Thinking on this some more...
L5R 5E duelist... fire ring 3, MA-Melee 3.
Success chances - needs two successes on 3 kept dice, with each strain marker also counting a success.
Potential to hit = infinite successes; if iterating a maximum of two explosions deep, that's a potential for 18s Max armor 6; 4 is normal for bushi, especially starting.
Black die success count in fire [0, 1, 0, 2, 1, 2*]= 6/6= expectation of 1 expected success
White [0,0,0,0,0,2,2,1,1,1,2*,1*]= 10/12. So we can expect 3 kept successes between success circles and strife flames.
When we count the explosions we add 1/6 and 1/36 to the black, for 7/6+1/36= 43/36 expectation., or 1 7/36.
White with explosions starts at 5/6, adds 5/36 and 5/216, so (180+30+5)/216= 215/216... So, we expect about 3 2/3 successes. On the correct Iajutsu, this means 5+2 damage, Not gonna be killing most of the time.
But there's an infinity there. It is unlikely for a starting PC to have maxed 12 fatigue, but it happens. So... on a low probability roll, it's possible to get up over 16 damage and zero out a PC on initial hit. I've seen it happen in duels.
(non-fire stance, those are B=[0,0,0,1,1,1*]=3/6 +3/36+3/216 and W=[0,0,0,0,0,1,1,1,1,1,1*,1*] =7/12+7/72+7/432 =252/432+42/432+7/432 =301/432 expected.
Now, in third ed...
a good starting character can be rolling 7k4 d10's, exploding, to hit... which Anydice can't calculate with the exploding.
TNTBH peaks for starting characters at about 35, so we'll use that. He's got a 10% chance of hitting himself if in full armor. If, however, it's a courtly duel... He's a 20 to be hit
Code:
output [highest 4 of 7d10]>35
output [highest 4 of 7d10]>20
10.10% and 95.3%, before explosions
Median damage is about 24, and the bushi is Earth 3... so about 4 lines gone. Still up. Peak is 30 without explosions. chance of instakill is about 5%... but exact calculations are a royal nightmare thanks to both XkY and exploding dice.
Both games have the potential to instakill their mirror, but it's not exactly likely. On the other hand....
A rank 4 Bayushi Saboteur in 3rd ed was able to get a rolled 10k10+55 to hit.... And Amaterasu is flat 10's for all rings, so was 75 TNTBH (extra 25 for kami in splendor)... He whacked Amaterasu-Kami from behind, with several free raises and a couple called raises to raise damage. So... in that campaign, Amaterasu was indeed replaced by a mortal, but it was a Bayushi Saburō. (Given the chance to guard the black scrolls, he decided to read them. «bleep»ing «bleep»-head Bayushi... wonderfully played in character for both PC and clan/school/family/setting.) He two-shotted her. KO strike 1, dead strike 2.
Now, in both editions, little nasty bakemono goblins are individually pretty wimpy... but they add together quite fast, and 3:1 odds are often dead PCs. (Done that in 5th ed more than once. But it wasn't exactly my fault - group was divided between "Let's fight" and "Take flight" - and the one "let's fight fought and died.
It's worth noting that PCs in 5th ed are not directly comparable to typical foes...