Chance-to-Hit vs Chance-to-Damage

[Rigben]

-zombies-
shamblers / no sprinters / str - normal / toughness - normal

-character-
no exp / custom occupation ( no traits ) / 5 fit,5 str / god mode

-Weapons-
M9 9mm - 50 rounds / JS-2000 Shotgun - 50 rounds / no moodle multi / no weather multi

-Other-
Realtime / Debug / 9 AM / No ground shots / Close range / Different locations (new zombies) / My Normal playstyle as a guns enjoyer / no mods

Chance to Damage
M9 - 0 aiming - 9 kills
M9 - 5 aiming - 22 kills (31)
M9 - 10 aiming - 32 kills (63)
Shotgun - 0 aiming - 32 kills (95)
Shotgun - 5 aiming - 35 kills (130)
Shotgun - 10 aiming - 47 kills (177)

Chance to Hit
M9 - 0 aiming - 12 kills
M9 - 5 aiming - 20 kills (32)
M9 - 10 aiming - 42 kills (74)
Shotgun - 0 aiming - 52 kills (126)
Shotgun - 5 aiming - 52 kills (178)
Shotgun - 10 aiming - 55 kills (233)

I attempted to be as un-biased as possible in these test. As someone who loves guns in Zomboid my end goal is simply enjoyment. While I did attempt to line up multi-shots for the shotgun they were pretty rare, as these tests were performed at a closer range then needed to regularly get multi-hits.

P.S.
You are not supposed to start the game with the pistol, in b41 you always wanted to start with a shotgun.

P.P.S.
Anyone can replicate my tests and do it however many times they want. I have provided all the settings and information required to do what I did short of my hand on your mouse.

If someone out there wants to provide better information I encourage it and am interested in the results from someone who can do all that stuff proper. This is literally just a Zomboid player who likes shooting stuff comparing old vs new.

[Skedastic]

Interesting. What are the numbers in parentheses?

[Rigben]

Originally posted by Skedastic:

Interesting. What are the numbers in parentheses?

Those numbers are the overall total kills for that character. This was done on 2 different saves so I used the ingame zombie kills number to keep track.

[FCino]

waiting Rainmaker comment before I give my opinion...

[Kaia]

Interesting post. But the sample size seems very low to begin with. Beyond that, i fundamentally think that the concept of a chance to damage is, on a game design level, inherently much less fun than the chance to hit.

Missing always feels meh, but your brain can connect dots and assume you just didn't aim right or you were too panicked/hurt/sleepy to make the shot.

Not damaging is worse because it just feels like BS.

[Rigben]

Originally posted by Kaia:

Interesting post. But the sample size seems very low to begin with. Beyond that, i fundamentally think that the concept of a chance to damage is, on a game design level, inherently much less fun than the chance to hit.

Missing always feels meh, but your brain can connect dots and assume you just didn't aim right or you were too panicked/hurt/sleepy to make the shot.

Not damaging is worse because it just feels like BS.

A sample size of 50 is average for most things not related to population.

[LeFrench]

i prefer the new chance to damage. i was hitting a lot. i was likely doing minimal damage most of the time, but its fun to hit. and i felt like i was hitting based on my putting the aiming reticule correctly. and one final point, if you miss in the old system.... you don't stagger a zombie.... if you "miss" in the new system you do stagger the zombie.

so.... this new system, despite the seemingly lower kills that you sample shows, felt a lot better.

plus... its more realistic, right? missing a lot seems a bit stupid considering how slow most zombies are. makes more sense, you'd hit a lot, but hit non-vital spots, staggering the zombie and finally you land that head shot.

[Kaia]

Originally posted by Rigben:

Originally posted by Kaia:

Interesting post. But the sample size seems very low to begin with. Beyond that, i fundamentally think that the concept of a chance to damage is, on a game design level, inherently much less fun than the chance to hit.

Missing always feels meh, but your brain can connect dots and assume you just didn't aim right or you were too panicked/hurt/sleepy to make the shot.

Not damaging is worse because it just feels like BS.

A sample size of 50 is average for most things not related to population.

It most definitely isn't average for anything pertaining to statistics, whether it is in population or game mechanics.

[Copingway]

from your testing, it seems they unintentionally nerfed firearm? Why would this happen? From what I understand this change should not affect the effectiveness of firearm, especially shotgun, which is always one-shot kill.
If this means a shotgun shot at close range will not always kill a zombie then this is very bad.

[Rigben]

Originally posted by Kaia:

Originally posted by Rigben:

A sample size of 50 is average for most things not related to population.

