The Worst Math Ever Used In Court

[@ShortHax]

Numbers don’t lie. People misusing numbers do

[@camicus-3249]

-There are lies, damned lies, and statistics.- There are liars, damned liars, and statisticians. (Sorry, statisticians lol)

[@alex.g7317]

Woah! A popular commenter with only one comment? Count me in!

[@prim16]

The numbers don't lie. And they spell disaster for you at SACRIFICE

[@stevethecatcouch6532]

Or, to coin a phrase, "figures don't lie, but liars figure."

[@kameyu]

Not to mention court normally uses evidences, not added probabilities. That's the tough reason why lots of rapists are still free today.

[@bobh6728]

Take the judge in the case. Take the chances of having his first name, last name, college attended, wife’s name, number of children, and years as a judge and you can “prove” he does not exist because the probability is so small that all of those characteristics exist in one person!!!!

[@kwxjibo]

Yes. Very unique situations and combinations of settings and events occur all the time, that's just how life is. Maths and odds can't actually prove anything in a situation like that.

[@HomicidalTh0r]

Hard to imagine this judge was real. A judge letting something like this fly in a courtroom seems mathematically improbable!

[@Stratelier]

@@kwxjibo Isn't there a known term for this? The "lottery paradox" or somesuch?

[@trspanda2157]

Jesus loves us all that's why he died for our sins,

[@HomicidalTh0r]

@@trspanda2157 Who's that?

[@pbs36]

This actually falls into a larger category, which is prosecutors and lawyers exploiting other peoples' ignorance (jurors, etc.). It can be a bluff (the prosecutor was spewing math knowledge he knew he didn't understand himself and could be incorrect), or they're convinced they know what they're talking about and that whoever they're trying to convince knows less than them.

[@AlDunbar]

Showing the errors made would be the job of defense counsel (a lawyer). If there is insufficient applicable case law, the lawyer should get a mathematician to testify.

[@normalchannel2185]

totally true. A judge probably left math after 10th (or middle school or whatever your country calls the class/grade in which you choose a bunch of subjects) and probability only gets touched on, not talked about in lenght before 10th math. its the same as a layer going to a mathamatecian and spouting a bunch of legalese to make him cancel a book

[@pullt]

1 in 10 black men have a beard lol Has professor ever met a black dude?

[@compositestechbb9087]

@@pullt bahahaha

[@pullt]

@@compositestechbb9087 would be a better video if Kevin didn't completely misuse statistics himself when saying 40% chance of a couple matching the description means it's 40% chance they weren't the couple.

[@rossjennings4755]

I am so disappointed in the math professor who was called as an expert witness. Assuming he really knew his stuff, he should have seen this abuse of the product rule coming from a mile away, and it was his responsibility to point it out.

[@MynameisBrianZX]

There are irresponsible people no matter how far they get in some career, and the responsible ones don’t randomly go to court or the media to spout unchallenged opinions.

[@seabassjames8222]

Unfortunately, the witness may not have been allowed to object unless they were asked if they have an objection

[@R3_dacted0]

They may not have known the qualities of the quantities involved. They may have just been given a set of probabilities and asked to find the probability of them all being true.

[@sadpee7710]

unfortunately a capitalist society doesn't function after what's ethical, it functions after what's financially rewarded. in the US legal system, buying a desired testimony rewards all parties involved. thus buying testimonies is an established foundation of the system. everyone working in a sector rely on it being the way it is right now, which is corrupt and unfair. if the system ceased being corrupt then the entire system would collapse. uneven judgment relying on unfair counseling relying on corrupt witnessing and so on. remove one and all other functions which have built around it won't function anymore. this expert witness doesn't really have a choice if they're to participate. to pad their resume and have a chance to exist they have to conform to "the way things are done" i.e. accepted legalized bribery. just the same as with all other sectors and industries

[@danielkeys8974]

I mean, that's true, but the math was mostly correct. Only that weirdly unnecessary part about the independence of beards and moustaches was obviously wrong. The main problem was the prosecutor just making up whatever claims he liked, and nobody calling him on it, to the point where the defense lawyer must have been asleep (or corrupt) to avoid asking at any point, 'Where did you get that number?' If you just make up premises, logic typically won't help you arrive at the truth.

[@Talsar624]

love how, had they been guilty, the accomplice got more jail time than the person who actually carried out the assault and robbery.

[@guyofminimalimportance7]

Well, that's Jim Crow South for you. Like he said in the video they probably let race effect his conviction.

[@incognito-px3dz]

he had prior conviction and women usually get easier sentences

[@smtandearthboundsuck8400]

Criminal record+male+black Kinda expected

[@gwilson314]

Who actually carried out the assault and robbery? We never found out. It still could have been the couple in question.

[@yourmum69_420]

@@gwilson314 re-read it

[@Arceaus98]

One thing that I also felt wasn't really touched on in the refuting arguments: "At the end of the alleyway, a man named John Bass witnessed a white blonde woman with a ponytail get into a yellow car, and drive past him. He saw that the driver was a black man with a beard and a mustache." That was the description given about the two people. The entire probabilistic argument presented was heavily based on the probability of a _couple_ matching those six requirements. *Where in the accusing descriptions does proof come up that the two people in John Bass' view were even a couple to start? If we assume John saw the two correct people, he saw the robber and her getaway driver. That does not make a couple.* I feel that was incredibly glossed over in the courtroom, almost more so than anything else.

