GOTM 42: Pre-Game Discussion

[Paul#42]

A sperm whale falling into the lake from the sky in turn 2? More likely just a bowl of petunias, as Ainwood's Improbability Drive goes into top gear.

As long as it's not two unchanged rockets falling down on us thinking: "Oh no, not again..." We don't want to know, why they would think this...

Maybe we are stuck (with on settler's leg eg.) on some unfertile planet just waiting... and feeling unchallenged

Guess this would be the Gotm to go for spaceship victory - if I'm not running out of time...

EDIT: please excuse missing accuracy - I read those books like 10 years ago in German language...

 

[eldar]

Hmm. Perhaps we have the ability to shape coastlines to our own liking...?

 

[Più Freddo]

As far as this start goes, settle in place and immediately begin a settler; use the worker to build a road toward the wines and also be exploring.
[...]
Since the capital isn't going to be working any land tiles until size 2, I'm going to move the worker right to the wines and start roading backwards, getting the early exploration from that.

For these goals, you're probably better off letting the worker chop a forest along the way. Use the 10 first shields in the city towards a warrior and fulfil your 3000 BC settler plan with the 10 missing shields replaced by the chop.

 

[leopalas]

Settler NW,N,NE and settle. Worker chop forest. One can still get the wines and the terrain north can only be better, maybe grassland.

 

[Skydance]

Rome is Commercial. Doh! But ... I thought Commercial civs got Bronze Working to start, and Scientific civs were the ones who got Alphabet. What's Rome doing with Alphabet?

Smith's is a lot later in the game than I like for my GA, though. Colossus (captured?) + Sun Tzu would be well-timed, but would require strong trading partners for AA tech in an Emperor game.

Turn 10's a tad optimistic! 5 turns just to get there, then another 12 turns to build the mine?

Is it 12 turns to build a mine? I've spent too much time playing Industrious civilizations ...

Unless there's FP to the south, this game is going to be famous for the number of slow starts.

 

[AlanH]

Is it 12 turns to build a mine?

Well, on the flat it's definitely 6 turns for regular civs before Replaceable Parts or Democracy. I believe it's twice that on a hill, which is where the grapes are. So that's 12 turns.

Sorry, the trait techs are the other way around as well. Alphabet for Commercial and Bronze for Scientific.

 

[Niklas]

I ran a test game with the same settings, not quite the same bad start though. My closest neighbour was Carthage, with the city of Carthage located on a hill. I lost 7 veteran/elite legions to a single (initially) regular Numidian Mercenary. I consider this an omen...

 

[Skydance]

Sorry, the trait techs are the other way around as well. Alphabet for Commercial and Bronze for Scientific.

Well, then ... that clears that up!

 

[ionimplant]

I ran a test game with the same settings, not quite the same bad start though. My closest neighbour was Carthage, with the city of Carthage located on a hill. I lost 7 veteran/elite legions to a single (initially) regular Numidian Mercenary. I consider this an omen...

can any one good with math calculate the odd of such thinig happening? 5%?

 

[denyd]

A Legionnaire is a 3 attack (3.3 for the elite I think)

Numidian Mercenary is a base 3 defense
and gets a 50% bonus for on a hill
and gets a 25% bonus for fortified
and gets a 50% bonus for a city or walls (I taking that for granted from his post)

That would be a total of 6.75 defense. So if it's Elite legion vs regular NM, it's likely the first only got 1 HP off (if that), the second could easily go 0 for 4 so now the NM is a 3/4 Vet NM, #3 Legion takes another HP off, but he promotes again to 3/5. I think he'd now get another 10% bonus for being elite (not 100% sure about this), the would be 3 attacking 7. The #4 legion could easily go 0 for 4 with those odds. If the remaining legions went 2 for 12 (not that rare considering only a 30% victory chance), the NM would be standing there shouting insults (probably something like "Your mother was a hamster and your father smelt of elderberries" ) at the other Roman units.

So it isn't that unusual for a normal RNG that he could withstand that kind of assault in that situation.

 

[Sabre]

According to TechCalc's combat calculator here are the odds of a Legion killing a fortified Merc on a hillside city before and after any promotions it recieves in battle:

Elite Legion vs Reg Merc: 37.1%
Elite Legion vs Vet Merc: 20.8%
Elite Legion vs Elite Merc: 10.8%
Vet Legion vs Reg Merc: 27.0%
Vet Legion vs Vet Merc: 13.6%
Vet Legion vs Elite Merc: 6.4%

Yuck. Any bad luck with your first attacker is going to make the job very hard. Say, you lost your only Elite on the first attack without doing any damage to the Merc and the Merc is then promoted. You are now left with a Vet Legion vs a Vet Merc with very good odds of the Merc winning and advancing to Elite. Hopefully, you've done some damage in that second attack, though I've seen enough stretches of bad RNG where I do no damage that getting to the point of Vet Legions vs an Elite Merc is possible, making it pretty likely that you could lose 7 Legions in the attack. So, while losing 7 would require some bad luck it's not that far out of the question. If I was in that situation I'd probably want 4-5 Legions for every Merc expected to be there, 2 in a normal town and 4 in Carthage. It may be a little bit of overkill, but nothing will stall out an attack like a SOD failing 3 tiles into enemy territory.

edit - Denyd beat me to the punch

 

[AlanH]

A Legionnaire is a 3 attack (3.3 for the elite I think)

It's always 3. Attack strength is not dependent on experience, and there's no 10% strength bonus for an elite, on either side. Attack strength relative to the opponent's defence strength determines the chances of hit point win/draw/lose on each attack during a battle. An elite's extra hit point just gives him one more to lose before he dies.

