A knight walks into a bar and gets into an argument with a samurai,

>>384046779
Fuck off furry

>>384046779
>75% dodge chance

Have you ever played a videogame before? That shit's broken

>>384046779
I abstain from answering because math hurts.

>>384046779
The knight

Panda wins

>>384046779

Realistically the battle never ends until one of them gets incredibly lucky. Samurai dosent do enough damage to get past knights defense. He relies on his special moves, which are rng. 25% of the time he'll do 50 damage, and 25% of the time he does 20.

Knight will practically never even hit him. He regens 10 life per turn, but he cant do any damage. Panda also has a shit ton of hp for no reason.

My guess is that EVENTUALLY the panda will win, but theyd be fighting for so long the bar will close by the time They're done. Maybe quicker if the knights potion lasts like 50 turns instead of infinite.

Knight does 20 damage per turn in average, so he kills the panda in 60 turns in average.

Panda deals 17.5 damage per turn in average, which means he kills the knight in an average of 40 turns without counting his HP gain per turn.

But since the knight cheated, the panda will need 63 turns in average to kill him. Knight wins.

>>384046779
Panda

File: 1301572477811.png (10KB, 429x410px) Image search: [Google]

10KB, 429x410px

>>384046779
>75% dodge chance
Well there you go. You might as well shave off that 75% from Knight's damage output.
Samurai wins.
OP bullshit,

>>384046779
is it xcom kind of dodge change or what?

>>384052569

Exactly this.

However, and panda can trigger BOTH his abilities at once, he will win if they activate at least twice.

Panda. Just.

Let's break this down into average time to kill the opponent. Normally the bear would kill the panda in 15 turns (1200/(100-20)) but with the 75% evade we need to multiply that by 4, as 3 times out of 4 on average the attack would miss, leading to an average time to kill of 60 turns.

The panda is trickier to calculate for, but let's average its damage over 4 turns. 2 of those turns, no damage would be done, 20 damage would be done one turn, 50 damage would be done another turn, and the bear would regenerate 40 hp over the 4 turns, leading to a net damage of 30 over 4 turns. This means that the panda would need 48-52 turns on average to kill the bear.

Remember, this would not happen every time as chance is a factor, this is only on average. This is also not taking into account that the panda could crit and ignore armour on the same turn, if it could then the panda would just win faster.

Out of 1,000,000 simulations the bear wins 96% of the time.
Simulation code here: http://codepad.org/vOjPRato

File: mm.png (263KB, 408x505px) Image search: [Google]

263KB, 408x505px

>>384046779
>mfw when the bear wins even though the panda has 75% dodge

>>384046779
>80 def vs 50 attack
>with 10hp regen per turn
After a long and annoying battle with many dodges, the knight wins because he is literally unkillable.

>>384055178
Whoops, there was a bug in my code that gave the samurai 75% chance to double attack and ignore defence.
Bear wins %99.9993

>>384055821
Your code is bugged if Panda doesn't win with a 75% chance to double attack and ignore defense.

56.25% to deal 100 damage
18.75% to deal 50 damage
18.75% to deal 20 damage
06.25% to deal 0 damage

That means the Panda an average of 69 damage per turn, or 59 damage including the potion. The bear only does an average of 20 per hit when you include dodge chance, and the panda has more hp than the bear.

My sim has the bear winning generally about 67-68% of the time. Doesn't matter if the Panda or the bear goes first in each turn.

>>384046779
Fuck, I forgot to explain that an attack can only either miss, ignore defense or deal (2*atk) damage, not a combination of them.

I'll repost it one last time after this thread dies.

>>384046779
Stat updates
-Knight gains shield(+40 defense, +90% chance to block, 2% chance to allow a counterattack parry)
-Fixed a glitch where the Samurai had more health than the Knight(Knight now has 1000 HP, Samurai has 850)
-Knight no longer regens hp, now has the special ability to drop his shield in exchange for weilding his sword 2 handed.(+15% chance to break the Samurais sword on clash.)
-Samurai dodge chance reduced to 30%
-Samurai given the ability to parry with counterattack(20% chance to parry, 5% chance to counterattack)

>>384058005
*normal hit, miss, hit that ignores defense or hit that deals (2*atk) damage

>>384058228
The Samurai is actually stronger than the knight, his gear is just crap and katanas suck

>>384058005
Doesn't even matter, because the Panda is already losing.