[@AlDunbar]

Clearly, the idea of them being an interracial couple was intended to bias the jury against them.

[@schrodingerskatze4308]

And the woman who was robbed only saw a blonde women. That blonde women probably didn´t even have a partner. Or a yellow car.

[@_invencible_]

also couldn't that same witness remember the model of the car?

[@Arceaus98]

@@_invencible_ I think it's a little more reasonable to not have that information. Not everyone knows model names just by visual and it's also possible he just wouldn't have processed quickly enough.

[@_invencible_]

@@Arceaus98 no, but they could show him the actual car and ask him if he recognized it. Edit: so he couldn't process the car model but he remembered that the driver was a black man with a beard and a mustache, come on

[@rocker223rock]

Prosecutors should face ramifications if they are solely responsible for a wrongful conviction. Especially if it is the result of fabricating evidence, and just making up probabilities is essentially fabricating evidence.

[@melvinseifried4647]

But the prosecutor isn’t responsible alone for the sentence they actually get. Just because a prosecutor uses dumb math you have to trust the judge to realize that and not make that sentence. (Imo) Maybe I understood something wrong though.

[@AlDunbar]

@@melvinseifried4647 the prosecutor would only be at fault if involved in fabricating evidence. Making up probabilities could qualify but only if it can be proven to have been done with the intent of being dishonest. How would that kind of dishonesty be proven, using statistical probability? Let that sink in for a moment.

[@snex000]

@@melvinseifried4647 Prosecutors are responsible for trying the case in the first place. However, in this case, if the attorneys didn't shred this nonsense on cross, then they should be in prison too.

[@stevethecatcouch6532]

But the prosecutor didn't make up the probabilities. The witness supposedly made up probabilities, but even that is doubtful. Given nature of the witness's testimony, it's more likely he just used examples of probabilities to make a point. He was not qualified to testify as to the actual probabilities because he was only a subject matter expert. The defendant's lawyer should have objected if the witness strayed from his expert testimony.

[@snex000]

@@stevethecatcouch6532 Do you not understand how the state prosecutes cases? They don't just call "random math guy" to the stand and then everyone gets surprised by who shows up and what he says. Prosecutors seek out experts months in advance and interview them for hours about what questions they will ask and whether they want to go forward with that person's testimony. The prosecutor knew what this guy would say ahead of time, knew if anyone else said it was nonsense, and still chose to put this guy on the witness stand. And the defense attorneys knew all this info as well, long before the actual trial. This is malicious misconduct, plain and simple.

[@LOLonHere]

"Things needs to be really dumb, before we get smart" Wise words, that's why no matter what anyone says, you're not useless.

[@anshumanagrawal346]

@microbial cat ya lol

[@maze7050]

I immediately thought of the Einstulung Effect

[@anshumanagrawal346]

@@maze7050 how so

[@aroncanapa5796]

Being smart is just knowing how to weave those dumb parts together

[@AbridgedAnime]

I should almost be smart now then

[@MrMaster841]

I love how the court's references for how they arrived to the 40% probability was citing literal intro to probability books

[@dreammaker9642]

They had me at 1/10 cars being yellow cause first of all even with basic level in probability you’d figure you’d notice if one in every 10 cars was yellow because there’d be yellow cars everywhere but the worse is how do you conclude the odds of all these events mean that couple must be the one… like that’s some « because trust me bro » level of source like you just going to make that argument with no reasoning ? We know there isn’t a coherent one because it’s all mathematical bollocks but how did a jury and a judge be that brain dead? Smells like corruption

[@i.Yallow]

​​@@dreammaker9642i agree with you but based on my (limited) research, the yellow cars may have actually been similar to 10% for the time period. theres an article by daily infographic that claims in 1971, 12% of cars were yellow. article is called Most popular car colors over time... shades of gray

[@ucantSQ]

Where was the defense on cross-examination!? You've gotta be a terrible lawyer to let this testimony fly. "No questions, your honor. I'd rather be fishing."

[@timothymclean]

The defense probably hadn't taken a math class since high school and had no idea how to argue against the immutable laws of mathematics. (Also, the odds that a probably-white lawyer would give 100% to defend a black man in the 60's aren't much better than one in twelve million.)

[@WooperSlim]

6:58 - Another way to think about it, the 1 in 12 million they calculated is supposedly the probability a random person met all those characteristics. But police weren't arresting random couples, but specifically those that match the description. That's why the probability that they are innocent is so much higher, because it isn't out of all couples, but out of all others that match that description.

[@mkilgore]

Yeah he kinda skipped over the actual prosecutors fallacy. The probability introduced in court was the probability of a *random* couple matching whatever parameters they set out. But you can't then flip that to determine the probability that this *specific* couple committed the crime, which is what they did. Even with odds of 1/12,000,000 there's so many people that it's likely there are multiple couples in the area fitting those parameters. And if there's Ex. 3 couples that match those parameters, then the odds this specific couple committed the crime is actually only 33%.

[@doomse150]

I'd guess the idea he tried to push is that the probability of a random couple matching this description is so low, that the one couple they did find is likely the only one in their general area, matching that description. Which is, obviously, still highly questionable.

[@mkilgore]

@@doomse150 Yes that's the idea, but keep in mind you can calculate the probability that they are the only couple, you don't need to guess. The catch is that it requires an estimate of the size of the population within which you found the couple, but without that information the "random couple" statistic is completely worthless anyway. The fallacy is assuming there's an actual correlation between the two without having to consider the population size. Imagine you win the lottery and I claim that you must have cheated because the odds of a random person winning the lottery were 1/1,000,000. That's the same kind of logic, and it's obviously a dumb argument because it's ignoring how many people entered the lottery, which is what really determines how likely it was that there would be a winner at all.