 

[AlanH]

I use a system based on the probability of each hit point being won or lost.

3 attack vs 6.75 defence gives 69% success per hit point to the defender according to the online combat calculator - set it to 1hp on each side.

With odds that bad you should assume that you will have to take 5 hp off the defender to kill him, even if he starts out as a conscript. Do the multiplication of 5 by 0.69 repeatedly, and you'll see that it will take 11 attacker hits to get above 90% chance of killing the defender. That's two dead elites. 13 attack hits will get you to 95% - an elite and 3 vets. For a 99% success rate you need to deploy 17 attack hits - 6 vet legions.

Niklas' 7 lost vets is 28 attack hits, and the chances of the defender surviving that long are down around 0.02%, so I'd say that was a pretty evil sequence of rolls of the RNG. Note that this was the Carthaginian capital. I have no evidence, but I have a gut feel that there are some extra hidden defender bonuses when (a) you are attacking the last defender in a city and/or (b) it's the capital. I'm probably just seeing things that aren't there, but I do tend to deploy a bigger force against a capital these days, in the expectation of a harder fight.

Even a rifle fortified in a hill city across a river would only have a 5% survival rate against 7 vet legions.

 

[ainwood]

The capital is considered a city for defense purposes (you can't build walls in it, for example).

 

[Tomoyo]

The capital is considered a city for defense purposes (you can't build walls in it, for example).

Yeah you can, actually. I've built them there in many AW games. They just don't show.

 

[ainwood]

Yeah you can, actually. I've built them there in many AW games. They just don't show.

I just read a thread on it, and yes, there is apparently no bonus (although I was sure I was unable to build walls the other day...)

BTW - the walls aren't there due to the defense bonus; they're there due to the bombard bonus (otherwise, a Civil defense in a town would make it appear 'walled'.)

 

[Skydance]

Was the city too large for walls, ainwood? I remember something obscure about not being able to put a wall around a city once it reaches a certain size.

Do the multiplication of 5 by 0.69 repeatedly, and you'll see that it will take 11 attacker hits to get above 90% chance of killing the defender.

I see what you're trying to do, Alan. You want to multiply 0.69 five times, to simulate the chances that the Defender will win 5 battles:

0.69 x 0.69 x 0.69 x 0.69 x 0.69 = 15.6%

That's the odds of the Defender winning 5 attacks in a row, though. The actual formula you need for the odds has to take into account that the Defense can fail 4 times (it takes 5 failures to kill him). I think you need to use that old factorial formula for computing this one. I wish I knew for sure.

The logic denyd followed is concrete, and seems reasonable to me. Here's another way to look at the problem:

Start with a 31% chance of losing a hitpoint (69% of not losing one).

1 attack = .31 hitpoints lost (on the average).
2 attacks = .62 hitpoints lost (on the average).
3 attacks = .93 hitpoints lost (on the average).

This passes a "reality check": if you've got a 31% chance of losing a single round of combat, you'd expect it to take an average of a little more than 3 attacks to lose 1 hit. Continue this line ...

17 attacks = 5.27 hitpoints lost (on the average)

So, in an average situation (50% of the time), it would take 17 attacks to kill a 5 hitpoint Defender. That's 17 attacks, 5 of which the attacker wins, leaving 12 attacks he loses. 50% of the time, you would kill 3 Vet attackers, and the fourth one would win.

If you kill 3 Vets 50% of the time, it just doesn't make sense that there would be a 0.02% chance of killing 6 Vets (particularly since we have a report of it happening). It does make sense that there would be a 25% chance of killing 6 Vets. I'm not saying 25% is the answer to the question; I'm just saying it has a good feel to it.

 

[ainwood]

Was the city too large for walls, ainwood? I remember something obscure about not being able to put a wall around a city once it reaches a certain size.

Not directly - I was actually doing some testing to see what it was that causes the wall graphics to appear on cities (where I found it was the bombard bonus, not the defence bonus). To test this, I actually gave barracks a bombard bonus. I was surprised to find that this prevented me from building a barracks in the capital (even though it was size 1). I'll double-check on this though.

 

[ionimplant]

in my test game, i lost 4 straight battles with barbarian (2 my warrior vs warrior and 2 my warrior vs horseman) and the barbarian unit came out full hp after three of them... what's the possibility for that!?
my 20K city ended up robbed three times by them (once i lost 1 pop and twice i lost the production) and it finally led to my race to oracle...
it seems i need to get enough military in my 20k city this time.

 

[Più Freddo]

I'm not saying 25% is the answer to the question; I'm just saying it has a good feel to it.

I thought "42" was supposed to be the correct answer?