>>384058453
Post calculations

>>384058421
Lmao no. Knights are trained from childhood to be swole af, samurai were largely archers and polemen with maybe a decade of training. Meat diet > Rice diet

>>384058680
Samurai are agile af though, that kind of mobility requires a healthy body. Good samurais are trained since their childhood as well. Good samurai > Good knight

>>384059423
A knight in full armor is as if not mors agile than a fully equipped US soldier wearing heavy gear. In fact they are just shy of the standard needed to be a firefighter.

>>384059423
>>384060034
Keep in mind that Samurai are agile in and out of bamboo and leather armor weighing at most 10 to 15 pounds.
Knights are that agile wearing full plate armor weighing in at 32+ pounds.

>>384058574
https://dotnetfiddle.net/4AYQ0X
did I get anything wrong here?

>>384058574
Alright.

The bear deals 80 damage when she hits. A 75% chance to dodge can be represented by dividing the 80 damage by 4 (since it only hits 1 out of 4 times). This means the bear deals an average of 20 damage per turn. Because the panda has 1200 HP, it's expected the bear will beat the panda somewhere around 60 turns (1200/20 = 60) for the bear to defeat the panda. (Of course, if the bear is super lucky, she can do it in as few as 15 turns (1200/80 = 15), or she might never hit the panda for all eternity. But the average gives us the answer of who is more likely to win.)

The panda deals a certain amount of damage based on his luck-based skills. If neither of his skills proc, he deals 0 damage because his attack is lower than the bear's defense. This happens 56.25% of the time (.75*.75 = .5625). If only one of his skills proc, he will deal either 20 damage from x2 crit (50x2 - 80 = 20) or 50 damage from ignoring armor. Either of these has a 18.75% chance of happening (.75*.25=.1875). If both skills proc, he will deal a devastating 100 damage blow which occurs 6.25% of the time.

The average damage of the panda is calculated by multiplying each of his possible attacks by the percent chance they happen.
(.5625 * 0) + (.1875 * 20) + (.1875 * 50) + (.0625 * 100) = 19.375

The panda deals an average of 19.375 damage per turn, but the bear chugged a regeneration potion that recovers 10HP each turn. This reduces the panda's average damage to 9.375 damage per turn. Because the bear has 700 HP, it's expected that the panda will beat the bear after around 75 turns (700/9.375 = 74.6667).

tl;dr On average it takes 60 turns for the bear to beat the panda and 75 turns for the panda to beat the bear. The bear is more likely to win.

All the non statistics-fags itt larping as mathematicians

Bear has a ~67% chance to win.

If he didn't chug that potion like the scheming liar bear that he is, then that drops to just ~7%.

>>384060886
this

>>384058005
This samurai nerf drops the Panda's win rate from ~33% to about ~12%.

>>384061682>>384061308
the fuck you talking about

the jap wins for not being a sneering piece of shit

if you change nothing in the op except the samurai's health to 1400 then the win rate becomes 50:50

>>384060886
Bear is more likely to win, but the call is closer.

Since the panda deals an average of 19.375 damage per turn (without considering the updated nerf in which he deals 17.5 damage per turn), he kills her in 37 turns without counting the bear's HP regeneration. This means that in 37 turns, he will deal 717 damage.

In 37 turns, the bear gains 370 HP, which means she has practically 1170 HP the panda will need to chug off from the start. 1170 HP / 19.375 damage per turn = 60.37 or 61 turns in average to kill the bear.

>>384052569
>Even with all the bullshit dodge, crit and ignore chances knight still statistically strongest
Weebs BTFO

Also:
>Katana
>Ignoring armor
Chose one and only one.

>>384046779
>Knight
ATK - DEF
= 100 - 20
= 80 Damage
On average, the Knight will hit every 1 out of 4 turns.
80 / 4
= 20 Damage per Turn
1200 / 20
= 60 Turns
>Samurai
1 in 4 chance will deal double damage
1 in 4 chance will ignore defense
2 in 4 chance will not deal enough damage
20 ATK + 50 ATK
= 70 ATK / 4
= 17.5 Damage per turn
700 / 17.5
= 40 turns
Knight will gain 400 HP in 40 turns.
1100 / 17.5
= 62.8571428571
That's also ignoring the fact that knight would have gained an additional 228 HP by then.
The Knight will win on average.