150 IPCs - Built the best fleet possible…

[Nukchebi0]

Math time:

Inf vs tanks: 10 inf for 6 tanks. 101/3=3.33 theorectical hits on defense. 61/2=3 hits on defense.

101/6=1.66 theorectical hits on offense. 61/2=3 hits on offense.

So yes, infantry suck for attacking. They are very good for defending though.

Now, subs vs destroyers:

3 subs for 2 destroyers.    31/3=1 theorectical hit on defense. The same for offense. 21/2=1 theorectical hits on defense, also the same for offense.

Now, for airplanes. 30/6=0 theorectical hits vs airplanes. Meanwhile 21/2=1 theorectical hits vs airplanes.

Considering 25% of the forces in a naval battle tend to be airplanes, we will take 75% of the anti-ship attacks and 25% of the anit-aircraft attacks and average them by dividing them by two.

1*.75=.75

.75+(0*.25)=.75

.75/2=.375

1*.75=.75

.75+(1*.25)=1

1/2=.5

Since .5 is 4/3 of .375, .375 is 3/4’s of .5.

Therefore, subs have only 75% of the effectiveness that destroyers do, making the destroyers a better buy.

[OLver]

@Nukchebi0:

3AC    48
4FGT   40
2DD     24
2BB      48

150 on the nose.

Subs suck because they don’t hit well, don’t defend well, and can’t hope to live an air assault.

Destroyers are much nicer, and having three loaded carriers will pawn anything which tries to attack.

The two battleships with soak up hits, amphibious invasion support, and destroy anything.

Fighters also have more versatile uses than subs.

What if ennemy has a large fleet of subs. Your toast. You don’t have enough fodder and those AC will probably die…

[Guest]

Hey Nukchebi0.

Couple thoughts:

1. Infantry are pretty good on offense.  How many infantry can 60 IPCs of tanks kill?  How many infantry can 60 IPCs of infantry kill?

2. If you build a lot of destroyers, you are telegraphing your move and you will be Rock-Paper-Scissored into submission.  Fighters beat destroyers easily.  Subs are great as fodder, so having plenty of them along with a carrier group or two is the way to go.  I think you have discounted them too much in your analysis.  In no instance would I have a destroyer-heavy navy for controling the water.

Peace

[Guest]

In answer to the first question…

The sim drops below 50% win rate above 14 INF being attacked by 12 ARM (46% for 15 INF, 60% for 14 INF)
For INF attacking INF, the win rate is at 50% for 20 INF on 14 INF, so slightly worse than 60IPC of ARM.  at 15 enemy INF, the win rate drops to 36%, well below the ARM

[Jennifer]

The problem with the whole tank vs infnatry analogy is that I’m not sending my tanks in against infantry without meat shields.  Especially if they are subject to counter attack!

Likewise, I’m not sending my carriers in against submarines without meat shields!

[Nukchebi0]

@OLver:

@Nukchebi0:

3AC    48
4FGT  40
2DD    24
2BB      48

150 on the nose.

Subs suck because they don’t hit well, don’t defend well, and can’t hope to live an air assault.

Destroyers are much nicer, and having three loaded carriers will pawn anything which tries to attack.

The two battleships with soak up hits, amphibious invasion support, and destroy anything.

Fighters also have more versatile uses than subs.

What if ennemy has a large fleet of subs. Your toast. You don’t have enough fodder and those AC will probably die…

Just attack them with airplanes, to weaken them.

[Nukchebi0]

@CrazyStraw:

Hey Nukchebi0.

Couple thoughts:

1. Infantry are pretty good on offense.  How many infantry can 60 IPCs of tanks kill?  How many infantry can 60 IPCs of infantry kill?

2. If you build a lot of destroyers, you are telegraphing your move and you will be Rock-Paper-Scissored into submission.  Fighters beat destroyers easily.  Subs are great as fodder, so having plenty of them along with a carrier group or two is the way to go.  I think you have discounted them too much in your analysis.  In no instance would I have a destroyer-heavy navy for controling the water.