[@malvoliosf]

Yeah, that’s the thing: the 12 million (if it were right, which isn’t, it’s just the off-the-cuff numbers of someone with incentive to get a conviction) is only significant compared to the number of people who might match it - so the number of people in the greater Los Angeles area. What are the odds that there are TWO pairs of people riding in yellow car, black bearded male, blonde pony-tailed woman, in all of Los Angeles? Uh, pretty high, I would think...

[@yourmum69_420]

@@malvoliosf the biggest thing of all though, which everyone seems to be missing out, is that we have no idea if the robbers was a pony-tailed woman with a black bf and a yellow car in the first place. We only know she was blonde

[@EpicBoss-]

When the people trying to prove that a minecraft speedrunner was or was not cheating are better at using probabilities than actual criminal prosecutors, you realize how awful our legal system is

[@joffles6516]

No the guy dream hired to try to make him seem innocent was worse

[@charlieanderson5490]

I mean what else is the guy going to do? He had a couple who vaguely matched a description but no connection to the crime. Of course you are going to try some bs.

[@JoniWan77]

@@charlieanderson5490 Not get them in front of the court if there is not enough evidence to support it. If you have to cheat to win a fight, you don't have to fight, don't fight. I am fairly sure in pretty much any country the prosecutors decide on who to take to court. Your argument only really holds water, if the prosecutor was forced to battle it out.

[@aniruddhvasishta8334]

@@JoniWan77 You may be misunderstanding, they were racists willing to go to any lengths to prosecute a black man. In their minds, it was a necessity to send this person to jail. The justice system has always been biased in this way.

[@siimad2988]

that is not a valid arguement. the team of prosecutors was maybe 10 or 15 at most, whereas the people 'prosecuting' dream were in the 100s, maybe 1000s. of course the arguements against dream will be better. more ideas, just as talented mathmeticians, and in greater quantity i study IT, and the same applies for open source programs vs proprietary. open source is generally beter due to the sheer quanrity of people involved.

[@warfjm]

Prosecutors wanting a win just to win is a miscarriage of justice. Their duty is to carry out justice. Anything outside of that duty is moraly reprehensible.

[@chair547]

That's not actually true. Our court system is adversarial by Design. Whether that's right or wrong is another question but that's how it is. Prosecutors are supposed to try to convince people and defense attorneys are supposed to try and acquit them. The problem is that the prosecutors are well-funded and the defense attorneys aren't

[@resolecca]

@@chair547 while what you are saying about the process being advertorial is true, that dosn't in anyway negate what @warfjm said about prosecuters just wanting to win for the sake of winning, nor does it mean that prosecuters don't regularly lie or make up evidence to do so. Yes you can still have an honest advertorial system (the one we currently have just isn't) but it is possible. Or how about both the sides work together to solve the crime and give the family justice, rather then just finding someone to pin it on to get the case off your books. Coz when you win to win everybody looses, innocent people sit in prison and murderers go free and no-one least of all the victims family gets justice. But that's just my humble opinion

[@paulmahoney7619]

Something interesting is that military JAG courts have a system where lawyers serve as prosecutor and defense in alternating periods. If we made it so that public prosecutors and public defenders would swap places every six months, I bet a lot of prosecutors would try and make some major reforms.

[@thetriathigamer1544]

Fr, they don't understand the misery they put these people behind just because they want the case to be over with, at that point just get a different, easier job

[@vyor8837]

Just wait till you look at the Rittenhouse trial, lying about how zoom functions work on video.

[@xl000]

They forgot to consider "probability that she didn't remember correctly", or "probability that she made up her testimony"

[@emilmullerv3519]

Did no one in the room seriously thought "hey, if there are 2 couples fitting this description in the state the chances would already be 50%"? And people complain about learning math in highschool

[@the_skips]

It's not gonna be 50% It's more complicated than that. That's like saying "there's a 50% chance that a meteor will hit me today, because it either will or it won't, these are the only options."

[@emilmullerv3519]

@@the_skips probability in the real world is normally knowledge dependent, if you don't know anything else about the couples, the best bet is a 50-50 chance on either one. That's my point. Of course the problem is also that I'm this technique for conviction assumes "who committed X crime" is random, which is obviously false, the only interpretation that makes it make some sense is an epistemic one

[@the_skips]

@@emilmullerv3519 oh. Fair enough

[@Nick30468]

I'm reminded of a teacher had I way back in elementary school when another student asked that stereotypical question during a math lesson "why do we need to learn this". She had responded (paraphrasing, it's been too many years) so people can't pull the wool over your eyes. Which is exactly what that prosecutor was doing. It's fair to say he knew full well he was using vague numbers that sounded good to lie to the court.

[@allanshpeley4284]

I was at the top of my class in statistics in highschool, but 20 years later I've forgotten most of it. This is yet another reason why we shouldn't be judged by our so-called peers.

[@aircloud1795]

But the pulling to wool carried out by the prosecutor is basically lying with statistics but he makes it sound legit

[@allanshpeley4284]

@@aircloud1795 Exactly why our legal system needs to change. If there were dedicated, experienced juries instead of random people selected from the population then that sort of trickery wouldn't be possible.

