The Norwegian Gambit

Nice article (and I even get quoted :-D)

I agree with nearly all of it, but i’d like to emphasize the following points:

1. It is crucial that the Russian stack on WR survives any G1 attack. Which means that R needs to get everything it can, including all the inf on Archangel. It is nice to have an spare inf to move to Karelia to stop the G armor on EE from blitzing to Norway but if R loses too many units on that attack then you should reinforce WR at all costs. You can always keep Caucasus fortified with 6 units (the 2 from Kaz plus new builds) - if the Germans decide to take a swing at it they will have to use their expensive armor to take it and then you can simply retake it on R2, along with Ukraine.
2. The rest of the strategy really depends on the Allied player’s choice, either KGF or with a US Pacific. One thing to point out though is that most likely the US will need a 2nd defense fleet, either on SZ2/8 to protect the transports shuck-shuck from E. Can to the UK from Axis bombers.

Interesting write-up.

I will give this a shot in low luck, and see how it goes.  It’s a different sort of game for me because I’m used to the Black Sea sub gambit, but if I went with the Norway too I would need to buy another fig to do it, and that’s too risky for consideration.

I’m reluctant to attempt this in dice because I’d be scared of bad luck at West Russia and then getting destroyed G1.  But against a strong enough opponent, I might conclude the risk is worth it.

It’s very difficult to evaluate this strategy in the context of dice games.  If the gambit works, it’s a totally different game then if the gambit doesn’t work.  Hence the need for low luck games to test it out.

Looking forward to playing this and having it played against me and seeing how it goes.

@Zhukov:  Low luck is not like dice in much the same way that a hamster is not like a large moon rock.  :lol:  Also, Granada specifies not buying a R1 fighter.

The validity of a strategy in low luck is completely irrelevant to its validity in dice, and vice versa.  If you propose a Russian triple in dice, your preferred excitement and/or sanity level probably includes things like bungee jumping naked from a suspension bridge to meet your in-laws for the first time.  If you propose a Russian triple in low luck, your penchant for excitement probably extends to Meatloaf Thursdays at the retirement home.  Which is not to say that Meatloaf Thursdays are not exciting; I am personally quite fond of meatloaf.

Saying the strategy “works” if it works in low luck means nothing for its validity in dice.

Another way of looking at it is this - let’s say that we have a six sided die, and decide to ignore 1 and 2 as being “low” and 5 and 6 as being “high”.  Basing our predictive model on this, we find that when we add the sums of these “low luck dice” when, say, 1000 dice are thrown, is approximately the same as the sums of regular dice.  That said, we smile and say that we are now Scientists with an Accurate Idea Of How Things Work.  Low Luck involves less of those evil, evil numbers, but gives us about the same results - right?

However, 66% of the results are being ignored in any throw of a REAL dice when compared to one of these “Low Luck” dice.

Besides that, Axis and Allies games in dice tend to go down extremely different paths when you get the equivalent of a 1,2,5, or 6.  From that point on, you start going down a whole new road that you never even see in the Low Luck scheme of things.

Suppose you have a door on your left, a door on your right, and a door in front of you.  Behind each of these doors are another three doors; beyond each of those doors are yet another three doors, and so on.  Low Luck says you’ll always go through the door in the center.  Actual dice say you pick the door randomly every time.  As you can imagine, the sort of view you see when going through a few doors in low luck and going through a few doors in dice tend to end up very differently.

I had been thinking of writing a Russian strategy article for a long time and after this post I think I now know what it should focus on, the several openings possible and some clues for mid and late game. Wait for news. :)

Nice read, Granada.

@football2006:

Nice read, Granada.

Yeah, Gran, I think they’re buying it!  :wink:

Juuust kidding.  It’s a long-standing joke between Gran and me.

@Bunnies:

The validity of a strategy in low luck is completely irrelevant to its validity in dice, and vice versa.

Saying the strategy “works” if it works in low luck means nothing for its validity in dice.

