Flack has a statistically better d2d than Grimm
Creator: the-map-making-shadow || First Published: 07/14/2016 || Players: 2 || Size: 24x25
Categories: None
Rating: 1.00 in 1 rating
For design map discussion or to get suggestions from other users, visit the AWBW Discord Chat!
Comments:
liandry (07/14/2016 03:44am):
I disagree. There are ways to mitigate Grimm's pathetic defenses, and a failed KO due to Flak d2d can deny his own
pushes, especially on heavy terrain/chokepoints.
the-deadly-shadow (07/14/2016 04:57am | Edited: 07/14/2016 04:59am):
I played a game to support my claims. I gave the unit on the right the first hit. I didn't
do one row, since Grimm always does 5 damage to tanks on silos. Here are my results:

On roads Grimm did 353 points of damage during his turn.
On silos Grimm would have done 250 points of damage during his turn.
On roads Flak did 343 points of damage during his turn.
On silos Flak did 259 points of damage during his turn.

I claim that both Flack and Grimm were a little unlucky. I calculated the expected average
and it should have looked like:

On roads Grimm should have done 355 points of damage during his turn.
On silos Grimm should have done 250 points of damage during his turn.
On roads should have done 347 points of damage during his turn.
On silos Flak should have done 255 points of damage during his turn.

I claim that on roads grim does at average 7.1 damage per attack, while Flack does around
6.97. So Grimm does around 0.13 attack more(around 2%). On silos Grimm does 5 damage and
Flack does at average 5.11 damage . So Grimm does -0.1 damage (around 2%). I believe the
gain to be linear, so Flak should gain around 0.8 damage per star. Which means that Flack
does around the same damage as Grimm in several situations. I have to admit, in many
situation Grimm is a very
little better, I made a miscalculation when I created the map.

This is about expected damage, not about failed KO's. I said Flak's d2d is statistical
better, in the sense that he does more damage. So you are right about the KO thing.

Something else that matters and favours Flack is infantry vs. tank battles. Because the
attack of infantry vs tanks/copters is mainly luck based, Flak's infantry do a better job
than Grimm's.

A further remark: If there is a comtower, Grimm gets even a little worse. However, Flack's
COP and SCOP really suck, so when taking that into account Grimm really gains the favour.

I didn't want to say Flack is better than Grimm. I only wanted to say that Grimm's d2d is
not so much better than a low tier CO like Flak.

By the way: congratulations on commenting on the map before it was published.
liandry (07/14/2016 07:59am | Edited: 07/14/2016 08:02am):
I saw the game on your profile (I stalk a lot) and commented here. You can play on an unpublished map?

I think the main concern here is the return fire; how much damage does each CO take from counterattacks?
EDIT: I want to see how you calculated the expected damage done by Flak's units.
Xmo5 (07/14/2016 09:58am):
I haven't looked into this in depth, but assuming your numbers are accurate and that Flak
is better than Grimm d2d based on damage, Grimm is more predictable and therefore a better
option d2d from a strategic standpoint because you know (approximately) what to expect
when planning your attack. It's the same reason why Sonja is probably the worst CO (except
in FoW, obviously)- her bad luck makes her attacks impossible to predict and if you want
to have a confident attack plan, you have to overcompensate and plan to do 10% more damage
than you need to, just to cancel out her *potential* for bad luck. Nothing is worse than
starting an attack and getting stuck partway through, leaving you vulnerable, exposed,
close to the enemy, and yet so committed that there's no way to salvage your efforts into
a halfway decent formation.
Nyvelion (07/14/2016 11:13am):
Maybe someone smarter than me could explain this better, but even if you were right,
wouldn't this only prove that Flak, which doesn't have a C in it, is better than Grimm
just with Tank vs Tank? And only when they are against each other? Grimm Tank vs Flak
Infantry on a road OHKO's better than the other way around, so if the test were done that
way, Grimm would look better.
A better test would use varying units/situations, and have both Grimm and Flak against Andy.
Xmo5 (07/14/2016 11:52am):
That's also a very good point. Yay for designed experiments!
liandry (07/14/2016 12:44pm | Edited: 07/14/2016 01:29pm):
Maybe we can use the damage caculator instead of making a map with all the (reasonable) matchups, no?
Repeated taps on the "calculate" button (I use a phone, I know I'm a pleb) indicates that Grimm does 7 (sometimes 8) HP
damage to Adder (fck Adder) while Flak does like 5-7 (8 with luck) in a tank fight on roads.