[@katherinegaymes]

@@allanshpeley4284 this reminds me of my two most recent jury summons appearances (didnt actually join a trial but...) and i was reading the rules on both the city and county website and was explicitly told that if i was an expert on the subject matter that that is not *allowed* to have any relevance on my judgment as a juror. For one, you cant really separate a person from their knowledge in a meaningful way. My own experiences will bias me in various unknown ways.

[@allanshpeley4284]

@@katherinegaymes Wow, that's really something. They might as well tell us we should forget everything we know about logic and reason while they're at it.

[@OptimusPhillip]

Lawyer: I will use probability to prove that they are the only people who could've done it! Me: Oh, are you going to pull municipal census data, vehicle registrations, or any number of real data sets that could give you numbers to work with? Lawyer: Nah, I'm just gonna name probabilities with denominators I rolled on a fucking d10.

[@TheKYLEdavid]

Yeah. The fact that he went with 1 in 10 cars being yellow is enough for me to know he made these numbers up out of thin air I can’t remember the last time I saw a yellow car

[@williwiebe]

@@TheKYLEdavid That part got me too. The only way I could fathom the numbers being that high are if they were including yellow school busses and taxis (if their municipality has those things) and then, the rest of the odds would have to be different. The odds of a black man working a lower status job such as taxi driver in a racial society is likely a lot higher and the odds of a person getting into a taxi in a municipality with enough taxis for 1 in 10 vehicles to be yellow has to be pretty high as well.

[@augustuscaeser5895]

@Balance of The hill probably that it was 1968 and there weren’t all that many interracial couples in Los Angeles. Obviously wouldn’t be true now but it may have been true then.

[@diamondportal77]

I think It was different in the 60s, more colorful cars were much more popular back then. So for all we know 1 in 7 cars are yellow.

[@evil001987]

Even if he did use proper census data and used appropriate numbers, he wouldn't even be able to prove that this is what the witnesses saw, they can missremember details. And even if those people were who the witnesses saw, doesn't prove they commited the crime only that they were at the site.

[@FreeDomSy-nk9ue]

I'm having a hard time trying to understand how these numbers 4:10 made it to court and actually won a case! What's the probability of that, given that the jury at least went to elementary school? Give me some Bayesian math!

[@the1exnay]

Probability is notoriously unintuitive and easy to mess up.

[@pullt]

would be a better video if Kevin didn't completely misuse statistics himself when saying 40% chance of a couple matching the description means it's 40% chance they weren't the couple who did the robbery

[@chrismanuel9768]

@@pullt 40% chance there was another couple that matched the exact description despite all the flaws, meaning a 40% chance there was another couple that could have been guilty, meaning a 40% chance this couple couldn't be made guilty on probability.

[@pullt]

@@chrismanuel9768 What you indicate is true What Kevin indicated....that there's a 40% chance they weren't the perpetrators.... is erroneous use of statistics

[@epajarjestys9981]

The probability of that is 3.

[@robertturley2974]

1 out of 12 million seems like plenty of room for reasonable doubt.

[@fen3311]

Well, if we're looking at raw numbers and ASSUMING all those numbers are 100% correct, then no. That is well, well beyond reasonable doubt to the point of certainty as far as court room requirements are concerned. The issue in this court case is the numbers were quite literally made up and required quite a few other assumptions that couldn't be or weren't proven.

[@hellohi2516]

Seeing Betty White at the beginning of this video caught me so off guard, especially since she had nothing to do with it.

[@RialVestro]

There's two MAJOR issues I have with this story and neither of them have anything to do with the math. 1. The woman who was robbed only said that she saw a blond woman. A completely different witness pointed to a blond woman in a pony tail getting into a yellow car with a black man. It is entirely possible that the two witnesses saw two entirely different blond women. There's no reason to assume that a woman with a ponytail, a yellow car, and a black significant other was even related to the crime since the victim never described any of those details. In fact it's entirely possible that the man was either knowingly diverting attention away from his accomplice or unknowingly spotted an entirely different blond woman in the area and assumed she was the same blond woman the victim saw. Blond is such a common hair color that it would be extremely easy for multiple witnesses of the same crime to give entirely different descriptions of who they believed the blond woman to be. 2. It makes absolutely ZERO sense for the woman they believed to have actually committed the crime to serve less time than the man they believed to be her accomplice. I don't know if it's race related, gender related, or both but I looked it up... The sentence for aiding and abiding is 3 to 10 years depending on the severity of the crime so his sentence was accurate. He would of only gotten 3 years anyway even if the case hadn't been over turned. The sentence for theft is 3 years, the sentence for assault is 10 years so the woman should of been looking at 13 years in prison. For her to get LESS than 3 years for assault and theft means she couldn't have already served her full sentence. I think this is a gender related thing cause it's pretty common for women to serve less time than a man would be for the same crime. That's the only reason why she could have served her time in less than 3 years when she actually should have been sentenced to 13 years. It's also possible for a sentence to been extended or reduced based on the person's behavior while in prison but that would in itself be a process of overturning their original sentence.

[@nuklearboysymbiote]

Your first issue was also my first instinct. But now that I've seen your second issue, it makes the story even more dodgy to me, and I would agree that we can infer from this that the justice system is so flawed and prone to bias, involving math in it only serves to taint math, and not help serve justice in any way

[@卵-n6e]

blondE!!! God its so hard to read when you're talking about a WOMAN who is BLONDE not BLOND lol

[@TheFinalChapters]

There's an even more glaring issue I saw someone else bring up: even if the blonde woman and the black man were in fact the robbers, who's to say they were a couple?