There’s some merit to what you are saying but these statements aren’t accurate.  Anybody who is skilled at low luck and dice can tell you low luck is a good testing ground for any strategy–to determine its overall soundness, in the case of average rolls.  IMHO, in order for a strategy to be ‘good’ it should meet the criteria of success given average results.  Isn’t that also the criteria for the ‘validity’ of a strategy?  Low luck shows us these average results.  So yes, low luck experimentation is extremely useful imho.

Granted, there are some strategies that work better in dice and vice versa.  KJF in Revised is a good example–it thrives on exploiting the sort of below 25% outcomes you never get with low luck.

It goes without saying there are additional complexities to a dice game that you don’t find in a low luck game.  I think becoming a good low luck player gives me additional perspective on these complexities and potentially makes me a better dice player.

Anyway, I only brought up the dice/low luck dichotomy to make a point about Granada’s case.

I don’t believe the Norway Gambit is a good move in dice, unless you have reason to gamble.  If elite players are capable of winning consistently with Ukr/Wr, why risk a Norway Gambit unless I’m up against one of those elite players?

But Norway could be promising with low luck dice and I’d like to test it further.  I just haven’t seen much of this because it’s been played against me primarily in dice games–where the outcome of the games is heavily dependent on those R1 dice.

@Zhukov44:

Anybody who is skilled at low luck and dice can tell you low luck is a good testing ground for any strategy–to determine its overall soundness, in the case of average rolls.

. . . in other words, playing Low Luck games is a good predictor for Low Luck games.

That bit about being skilled at low luck and dice is just tooting your own horn.  In my opinion, if you really understood the difference between low luck and dice, you wouldn’t be claiming that testing a strategy in low luck has any say towards its validity in dice.  Yeah, that IS inflammatory, but when you make such a strong claim, I leap in with equally strong words for the counter!  (yay internetz!  do not try this in real life, kids . . . you’ll put an eye out . . . real life disagreements are usually best avoided . . .)

Playing dice means doing risk management - in particular figuring out what battles are high risk high reward, and figuring the possibilities of failing at make or break battles.  This simply does not come into play in low luck games.  All you do in low luck is count up hits and casualties; you know the result of a battle with extremely high probability when compared to dice, you need only commit relatively minimal forces when compared to dice, and you absolutely never have to figure on contingency plans in case of poor round 1 dice in low luck - at least HARDLY so when compared to dice.

On the other hand, the ease of predicting outcomes in low luck means a player needs to do all the hit and casualty counts for at least one entire round in advance, because the opponent will easily be able to exploit any openings.

How about a bit of substantiation for what I’m saying?

Let’s say you have a Low Luck battle of 32 tanks against 30 tanks for a capital.  Under Low Luck rules, the attacker ALWAYS wins.  Under dice, though, the attacker loses about 25%.    After the first round in dice, the odds are extremely high that neither attacker nor defender got the exact “average” number of hits, and that drastically changes the numbers going into the second round, and so on and so forth.  So in dice, the attacker needs to look at board position.  If the attacker is probably going to lose, the attack should be made, for a 75% chance of a high IPC swing.  If the attacker is winning anyways, the attack probably should not be made, so the 75% chance of a high IPC swing can be increased to 80% or more.  If the attacker is winning in some places and losing in others, the attacker needs to size up the situation.  But in any case, in dice the attacker needs to plan for what will happen if the first round of combat doesn’t go well, which is something the attacker NEVER needs to do in Low Luck.

Try carrying out a few dummy battles in TripleA of 32 tanks vs 30 tanks using dice.  You will very quickly see exactly what I mean, especially when you start throwing a few odd assorted destroyers/carriers/transports/infantry/fighters/artillery on the board to complicate the situation.  Under Low Luck, it’s a no-brainer.  Under dice, you have to consider contingency plans.

How about another example?

In Low Luck, 2 infantry 2 bombers attacking 6 infantry is not a monumentally stupid idea, because under low luck, the defenders always get exactly 2 hits.  But in dice, there’s a pretty good chance that attack loses a bomber.  Repeatedly doing 2 infantry 2 bombers vs 6 infantry in a dice game will lose the attacker games.

