Is there something I am not understanding about fire starting?

[๖ۣۜMk137]

Hello,

I tried to start a fire 6 times using a stick, matches, and tinder plug. In all instances my chance of success (according to the game) was 75%. The fire failed to start 5 out of 6 times. If my chance to start a fire was actually 75%, the odds of this happening are insanely low (less than one tenth of one percent).

Is there something I am missing about the % chance the game shows for starting a fire?

Edit: I was wrong. All empirical evidence presented to date suggests that the % chance given by the game is completely accurate. I really did just get extremely unlucky. I hope that this thread will put this issue to rest for another couple years until another unfortunate soul like me comes along. Maybe that person will have experienced 1 in 100,000 odds. :)

Thank you to everyone who contributed to this thread and entertained my suspicions. This is honestly one of the most engaging and intelligent communities I have encountered on steam, and I am really happy to have found this game. Until next time.

[๖ۣۜMk137]

Originally posted by Tekel1959:

You start over every match. Bad runs do happen, and 75% makes no statement about the potential distribution of the expected 25% failures.

Every time is a 75% chance. So yes, the chance of failing a 75% success event five times in a row (which is what happened) is less than one tenth of one percent.

Imagine if you flip a coin five times. 50% chance of heads and 50% chance of tails on each try. But, the probability of getting heads all five times? To calculate that you take (1/2) * (1/2) * (1/2) * (1/2) * (1/2). That equals 1/32.

Likewise, I take the chance of failing the fire, which is 1/4. I go (1/4) * (1/4) * (1/4) * (1/4) * (1/4). That equals 1/1024. I don't think I got those odds (although I guess it IS possible), so I ask if there is something I'm missing.

[IFIYGD]

Your odds of starting a fire are per attempt, not cumulative. So if you have a 75% chance of success, you still have a 25% chance of failure each time. And RNG is a fickle thing. Bad dice rolls happen, and can/do happen one after another in this game. You can have runs of 19 failures followed by runs of 10 successes. We tend to forget the long strings of successes, and remember the long strings of failure.
Chances are per attempt, not cumulative.

If you have multiple failures, i have always found that leaving the stove or campfire location i am trying to start, going elsewhere, and trying somewhere else, or coming back and trying the same location later. never wait to start a fire until your life depends on it, if you can help it. RNG will happily kick your patooty at times like that. The game wants you dead. If you wait until you are already almost dead, it will get its wish most of the time. :)

[๖ۣۜMk137]

Originally posted by IFIYGD:

Your odds of starting a fire are per attempt, not cumulative. So if you have a 75% chance of success, you still have a 25% chance of failure each time. And RNG is a fickle thing. Bad dice rolls happen, and can/do happen one after another in this game. You can have runs of 19 failures followed by runs of 10 successes. We tend to forget the long strings of successes, and remember the long strings of failure.
Chances are per attempt, not cumulative.

If you have multiple failures, i have always found that leaving the stove or campfire location i am trying to start, going elsewhere, and trying somewhere else, or coming back and trying the same location later. never wait to start a fire until your life depends on it, if you can help it. RNG will happily kick your patooty at times like that. The game wants you dead. If you wait until you are already almost dead, it will get its wish most of the time. :)

This doesn't make any sense to me. I understand that the chances are independent of each other. It still happened. 19 failures in a row is so astronomically unlikely that it should never happen in 1000 years, if the chance is truly 75% each time.

https://www.mathsisfun.com/data/probability-events-independent.html

The chance of 19 consecutive failures, by the way, is 0.00000000035%, or (1/285,714,285,714). One in about 286 billion.

[Martial.Lore]

Next time you light a fire, pull a torch from it. Light your next fire with the torch. Lighting a torch never fails and you won't waste matches on failed fire starting.

[IFIYGD]

There are thousands of thread about this over the years, here and on the online forums. No, literally, thousands of threads.
Using cumulative statistics to try to understand static chances does not work. Random Number Generation, which actually *can be* Random, making the type of statistical reasoning you are using essentially useless. The stats you are trying to figure do not take RNG into account at all. Randomness is dynamic, and does not work well with static statistical maths.

If you have multiple failures, leave the location, do something else, and come back to it, and you may get a much better dice roll the next time. Any number larger than 0 = a chance. And if RNG is not smiling on you, you may get that fail chance coming up face-up each time.


Seriously, use the search box on the main forums page and search "Firestarting chance" or similar keywords. This horse has been beaten so many times that horses have outright refused to be added to the game because of it.

[rhwilson1953]

Originally posted by IFIYGD:

<snip>

If you have multiple failures, leave the location, do something else, and come back to it, and you may get a much better dice roll the next time.

<snip>

You don't need to actually leave the location to be successful. After a failure just taking a step to the side or forward/backward can make a difference. I have had 5 failures at lighting the fire in succession and even crouching has worked for me.

[Cal Thang]

Next time, just use a book + firestriker to start a fire

[๖ۣۜMk137]

Originally posted by IFIYGD:

There are thousands of thread about this over the years, here and on the online forums. No, literally, thousands of threads.
Using cumulative statistics to try to understand static chances does not work. Random Number Generation, which actually *can be* Random, making the type of statistical reasoning you are using essentially useless. The stats you are trying to figure do not take RNG into account at all. Randomness is dynamic, and does not work well with static statistical maths.