[@celestialtree8602]

@@卵-n6e Both have the exact same meaning, and using them interchangeably is becoming more common. It's not necessarily wrong (though it still may be an issue in highly formal writing), just language evolving.

[@EebstertheGreat]

@@卵-n6e You probably learned that rule once and have decided to try to lay it on everyone else ever after, even though it isn't a real rule. English is not French. We don't have gendered adjectives.

[@oogrooq]

Odds of anyone having a testicle = 1/2. Odds of having a breast = 1/2. Therefore, odds of having both a testicle and a breast = 1/4. QED.

[@billweasley1382]

Sorry, they aren't independent events. - "What is an Independent Event? An independent event is an event that has no connection to another event's chances of happening (or not happening). In other words, the event has no effect on the probability of another event occurring."

[@fos1451]

@@billweasley1382 that’s literally the joke

[@fos1451]

That’s actually a smart example

[@starboundsingularity]

the chance of having TWO testicles or TWO breasts is also 1/4 /silly

[@re4perthegamer]

the chance of having neither = 1/4 source: trust me bro

[@ripstick45]

It's really interesting how dangerous math can be at times. Did you know that 80% of statistics are made up on the spot?

[@thefloormat3297]

Did you know that 80% of cheezitz boxes are bought by me?

[@lizardbrain5962]

I see person attempting to make a Cheese-Them, as all should.

[@fos1451]

I will be honest, when hbomberguy said that I immediately believe it for some reason

[@jmanpolo5611]

Thought it was 75%

[@Hensley_Jb]

🤤😭

[@feyetho9524]

love that Kevin used a photo of recently deceased Betty White as the "little old lady". This video was released 5 days before her demise. Little macabre

[@strikerz360]

i came to the comments to search for anyone mentioning this, it’s kinda spooky lol

[@MrKhaosBlaze]

Probably the worst timing to use Betty White for the victim.

[@Burbie]

Really love this series , i think they could be really well going forever not just for now and always sprinkled in between!!

[@AbsolXGuardian]

Yeah. I'm sure there are way more math related true crime cases. If not, it could be expanded to math being perverted for political discourse or guiding public policy

[@amiyakumarmaity5077]

Nobody going to talk about how this comment is posted 1 Hour before this video got posted

[@Burbie]

@@amiyakumarmaity5077 i posted this comment a min after the video The comment and the video are both 20 hrs ago for me

[@trspanda2157]

Jesus loves us all that's why he died for our sins

[@CoWulfse]

Yesyesyes

[@SojournerDidimus]

While the math is bad (or more exactly, the numbers used in it), there is a much much *much* larger issue with the trail as you described it! The profile was compiled by using three witnesses to things that may be completely orthogonal! The guy from the gas station might be talking about an entirely different couple than the man at the alleyway, making the color of the car irrelevant altogether. Combined with the notion that you might want to commit a robbery in not-your-own neighborhood makes it all the more likely that they were in fact *not* the same couple as the couple at the gas station.

[@glumpfi]

Isn't it even worse? I mean, the only thing the lady remembered was a blond woman - how can the other guy be sure that it was exaclty that woman that got into the yellow car? And what about mistakes in perception and memory?

[@AlDunbar]

Yeah, that and eyewitness testimony being notoriously unreliable for a number of reasons.

[@ellicerslavic]

I remember at a science museum they showed a video of a crime, then asked questions about it and I got most of them wrong because like you say, memory isn't exactly accurate. And if the blonde lady even did the deed, why did the black guy get a longer sentence than her?

[@d0x2f]

seems like a harsh sentence for purse snatching in any case.

[@JKBDTS]

Ok, but why did the woman get the shorter sentence if she was supposedly the one doing it directly and the husband was only guilty of accessory?

[@blueyindustries8503]

Because it’s the 1960s and she’s a white woman, and he’s a black man. It doesn’t take a detective to figure out that mystery.

[@TickedOffPriest]

The crime for a prosecutor knowingly lying should be whatever the defendant was going to get.

[@michaelmcevoy9278]

Assuming you’re a Roman Catholic, we have some foundational disagreements. But we agree about this foundational issue. I wish our criminal justice system agreed with us.

[@_BangDroid_]

Imagine being the guys who's legacy is ruining the reputation of math in law.

[@Trancefreak12]

I wouldn't say that the reputation of math in law was ruined, but rather that a much higher standard of math is now required as a result. That's a good thing.

[@Vearru]

This seems like it’s just about the most ridiculous mistrial I’ve heard of. The argument is as good as “If someone won the lottery in the state this is better than any other evidence that this person is guilty.” It’s entirely nonsensical.

[@erickpoorbaugh6728]

It's like arresting a lottery winner for fraud based solely on the fact that they won and the odds on them winning legitimately are astronomical.

[@danielburger2550]

Two people into jail for 40$? Even if they were guilty... just wtf???

[@username5155]

I love how they put a chance on them being an interacial couple despite the fact that there was no proof the criminals were one

[@samarthagarwal7529]

The first two parts of the Collins Test are literally just do math properly, and the other two are don't create random garbage based on whatever numbers you have.

[@justseffstuff3308]

Yep. All of that just seemed so bizarre, like- why is that not just a basic expectation in the court for everything already?!

