Maintenance for the week of April 6:
• PC/Mac: No maintenance – April 6
• PC/Mac: No maintenance – April 6
Lockpicking chance
Bad_Company
✭✭✭✭
Hello everyone,
I'm posting this thread because I think there's something wrong with the lockpicking chance when you try to force the lock. I thought something was off even before the release of TG (and the new lockpicking passive), so I decided to test it. Without any point spent into "Light Fingers" I have 15% chance to force any simple lock. On 25 "simple" chests I forced the locks with the following results:
The mean ratio should be around 6-7 lockpicks:1 simple lock, while the ratio I experienced was more akin to 13-14 lockpicks per lock.
Has anyone else had issues like me? What do you guys think?
I'm posting this thread because I think there's something wrong with the lockpicking chance when you try to force the lock. I thought something was off even before the release of TG (and the new lockpicking passive), so I decided to test it. Without any point spent into "Light Fingers" I have 15% chance to force any simple lock. On 25 "simple" chests I forced the locks with the following results:
- 3 times with 1-2 lockpicks
- 6 times with 3-5 lockpicks
- 5 times with 6-10 lockpicks
- 11 times with 11+ lockpicks (with 1 trials involving the use of 27 lockpicks and 20+ lockpicks for 4 trials)
The mean ratio should be around 6-7 lockpicks:1 simple lock, while the ratio I experienced was more akin to 13-14 lockpicks per lock.
Has anyone else had issues like me? What do you guys think?
My characters (EU PC):
Leopardo Di-Caprio (Khajiit Templar) || Matthew Makehoney (Altmer Sorcerer) || Luck-Luster Burt (Redguard Dragonknight)
Clint Histwood (Argonian Templar) || Martin Uber Ping (Redguard Sorcerer) || Louis Farmstrong (Imperial Nightblade)
Anthony Hotskins (Altmer Nightblade)
0
-
Volkodav✭✭✭✭✭
✭✭I have never had a problem unlocking a lock.Ive lost maybe one during a try,but never more than that.0 -
Lysette✭✭✭✭✭
✭✭✭✭✭Bad_Company wrote: »Hello everyone,
I'm posting this thread because I think there's something wrong with the lockpicking chance when you try to force the lock. I thought something was off even before the release of TG (and the new lockpicking passive), so I decided to test it. Without any point spent into "Light Fingers" I have 15% chance to force any simple lock. On 25 "simple" chests I forced the locks with the following results:- 3 times with 1-2 lockpicks
- 6 times with 3-5 lockpicks
- 5 times with 6-10 lockpicks
- 11 times with 11+ lockpicks (with 1 trials involving the use of 27 lockpicks and 20+ lockpicks for 4 trials)
The mean ratio should be around 6-7 lockpicks:1 simple lock, while the ratio I experienced was more akin to 13-14 lockpicks per lock.
Has anyone else had issues like me? What do you guys think?
Hm, when I take the first 3 with 1.5 each, the second with 4 each, the third with 8 each and the last group with 15, the result is more like 9 something per lock - that is not that far off for a relatively small sample.1 -
Bad_Company✭✭✭✭
Exactly. I think the problem lies with the displayed/intentional chance to force a lock. E.g.: the displayed chance is 15%, but the actual chance is somehow lower. Could this be the case?My characters (EU PC):Leopardo Di-Caprio (Khajiit Templar) || Matthew Makehoney (Altmer Sorcerer) || Luck-Luster Burt (Redguard Dragonknight)Clint Histwood (Argonian Templar) || Martin Uber Ping (Redguard Sorcerer) || Louis Farmstrong (Imperial Nightblade)Anthony Hotskins (Altmer Nightblade)0 -
Bad_Company✭✭✭✭Bad_Company wrote: »Hello everyone,
I'm posting this thread because I think there's something wrong with the lockpicking chance when you try to force the lock. I thought something was off even before the release of TG (and the new lockpicking passive), so I decided to test it. Without any point spent into "Light Fingers" I have 15% chance to force any simple lock. On 25 "simple" chests I forced the locks with the following results:- 3 times with 1-2 lockpicks
- 6 times with 3-5 lockpicks
- 5 times with 6-10 lockpicks
- 11 times with 11+ lockpicks (with 1 trials involving the use of 27 lockpicks and 20+ lockpicks for 4 trials)
The mean ratio should be around 6-7 lockpicks:1 simple lock, while the ratio I experienced was more akin to 13-14 lockpicks per lock.
