Micro Challenge II

[Mantic0re]

WoW. Reason number 181 why I'll never be an immortal/deity player.

Very educational post.


Maybe like chess instructional book, these can be grouped into openers (micro and moves that support multiple plans) and closers (games that focus on grand strategy and end game) that combine can become the BTS equivalent of Grandmaster teaching drill.

 

[TheMeInTeam]

WoW. Reason number 181 why I'll never be an immortal/deity player.

I've won quite a few deity games without planning so carefully. You'll win higher % if you do, but you can win on deity well over "1/#players" even if you don't.

A big part of the reason for it is that AI abuse > micro as a determinant in victory. However, Micro > AI abuse alone in having a very high deity win % and getting the truly fast victory dates.

 

[on_land]

Really enjoyed this as I'm bad at micro and these are helpful (as mantic0re says, like chess problems). Got the normal after many tries, but always one or two turns late on the nightmare. After reading Snaaty's solution i see my mistake was

Spoiler :

when i'd grown back to size 2 while chopping the granary, i would work the 3F corn, thinking food --> more whip power, but duh it's going to be 1 pop whip no matter what. so work hammers to do the whip (and the following whip chain) sooner.

 

[bestsss]

on_land, an obvious tip: if you wish to whip granary you wish to have (exactly) half of the food bar filled, not more.

 

[kossin]

on_land, an obvious tip: if you wish to whip granary you wish to have (exactly) half of the food bar filled, not more.

It's a decent rule of thumb but not optimal.

 

[toob]

It's a decent rule of thumb but not optimal.

Trying to reason this out ...

The strength of a granary is that it stores food, up to 1/2 of the food bar.

A granary stores all food put into it, with a cap of 1/2 the food bar. This is different from, stores 1/2 the food put into it.

So, at a given population (usually post-whip, does anybody slow build granaries if running Slavery?), if >1/2 the food bar is already full, there will be <1/2 the next food bar stored. (There's a fencepost issue here on Epic, which has odd food bar maximums, but let's assume it away.)

If <1/2 the food bar is already full, at the time the granary completes, then food will go into the granary but not be stored, because it exceeds the granary max capacity.

My formula for whipping a granary is at food_stored = 1/2 max_food_at_post_whip_size - net_food_at_post_whip_size - overflow_from_pre_whip_size. (overflow is "how much food would exceed the food bar on the turn it passes the food bar", 29/30 + 1 = 0, 29/30 + 6 = 5) This is for the turn I click the whip button, not the turn the granary completes.

I think this is the "rule of thumb" you mention, but you say it's not optimal. What's optimal? Are you talking about mm'ing max overflow hammers too?

 

[kossin]

The granary stores food on the turn of growth as well.

Simple example: you have +11F surplus at size 1 (granary can store 11F).

Would you rather complete your granary at 11/22F or 12/22F? (assume same turn)

11 + 11(surplus) + 11(granary) - 22(growth) = 11F
12 + 11(surplus) + 11(granary) - 22(growth) = 12F

If it were possible to do so on the same turn, the optimal food count would be 21F when the granary starts gathering.

 

[toob]

The granary stores food on the turn of growth as well.

Simple example: you have +11F surplus at size 1 (granary can store 11F).

Would you rather complete your granary at 11/22F or 12/22F? (assume same turn)

11 + 11(surplus) + 11(granary) - 22(growth) = 11F
12 + 11(surplus) + 11(granary) - 22(growth) = 12F

If it were possible to do so on the same turn, the optimal food count would be 21F when the granary starts gathering.

21 + 11(surplus) + 11(granary) - 22(growth) = 21F.

Or, food_at_curr_size + net_food_at_curr_size + 1/2 max_food_at_curr_size - max_food_at_curr_size. Where curr_size is on the turn the city production bar says 1 turn left on granary (this could be after clicking the whip button).

Rearranged, you want food_at_curr_size = max_food_at_curr_size - net_food_at_curr_size - 1/2 max_food_at_curr_size. In the +11F example, this means you want 21F in the food bar. With wet corn, +4F, you'd want food_at_curr_size = 22(max_food_at_curr_size) - 4(net_food_at_curr_size) - (1/2) 22 (max_food_at_curr_size) = 7. This is after you click "whip", before you click "end turn".

Thanks for the quick (and awesome) response. Do I have this right?

 

[kossin]

With +4F food surplus, size 1 granary takes 11/4 = 3 turns to fill.

(22-1) - 4*(3) = 9F when whipped
(maxFood -1) - (surplus)*(turnsToFill)

9F (whip)
13F(gathering)
17F (4 stored)
21F (8 stored)
Growth (11 stored)

It gets more complicated if you have a good tile you're not actively working

 

[Mitchum]

It's amazing how much difference the correct micro can make in this little challenge. My first go, I only had 4 pops and hadn't even finished the LH or warrior yet. I build the second wb before the granary = mistake.

Nice work, Kossin! Although, in xOTM games, I don't get 2 or 3 tries at the opening...

 

[kossin]

Test games/spreadsheets are there just for that

 

[nona_mouse]

@kossin

Spoiler :

Just got back from the cottage.