Peace

1. Yes, in numbers, but they can’t move two and can’t blitz. Tanks provide more reliable defense. Maybe I should start buying infantry, though.

2. I have everything in my navy, so there is no Rock-Paper-Scissors opportunity.

3. Jennifer is right, though. Never send in expensive units without meat shields. The BB’s take care of that for the navy, and other infantry would come in with the tanks.

[Guest]

BTW:  On the 60 IPC land units on attack…
60 IPC ARM takes out 14 INF
60 IPC INF takes out 13 INF
60 IPC as 12 INF, 6 ART takes out 17 INF, as does 10 INF, 5 ART, 2 ARM

It is not an “either/or” question, but rather having the correct MIX.

[Guest]

There were really two main things I was replying to:

Therefore, subs have only 75% of the effectiveness that destroyers do, making the destroyers a better buy.

I think that is not really a correct assessment.  The most destroyers I would ever want is 1, and even that is questionable.  The reason is that the early strike ability of the sub is very over-rated.  With a single opposing btl, you need 4sub before your early hit is mathematically significant.  Even then, opposing subs can be taken as casualties and still fire, so you’re not gaining much - odds are that sub would have died in the first round anyway.  There is almost no large naval battle where the early sub shot will be a big difference maker.  But if my opponent has 10 subs and I have fewer than 2btl, I’d want 1des.  That’s about the only time.

So yes, infantry suck for attacking.

I think Switch has shown that inf are about as good as tanks $ for $.  After spending 60 IPCs, the tanks have only managed to kill one additional unit.

This all comes back to the three tactical fundamentals:

Count - Who has more units?  This is usually the most basic and most important assessment.  Look at the R1 Baltic battle if Germany buys a single tra: you’d have 2ftr 1bmr vs. 2subs 2tra 1des.  The count wins even though the oPunch is 10 and the dPunch is 5.  The oPunch is DOUBLE the dPunch, but because the dCount (5) is almost double the oCount (3), the defender has a better chance to win the battle.

Punch - Who has the best chance to inflict the most damage per round?  Punch is found by adding the “hits-on” values for units.  So on offense 3inf 1tnk has oPunch = 6; on defense that same group has dPunch = 9.  Punch is related to count, but not directly.  Obviously four attacking bombers bring more punch than four attacking inf (oPunches = 16 vs 4, counts = 4 vs 4).  You want punch; but take care of count first.

Skew - The strongest tactical groups match high values and low values to make use of “fodder”.  You don’t want “even” groups of all 2s or all 4s - a skewed mix of 2s and 4s is better.  So Switch is on the mark when he says you need a mix of units to be effective.

If you really want some exciting math, check out the Caspian Sub Policy Paper #05 about tactics.  It covers what Switch was demonstrating as well as a few other tricks like Count Equivalents.

http://games.groups.yahoo.com/group/Caspian_Sub/

Peace

[Guest]

You know, we agree on so many things CS, but disagree on so many others.

I honestly think our differences stem from WHERE and HOW we play most of our games.

[Guest]

The only thing I would say as regards your post CS is that I would switch points 1 and 2, in all but the smallest battles.

Once you reach a certain level of participating units on both sides, count is reduced in terms of how important it is in a battle.  For small battles, yes count is the MOST important point, but above a certain number it is not as critical.

For example, 1 BOM on 4 TRN.  Punch is even at 4.  But due to the count issue, the BOM only wins less than 10% of the time.

But on mid to large sized battles, punch moves ahead of count as a decisive factor (as noted above in that a smaller number of mixed INF/ART/ARM has a stronger attack than a much higher number of INF.

As for skew…
Not sure if I am reading how you are using skew correctly.

I try to match high-punch unit to high-punch unit.  For example, in a Naval battle I want BOM and BB to counter BB and FIG.  I want FIG and DST to counter AC and DST.  I want AT LEAST an equal number of 3’s and 4’s (or two 3’s to counter a 4) plus my fodder units (0, 1, 2’s) in order to ensure a high probability of victory.  Is that what you mean by skew?