[@prosamis]

The probability question the prosecutor was answering was "if you pick a random person, what are the chances they satisfy all these conditions?" when the actual question to be answered is "What's the probability the lady of this couple is the thief?" That's when, should the numbers be right, we get a much clearer idea of what we're dealing with. I believe probability like that can be used as a reason to call people for questioning but definitely NOT as evidence

[@charlieanderson5490]

It is used as evidence. Finger prints are evidence right? Its actually not impossible for 2 people to have the same fingerprints. 1 in 64 billion.

[@Aeivious]

@@charlieanderson5490 thats entirley different. the probabilities brought up in this case only narrows it down to a small few, fingerprints narrow it down to one person and thus can be used as evidence.

[@charlieanderson5490]

@@Aeivious finger prints don’t narrow it down to one person because there is always the chance that 2 people have the same fingerprints

[@juanausensi499]

​@@charlieanderson5490 Fingerprints are also being misused lots of times. The probability of two people having the same fingerprints is irrelevant, what counts if the likelihood of an expert telling apart different fingerpint impressions, moreso when those impressions can be partial, faint or degraded. Some studies (ones based on exonerations of previously convicted people based on fingerprints and others based on testing experts with random sets of fingerprints) have shown that the average expert has a 2/3 chance of getting it right.

[@senseisecurityschool9337]

Yes, they are completely different questions. As different as: How likely is it that a specific flamingo is a bird? Vs How likely is it that a specific bird is a flamingo? The questions sound very similar, but they are completely different and largely unrelated.

[@ethanpederson]

I’m loving this new style of videos you’re doing Kevin!

[@karynjohnson]

First reply lol

[@dr.doppeldecker3832]

@@karynjohnson ?

[@enga-wh8qp]

In some cases it would really be better to let a guilty person get away with their crime, than risking to lock away an innocent person if the evidence is just not good enough. Chances are high the guilty one will commit another crime and finally get caught for good.

[@ark5458]

This kinda makes sense, but aren't we risking a future victim from the same criminal?

[@ark5458]

(also off topic but weathering with you ♥️♥️♥️♥️)

[@Aeivious]

@@ark5458 if you don't have evidence without a doubt then there's nothing you can do, locking up a suspect based off of faulty or incomplete evidence doesn't help anyone.

[@enga-wh8qp]

@@ark5458 Yeah that's true and sad, but we're also at the same time risking an innocent person getting locked up. Guess it depends on what you find morally better.

[@zachhalverstam2804]

Him using Betty White as the old woman really aged poorly

[@spacemantiss]

Very poorly.

[@GrildCheese592]

so did she

[@spacemantiss]

@@GrildCheese592 wow ok.

[@dogge929]

Lol

[@VelkanAngels]

Why? Literally every human in existence is either dead or WILL be dead.

[@romangonzalezadrianmaurici6302]

-But when I will ever use any of these maths? I want to be a lawyer anyway!! A few years later...

[@Krekkertje]

The problem is that in a typical courtroom the only person who understands maths is an expert witness that has been brought in to fabricate some result. Unless the person on trial happens to be a mathematician or engineer.

[@ItsThatSheep]

Their lawyer should of shut this argument down easily, it’s unfortunate that either no one listened or the lawyer didn’t bother.

[@tidus9942]

@@ItsThatSheep yea, a good defense lawyer would have countered it. Unlike TV shows the prosecution cant just surprise the court with this. it has to be introduced and the defense would have been able to counter argue against it. Hell, a good enough defense attorney would have been able to get a mistrial with prejudice over this, at least if the judge was semi competent at all.

[@MadDragon75]

I've been subscribed for years Kevin. You never disappointed me. I don't know what you do as a day job but I know there's a special place you can thrive in others cannot. Thank you for years of quality programming.

[@tapiocaweasel]

I think this is his day job

[@MadDragon75]

@@tapiocaweasel I was thinking the same thing too, but I wasn't going to assume that. He is definitely in the most fitting environment for him.

[@NeverlandSystemPunkGirlChloe]

OMG that this judge ALLOWED THIS is insanity.

[@sandermesila4904]

"Probability and statistics are the math of lies." That's what my high school math teacher said, not word for word though. Real life is subject to so much more chaos and factors, many of which cannot be accounted for in any equasion. How using probability is even legal in a court case is beyond me. Lucky me, I dont have to live under Common law. Long live the German-Roman justice system.

[@ribbonsofnight]

A very poor and narrow vision of statistics

[@mikip3242]

1) He made up the input numbers of his formula or at best never provided the sources for those input values, thus never gave evidence. 2) He used the wrong mathematical model by assuming each thing was totally uncorrelated, which is absurd. Not only a man with a mustache can have a larger chance of having a beard, but also a black man has a larger chance of having a beard than a white one or a woman, a white woman has a larger chance of having a ponytail than a black one or a black man. Almost all values of his product estimate are really correlated and should be treated as conditional probabilities. Each of this correlations is an alternative argument against his approach and they sum up. 3) Even if the formula applied (which does not), even if the input values are correct (which are not), we still need to explain some evidence they did it, perhaps finding the money, something left behind etc... Patter recognition is not evidence it just means that you have one of many possible matches. It mean you can begin investigating and having candidate theories not that you can conclude the entire process. 4) All this "math" relies on the testimonies being 100% accurate and truthful. They don't need to be lairs to be wrong. They can be biased, confused, the information can be of low quality (considering they were unable to recognise them it seems reasonable). Would you destroy a lofe just because two random people say they think they did it? Then why having a process and a judge at all, testifiers can judge it by themselves!

