Posts made by Magister

I had been a fan of efficient US effort towards Europe, as detailed in Caspian_Sub Policy Paper #2
US1 (42 IPC): build in EUS 3tra, 6inf (or even 3tra, 2inf, 3arty)
US2 (40 IPC): build in EUS 1tra, 4inf, 4tnk
US3 (38 IPC): build in EUS 1tra, 4inf, 2arty, 2tnk
Planes fly to Europe as well.
And even more, getting to 5+5 transports to Europe (35 IPC = 5 inf + 5 arty), once even 6+6 (adding some IPC in Europe). Building US planes instead of inf+arty and a little tanks is less effective.

But I found all this the most vulnerable to Japanese “coastal guerilla” as described above. Alaska may be only a staging point, but if US gets negligent then Japan may continue building a serious base (ftr’s landed in Alaska, further landing a ‘screen’ in WCA). All without committing really strong forces.

The real point is how US may prevent it with minimal sacrifice (IPC investment or delay to Europe).
Then yes, I found keeping one extra turn production’s investment (or 3/8 if using tanks) in building in WEU instead of EEU does protect nicely. Inf WEU > WCA > ECA, tanks WEU > ECA. Japan cannot win WCA, ALA is empty but swapping it may be pointless for Japan and so deterred.
Else keeping enough US defenders in both WCA and ALA may make the same Japanese landing threat “multiplied”. Keeping 6-7 inf in each ties up too much from the KGF effort.

What if as US I prefer to KGF with more inf+arty rather than less inf+tanks ? is the weakening of deeper counterattack potential vs Jap landings reason enough not to do that ?

More important - is there a cheaper US sufficient deterrence towards Japan ?

Sorry I hadn’t followed all this topic [Am I the only where these forum pages load way too slowly ?].

But I have another situation that may be considered a “Nit picky German Economizer”.
Germany has a mass in EEU, Russia in Caucasus. Between them it’s no man’s land.
Germany has only 4 fighters (one lost R1 in UKR, one lost G1 sinking the Med British BB) and no bomber (previously lost to AA in strat bombing) to swap 3 territories.
Suppose Russia has 2 fighters.

A) A neighboring area is empty but enemy-controlled. Is it better to blitz with a tank (and back!) [enemy can do the same] or advance with 1 inf there, or more ? Does it make a difference if it’s UKR for 3 IPC or BEL/KAR for 2 ? That difference is 1 IPC, or less (due to forced enemy reaction that may regain the same IPC) ?

B) All 3 neighboring areas have 1 inf. Is it worth to attack two of them with only 1inf,1ftr ? That’s what I usually and systematically do, and I recognized as “nit picky economizer”. Or attack only two areas with 1inf,2ftr each ? is 2inf,1ftr worth the try ? (or even 1inf,1art,1ftr what I often use as Russia against 2inf) ?

On the other hand, suppose something quite horrific may happen if area remains traversable by enemy (Like an US tank horde in KAR adding to an attack on GER). How mandatory is to use more inf instead of fighters ?

For pure defense (standing somewhere and not being defeated) pure INF is still the best for the money.

INF: at $3, 1 “body to die” and 2 punch is the baseline.
ARTY: $4, same defensive value. A “waste” of $1.
TNK: $5 and 3 punch. Added in small numbers to inf (that is, when they don’t get to die, only the accompanying inf) only the added punch matters. So 1 TNK does the thing of 1.5 INF = $4.5. A “waste” of $0.5. May be well worth due to the flexibility.
If tanks are alone, their defensive advantage over an inf+arty mix is SQRT(3/2) ~1.22x from the Lanchester theory. So worth $3.66 in equivalent inf, so a “waste” of $1.34. In between (large number of tanks with some inf to die first) the “waste” is in between.
FIG: $10 and 4 punch, in small numbers worth 2 INF = $6. A “waste” of $4. Mostly not worth building for land defense only
(sometimes it does, when total IPC isn’t as limited as local factory capacity e.g. defending India).

So in the final defense of Germany, if I have 2 IPC more than a multiple of 3, I build a tank on top of the pile of inf. If I have 1 IPC more, I may choose to build 2 tanks (and forego 3 inf) OR to build only inf and save 1 IPC for next round where it may mean an extra inf, or a inf converted to tank.