. . . and another example

Under dice, a Russian triple attack will usually fail at one or more points, leaving the attacker open to a nasty German counterattack.  But under low luck, none of the attacks in a Russian triple need “fail”, because the results are so rigidly controlled.

More specifically, under dice, you can do three Russian attacks on R1 with around 62%, 62% and 82% (something like that) chances of success.  If you succeed at all of them, you’ll do great, but a little dice deviation at any, and you fail.  It’s not a question of failing all the Russian attacks, just one leaves you open to a counter, and with 38%, 38%, and 16% chances of failure, there’s good odds the Russians will fail at at least one of those points.

But under Low Luck, the deviations are slashed, so it’s just a question of degrees of success.

–

For a real world example - let’s say you have a betting game in which you have 10 coins and flip them all.  The single best chance is that there will be 5 heads and 5 tails.  Let’s say you can choose between two bets; 5 heads and 5 tails which pays out at 2 to 1, and any other result which also pays out at 2 to 1.

Run this game under “low luck”, and you get real rich real fast if you always bet on 5 heads 5 tails.  If you bet on “anything else” under “low luck”, you’ll lose all your money instantly.

Run this game under “dice”, and you’ll lose all your money in time if you always bet on 5 heads 5 tails.  If you bet on “anything else”, though, you make money hand over fist.

Considering the best strategies for this game are literally opposite between “dice” and “low luck”, how can anyone say that low luck strategies are good for dice and vice versa?

I give the PEOPLE’S ELBOW to that!

http://www.youtube.com/watch?v=L_3Zi6t7W4s

Re: the old shingle about dice vs. low luck:

I don’t claim greater skill is necessary to play one or the other.  I just say different skill sets are required.

But I will say that you don’t hang fuzzy “low luck” on your rearview mirror.

Bunnies I already conceded that there are many different factors and levels to a dice game that aren’t present in a low luck game.

My point was that low luck is an excellent testing ground for the overall effectiveness of a grand strategy against average dice.  Nothing you said in your post indicates I am wrong about this.  Your examples are about battle-to-battle tactics and are beside the point.  Deviations in dice are also beside the point–of course a good dice player needs to take dice deviations into account, and that a dice game requires a more engrossing analysis of multiple risk factors.  Yes, tactics in dice are considerably more complex. But that’s tactics–not strategy.  A good strategy is going to take the law of averages into account–and low luck games will give you a good idea what the law of averages is.

Anyway this is a pointless tangent and I don’t want to distract from Granada’s thread.

@Granada:

But I see Zhuk’s point. Why should I risk 20 % of an uphill battle (and I have just yesterday saved a game at R23 after a failed Norwegian with both Russian figs killed R1: he needed all 4 hits to do it and he got them)? If I am confident I will trash Axis anyway, no need to take the risk. So maybe I lack Zhuk’s confidence. Or you may just call me an adventurer, but I simply like to live with the risk. For me it is an integral part of the game.

The thing is, the chances of something going bad are greater than 20%.  That is, there is the probability of clearing Nor with 1 fig or less or not taking it (approx 20%) in addition to the probability of taking WR but losing 4 or more inf (approx 20%).  So taken together the probability of something going bad in this opening is more like 30%, and the consequences depend on the severity of the dicing.

Of course, if you are playing someone who you know would otherwise have your number, then a 70/30 risk could very well be worthwhile.

One more thing I wonder about the Nor gambit is what to do with the German sub.  I guess I usually take out the lone UK tran, but maybe it’s a better move to go for the USA fleet (50/50 odds).  I still kinda like going for the lone trn because so many players sacrifice the USA fleet USA1 anyway.

I tried the Norwegian Gambit last night with LL. I hadn’t tried it myself for a while (but I’ve played a few times against it) and decided to have a swing at it.

Russia bought 3 inf, 1 arm, 1 ftr. Took Norway with 1 inf, 1 arm remaining, landed 1 ftr on Karelia. Ended turn with 10 units in WR and 6 on Caucasus. G replied with a bomber and inf buy and conquered Caucasus, Norway and Karelia, plus Egypt for a total of 46 IPCs.