[@Beertraps]

I disagree with 3). If the probability is low enough I feel that it is justified to convict somone.

[@TheMoonRover]

5) The (incorrect) calculation is the probability of a randomly chosen person matching the description, but they didn't arrest a person at random; they arrested someone matching the description. What they want is the probability of someone who matches the description actually being the culprit. That distinction is the main reason for the vast difference between that 1 in 12 million figure and the 40% produced by the appeal. It's like winning the lottery. The odds of YOU winning it are incredibly small, but so many people enter that SOMEBODY wins it almost every time. They picked a winner and tried to calculate the odds of them having won.

[@samuelallanviolin752]

@@TheMoonRover Yeah exactly, for some reason I'm seeing very little of this explanation and in my opinion this is what people should be focusing on - discussing conditional probability and the difference between the probability of A given B and the probability of B given A (A = couple is guilty, B = couple matches description). On trial they presumably calculated P(B) but not P(A given B)

[@iluvgtasan]

The fact that I can see this bullshit logic but a judge didnt makes me lose faith in the justice system altogether.

[@DBZHGWgamer]

A judge can't strike evidence unless an objection is made.

[@charlieanderson5490]

You have only lost faith in a arguable racist jury, didn't really have anything to do with the judge. Plus, the prosecutions case wasn't totally terrible, it is relatively unlikely to have a matching couple like that. The real problem was that 1 in 12 million was viewed as proof when it isn't nearly high enough.

[@Moleoflands]

@@charlieanderson5490 or even an accurate figure

[@charlieanderson5490]

@@Moleoflands it’s probably pretty close

[@enochliu8316]

@@charlieanderson5490 No, the judge did have a role in this: When [the] motion [to strike] was made at the conclusion of the direct examination, the court denied it, stating that the testimony had been received only for the "purpose of illustrating the mathematical probabilities of various matters, the possibilities for them occurring or re-occurring."

[@tim40gabby25]

The sudden infant deaths case was the most egregious, as the outcome was so terrible for the person falsely convicted

[@cmonster4926]

The use of Betty white’s picture has aged well

[@DarthCrustyYT]

betty white died after this video

[@pismodude2]

Ah but the Zodiac Killer is suspected to be a white man who wears squared glasses, layered shirts, jeans, and has a high IQ, speaks English, and has never been arrested for his crimes. Among all humans, the odds of this are 1/10 • 1/2 • 1/30 • 1/500 • 1/2 • 1/10 • 1/2 which means the odds are 1 in 12 million. Yet this KZbinr fits the description perfectly. Should we really be trusting the Zodiac Killer to explain what evidence should or should not be allowed in court?

[@AlDunbar]

Two errors in your logic: The defendant might match all of those parameters, but it is only "suspected" that the actual criminal does. If the actual criminal does not then the stats mean nothing withbrespect to the defendant. How is the actual zodiac killer saying anything about the evidence when that person's identity is not known?

[@flamingmonkays]

One more error: We already know it's Ted Cruz.

[@re4perthegamer]

third error: 8 billion people in the world, thats a lot of zodiac killers@@AlDunbar

[@Islingr]

Your usage of the Betty White image for the old lady is pretty improbable.

[@afonsomonteiro2003]

I'm so confused, I can't believe people overlooked that the prosecutor's argument can also be interpreted as "there's a 1 in 12.000.000 that this could happen". Wouldn't that hurt the case? 💀

[@pretzelbomb6105]

The basic logic was “The crime was committed by a pair matching this set of descriptors. There is a 1-in-12,000,000 chance that any two people would fit this description. Therefore, in a city of ~7.5 million, it is statistically impossible for there to be another couple matching these descriptors. By process of elimination, it must have been these two!” If you don’t know any of the case details, the source of the odds, or how probability and statistics actually work, it sounds fairly reasonable.

[@marixsunnyotp3142]

​@@pretzelbomb610512,000,000 is not even 2 times of 7.5 million Also probability cannot prove, it can only point to someone that is possibly but not certainly responsible, and getting innocent people into jail is so much worse than letting go of criminals(especially American jails where (at least according to what I heard about) you get beaten up and enslaved)

[@raymints9583]

Portraying Jimmy (aka Saul) as a prosecutor and not as a defender. 💀

[@charliesmith4072]

Fifty years ago in law school I came across the appellate decision and loved it.

[@davidtruett8255]

Betty White died today and yet she still shows up in whatever I'm watching 😭 RIP to a legend

[@Furnus105]

Really odd that this video had a picture of Betty White as the victim in it, and then a few days later she died.

[@LordOakrum]

COINCIDENCE?!

[@aidandruck2423]

Incidents like this and the video editing in the Rittenhouse trial serve as a reminder of why "beyond a reasonable doubt" needs to remain an extremely high standard, especially if we base our legal system off of the philosophy of Blackstone that it's better that 10 guilty men escape justice than 1 innocent be wrongly punished.

[@aram00001]

full list of mathematical errors made by prosecution: 1. making up prevalence statistics for the attributes listed 2. introducing mustache as a separate variable from having a beard 3. assuming the robbers were romantically involved 4. assuming that there are no other two people with such a description in the LA area 5. assuming that the robbers were local to LA and not visiting temporarily from out of the city 6. assuming that the attributes involved are completely unrelated, since one of the attributes could more often accompany a second attribute (blonde white women could be more likely to own yellow cars).

