| Author |
Thread Statistics | Show CCP posts - 0 post(s) |
Mangold
Freelance Unincorporated
Ushra'Khan
   |
Posted - 2006.11.15 12:37:00 -
[1]
 Originally by: Trem Sinval
The numbers, Mr. Sulu!
Crappiest racial jammer: 6.
Vindi sensors: 24.
Chance to jam (any cycle): 6/24, 1/4, 25%
Chance to jam 4 consecutive times: (1/4)^4 = 0.39% chance
-----
- Trem
Erm, your calculations are not correct. Running 4 jammers on a single target (with your numbers) will give a 68% chance of jamming every time. That is the same as a 32% chance of failure.
Fitting a ECCM (60% added sensor strength) will change that to 49% chance of successful jam i e 51% chance of failure.
|
Mangold
Freelance Unincorporated
Ushra'Khan
|
Posted - 2006.11.15 23:05:00 -
[2]
Originally by: Trem Sinval
Edited by: Trem Sinval on 15/11/2006 22:22:56
Originally by: Mangold
Erm, your calculations are not correct. Running 4 jammers on a single target (with your numbers) will give a 68% chance of jamming every time. That is the same as a 32% chance of failure.
4 jammers vs. 1 ship:
Each jammer having a 1 in 4 chance to hit the jam, statistically speaking at least one of your jammers should keep the target jammed all the time (the extrapolated probability being (1/4)*4, or 100%). If you can take 4 stabs at a contest in which every fourth person wins, you'll always win.
No. This is where you are wrong. The other calculations may be correct (haven't looked at them tbh and I don't belive there is anything wrong with them).
You cant look at it that way. Compare it to throwing snowballs at someone. If the first snowball has a 25% of hitting and the 2nd, the 3rd and the 4th one has the same probability how big is the chance that one of them hits the target? It's not 100%.
|
Mangold
Freelance Unincorporated
Ushra'Khan
|
Posted - 2006.11.15 23:18:00 -
[3]
Edited by: Mangold on 15/11/2006 23:28:47
Originally by: Locke DieDrake
Originally by: Mangold
Originally by: Trem Sinval
Edited by: Trem Sinval on 15/11/2006 22:22:56
Originally by: Mangold
Erm, your calculations are not correct. Running 4 jammers on a single target (with your numbers) will give a 68% chance of jamming every time. That is the same as a 32% chance of failure.
4 jammers vs. 1 ship:
Each jammer having a 1 in 4 chance to hit the jam, statistically speaking at least one of your jammers should keep the target jammed all the time (the extrapolated probability being (1/4)*4, or 100%). If you can take 4 stabs at a contest in which every fourth person wins, you'll always win.
No. This is where you are wrong. The other calculations may be correct (haven't looked at them tbh and I don't belive there is anything wrong with them).
You cant look at it that way. Compare it to throwing snowballs at someone. If the first snowball has a 25% of hitting and the 2nd, the 3rd and the 4th one has the same probability how big is the chance that one of them hits the target? It's not 100%.
Um, yes it is.
If I have a 1 in 4 chance of hitting, and four chances to try it. Then I have a 100% chance of hitting.
But here is where those numbers break apart. Just because I have a 100% chance to hit, doesn't mean that I will actually hit.
Probabilities are like that.
Probabilities are difficult. Let me try to explain my point with jamming.
If you look at the lottery with 4 tickets as an example you are not allowed to buy all 4. You can buy one ticket for one lottery (with 25% chance of winning) and another ticket at another lottery (with another 25% chance of winning the OTHER lottery). Keep buying tickets in 4 different lotteries and you wont have a 100% chance of winning one of them. You will still only have a 25% chance of winning each of them. You will actually have a 75% chance of not each winning each one of them.
Lets look at the numbers: 4 times with 75% of failure that will equal 0.75^4 = 31% chance of failure (i e not winning in any of the 4 lotteries) which is the same thing as 69% chance of winning one of the lotteries if you bought 1 ticket in all 4.
|
Mangold
Freelance Unincorporated
Ushra'Khan
|
Posted - 2006.11.16 09:30:00 -
[4]
Originally by: dalman
You obviously doesn't understand statistics at all.
"4 jammers means 4 tickets in a single lottery (the lottery necessarily having 4 possible outcomes, 3 of which do not result in a jam). "
This part is what cause your failure.
Because it is not correct.
A failed jam doesn't reduce the strength of a ship. A raven has a strength of 22, no matter.
If your first 3 jammers fail, the ship still has a 22 strength and you only have a 25% chance of jamming the target.
Even if you have 8 jammers fitted, there's still a 0.75^8=10% chance that none of them will jam the target.
The chance does NOT approach 100% as we approach infinity. It approach the distribution I've explained already. The average number of jams does though. Which is very different.
Like I said. Probabilities are difficult. dalman obviously has some knowledge on this topic.
What i tried to explain earlier was that no matter how many jammers you throw at a ship you wont get a 100% chance of jamming. There will always be a chance of failure. The risk of failure will be smaller the more jammers you use but it will still be there.
The example with a lottery with 4 tickets is flawed as you cant buy all 4 tickets in the same lottery. Let me try to explain.
Jammer 1 = lottery 1.
You buy a ticket with 25% of winning that lottery. You cant buy another ticket in it as it would be the same thing as activating your jammer on a target and realising that it's not working and then magically travel back in time and try another time. Each jammer is a new lottery.
|
| |
|