[Guest]

Hey Switch.

What you layed out in your post is a very good way to quickly compose an attack; for every 3 on defense, have a 3 on offense.  That’s basic matching.  It’s a good, fast counting skill.

But that’s not really what skew is.  A skewed force has a significant split between high values and low values.  It is measured within the force, not compared to the other forces.  Examples:

Attacking groups:
10inf - Perfectly even; not skewed at all.  You have 10 1s.
8inf 2art - Somewhat skewed.  6 1s and 4 2s.
5inf 5tnk - Highly skewed.  5 1s and 5 3s.

In simple terms, a skewed force has fodder at the low end to protect the heavy damage dealers at the high end.  The punch of a heavily skewed force dimishes more slowly than the punch of an even force because your early casualties represent less of the force’s overall punch.

Where does this matter?  Suppose you have 2inf 1art 1tnk.  You have two territories you want to attack.  Both have 1inf on them.  You need to take them both, but one is slightly more important than the other.  How do you divide your troops?

Look at these divisions:
[Force A] 1inf 1art : oPunch = 4 oCount = 2
[Force B] 1inf 1tnk : oPunch = 4 oCount = 2

For count and punch the scores for the forces are equal.  Which group should go to the more important target?  You should send Force B to the more important battle.  The reason is that after 1 hit Force A loses 50% of it’s punch (2) while Force B loses only 25% of it’s punch (1).  That uneven distribution of hitting power is the skew; Force B is skewed.

Force B wins the battle 90% of the time, Force A wins 85% of the time.

The concept of skew also helps when sizing up stacks.  Here are two more examples with Force C attacking Force D:
[Force C] 5inf 5tnk : oPunch = 20 oCount = 10
[Force D] 10inf : dPunch = 20 dCount = 10

So those groups are even on punch and count.  Who wins?  The skewed group; Force C wins 63% of the time even though the forces have equal count and punch at the start of the battle.

Hopefully that makes it clear.  Did it?

Peace

[Guest]

Yes, now I know what you mean by skew… what Jen calls “meat shields” to preotect the high value units, and what i refer to as punch loss rate.

[Jennifer]

Personally I always viewed subs to act much as they did in the real war.  Solitary navies designed to sink solo units left alone in striking range.  ie) Sub SZ 8 vs TRN SZ 1 on Germany Rd 1.

It’s just too expensive to get a significant portion of them to use in a large naval engagement.  You’re better off buying Bombers or destroyers for that.  (I actually like to have a 2nd BB, it’s a mental thing, I can’t really justify it vs 2 DDs)  That’s assuming you plan on attacking the enemy.  If you want defense, carriers and fighters.

[newpaintbrush]

I use subs for two reasons.

1.  Battle fodder.  They cost 8 IPC and attack at 2.  “Infantry o’ the Sea”

2.  Blocker.  It’s the cheapest naval unit you can buy except for transport, and if you use a transport, you can’t submerge after an air attack.  If you use a sub, they have to commit a battleship to soak up a hit, or risk losing a naval unit - and if they send NO naval units, the more air they send, the less air they send against other targets; the less air they send, the more likely that the sub survives.

I don’t quite see how getting bombers is better.  Fighters are usually easier to use with carriers, and destroyers are not as cheap as subs.  Destroyers can defend against air, but I figure that’s what carriers are for.

[DarthMaximus]

@CrazyStraw:

Hey Switch.

What you layed out in your post is a very good way to quickly compose an attack; for every 3 on defense, have a 3 on offense.  That’s basic matching.  It’s a good, fast counting skill.

But that’s not really what skew is.  A skewed force has a significant split between high values and low values.  It is measured within the force, not compared to the other forces.  Examples:

Attacking groups:
10inf - Perfectly even; not skewed at all.  You have 10 1s.
8inf 2art - Somewhat skewed.  6 1s and 4 2s.
5inf 5tnk - Highly skewed.  5 1s and 5 3s.