If you have multiple failures, leave the location, do something else, and come back to it, and you may get a much better dice roll the next time. Any number larger than 0 = a chance. And if RNG is not smiling on you, you may get that fail chance coming up face-up each time.

Seriously, use the search box on the main forums page and search "Firestarting chance" or similar keywords. This horse has been beaten so many times that horses have outright refused to be added to the game because of it.

So are you saying...my chance might not actually be 75% each time based on the RNG seed? That is what I was asking, if there is something about the % chance given that I am not understanding.

[rhwilson1953]

Originally posted by AfLIcTeD:

Originally posted by rhwilson1953:

You don't need to actually leave the location to be successful. After a failure just taking a step to the side or forward/backward can make a difference. I have had 5 failures at lighting the fire in succession and even crouching has worked for me.

Was it taking that step to the side that worked or was it just RNG in your favour?
Doesn't matter if you step to the side, leave the location or do back flips. RNG is RNG.

I was explaining that if you are having successive failures when trying to light a fire then you don't necessarily need to leave the location. Heck, even selecting a different type of tinder or dropping something resets the RNG. Though the RNG is supposed to reset every time you try the same task from the same position and stance.

If you need to get a fire going because your about to die of hypothermia and your fire keeps failing to light, what have you got to lose if you do something as simple as moving one step.

How many people think you have a 50-50 chance at a coin toss is either heads or tails? You can influence the toss if you flip with the coin starting in the same position and use the same amount of force when tossing.

[๖ۣۜMk137]

I’ve been looking through many threads and articles as was suggested and it seems like basically no one actually understands how RNG works in this game. But it quite clearly affects the game in a way that the odds given (in my case 75%) are simply not true to what is happening.

Many many players have experienced this. I would be quite curious to see the numbers from 1000 attempts each by 1000 players. Whatever is going on, the numbers provided by the game are simply wrong in some cases as I suspected, which is why I created this thread.

I understand now that it’s the pseudo-random number generation, but I would quite like to know more about why exactly the RNG causes these statistically impossible outcomes. I will keep looking and share if I find a good explanation from someone else.

Edit: many people are also suggesting that other factors such as fatigue and condition affect the probability, but that it is not reflected in the % chance.

Edit edit: many other people saying that to get some outcome with a small sample size means nothing as long as the outcome approaches the correct % chance with a very large sample size. This is essentially an argument that something like (with coin flips): HHHTHTTTT has the same chance of happening as HHHHHTTTTT and the same chance as THTHTHTHTH. Meaning that if I kept lighting fires I could get FFFFFSSSSSSSSSSSSFFSSFSS.

I think in general there is a lot of confusion from me and everyone else regarding this topic. No one seems able to give any kind of definitive answer, either about probability or the game mechanics.

[konzacelt]

My current personal guide:

95% = 90%
85% = 70%
75% = 50%

This way you're never really let down by the results.

Also, I find it interesting that some people say each successive attempt has nothing to do with the previous. Yet doing something else unrelated can somehow "reset" the bad luck? That is some pro backhand shenanigans if true.

[Nogen]

If a roll a D4 multiple times and get the same number 5 time in a row, I would be not be all over the place, saying that the dice is garbage. I would be a bit surprised, since it is not a common thing, but I would move on. And I am pretty sure you too. It was just bound to happen. It is unlikely, but not that extraordinary.

I don't understand why people think that odds are off. Why a game developers would say that succeeding a certain task is 75% of success, but deep in the code of the game, they fooled us and put the real odd much lower? If they judge that this specific task is 50% of success, why they would tell us otherwise? It doesn't make any sense.

Or what? Hinterland Studio is not able to program a RNG and check the value against a target? They're literally dozens of code to do so to grab right on the internet. Doesn't make any sense either.

So get yourself together and move on.

[Nogen]

I know.

The thing is; pseudo-random generated numbers act, in practice, like true random number for the purpose they're assigned to. Hence the"pseudo" prefix.

Also, it doesn't answer to the interrogation I made in my previous post;

- Why a game developers would lie on the odds?

- Do you think that a team who code an entire game are not able to program a code that can generate a random number (or pseudo-random) that needs to be checked against a certain threshold?

[Imran Khan]

I don't think the % you see is the whole story. You may see a 90% chance to start a fire with a log and it never works. Then change to a stick, also with 90% chance, and it works. I think the wind is a major factor, but doesn't show up in the %.

[๖ۣۜMk137]

Originally posted by Nogen:

If a roll a D4 multiple times and get the same number 5 time in a row, I would be not be all over the place, saying that the dice is garbage. I would be a bit surprised, since it is not a common thing, but I would move on. And I am pretty sure you too. It was just bound to happen. It is unlikely, but not that extraordinary.

I don't understand why people think that odds are off. Why a game developers would say that succeeding a certain task is 75% of success, but deep in the code of the game, they fooled us and put the real odd much lower? If they judge that this specific task is 50% of success, why they would tell us otherwise? It doesn't make any sense.

Or what? Hinterland Studio is not able to program a RNG and check the value against a target? They're literally dozens of code to do so to grab right on the internet. Doesn't make any sense either.

So get yourself together and move on.

I don’t appreciate you implying that I am all broken up about the fire starting odds. I know it’s just a video game and doesn’t matter, but I think it’s an interesting phenomenon and a cool puzzle to solve. I want to understand the game better because it’s a game I love.

Also there is a difference between unlikely and a near impossibility.