EDIT: iirc Grimm has 130/75 stats (based on musings from the forums anyway) while Flak has (85-125)/100 stats
(averages to like 105/100). No way Flak's better. However, any bland CO vs Grimm has a 25% (?) atkboost anyway, so
maybe that's why Flak is seemingly consistently merely slightly weaker on the offense (vs Grimm) than Grimm.
(Flak's stats vs Grimm is ~(106.25-156.25)/100, which should average to 131.25/100 yay results)
a9977321 (07/16/2016 06:03am | Edited: 07/16/2016 06:33am):
Grimm: 130/80, luck=0-+10
Flak: 100/100, luck=-10-+25
For 100/100 COs(like Andy), the damage of tank vs tank=55%
Case1 Road battle damage
Grimm: 130*55=71.5=71, luck=0-10, damage=71-81 EX=76
Flak: 100/0.8*55=68.75=68, luck=-10/0.8-25/0.8=-12.5-31.25=-12-31, damage=56-99
EX=77.5
Case2 Silo battle damage
Grimm: 130*55*0.7=50.05=50,luck=0-10*0.7=0-7, damage=50-57 EX=53.5
Flak: 100*55*0.9=49.5=49, luck=-10*0.9-25*0.9=-9-22.5=-9-22, damage=40-71 EX=55.5

Calculations are based on the mechanics I know. There are some differences between
the calculation and the result of damage calculator, but the diffferences are not too much.
One thing important is that Flak does luck damage, and it is quite different from fire power
boost.

EDIT: I got it wrong when calculating the defense bonus. I will calculate that again.
The damage fomula: (http://awbw.wikia.com/wiki/Damage_Formula)

EDIT: Corrected version of calculations
Grimm:130/80. Luck=0-+10
Flak: 100/100. Luck=-10-+25
Tank basic damage=55
Case of roads:
Grimm: 55*1.3=71.5=71, luck=0-+10, damage=71-81 EX=76
Flak: 55*1.2=66, luck=-10*1.2-+25*1.2=-12-+30， damage=54-96, EX=75
Case of silos:
Grimm: 55*1.3*0.7=50.05=50, luck=0-10*0.7=0-+7， damage=50-57, EX=53.5
Flak: 55*0.9=49.5=49, luck=-10*0.9-+25*0.9=-9-+22, damage=40-71, EX=55.5
liandry (07/16/2016 08:29am):
It's -10 luck? Snap. Maybe Flak is better d2d after all.
a9977321 (07/16/2016 10:00am):
What we care about is not expectation.... In strategy-heavy battles any bad luck can be
fatal. One falls and the whole campaign shall fall.
the-deadly-shadow (07/17/2016 05:07am):
about the luck, as far as I know, 0-10 luck means a random number from 0-9. I believe
having 10 is impossible. If it was, the Grimm tanks on the roads would do 8 HP damage with
9 and 10 points of luck. So that should happen 2 out of 11 times. That contradicts the
result quite a lot (3 as a result, while 9 were expected).

So the maximum damage should be a little lower. That favours Grimm a little. I don't know
if Flak's units ever gets 25 points of luck.

When we are talking about such a small difference, we actually should know about the
rounding off to get a better idea. I think that 71.5% is 71.5% not 71%

My source: http://awbw.wikia.com/wiki/Damage_Formula
observerandhost (07/17/2016 12:11pm):
That's true.
Flak is a luck CO and the specialty of luck makes luck bonus stronger than fire power bonus
in terms of expectation, as luck damage makes all your infs useful against megatank.
If Flak does not have bad luck but have same damage expectation, Flak can surely be
high tier.
liandry (07/17/2016 12:24pm):
Flak without bad luck > Nell, so yeah.
Nyvelion (07/19/2016 01:41pm):
Except on COP or SCOP.
Though Flak's SCOP is so bad I'd be afraid to use it, since he can randomly do next to no
damage at a crucial time, while Nell can still take out a Megatank with a 1HP infantry.
the-deadly-shadow (09/16/2018 03:54pm):
Dear Nyvelion.

One day you will read this comment. I guess that is the way you are. I found out that for
Flak, the luck values aren´t sampled from a flat distribution, but the luck values are
sampled from a distribution like X-Y, where X and Y are distributed flat. That means that
extreme values are quite unlikely.
Tmi489 (10/28/2021 11:18am | Edited: 10/31/2021 10:38pm):
Dear Nyvelion.

One day you will read this comment. I guess that is the way you are. I found out that luck values are dependent on health.
This
means that a 1HP unit can only deal up to 10% luck damage.

Also, hi! I'm sure that you've heard the news by now. Welcome to the new era, though I'm not sure if you'd like a certain big
face in AW right now...



Advance Wars is (c) 1990-2001 Nintendo and (c) 2001 Intelligent Systems. All images are copyright their respective owners.