Beautiful! Somehow I never considered that I was losing a raw yield when working the fish over the grass hill, I'd gotten up to 24 but was stuck always at size 4 on turn 36, often a single food short.

I see III is up, it'll be something to continue relaxing for the rest of tonight I think. .

 

[on_land]

Awesome. Been years since I did any algebra, but I think I have this formula down, optimal whip granary at (food to grow - 1) - (surplus * turns to fill granary), and err on earlier side. In the 11 surplus example, whip at 10F (to have 21F at gathering); 9F whip is fine but not optimal, while 11F is total fail (growth before gather). I understand it as main goal is to grow and finish filling the granary on the same turn--and given this, the closer to the growth threshold the better.

 

[AFS]

Wow, this was very fun.

I only managed to get Normal after 4 or 5 tries, and there's no way I'll ever archieve Hard or Nightmare. I'm not much of a micro guy, I just try to play as fast as possible and letting the governor take care of the tiles, but I must admit, this motivated me to actually learn to use micro from now on.

 

[DanF5771]

great series, Kossin

t37 some hammers into whatever and 2food short of pop 6

that would mean 28/30 - is that really possible??

my best attempt gets me to 26/30 and 4 into Galley (or whatever) or 25/30 and 7 into Galley

Spoiler :

t0: work Grassland, build Barracks for +2
t1-t3: FGH Worker
t4: work Clams +2 build WB
t5-t10: Corn +3, grow on WB
t11: size 2 - Worker on top of build queue!!! - revolt to Slavery
t12: whip WB, work Corn!
t13: overflow into Worker
t14-t17: build Warrior (t17 Worker loses one due to decay)
t18: size 2 - continue Worker
t19: whip Worker @30
t20: overflow into Granary
...
from here pretty much same as Kossin

 

[Mec AntiKythera]

Thanks for the challenge Kossin, it is a great way to learn. It took me many attempts to find a solution (for normal) and strangely it was a very different one than the way to Nightmare victory.

Spoiler :

working highest food tile build worker and tech pottery, tech BW and finish worker, chop forest into granary and revolt to slavery, whip granary while teching fishing then sailing, improve corn and build WB, whip WB and send to fish, work fish and corn and build WB while worker mines grass hill, at size 3 work clams+fish+corn build LH, whip LH and grow back to size 4 on turn 37 thus exactly meeting the requirement for Normal. In second run through improving corn before chopping forest into granary yields exactly the same result.

While this works for Normal it will not get me to Nightmare. Strangely, my first attempts, all failures, approximated the optimal solution but didn't even get me to Normal. Exact timing is apparently that important.

Clearly I have a lot of MM to learn, but now I have a way to do just that. Keep up the good work! Thanks as well to all other who posted details of their solution, it is appreciated.

 

[kossin]

great series, Kossin



that would mean 28/30 - is that really possible??

my best attempt gets me to 26/30 and 4 into Galley (or whatever) or 25/30 and 7 into Galley

Spoiler :

t0: work Grassland, build Barracks for +2
t1-t3: FGH Worker
t4: work Clams +2 build WB
t5-t10: Corn +3, grow on WB
t11: size 2 - Worker on top of build queue!!! - revolt to Slavery
t12: whip WB, work Corn!
t13: overflow into Worker
t14-t17: build Warrior (t17 Worker loses one due to decay)
t18: size 2 - continue Worker
t19: whip Worker @30
t20: overflow into Granary
...
from here pretty much same as Kossin

Spoiler :

Interesting, it's actually possible to complete all tasks by T36

However, you're left with an extra 6 turns of whip which means it would not be as good for continuing the game (not that this is the goal here, but for anyone looking into improving micro).

 

[Nials]

Great series here, kossin!

Spoiler :

I completed the normal challenge by going Fishing -> BW -> Pottery -> Sailing.

Paris built Worker until Fishing was in, then switched to a WB. The worker was finished after the first WB and then I started another WB, followed by a whipped Granary and Lighthouse.

Worker did Corn -> Mine on grassland hill. The chop went into either the Granary or Lighthouse, I forgot which

Going to give the harder challenges a go now.

EDIT: So close to beating the Nightmare challenge, only 3 hammers short!

 

[rfcfanatic]

I tried it several times, but I never accomplished the goals before turn 40-41. Organized Lighthouse might be the key, but I'm not sure how to use it in a right way.

 

[Mec AntiKythera]

The lighthouse isn't the key as far as I am concerned. Getting some high-yield tiles online quickly, and always working them -- don't whip away pop needed to work them!! -- is absolutely crucial. Next priority is the Granary. Keep trying, if you can move from T40 to T37 you have learned some important lessons.

I do have one question for you masters though:

Spoiler :

One detail is still a bit of a mystery for me. After looking at the video solutions I only have a rather unclear idea about the switch from the 3-food tile to the 2-food+1hammer tile. I am assuming that the extra food would be wasted in the whip (not saved by the granary) so it is simply better to use it as a hammer than as a food. Is this correct? Anyone care to spell this out in detail?