For a defensive campaign, all types are good at their thing. e.g. the Novosibirsk deterrent position has inf+arty with tanks+fig back in Moscow. But if deterrence is truly put to the test (Japan enters Yakut or Sinkiang or Kazakh in some force) it’s the attack that matters.
Strategic defense, tactical attack. But here tactical attack, even threatened is helped more by the virtual multiplication - same force threatening several approaches.

Several units bid in same area are allowed ?

In a ‘Tournament of Champions’ file I’ve seen 3 inf in UKR are enough to deter Russian attack and give Germany a lasting edge.

In the Ladder (1 unit per area!) my recent favorite is 1 tank in Libya (or Algeria) and Japanese 1 arty in Indochina. Helps greatly with the usual J1 fighter crisis, to the point that Allied don’t dare so many threats at once, so the arty is a bit superfluous (I’ve attacked China with 5inf,1arty, 3 ftr) and taken empty India.

A nice paradox - Without the arty, it’s badly needed. With the unit it’s not needed any more…

The old “Lords of Midnight” (anyone remember the venerable Sinclair ZX Spectrum ?) used a square grid with diagonal movement allowed at 1.5x cost. This made the “circle” (locus of points at equal distance) an octagon with lower distance, area and shape errors than

• a hex grid (where the ‘circle’ is itself a big hexagon)
• square grid with diagonal moves prohibited (‘circle’ = tilted square)
• square grid with diagonal moves allowed at same cost as straight moves (‘circle’ =square. This is the most UGH !)

Also congratulations Jen, you’re the first lady I ever read to prefer hex map wargames !! Now, you mentioned Avalon Hill… ;-) maybe you would like Gamers OCS more ? ;-)

An example of great games on same topic (Pacific WW2): Victory in the Pacific (areas) and Fire in the Sky (big hexes). At heart they are surprisingly similar, but the second is really deeper and more accurate, allowing less aberrant “strategies” really unfeasible because of logistics. A&A has it even worse.

Now, I don’t know any fast-enough-playable game of whole WW2 outside A&A ;-) I don’t talk of World in Flames or other such monstrosities…

With normal dice (not Low Luck) it’s impractical to just “strafe” Ukraine because of the large variance - and you all are right - the risk of “too good dice” and leaving tanks stranded against counterattack. Better go to the end, kill the fighter, take the 3 IPC. German T1 counterattack may be limited and/or exposing their tanks too, due to lack of fighters (at most 1 after pressing Med and Egypt missions).

With the best forces available there:
RUS 3i,1a,3t,2f = 21 firepower, or 3.5 killed on average. May kill an arty, is it worth that ? If it kills the tank, then it’s “too close” to that “too good”.
GER 3i,1a,1t,1f = 15 firepower, or 2.5 killed on average

And if Russians send less, say one tank less, then it’s 18:15, too ‘fair’ attrition. Germany may like that ;-) Say it goes 3:3 in the first round, then what ? To stay one more round is “too good”, to withdraw is just equal attrition.

On the other hand, with Low Luck I send 3i,1a,2t,2f (one tank less, sent to WRU instead) with good chances to gain it with 2arm left.

Clear Baltic of German fleet, then send UK+US troops to Norway, when possible advancing them to Karelia, Eastern Europe then Germany itself ! It can be slow but it’s certain. If possible, take Norway with US (or put a British “shield” before it in Karelia) to build a factory and add to the ship logistics.
Raid West Europe if weak enough, and don’t skip any reasonable opportunity for direct landings (UK sacrifice+US) on Germany if not strong enough. Even if wiped by counterattacks, the income shock (German cannot build next turn, US builds more) can be hugely helpful !
All this, while helping Russia survive against Japan. Just make your certain take of Berlin earlier than his/her certain take of Moscow ;-) Care for economics helps this in the longer run.

There are still situations with enough luck left in LowLuck. Especially naval battles, where 1-2 submarines roll dice separately from the rest of ships/air. Your 2 subs missing (2/6 chance) and opponent 1 sub hitting (2/6) also with victims removed before normal fire, can cascade seriously into a disaster in further rounds of firing, not far from ‘regular’ dice.
Obviously, battles between lone units (DD vs tra in Kwantung, sub vs tra) are exactly the same risk.

4 German fighters guarantee killing the Med BB (but not 1 sub, 3 fighters !).

With Low Luck, the 3-area Russian attack (UKR-WRU-BEL) is also assured of working (but with not a few casualties).

