve all seen how easy it can be to dud out with a Russian Yak or miss a critical deuce with a Russian artillery piece. I would urge caution in the first round and sending them immediate support from all quarters. You’ve got 4 UK fighters and 1 US fighter that can back up the Russian position, provided you protect Szech, and Stalin needs basically all of them to hold against the initital German advance.
I’m a big fan of the six unit buys with Russia
4 inf 2 tanks or 6 artillery. The former if you want to try to hit Ukraine, the latter if you don’t. The second round I generally do the same, and purchase strong attack/counter-attack capable units. All infantry is kind of a worthless buy in the first two rounds, it just gives Axis a green light to move forward and stack since they know you can’t threaten them with anything.
Here’s a good example, say you have 26 ipcs as Russia, which is fairly standard in early rounds. How would you spend it?
Would you buy 6 Infantry and 2 artillery?
or 2 infantry and 5 artillery?
Your instinct from classic will tell you to spam infantry, but the latter buy is actually much better in v5. Basically you are sacrificing a single hit point on defense fodder, to effectively double your counter attack potential.
After the first 2 rounds, UK should have provided some cover to make up the difference, and that’s when you start spamming infantry, which is now more effective since they’re activated by extra artillery you invested in early on. You can make the same argument with tanks. Sacrificing a hitpoint to get the extra reach out of Moscow on counter attack. in either case, at least Stalin will be able to fight if you invest in more artillery or armor. The Alternative is just defensive Soviet infantry turtle, and since UK USA take so much longer to move out, Russia will usually just get slammed before the Western Allies can prop them up, and then Russia won’t have enough offensive units to trade territory anymore.
Agreed on the artillery buy; it not only gives you more offensive hit points than the equivalent infantry buy, it amps up the existing infantry so that Germany has to be on it’s toes. I would usually buy 4 inf 3 art however I think the 6 art buy is stronger in the first round.
I was looking at options for the first German round and was wondering if it makes more sense to take out the USA destroyer and 2 transports than use the subs vs the cruiser. Go ahead and expect a fighter loss vs the cruiser and buy an extra bomber on the first round instead of 4 inf/3 art?
Why am I thinking about Axis moves?
The AI usually leaves the India carrier alone once you move it out however it is within range if the fighters want to hit it and the carrier and battleship decide to move a couple of spaces towards Africa.
Yeah I think it is more disruptive to kill those US transports than the cruiser. Most will double up to make certain of it. This sometimes has the added bonus of annoying the UK player into bringer their bomber back to Canada in an attempt to clear sz 11 of subs.
For the UK carrier to safely swing around Africa it can sometimes help to leave the Egypt fighter behind to protect it from the 2 Japanese fighters in range. It’s not full proof, but Japan has a lot of uses for its fighters and might not want to take the risk. Especially if you are able to fly the London fightets and bomber around in range to cover sz 34 on counter attack.
I posted some games in the AI dev thread showing some pearl openings for Japan. Its fairly simple for the Japanese to sink the US carrier deck if they want to. The way to run that attack is with 2 fighters 1 bomber the sub and the cruiser. Basically you leave the Caroline carrier where it is, since you don’t need to actually move it until non com. As long as you take the first available Japanese air casualty on the fighter that has only 1 movement left, you can usually sink the USA carrier, without ever risking your own.
Then land the surviving bomber and fighter on wake. And move your carrier to sz 60 to threaten counter attack if USA takes the bait, tries to hit wake and throws away their battleship That’s a move I’ve seen Axis do once or twice, and it can definitely give the Allies some headaches. So as USA you can’t necessarily plan to have the transports in sz 11 or the pacific fleet in sz53, as both can be destroyed if the Axis player is determined.
And where, if I may ask, is the AI development thread?
cool… thanks! I have a lot to learn… just call me grasshoppah…
Anytime man, I’ve been playing a lot of v5 lately vs HardAI using the latest jar and thinking about things more in general terms for the endgame. I posted some more saves you might want to check out for ideas.
From an Allied strategy standpoint, there are basically 2 relatively safe approaches, which work well against the AI and also against humans. I say “relatively” because this is clearly a dice game, and a human may do things which the AI currently does not (like engage in heavy factory buys, or use SBR and carriers, or can-open and set up for multi round attacks.) There are many points at which the game can just come undone for Allies which is why many bid for them. Against the AI OOB I think its very possible to win, provided you are willing to play it long enough. For more experienced players, it can be entertaining to give the Axis a bid as well and to forego SBR (since the computer doesn’t use it.)
But back to strategy, the two main approaches I see are…
1. Tank wall
2. Air wall
Both approaches use units of the other type too, its more a matter of which to magnify first (buy more of early on, either tanks or fighters.)
The first (Tank Wall) seems to work especially well under KGF conditions. The main objective is to hold India for as long as possible and produce as many tanks as you can there. The real use of the tanks is not so much to do with India defense as it is the endgame launch to Karelia, so Allies can hold the north. Usually in KGF India is eventually evacuated and ideally traded for a few rounds thereafter, so that the India tanks can rush west towards Berlin. The tank wall is particularly effective against Germany in that it allows you to shoot north then stack Baltic states to get the line on Berlin. Until then, they can be positioned in Caucasus to threaten a fairly wide area especially Karelia/India/Szech/Transjordan. Ultimately to make good on the money you spent to build them, these tanks need find their way into an attack on Berlin.
The second (Air Wall) seem to work especially well under KJF conditions. Here the main objective is again to hold India for as long as possible, by building ground there and then transiting fighters over from UK via W. Russia/Arch or if necessary Gibraltar to Caucasus. More than India defense, the primary use of these fighters is to hold Moscow and the Center of the gamemap, and especially to keep the IJN from comfortably coasting into Africa and the Med sz 34. Basically you’ll be wheeling the British fighters back and forth between India and Moscow, trying to prop them both up while America dedicates the full weight of their economy to steeling the money islands away from Japan. This air wall move works fairly well for Japanese containment and for giving the USA the income it needs to be truly effective. Its easier to take away the money island and defend them from recapture than it is to take Tokyo though. To actually knock Japan off the board you either have to invest in factories for the money islands and push up French Indo China, or invest heavily in transports and troops out of W. USA for a dead drop on the home island.
For either approach, someone has to buy fighters to keep Moscow afloat and someone is going to have to buy bombers for transport sweeping, whether that’s USA or UK kind of depends on which sort of wall you go for early, and how early you want to try and set up drops out of UK. Also, just like in Classic, its fairly critical for Allies to retain control of Africa in the long term. You can trade it for a while, but a minimal investment from USA/UK to retain control of this region early on is pretty key to keeping UK viable and preventing Axis from collecting too much too soon.
The center wedge, keeping Moscow/Caucasus/India viable is usually the optimal approach, there is a certain point generally around 50 units at Moscow, when it starts to become increasingly difficult for Axis to take it without a 1-2 punch, which is the whole idea behind the wedge. Just keeping Axis nations (whether G or J) at least one space off the capital. If you can trade the area around Moscow, and then trade Moscow itself for one of the Axis capitals at an attack advantage, that’s basically what you’re shooting for. The longer you can hold Russia the better chances you’ll have to set that up.
KJF/KGF are also conceived more as general investment directions than strict pursuits. Unlike some earlier boards, it a bit harder in v5 to go 100 percent in one direction. Fequently you have to bounce your air back and forth, or make at least a small investment in the minor theater of operations (whether that’s Atlantic of Pacific) to prevent the Axis from going monster. But you want to develop a clear thrust one way or the other by the fifth round, otherwise it just takes forever to knock out a capital.