It most definitely isn't average for anything pertaining to statistics, whether it is in population or game mechanics.

Well, I am not really educated in math stuff so I just got my information from a Universities website. Technically it would be 100 cause of M9 / shotgun at matching aiming levels.

"For a sample size of ten, the result is not statistically significant. However, as the sample size increases, the confidence intervals narrow. Once the sample size is 50, the null hypothesis falls outside the interval - the result is statistically significant. But here the effect estimated - the correlation - has exactly the same magnitude. The studies haven't estimated a different effect; rather they've estimated the same effect with different precision. The different sample sizes here mean that there is different potential for finding small P-values."

Unlike most people my opinions are not set in stone so if someone cause prove me wrong I will likely change my stance. Right now though as someone who likes shooting in Zomboid the new system "feels" like a nerf to my playstyle.

[MrMinMax]

So if I am aiming for the head, I will hit the head, but then "chance to damage" decides if the damage actually happened?

[Rigben]

Originally posted by MrMinMax:

So if I am aiming for the head, I will hit the head, but then "chance to damage" decides if the damage actually happened?

They are two different sandbox settings, chance to damage (new/default) and chance to hit (old).

Chance to damage means you will basically never "miss" if you are aiming anywhere near a zombie but it has a chance to just stagger them for no damage.

Chance to hit means a miss is a miss but for me often means if I hit I kill. I always aim for the head of what I am shooting and with melee aswell. So giving an opinion on hit location on zombies based on crosshair I don't have any real info on aside from I get alot of headshots.

[Nordil(Hun)]

Originally posted by Rigben:

Originally posted by Kaia:

It most definitely isn't average for anything pertaining to statistics, whether it is in population or game mechanics.

Well, I am not really educated in math stuff so I just got my information from a Universities website. Technically it would be 100 cause of M9 / shotgun at matching aiming levels.

"For a sample size of ten, the result is not statistically significant. However, as the sample size increases, the confidence intervals narrow. Once the sample size is 50, the null hypothesis falls outside the interval - the result is statistically significant. But here the effect estimated - the correlation - has exactly the same magnitude. The studies haven't estimated a different effect; rather they've estimated the same effect with different precision. The different sample sizes here mean that there is different potential for finding small P-values."

Unlike most people my opinions are not set in stone so if someone cause prove me wrong I will likely change my stance. Right now though as someone who likes shooting in Zomboid the new system "feels" like a nerf to my playstyle.

This actually does not work like that:)

The thing you could calculate from this is actually the chance to get this kind of result, if you know the "hit chances". Which if i recall well for aiming skill level 0 is 50%.
The hit chance of 50% (if no other malus or bonus is affecting), means that if we would be shooting infinitely then the hit chance would be gathering around 50%.

What you could do though, is calculate different distribution chances for your shots.
Like how likely it is that you hit 40 times out of 50 shots when the hit chance is 50%.
(With 4 out of 5 shots:
0-1-1-1-1 1-1-1-1-1
1-0-1-1-1 0-0-1-1-1
1-1-0-1-1 0-0-0-1-1
1-1-1-0-1 0-0-0-0-1
1-1-1-1-0 0-0-0-0-0
The first column is the good one so you have only 5 options, out of 2^5. If you want to shoot at least 4 times succesesful then you have 6 options. If my calculations are correct).

[Rigben]

Originally posted by Nordil(Hun):

Originally posted by Rigben:

Well, I am not really educated in math stuff so I just got my information from a Universities website. Technically it would be 100 cause of M9 / shotgun at matching aiming levels.

"For a sample size of ten, the result is not statistically significant. However, as the sample size increases, the confidence intervals narrow. Once the sample size is 50, the null hypothesis falls outside the interval - the result is statistically significant. But here the effect estimated - the correlation - has exactly the same magnitude. The studies haven't estimated a different effect; rather they've estimated the same effect with different precision. The different sample sizes here mean that there is different potential for finding small P-values."

Unlike most people my opinions are not set in stone so if someone cause prove me wrong I will likely change my stance. Right now though as someone who likes shooting in Zomboid the new system "feels" like a nerf to my playstyle.

This actually does not work like that:)

The thing you could calculate from this is actually the chance to get this kind of result, if you know the "hit chances". Which if i recall well for aiming skill level 0 is 50%.
The hit chance of 50% (if no other malus or bonus is affecting), means that if we would be shooting infinitely then the hit chance would be gathering around 50%.