Leaving medium-sized forces (say 5 units) is a blunder now, as enemy can bring a raid force to take out certainly 4 units (e.g. 3inf,3art,4tnk) then retreat. With Regular Luck, that defender was protected by the cloud of uncertainty, and the raider could not bring too much offensive power for the risk of doing “too well” (!) killing all defenders and being stuck with expensive units in a dead zone.

Small deadzone swapping is more effective. Either 1inf+1ftr or 1inf+1tnk hit at one 4/6 ~ 67% instead of separate 1/6 and 3/6 (overall chance of at least one hit 21/36 ~ 58%). 1inf+1art is still 4/6 while with Regular dice it would have been 20/36 ~ 56%).

Grouped transport ships (3-4) are a better deterrent against lone fighters or bombers. The average is the same, but with Low Luck it’s concentrated into getting 1 hit, while with regular dice some of the hits were “wasted” by overkills of 2, 3… hits.

Other significant differences in strategies, openings etc. ?

Remember when… Germany fell to Russian troops in a supreme 1-2-3 Allied effort. Just after that, Japan took Moscow, capturing the income of German territories too. As Allies had insufficient troops to retake Moscow and HOLD it, Moscow became a sort of Japanese no man’s land, because if Japan retook Moscow they would retake all German income again…

I agree with Switch about the “art” in gaming. For me math crunching is mainly the preparation to reach some quantitative then qualitative conclusions. After they were well ‘chewed’ and assimilated, one may just forget the numbers and go with the real art !

“Culture is what remains after people forgot all they learned”

Jen’s way of looking at losses is better expressed as:
-6 IPC (or whatever the cost of sent troops) is certain
+2 IPC income * probability to gain area (say 90%). If enemy doesn’t counterattack and win this becomes a permanent flow !

• enemy force losses as expected value (At most their value, often less because of partial defeats and attacker retreats.E.g. 1inf 1ftr takes 1 loss then ftr retreats)
• own land forces retreated expected value (E.g. You started with 2inf 1ftr vs 2inf and took a loss without effect. Would you continue with 1inf 1ftr vs 2inf ? If not, the 1inf was NOT lost)
• this cost from the opponent’s side should they want to counterattack. This is also an expected value
  (probability to survive 1 inf * cost to clean 1 inf + prob to survive 2 inf * cost to clean 2 inf + …)

Another reason to gain an area (besides preventing deeper tank action): to prevent enemy use as a fighter base that may allow them to advance the main force there. Maybe without fighters they aren’t safe enough, with fighters they are.

Interesting experiments !
Now my 10% for tanks come from their value in a SINGLE battle - that is, they do not need a matching inf to fire at 3, whereas an ‘excess’ arty fires at 2.
Of course the speed-2 of tanks helps. Not within each ONE battle but by giving more CHOICE among battle positions.
Even speed-1 inf+arty can have choice (e.g. a Novosibirsk position projects dead zones to Yakut, Sinkiang and Kazakh).

Speed 2 of tanks in traveling from home factories allows doing what I call the ‘Cumulative Charge’ effect. For some turns, they produce slow troops, then switch to tanks only. 2-3 turns away, they arrive simultaneously on the front, overwhelming enemy or making them retreat. But after the advance a similar-sized ‘hollow’ ensues.
Usually Germany does it, but Japan can too - factories in Asia don’t HAVE to build tanks only. They can switch according to similar tactical needs.

I’d wish to discuss, prepare numbers and arguments for a more definitive article on swapping contested Dead Zones.

I’m assuming two large main forces that do not enter direct contact, each fearing counterattack from the enemy main mass. But the area(s) between them are attacked each turn. At stake there’s the income of that area (may be 3 IPC like Ukraine or 2 - many others on the Russian front or even 1 - Evenki, Persia) and mutual attrition trying to improve force ratio for a future main battle or advancing/retreating the main position relative to factories (Doppler effect: for a turn retreater gets double production flow, advancer nothing).

Measures which I would compare would be:

• Own forces committed (land)
• Average IPC losses for each side
• Probability for land survivors to gain area with income (May mean more, such as preventing tanks blitzing through)
• Probabilities of various levels of survivors - that may be similarly counterattacked on enemy turn.

A stopping threshold is necessary. For instance, 1ftr alone is not economical to continue against 1inf. But 1tnk usually yes.