Has anyone else had issues like me? What do you guys think?
Hm, when I take the first 3 with 1.5 each, the second with 4 each, the third with 8 each and the last group with 15, the result is more like 9 something per lock - that is not that far off for a relatively small sample.
I came up with around 13 lockpicks because of the number of total lockpicks spent (I think it was around 310, more or less).My characters (EU PC):Leopardo Di-Caprio (Khajiit Templar) || Matthew Makehoney (Altmer Sorcerer) || Luck-Luster Burt (Redguard Dragonknight)Clint Histwood (Argonian Templar) || Martin Uber Ping (Redguard Sorcerer) || Louis Farmstrong (Imperial Nightblade)Anthony Hotskins (Altmer Nightblade)0 -
Lysette✭✭✭✭✭
✭✭✭✭✭Bad_Company wrote: »Bad_Company wrote: »Hello everyone,
I'm posting this thread because I think there's something wrong with the lockpicking chance when you try to force the lock. I thought something was off even before the release of TG (and the new lockpicking passive), so I decided to test it. Without any point spent into "Light Fingers" I have 15% chance to force any simple lock. On 25 "simple" chests I forced the locks with the following results:- 3 times with 1-2 lockpicks
- 6 times with 3-5 lockpicks
- 5 times with 6-10 lockpicks
- 11 times with 11+ lockpicks (with 1 trials involving the use of 27 lockpicks and 20+ lockpicks for 4 trials)
The mean ratio should be around 6-7 lockpicks:1 simple lock, while the ratio I experienced was more akin to 13-14 lockpicks per lock.
Has anyone else had issues like me? What do you guys think?
Hm, when I take the first 3 with 1.5 each, the second with 4 each, the third with 8 each and the last group with 15, the result is more like 9 something per lock - that is not that far off for a relatively small sample.
I came up with around 13 lockpicks because of the number of total lockpicks spent (I think it was around 310, more or less).
yeah, that would be 12.4 then.
Edit: it is not that simple to compute, how the outcome has to be distributed, because it is a case of conditional probability. To check it would require to have the exact numbers and compare it with the probability table for any such event and see if the outcome is similar to the theoretical distribution.Edited by Lysette on March 21, 2016 12:44PM0 -
Divinius✭✭✭✭✭
✭Random chance is random.
I took estimates (as Lysette did) based on your numbers, and came up that you are averaging about a 9% chance. Since each attempt needs to be considered separately, you have to divide the successes by the total number of attempts.
You did 25 chests, so that's exactly 25 successes. Since you didn't provide exact numbers I estimated about 248 failures, for a total of 273 attempts. 25/273 = 0.091575, or about 9.16%.
Without exact numbers, there's error in that calculation, but yes, that particular set of chests came up with an average about almost 6% less than the 15% that it indicated.
But again, random is random.
0 -
Nestor✭✭✭✭✭
✭✭✭✭✭Its 15% per attempt not an average of 15% over time. It means your going to fail on average 6 times out of 7, roughly, per attempt. This does not mean that after 6 or 7 attempts your going to force a lock.
What happens on the previous attempt has nothing to do with subsequent attempts, the failure rate is still the same. That can easily add up to 20 or more picks being used on one chest to force a lock.Enjoy the game, life is what you really want to be worried about.
PakKat "Everything was going well, until I died"
Gary Gravestink "I am glad you died, I needed the help"0 -
Lysette✭✭✭✭✭
✭✭✭✭✭Random chance is random.
I took estimates (as Lysette did) based on your numbers, and came up that you are averaging about a 9% chance. Since each attempt needs to be considered separately, you have to divide the successes by the total number of attempts.
You did 25 chests, so that's exactly 25 successes. Since you didn't provide exact numbers I estimated about 248 failures, for a total of 273 attempts. 25/273 = 0.091575, or about 9.16%.
Without exact numbers, there's error in that calculation, but yes, that particular set of chests came up with an average about almost 6% less than the 15% that it indicated.