The thing with this strat is how much it turns the Eastern front into a meat grinder for units. If G had bought instead 5 armor + 5 inf (my usual counter for the Gambit) it could have made things harder for the Russians. Instead, what happened was:

• G permanently lost Norway on R2 to the UK
• G bought 1 plane each round, ending up with 7 fighters and 6 bombers
• Russia was a bit low on units at the beginning of R3 since it hadn’t placed any on the Caucasus and had to kill 3 German armor + art on R2 to retake it, plus attacking Kar, BR and Ukr
• But since the German army had also taken such a blow on the 2 first rounds the Allies were able to keep the Eastern front under control, with UK/US landings and prevent the Germans from advancing.
• Meanwhile the Russians turned most of its production to holding Japan, along with UK/US assistance, mainly armor landed on the Europe
• Game ended with the Axis conceding on round 12 - the German airforce never had a chance of sinking the main Allied fleet and the while the Japanese fleet ended up sinking the US transport fleet it was also obliterated on the US counterattack. Meanwhile the Allies were earning 1/3 to 1/2 more than the Axis, so the writing was on the wall.
• It was an unusual game on Asia: the Russians had managed to send a huge army into Sinkiang, forcing the main Japanese army to retreat from Yak into Buryatia to protect their Manchurian IC but losing temporarily Kwantung and FIC.

All of this to say that the sucess of the Gambit also depends on the German response: if G attacks Caucasus and follows with a large armour buy then they may have a shot at controlling Karelia or Ukraine by G3 since there won’t be much Russians left after the 2 initial turns.

@Zhukov44:

A good strategy is going to take the law of averages into account–and low luck games will give you a good idea what the law of averages is.

well i’m convinced.  for a strategy to be really considered viable i think it would have to work in low luck.  if it doesn’t work in LL, then you don’t have any reasonable expectation for that strategy to succeed.  i don’t expect the opposite would be true, however.  therefore a reasonable position is that LL could be a sort of litmus test, failure in low luck meaning don’t bother in dice.

My post above was based on 2 aa calcs

http://frood.net/aacalc/?mustland=0&abortratio=0&saveunits=0&strafeunits=0&aInf=9&aArt=2&aArm=3&aFig=&aBom=&aTra=&aSub=&aDes=&aCar=&aBat=&adBat=&dInf=3&dArt=1&dArm=1&dFig=&dBom=&dTra=&dSub=&dDes=&dCar=&dBat=&ddBat=&ool_att=Bat-Inf-Art-Arm-Tra-Sub-SSub-Fig-JFig-Des-Bom-HBom-Car-dBat&ool_def=Bat-Inf-Art-Arm-Tra-Sub-SSub-Bom-HBom-Des-Fig-JFig-Car-dBat&battle=Run&rounds=&reps=5000&luck=pure&ruleset=Revised&gameid=&password=&turnid=&territory=&round=1&pbem=

WR
  5.84% 14: 9 Inf, 2 Art, 3 Arm. no units. : 0 IPCs
  21.34% 13: 8 Inf, 2 Art, 3 Arm. 1 Inf. : 3 IPCs
  29.14% 12: 7 Inf, 2 Art, 3 Arm. 2 Inf. : 6 IPCs
  23.98% 11: 6 Inf, 2 Art, 3 Arm. 3 Inf. : 9 IPCs
  12.78% 10: 5 Inf, 2 Art, 3 Arm. 4 Inf. : 12 IPCs
  4.96% 9: 4 Inf, 2 Art, 3 Arm. 5 Inf. : 15 IPCs
  1.38% 8: 3 Inf, 2 Art, 3 Arm. 6 Inf. : 18 IPCs
  0.46% 7: 2 Inf, 2 Art, 3 Arm. 7 Inf. : 21 IPCs
  0.08% 6: 1 Inf, 2 Art, 3 Arm. 8 Inf. : 24 IPCs
  0.02% 5: 2 Art, 3 Arm. 9 Inf. : 27 IPCs
  0.02% 4: 1 Art, 3 Arm. 9 Inf, 1 Art. : 31 IPCs

