After deadly flooding in Texas, Republicans used an emergency legislative session called to tackle the state’s most pressing issue: where trans people pee.
They’re trying to use attacking trans people just minding our business to cover up GOP having sex with kids while letting dozens of other kids get murdered by psychos or murdered by mismanaged disasters. Tx is perfectly representative of entire GOP - unaccountable, useless, and godless scum
Laws like this are doomed. “You Can Aways Tell!” Right, silly of me to overlook that. I’d like to introduce you to an interesting bit of statistics called Bayes Theorem.
There is a famous theorem that connects conditional probabilities of two events. It's named Bayes' theorem, and the formula is as follows:
P(A|B) = P(B|A) * P(A) / P(B)
You can ask a question: "What is the probability of A given B if I know the likelihood of B given A?". This theorem sometimes provides surprising and unintuitive results. The most commonly described examples are drug testing and illness detection, which has a lot in common with the relative odds of finding some unique state in the population. Let's stick to the second one.
In a group of 1000 people, perhaps 10 of them are transgender persons. Everybody is transvestigated using some magical technique, which somehow shows the actual result in 95% of cases, an absurdly high accuracy, but there’s a point to this. Next, let’s find the probability of a person being actually transgender if their magical investigation result is positive.
Without thinking, you may predict, by intuition, that the result should be around 90%, right? Let's make some calculations and estimate the correct answer.
1. We will use a notation: C – cisgender, T – identified as transgender, + – test positive, - – test negative.
2. Rewrite information from the text above in a way of probabilities: P(C) = 0.99, P(T) = 0.01, P(+|T) = 0.95, P(-|T) = 0.05, P(+|C) = 0.05, P(-|C) = 0.95.
3. Work out the total probability of a test to be positive: P(+) = P(+|T) * P(T) + P(+|C) * P(C) = 0.95 * 0.01 + 0.05 * 0.99 = 0.059.
4. Use the Bayes' theorem to find the conditional probability P(T|+) = P(+|T) * P(T) / P(+) = 0.95 * 0.01 / 0.059 = 0.161.
Hmm... About 16%, not 90%! it isn't that high, is it? It turns out that this kind of paradox appears if there is a significant imbalance between the number of people in two distinct groups. Trans people are relatively rare, altering the probabilities in a very non-intuitive way.
Even if the “Transvestigators” have a 95% accuracy in identifying individuals as trans or cis, trans folks are uncommon, and they will be wrong 5 out of 6 guesses.
Enforcing laws aimed at transgender folks is going to be hard, with lots of false arrests. Your attempts at restriction will sweep up more folks who are not transgender than are, by a substantial margin. There will be litigation. There are so very many underemployed lawyers in the United States, and they will be delighted to go for 30% of a sure thing.
I’m just an old engineer. Understanding how rare or low probability events interact was part of the job. Often people, especially legislators and politicians, don’t understand how various statistics can combine to produce extremely non-intuitive results. The case of trying to spot rare items, even with a very accurate detection system amplifies the “mis-detections” of non-rare items as being rare. With a less-accurate “rare ones” detector, the actual rare ones vanish into the vastly larger population of mis-detections.
This leads to lots of women being accused of the terrible offense of “peeing while trans”, and suffering verbal or physical abuse, in spite of their not being trans.
Another way, a bit less mathematical but expressing the same idea, is that we have to allow for the effect of wrong guesses, both false positives (cis folks ID'ed as trans) and false negatives (trans folks ID'ed as cis).
No matter the rate of false IDs, because there are far more cis folks than trans folks, the false positives will swamp the false negatives, making correct IDs over the whole group far less likely. Which is reflected in the calculations above.
There will be a chance for public comments before/during the committee hearing for the Texas House version of the bill (or they may use the same bill). I think the online comment form is restricted to Texans. I need to check on what has been scheduled about this, but haven't yet. If you are in Texas, call your state rep. I pointed out to my state senator that this was going to lead to vigilantes deciding that people were breaking the law.
Gender-nonconforming cis people will be targets of harassment due to these bans a long time before many trans people will. There are many more such cis people than trans, and people who are trans (by contrast) can often be very gender-conforming and blend in relatively easy. Which raises the key point: bathroom bans are as ludicrous as they are cruel. Any mechanism at enforcement is just going to cause huge problems. I guess everyone will need to carry their birth certificate with them when they’re out in public in Texas now?
What else would you expect from Texas- the Texas republicans will everything and anything to hurt all Texans from kids to adults - the only thing they care about is power - if Texans want a better government they need to rise up protest in big numbers and VOTE THEM OUT
They’re trying to use attacking trans people just minding our business to cover up GOP having sex with kids while letting dozens of other kids get murdered by psychos or murdered by mismanaged disasters. Tx is perfectly representative of entire GOP - unaccountable, useless, and godless scum
Laws like this are doomed. “You Can Aways Tell!” Right, silly of me to overlook that. I’d like to introduce you to an interesting bit of statistics called Bayes Theorem.