I think of variants like:

Attack 1 inf with:
1inf 1ftr
1inf 1bmb -> I like that !
1inf 2ftr
1inf 1arty
1inf 1tank
2inf 1ftr
2inf 2ftr
2inf 1bmb
2inf 1arty
… etc

Attack 2inf with:
2inf 1ftr
2inf 2ftr
2inf 1bmb
2inf 1arty
2inf 1tnk

I’ve also read proposals like massing 3-4 inf in each area (especially by the side with more air: Germany, Japan) to accelerate side-attrition against Russia.
At what size will these forces be not ‘small’ enough and be subjects to the risks of one ‘large’ battle ?
I suppose the answer lies in the variability. A small force (3 units) cannot be strafed by a medium one (say 6) with economic advantage, and without enough risk to wipe it in the first round, so exposing more forces to counter-counterattack. A larger force (say 6) can be more safely strafed.

Low Luck makes impossible this last strategy, since a ‘strafe’ can be calculated to take out an exact number of enemies.
E.g. enemy has 4 inf -> attack with 3inf 1arty 4tnk. Punch 18. Eliminates always 18/6 = 3 units, no little and no less.

To clarify:
Yakut as starting position: it’s mainy a threat to strafe, not a full counterattack advancing to the coast (But may risk to become so if it goes ‘too well’)
1R: 10 inf 2 ftr threatens max Japanese force in Buryat which can be only 3inf 1tnk. (Assuming Kwangtung transport sunk 1UK) So they don’t try.
2R: 10 inf 2 ftr threatens max Japanese landing which can be 6inf 1arty 1tnk. So they do it one area back, in Kwangtung.
3R: cannot attack anymore, if Japan concentrate in Buryat 2 turns worth of transports. But they can still hold. 1-2 tanks may arrive to improve defense.
4R Russians have to retreat.

Novosibirsk as medium-term holding position. The unapproachable dead zones would be Yakut, Sinkiang and Kazakh. Tank reserve can be in Moscow, threatening Ukraine and Persia too.

On longer term building race: agree that 8 units/turn is only a start, and at one time Japan would NEED a factory or two.

Factory in Novosibirsk: I think it’s necessary once Japan is solid in place. Cheaper inf/arty can be built and defend right now, and attack next turn. Tanks in factories (Kwantung, FIC etc) need 3 turns from income to battle, and are relatively more expensive. Forces from Japan need 5 turns.
And for a time, the two flows cumulate: new forces built in Novo, old flow still coming for 4 more turns before being reduced (not a lot, because of the total income).

Factory in Caucasus: it’s very nice, the problem is that at this stage both sides have a stalemate of 80-100 units. If this is all, Russians still win with 6-8 units/turn built compared to Japan’s 4. What I was talking about was the smaller new force gathering in Yakut etc. from further landings. It may get to 20-30 units or so, but surely it would be suicide to put it in Novo next to the 100 unit Russian mass strafing every turn. And if waiting until second Japan army is strong enough, at that time the Russian mass may have overtaken the Caucasus besiegers. To unify back in Kazakh at the last moment would abandon the factory, and the east force needs to come from Sinkiang (since Novo is a dead zone). Right ?

So back to a simple frontal siege from Novosibirsk.

An interesting case can be made for swapping or keeping flank areas (Evenki, Kazakh, Persia etc) with more forces. Like 3-4 inf each. If Russia wants to swap them, they can do it only one at a time, and send some forces (some of them offensive, like arty) each turn. At this stage, accelerated attrition works for Japan, and they have more fighters. Or Russia may ignore Japanese presences, and the income difference helps Japan again.

[DarthMaximus said]
I think IPM is still valid.
Simply put if you are buying 10 inf (30 ipc) per turn it is going to take your opponent at least 40 ipc to defeat you.

Right, it’s mostly about how Defender’s advantage changed from Classic to Revised. The addition of arty decreased it.
I define here Defender’s advantage as the ratio of IPC costs for attacker to be ‘equivalent’ to defender (both grind to zero with average luck). Any better for one side gets a cascading victory.

Pure infantry is still the best at pure defender.
Inf vs inf: the advantage is ~1.41 (SQRT 2) from Lanchester’s theory. Would need slightly more than 14 inf to defeat 10 inf. And this is mutual - if the other side wants to attack the 14 inf, they would need 20 inf. A mutual advantage of 2 x.

A simple mix to defeat pure inf would be Inf+Arty in equal proportions.
5 inf+5 arty (35 IPC) are exactly equal to 10 inf defending (30 IPC). Each unit hits at 2.
The defender’s advantage would be 35/30 = 1.166 x.