In simple terms, a skewed force has fodder at the low end to protect the heavy damage dealers at the high end.  The punch of a heavily skewed force dimishes more slowly than the punch of an even force because your early casualties represent less of the force’s overall punch.

Where does this matter?  Suppose you have 2inf 1art 1tnk.  You have two territories you want to attack.  Both have 1inf on them.  You need to take them both, but one is slightly more important than the other.  How do you divide your troops?

Look at these divisions:
[Force A] 1inf 1art : oPunch = 4 oCount = 2
[Force B] 1inf 1tnk : oPunch = 4 oCount = 2

For count and punch the scores for the forces are equal.  Which group should go to the more important target?  You should send Force B to the more important battle.  The reason is that after 1 hit Force A loses 50% of it’s punch (2) while Force B loses only 25% of it’s punch (1).  That uneven distribution of hitting power is the skew; Force B is skewed.

Force B wins the battle 90% of the time, Force A wins 85% of the time.

The concept of skew also helps when sizing up stacks.  Here are two more examples with Force C attacking Force D:
[Force C] 5inf 5tnk : oPunch = 20 oCount = 10
[Force D] 10inf : dPunch = 20 dCount = 10

So those groups are even on punch and count.  Who wins?  The skewed group; Force C wins 63% of the time even though the forces have equal count and punch at the start of the battle.

Hopefully that makes it clear.  Did it?

Peace

Yes, you definitely need a “skewed” force but the 1 to 1 skewed force in your example is extremly inefficient.

The problem with some of this is, you leave out cost.  Which should be in your write up #4.

It is IMMENSELY cheaper to defend.

Regardless of what units you buy, as the battles grow (say minimum 10 inf on a ter), you need to out spend your opponent 4:3 in order to take.

Which coincidently is the cost of 5 inf, 5 arm vs. 10 inf.  (40 IPC vs. 30 IPC).
However, it is not fair to compare:
10 inf vs. 8 inf, 2 rt, vs. 5 inf, 5 arm b/c the costs are 30 IPC vs. 32 IPC vs. 40 IPC.

What did player 1 and 2 do with their extra 10 IPC and 8 IPC respectively?

My hunch would be 7 inf, 2 rt, 2 arm would be just as effective as 5-5, due to the 3-1-1 (4-1-1) ratio.

I wrote about this in another thread a while ago in the Classic Forum (I think maybe it was in revised).

I’d have to dig it up but I believe the best ratio (inf/rt/arm) in terms of cost and off/def punch is 3:1:1 to 4:1:1, where you match rt and arm 1:1 but have 3-4 times as many inf.

But I found that the 4:3 offense to defense cost ratio is pretty darn solid in terms of taking a territory.

[Guest]

3-1-1 sounds about right Darth.  And based on your track record, I would also say well proven.

You need punch, that is what the ARM/FIG is for.  You need to expand the chance for hits, that is what the ART is for (1 ART for 4 IPC  and 1 INF for 3 gives you a 100% increase in hit probability over 2 INF, for only 1 extra IPC).  And you need unsupported INF as the most cost effective way to absorb enemy fire (and provide for a few possible hits as well).

A defender relying on INF loses defensive punch pretty evenly, just a few ARM or FIGs added to an INF stack makes a MASSIVE increase in defensive power.  likewais, an attack with no unsuported INF loses punch in a hurry, expsoing more costly units to defender fire.

A 3-1-1, if otherwise matched with the enemy forces (force count, total dice value) will usually outlast a 1-1 enemy defense, and almost certainly a 1 enemy defense.

[jsp4563]

2 CV (32)
4 Fgt (20+2 free)
2 BB (48)
2 DD (24)
3 SS (24)
4 TP (free)

Save $2

Offensive Punch: 34
Defensive Punch: 62
Hit absorbtion: 9

[DarthMaximus]

I know this is a naval thread but I work on that analysis later, but I wanted to finish up some things on the land units.  I’ve spent a lot of my day at work pretending to work while I worked on this.  :-D

@ncscswitch:

3-1-1 sounds about right Darth. And based on your track record, I would also say well proven.

