There is actually an entire industry focused on testing security measures to ensure they work. It is called penetration testing. For something like the TSA, they’ll do audits where test passengers are sent through with contraband. Sort of like secret shoppers who evaluate a retail store by pretending to shop there. In one particular audit, they only successfully caught 3 out of 70. Some audits estimate a 95% failure rate.
The audits have consistently found that TSA’s catch rates are lower than random searches, by a wide margin. As in, they’d be better off not searching everyone, and just doing randomized searches on ~10-15% of passengers. That random “10-15% of passengers get a full search” system would catch more than the current “search everyone but miss 95% of contraband” system.
They could literally just roll a d8 die for each passenger in the line, and on a 1 they initiate a full search. And that would be more effective than their current methods.
Here is a write-up about one of the old 2015 audits. And here is one from 2017. And it’s worth noting that new audit results aren’t readily available, because the TSA started classifying their results around 2017 instead of releasing the numbers, when David Pekoske was installed as administrator. Because that definitely screams “our numbers are improving!”
Basically, a thorough search of 10-15% of passengers would more accurately catch threats, when compared to searching everyone with a ~95% miss rate. There are even systems designed to randomly select people for searches. Usually used in jobs where employees are subject to searches/drug tests as they’re arriving/leaving. For instance, if a company needs to drug test 5% of their employees every day, they can set their random selector to ping on 5% of people as they’re arriving.
They’re usually triggered automatically by walking across a mat, by employees badging through a controlled access door, or via a button push as security buzzes you in. But it could also be configured to be triggered based on something like ticket scans for passengers. Passengers get their ticket scanned, the random selector system randomly selects the pre-programmed percentage of passengers, and they’re the only ones who get pulled aside. All the rest are free to continue to their gate as usual. That way there are no accusations of random searches being discriminatory, because the random selector system is doing the choosing based on the defined percentage.
There is actually an entire industry focused on testing security measures to ensure they work. It is called penetration testing. For something like the TSA, they’ll do audits where test passengers are sent through with contraband. Sort of like secret shoppers who evaluate a retail store by pretending to shop there. In one particular audit, they only successfully caught 3 out of 70. Some audits estimate a 95% failure rate.
The audits have consistently found that TSA’s catch rates are lower than random searches, by a wide margin. As in, they’d be better off not searching everyone, and just doing randomized searches on ~10-15% of passengers. That random “10-15% of passengers get a full search” system would catch more than the current “search everyone but miss 95% of contraband” system.
They could literally just roll a d8 die for each passenger in the line, and on a 1 they initiate a full search. And that would be more effective than their current methods.
Sorry, but that doesn’t sound even remotely realistic. Source?
Here is a write-up about one of the old 2015 audits. And here is one from 2017. And it’s worth noting that new audit results aren’t readily available, because the TSA started classifying their results around 2017 instead of releasing the numbers, when David Pekoske was installed as administrator. Because that definitely screams “our numbers are improving!”
Basically, a thorough search of 10-15% of passengers would more accurately catch threats, when compared to searching everyone with a ~95% miss rate. There are even systems designed to randomly select people for searches. Usually used in jobs where employees are subject to searches/drug tests as they’re arriving/leaving. For instance, if a company needs to drug test 5% of their employees every day, they can set their random selector to ping on 5% of people as they’re arriving.
They’re usually triggered automatically by walking across a mat, by employees badging through a controlled access door, or via a button push as security buzzes you in. But it could also be configured to be triggered based on something like ticket scans for passengers. Passengers get their ticket scanned, the random selector system randomly selects the pre-programmed percentage of passengers, and they’re the only ones who get pulled aside. All the rest are free to continue to their gate as usual. That way there are no accusations of random searches being discriminatory, because the random selector system is doing the choosing based on the defined percentage.
Fine