Actually, the optimal mix for large attacking forces would be in the proportion of 60% inf, 30% arty, 10% tanks (calculated with a detailed, mathematically quite ugly model). 6+3+1 would cost the same 35IPC.
I’ve tested on simulators, for large forces (30+15+5) this is very slightly better than 25 inf 25 arty. But 6 inf 3 arty 1 tnk is very slightly worse than 5inf 5 arty to attack 10 inf (or 5 inf 5 arty). In general, most optimization problems have a ‘flat optimum zone’ around which small variations in decision result in extremely small variations in value.

If opposed by a similar force, the 50-50-0 force would have no defender’s advantage. The two forces each would cost 35 IPC and fight the same. On defence, the 6+3+1 would be a bit superior due to the 3 firepower of tank. (Roughly like 0.5 inf more, but without its staying power).

What do you feel it’s the defensive value of an AA gun ?
for instance, when to bring one extra tank instead on those scarce transports ?

Obviously, if enemy has 6 planes to attack, on average one would be lost in each attack, which is nearly as valuable as an extra tank, and better economically (they lose 10 instead of 5 IPC). AA increases variability (they can lose 0 or 2 planes; in good cases they can press the attack, in bad ones they may stop).
But if enemy has 3 or 4 planes ? it would depend on how many times the enemy would try to attack the area, but in the short term a tank would be better…

Some observations on the Japanese front:

• Russia can mass 10 inf in Yakut. Combined with 2 ftr threatening from Moscow, it makes impractical to even land in Buryat until J3 (when uniting with forces landed in Manchuria on J2).

• A Russian inf+ARTY mass in Novosibirsk is very effective at gaining more time, or punishing not-too-strong approaches (either from east or south).

• Is the southern approach (Indochina, India, Persia, Caucasus) viable ? I found it very hard to do together with a northern approach towards Novosibirsk. The same Russian tanks and fighters can threaten both, effectively ‘doubling’ their presence. Maybe press only the southern route ?

A steady moderate flow is possible to both routes: 4 transports, 2 are off FIC and 2 of Japan. Each turn, the FIC ones move to Japan, load and unload 4 units to Buryat; the Japan transports load 4 units and unload 4 units to FIC.

Later, after much struggle… Japan succeeded to bring the mass in Novosibirsk and have it survive. The flow of reinforcements from Japan >tra >Buryat >Yakut is very ‘inertial’ and ‘slow’ (and other routes are slower). UK and/or US sent some help to Moscow, and can send more. Now what to do to crack Moscow before Allies crack Berlin ?

• Build a factory in Novosibirsk to add 2 ARTY /turn ? The foregone income means less inf… in 4 turns !
• Swing the Japanese mass west to Kazakh then Caucasus for more local building ? Then the ‘traditional’ flow of reinforcements may get too weak to survive in Novo’ against the ever-increasing Russian mass. Essentially Japan needs then 2 defensive armies, and if they are that strong they would win anyway ;-). Or the new Novo army to be used as ‘bait’ - if Russians attack it, they weaken themselves against the Caucasus army ?

Arty is a good offensive boost where one does not need the speed of tanks, and there is already plenty of infantry.
For the price of ‘upgrading’ one inf to a tank, one can upgrade 2 inf to 2 arty.

say 10 inf 6 arty are much better than 13 inf 3 tnk.

Practically, a great use is the threatening position in Novosibirsk: inf and many arty, with tanks and fighters back in Moscow. May make it difficult for Japan even to come close. Much better than just a defensive position waiting to die.

But for a final defense (say Germany) 1inf 1tnk are better than 2arty. The extra cost is wasted.

Good… then a working definition of Skill would be “that thing that would need less Luck” ??

I think we should look at the much-more-analyzed Backgammon. A very similar Skill vs. Luck analysis is at:
http://www.bkgm.com/articles/Zare/HedgingTowardSkill.html
Several software (GNU, Snowie, Jellyfish etc) using genetically-trained Neural Networks succeeded to estimate ‘values’ of positions in terms of ‘equity’ or probability to win to resolutions of 0.001 (not sure if they are really that ‘right’ or non-biased, absolutely or comparing variants). Then analyzing ordinary human play, they found errors cumulating to winning and losing each game several times over ;-) so it becomes essentially a contest of making less errors !

Essentially the question in A&A and any other games is: how “easy” is for real humans to do mistakes with effects comparable to the normal fluctuations of luck ?

[Backgammon] is a game of skill and luck.
If one wins it’s skill, if one loses it’s luck.