But again, random is random.
on the other side - the probability that the lock will not break 4 times is 0.85^4 = 52.2% - so about half of the attempts should take at most 5 attempts - that would be 12.5 - but just 9 are actually below that threshold.0 -
Lysette✭✭✭✭✭
✭✭✭✭✭Its 15% per attempt not an average of 15% over time. It means your going to fail on average 6 times out of 7, roughly, per attempt. This does not mean that after 6 or 7 attempts your going to force a lock.
What happens on the previous attempt has nothing to do with subsequent attempts, the failure rate is still the same. That can easily add up to 20 or more picks being used on one chest to force a lock.
while each attempt has a 15% chance, a sequence of attempts follows the rules of conditional probability, which determines the whole outcome of the sequence - read this
https://en.wikipedia.org/wiki/Conditional_probability0 -
Lysette✭✭✭✭✭
✭✭✭✭✭Its 15% per attempt not an average of 15% over time. It means your going to fail on average 6 times out of 7, roughly, per attempt. This does not mean that after 6 or 7 attempts your going to force a lock.
What happens on the previous attempt has nothing to do with subsequent attempts, the failure rate is still the same. That can easily add up to 20 or more picks being used on one chest to force a lock.
while each attempt has a 15% chance, a sequence of attempts follows the rules of conditional probability, which determines the whole outcome of the sequence - read this
https://en.wikipedia.org/wiki/Conditional_probability
Edit: example - the probability of a lock breaking on exactly the 5th attempt is 0.85^4*0.15 = 7.83%0 -
Divinius✭✭✭✭✭
✭
Exactly. Whichever way you calculate it, he got less than the 15% chance for those particular 25 chests.Random chance is random.
I took estimates (as Lysette did) based on your numbers, and came up that you are averaging about a 9% chance. Since each attempt needs to be considered separately, you have to divide the successes by the total number of attempts.
You did 25 chests, so that's exactly 25 successes. Since you didn't provide exact numbers I estimated about 248 failures, for a total of 273 attempts. 25/273 = 0.091575, or about 9.16%.
Without exact numbers, there's error in that calculation, but yes, that particular set of chests came up with an average about almost 6% less than the 15% that it indicated.
But again, random is random.
on the other side - the probability that the lock will not break 4 times is 0.85^4 = 52.2% - so about half of the attempts should take at most 5 attempts - that would be 12.5 - but just 9 are actually below that threshold.
Once can argue that 300 attempts is too few for a statistically accurate result, especially where pseudo-random number generators are concerned, but his "feeling" that he was getting less successes than he should for that test is confirmed.
Repeat the test, record exact numbers for the attempts used for each chest, and calculate it again. The more data you can average, the more likely that your cumulative results are accurate.
That all said, who ever forces chests anyway? Unless your success rate is very high, it's faster just to pick it normally..1 -
Lysette✭✭✭✭✭
✭✭✭✭✭
Exactly. Whichever way you calculate it, he got less than the 15% chance for those particular 25 chests.Random chance is random.
I took estimates (as Lysette did) based on your numbers, and came up that you are averaging about a 9% chance. Since each attempt needs to be considered separately, you have to divide the successes by the total number of attempts.
You did 25 chests, so that's exactly 25 successes. Since you didn't provide exact numbers I estimated about 248 failures, for a total of 273 attempts. 25/273 = 0.091575, or about 9.16%.
Without exact numbers, there's error in that calculation, but yes, that particular set of chests came up with an average about almost 6% less than the 15% that it indicated.
But again, random is random.
on the other side - the probability that the lock will not break 4 times is 0.85^4 = 52.2% - so about half of the attempts should take at most 5 attempts - that would be 12.5 - but just 9 are actually below that threshold.
Once can argue that 300 attempts is too few for a statistically accurate result, especially where pseudo-random number generators are concerned, but his "feeling" that he was getting less successes than he should for that test is confirmed.
Repeat the test, record exact numbers for the attempts used for each chest, and calculate it again. The more data you can average, the more likely that your cumulative results are accurate.
That all said, who ever forces chests anyway? Unless your success rate is very high, it's faster just to pick it normally..
Well, the chance is high that the pseudo-random generator is a "multiply-with-carry" which has a very high performance and a quite long period - so it is unlikely that its results are not like random.
https://en.wikipedia.org/wiki/Multiply-with-carry
I use this kind of random generator because it can produce about 80-100 million numbers per second, if implemented natively - it has a very high reliability.0