http://frood.net/aacalc/?mustland=0&abortratio=0&saveunits=0&strafeunits=0&aInf=3&aArt=&aArm=1&aFig=2&aBom=&aTra=&aSub=&aDes=&aCar=&aBat=&adBat=&dInf=3&dArt=&dArm=&dFig=1&dBom=&dTra=&dSub=&dDes=&dCar=&dBat=&ddBat=&ool_att=Bat-Inf-Art-Arm-Tra-Sub-SSub-Fig-JFig-Des-Bom-HBom-Car-dBat&ool_def=Bat-Inf-Art-Arm-Tra-Sub-SSub-Bom-HBom-Des-Fig-JFig-Car-dBat&battle=Run&rounds=&reps=5000&luck=pure&ruleset=Revised&gameid=&password=&turnid=&territory=&round=1&pbem=

Nor
  1.98% 6: 3 Inf, 1 Arm, 2 Fig. no units. : 0 IPCs
  12.48% 5: 2 Inf, 1 Arm, 2 Fig. 1 Inf. : 3 IPCs
  23.46% 4: 1 Inf, 1 Arm, 2 Fig. 2 Inf. : 6 IPCs
  25.1% 3: 1 Arm, 2 Fig. 3 Inf. : 9 IPCs
  17.78% 2: 2 Fig. 3 Inf, 1 Arm. : 14 IPCs
  8.12% 1: 1 Fig. 3 Inf, 1 Arm, 1 Fig. : 24 IPCs
  11.08% 0: no units. 3 Inf, 1 Arm, 2 Fig. : 34 IPCs

So yeah my numbers for Nor were 8% (1 fig remains) plus 11% (defeat)=19% for Nor, plus 13% (4 inf die) plus 5% (5 inf die) plus 2% (6+ ind die) for WR.

I’m not sure how to do the numbers to determine the probability that one of the battles will go bad (eg defined as taking Nor with 1 unit or less or losing 4 or more at WR), but I believe its at least 25%, possibly as high as 30%.

It may be that we have different criteria for the success of Nor also.  IMO if Russia loses all its units or all but 1 unit at Nor the attack was not a success and Russia would have been better off with WR/Ukr.

Also, if Russia has 10 units in WR, then Germany has a 76% chance of clearing it with 2 figs, and 89% with 3 figs, 94% with 4 figs.  So I think 10 is borderline–its definitely not safe with 9 units.

Good post by Zhukov about the probability breakdown.

@ Granada:  I already agree on R1 NOT purchasing a fighter - in a close game, it’s best to concentrate on ground units.  But why 3 inf 3 tanks on R1?  Why not, say, 5 inf, 1 art, 1 tank?  For battles with 1-3 ground units on each side, artillery are like cheap tanks - for example, inf/art vs 1 inf is about 5% worse overall than inf/tank, but is also 1 IPC less expensive, which adds up pretty quickly.

Is it specifically because you’re trying to avoid a G1 capture and hold of Karelia?

Another thing me and Zhukov were discussing the other night is that the survival of the UK BB also means that G will survive with other units:

• The sub on SZ8 can be sent to SZ1 to sink the transport (-7 IPC for UK) while the sub survives (+6 for G)
• The fighter on Ukraine survives (0 IPC gained since G would lose a fig anyways) and can be used with the bomber to attack Egypt, allowing in theory for at least 1 more German ground to survive the assault
• The 2nd Russian fighter dies (-10 IPC for R)

So, according to average results the situation is essentially the same regarding gains/losses. Tactically, if the Germans retake Norway the UK may be able to take it, unless Germany buys 1 bomber and manage to land 4 fighters on W. Eur and the other bomber in range on SZ3. In that case it is impossible for the UK to take Norway without risking the destruction of its fleet (assuming a 1 AC, 2 DD buy).

@Granada:

I am looking forward to reading Hobbes’ piece on the Russian strategy.

The 1st part was posted today to the Articles section. It is an overview of its objectives and the map board. 2nd part dealing with tactics may take longer…