There is a famous theorem that connects conditional probabilities of two events. It's named Bayes' theorem, and the formula is as follows:
P(A|B) = P(B|A) * P(A) / P(B)
You can ask a question: "What is the probability of A given B if I know the likelihood of B given A?". This theorem sometimes provides surprising and unintuitive results. The most commonly described examples are drug testing and illness detection, which has a lot in common with the relative odds of finding some unique state in the population. Let's stick to the second one.
In a group of 1000 people, perhaps 10 of them are transgender persons. Everybody is transvestigated using some magical technique, which somehow shows the actual result in 95% of cases, an absurdly high accuracy, but there’s a point to this. Next, let’s find the probability of a person being actually transgender if their magical investigation result is positive.
Without thinking, you may predict, by intuition, that the result should be around 90%, right? Let's make some calculations and estimate the correct answer.
1. We will use a notation: C – cisgender, T – identified as transgender, + – test positive, - – test negative.
2. Rewrite information from the text above in a way of probabilities: P(C) = 0.99, P(T) = 0.01, P(+|T) = 0.95, P(-|T) = 0.05, P(+|C) = 0.05, P(-|C) = 0.95.
3. Work out the total probability of a test to be positive: P(+) = P(+|T) * P(T) + P(+|C) * P(C) = 0.95 * 0.01 + 0.05 * 0.99 = 0.059.
4. Use the Bayes' theorem to find the conditional probability P(T|+) = P(+|T) * P(T) / P(+) = 0.95 * 0.01 / 0.059 = 0.161.
Hmm... About 16%, not 90%! it isn't that high, is it? It turns out that this kind of paradox appears if there is a significant imbalance between the number of people in two distinct groups. Trans people are relatively rare, altering the probabilities in a very non-intuitive way.
Even if the “Transvestigators” have a 95% accuracy in identifying individuals as trans or cis, trans folks are uncommon, and they will be wrong 5 out of 6 guesses.
Enforcing laws aimed at transgender folks is going to be hard, with lots of false arrests. Your attempts at restriction will sweep up more folks who are not transgender than are, by a substantial margin. There will be litigation. There are so very many underemployed lawyers in the United States, and they will be delighted to go for 30% of a sure thing.
I really, really appreciate your laying out the math of this.
I’m just an old engineer. Understanding how rare or low probability events interact was part of the job. Often people, especially legislators and politicians, don’t understand how various statistics can combine to produce extremely non-intuitive results. The case of trying to spot rare items, even with a very accurate detection system amplifies the “mis-detections” of non-rare items as being rare. With a less-accurate “rare ones” detector, the actual rare ones vanish into the vastly larger population of mis-detections.
This leads to lots of women being accused of the terrible offense of “peeing while trans”, and suffering verbal or physical abuse, in spite of their not being trans.
Another way, a bit less mathematical but expressing the same idea, is that we have to allow for the effect of wrong guesses, both false positives (cis folks ID'ed as trans) and false negatives (trans folks ID'ed as cis).
No matter the rate of false IDs, because there are far more cis folks than trans folks, the false positives will swamp the false negatives, making correct IDs over the whole group far less likely. Which is reflected in the calculations above.
What can we do to address this horrible action?
There will be a chance for public comments before/during the committee hearing for the Texas House version of the bill (or they may use the same bill). I think the online comment form is restricted to Texans. I need to check on what has been scheduled about this, but haven't yet. If you are in Texas, call your state rep. I pointed out to my state senator that this was going to lead to vigilantes deciding that people were breaking the law.
HB32 is the companion bill to SB7. It's listed as identical. I don't see it scheduled for a hearing yet.
Gender-nonconforming cis people will be targets of harassment due to these bans a long time before many trans people will. There are many more such cis people than trans, and people who are trans (by contrast) can often be very gender-conforming and blend in relatively easy. Which raises the key point: bathroom bans are as ludicrous as they are cruel. Any mechanism at enforcement is just going to cause huge problems. I guess everyone will need to carry their birth certificate with them when they’re out in public in Texas now?
"Never let a crisis go to waste..."
*𝘥𝘪𝘴𝘵𝘢𝘯𝘵 𝘴𝘰𝘶𝘯𝘥𝘴 𝘰𝘧 𝘷𝘰𝘮𝘪𝘵𝘪𝘯𝘨, 𝘴𝘤𝘳𝘦𝘢𝘮𝘪𝘯𝘨, 𝘢𝘯𝘥 𝘩𝘢𝘪𝘳 𝘵𝘦𝘢𝘳𝘪𝘯𝘨 𝘰𝘶𝘵*
The very shock of using a session for flood relief for this crap…
What else would you expect from Texas- the Texas republicans will everything and anything to hurt all Texans from kids to adults - the only thing they care about is power - if Texans want a better government they need to rise up protest in big numbers and VOTE THEM OUT
When I began writing Fixing the System about teens fighting a bathroom ban I didn't expect it to be so relevant.
Yes, it does seem like the elected Republicans are using a non-issue to further deflect from their own failures! So disgusting.
Nothing but dumb red meat for dumb red voters.
TX republican officials and lawmakers know that such a law is extremely hard to enforce and leads to even worse optics, as has been the case in FL
These pols need to get a life!