What you could do though, is calculate different distribution chances for your shots.
Like how likely it is that you hit 40 times out of 50 shots when the hit chance is 50%.
(With 4 out of 5 shots:
0-1-1-1-1 1-1-1-1-1
1-0-1-1-1 0-0-1-1-1
1-1-0-1-1 0-0-0-1-1
1-1-1-0-1 0-0-0-0-1
1-1-1-1-0 0-0-0-0-0
The first column is the good one so you have only 5 options, out of 2^5. If you want to shoot at least 4 times succesesful then you have 6 options. If my calculations are correct).

If only this actually made sense to me heh. I did turn off everything that could externally effect aiming so the numbers I did provide are as reliable as I could make them. Would a sample size over 50 rounds per gun per aiming skill change things? I dunno, maybe someone who has more knowledge and cares more will give us different information.

[Nordil(Hun)]

Originally posted by Rigben:

Originally posted by Nordil(Hun):

This actually does not work like that:)

The thing you could calculate from this is actually the chance to get this kind of result, if you know the "hit chances". Which if i recall well for aiming skill level 0 is 50%.
The hit chance of 50% (if no other malus or bonus is affecting), means that if we would be shooting infinitely then the hit chance would be gathering around 50%.

What you could do though, is calculate different distribution chances for your shots.
Like how likely it is that you hit 40 times out of 50 shots when the hit chance is 50%.
(With 4 out of 5 shots:
0-1-1-1-1 1-1-1-1-1
1-0-1-1-1 0-0-1-1-1
1-1-0-1-1 0-0-0-1-1
1-1-1-0-1 0-0-0-0-1
1-1-1-1-0 0-0-0-0-0
The first column is the good one so you have only 5 options, out of 2^5. If you want to shoot at least 4 times succesesful then you have 6 options. If my calculations are correct).

If only this actually made sense to me heh. I did turn off everything that could externally effect aiming so the numbers I did provide are as reliable as I could make them. Would a sample size over 50 rounds per gun per aiming skill change things? I dunno, maybe someone who has more knowledge and cares more will give us different information.

Okay let me rephrase it:)
M9 - 0 aiming - 9 kills

The first thing is are these 9 kills from 9 shots? Or are they from say 24 shots?
Because now we would need to add another variable.
0-does not hit
1- hits.

This way we could determine what is the chance of having 24 shots hit their target out of 50 shots. (There is a formula for this, but i don't know it by heart anymore:P)
Then we would need to make a separate calculation for the HITS.
How many were killing shots?
(So out of the 24 we would have 15 non killing shots and 9 killing shots. This would allow us to calculate what the amount of bullets was required to kill the zombies.
Now we could a) calculate what hit and kill ratios will give us the same killing ratio, as in what are the different kind of distributions, or we could
b) calculate how much chance is there to get this exact kind of hit and kill ratio:P)

[Rigben]

Originally posted by Nordil(Hun):

Originally posted by Rigben:

If only this actually made sense to me heh. I did turn off everything that could externally effect aiming so the numbers I did provide are as reliable as I could make them. Would a sample size over 50 rounds per gun per aiming skill change things? I dunno, maybe someone who has more knowledge and cares more will give us different information.

Okay let me rephrase it:)
M9 - 0 aiming - 9 kills

The first thing is are these 9 kills from 9 shots? Or are they from say 24 shots?
Because now we would need to add another variable.
0-does not hit
1- hits.

This way we could determine what is the chance of having 24 shots hit their target out of 50 shots. (There is a formula for this, but i don't know it by heart anymore:P)
Then we would need to make a separate calculation for the HITS.
How many were killing shots?
(So out of the 24 we would have 15 non killing shots and 9 killing shots. This would allow us to calculate what the amount of bullets was required to kill the zombies.
Now we could a) calculate what hit and kill ratios will give us the same killing ratio, as in what are the different kind of distributions, or we could
b) calculate how much chance is there to get this exact kind of hit and kill ratio:P)

Ah ok I did not track things that deep. Basically I gave myself a m9 with 3 mags + 5 rounds / shotgun with 50 rounds moved to a new location for each 50 and played as I would normally. The idea to doing it this way was to simulate my actual gameplay (actually aiming with the intent to kill not just blasting). In regards to the ranges for this test they would be approx 5-8 feet distance (when they start the fast shamble) for the sake of consistency and not wasting bullets.

So how does that work out for the shotgun seeing as those numbers to me are the more informative ones? I just threw the M9 in there because they start you with it currently.