You need punch, that is what the ARM/FIG is for. You need to expand the chance for hits, that is what the ART is for (1 ART for 4 IPC and 1 INF for 3 gives you a 100% increase in hit probability over 2 INF, for only 1 extra IPC). And you need unsupported INF as the most cost effective way to absorb enemy fire (and provide for a few possible hits as well).

A defender relying on INF loses defensive punch pretty evenly, just a few ARM or FIGs added to an INF stack makes a MASSIVE increase in defensive power. likewais, an attack with no unsuported INF loses punch in a hurry, expsoing more costly units to defender fire.

A 3-1-1, if otherwise matched with the enemy forces (force count, total dice value) will usually outlast a 1-1 enemy defense, and almost certainly a 1 enemy defense.

Okay, I ran some analysis and I used a LL system for hits (and simplicity), both attacker and defender hit on 3 and above.

Also I used 60 IPC for total IPC to spend b/c it is easily divisible by 3, 4, and 5 and allows enough units for a couple rds of battle but not too many.

I did three groupings:
#1 -  Inf/Rt/Arm
#2 -  Inf/Rt
#3 -  Inf/Arm

And tried to find the optimum buy for each grouping.
For #1, I started with 20 inf (60) then moved down in the following increments:
20/0/0
17/1/1
14/2/2
11/3/3
8/4/4
5/5/5
2/6/6

And I had each category attack all the other categories.  So for example, I did 20 inf attack 20 inf, then 20 inf attack 17/1/1, then 20 inf attack 14/2/2…etc.  then I did the same for 17/1/1 and 14/2/2…etc.

The combo that did the most damage to the most categories was the grouping of:  11/3/3.
It did the most damgae but tied with the 5/5/5 in the category of attacking 20 inf and 17/1/1.  Both left 8 inf and 7/1/1 behind in each respective category.
but the 11/3/3 eeked out slight wins in the other 5 categories for this group.

Now for the Inf/Rt groupings using the same method I tried:

20/0/0
16/3/0
12/6/0
8/9/0
4/12/0
0/15/0

And here I came out with both the 12/6/0 and 8/9/0 grouping each winning 3 categories, with the 12/6/0 doing the most damage (or taking with the most units) against 20/0/0, 12/6/0, and 0/15/0 and the 8/9/0 grouping doing better against 16/3/0, 8/9/0, and 4/12/0.

Now for the Inf/Arm groupings I used:
20/0/0
15/0/3
10/0/6
5/0/9
0/0/12

And this was pretty clear cut as the 10/0/6 grouping did slightly more damage to each category.

Now finally I took the best from each category to squared off, so I had:

11/3/3
12/6/0
8/9/0
10/0/6

And did the same thing and found that:
The 11/3/3 won 2 and tied in 1
The 12/6/0 tied in 2
The 8/9/0 tied in 1
The 10/0/6 performed the weakest against the others.

So what does this all mean?

I think certainly in could be used to back up the idea of keeping the 3/1/1 to 4/1/1 ratio for land units as both being good offensively to take on almost any style of army and being pretty good on defense too.
Although, the mostly inf breakdowns whether it was 20/0/0, 17/1/1, or another mix with a high % of Inf dominated the defensive aspect.  You really can beat Inf for defense at the cost of 3 IPC.

I was kind of surprised by the 8/9/0 one, but I guess it makes sense since you instantly double the power of inf, but I still think the 1-1 is not the way to go, but the 12/6/0 also did pretty good.  I’ve never really given rt a fair chance but I may consider it more in the future if I don’t need the rapid movement.  Maybe load up on rt in the early rds and save armor for the mid to late rds when you may need quick movement.

I hope you guys can follow this, and now to work on the Naval breakdown when I have some time.  :-)

[Imperious Leader]

Darth i feel that the best combination for land units is Infantry and artillery at specific ratios of say 1.5 to one or 2 to 1 Infantry. The idea is to never get into the kill zone involving artillery and use the extra infantry beyond artillery as fodder. Financially this i feel is the best buys all around.