[@nwolinsP]

Putting folks in prison wrongly needs to result in draconian punishment. Judge and jury included.

[@nagranoth_]

As soon as someone calls something the "new math" you need to run away... And that's even ignoring how unreliable eye witnesses are known to be. You cannot base statistics on flawed memories...

[@thomasgaines7988]

The usage of Betty white did not age well…

[@syvulpie]

Neither will she ever again.

[@TripPy_Poly]

statistic don't lie but you can lie with statistic

[@Mark73]

Did the defense attorney not ask the math professor if a person should be convicted of a crime based on this method if there is no other evidence of their guilt?

[@connorwilcox146]

Very unfortunate timing to use a picture of Betty White

[@Schultzie580]

The Betty White placement did not age well

[@arttukettunen5757]

Math should not be hard evidence. Instead, it should only be used as a helpful tool, to raise or decrease suspicion and decide what further investigation is needed

[@BariumCobaltNitrog3n]

It's used every day for fingerprints and DNA. A large sample takes the place of looking at every single person. But even then the chances of two people having a single matching fingerprint is a non-zero number.

[@PandaOnSkis]

Rip Betty white

[@edwardswartz8471]

One of the most shocking things is how Malcolm got a longer sentence than Janet even though she is the one who committed the crime

[@urvanrry2348]

No way it wasn’t racism 💀

[@Ducktor]

Males always get more prison time than females

[@defaulter264]

@@urvanrry2348 it was female privilege

[@khulhucthulhu9952]

Painful use of Betty White😓

[@IABITVpresents]

RIP Betty White

[@callbettersaul]

Guys... it was me... I stole the purse. I'm sorry.

[@irrelevant_noob]

... you're the blonde woman with a pony-tail?

[@wilkinscoffee4228]

That was the worst math ever. Of all time.

[@im_snoopyrs8947]

The betty white thing didnt age well.

[@baso53]

Bro, literally the worst time to use a Betty White photo. She died a few days after uploading of this video

[@TurtleneckTim]

0:09 Rest In Peace Betty White

[@First-Name--Last-Name]

L bozo

[@davidjames6294]

That Betty white reference has changed as of today

[@gigabytemon]

Rest in peace, Betty.

[@3crownedprince940]

I love how interesting this can get and how you are pretty much an information page but its simplified but not overly simplified so it can be enjoyable as well as the music and the mood set in each section

[@BrokenPuzzle03]

The fact he used betty white's picture and her dying almost a month after this video's release is a bit scary...

[@harmonicarchipelgo9351]

They must have had a lousy defense lawyer (probably a public defender). I don't expect the lawyer to understand Baye's theorem, but there were a lot of ways to discredit the case against them.

[@DBZHGWgamer]

For all you know the defense destroyed the argument but the jury was too racist to care.

[@enochliu8316]

They had a lousy judge deny the motion to strike. From the appeals judgement Objections were timely made to the mathematician's testimony on the grounds that it was immaterial, that it invaded the province of the jury, and that it was based on unfounded assumptions. The objections were "temporarily overruled" and the evidence admitted subject to a motion to strike. When that motion was made at the conclusion of the direct examination, the court denied it, stating that the testimony had been received only for the "purpose of illustrating the mathematical probabilities of various matters, the possibilities for them occurring or re-occurring."

[@enochliu8316]

I.E the defense lawyer did object to the testimony, but the judge allowed it to be given to the jury.

[@harmonicarchipelgo9351]

@@enochliu8316 If the testimony was allowed, doesn't that mean that the lawyer could cross-examine the witness and expose the complete disconnect from reality?

[@tidus9942]

@@harmonicarchipelgo9351 they could and most likely did but jury's are normal people so it doesn't always matter. This is on the judge. it should have been thrown out.

[@zander3902]

Using Betty white as the old lady didn't age well 😶

[@coreyellis5591]

Betty White died today and I had to see her in this video 😭

[@abelsm6270]

To recap, Scarlett Johansson cut off Betty White's hand in an alley way and drove off in a Yellow car from the Disney-Pixar movie "Cars" with Idris Elba as her get away driver. According to Larry David.

[@lulairenoroub3869]

Real unfortunate timing with the Betty White pic

[@7thesage853]

The Betty White pic has aged significantly in such a short amount of time

[@LegendStormcrow]

I worked as a correctional officer for 12 years. This kind of BS is part of the reason being firm, fair, and consistent is so important. Being in prison alone is hell, no need to make it worse on a potentially innocent person.

[@Aeronor2001]

Ohh, that picture of Betty White to start us off :(

[@MrJinxmaster1]

Chance of a mixed race couple being in a car as 1/1000 is so funny

[@Mythikal13]

The odds of you existing as you are is 1 in quintillions or something, 1 in 12 million odds those two people were who they were was a little off

[@dataandcolours]

This was well described. Well done. May I just add that there is one even bigger mistake they made that is not pointed out here that is often omitted: The last 1 in 1000 is more or less a summary of the previous scenarios they have already calculated. So even if we ignore that they assume sufficiently close to independent event when it wasn't, it would still be way off. This is to me the most scary part of all the errors. It's like saying the probability of Real Madrid winning Champions League this season is P(Real Madrid winning Champions League) = P(winning round of 16)*P(winning quarter final)*P(winning semi-final)*P(winning the final)*P(winning Champions League) = 0.5*0.6*0.6*0.55*0.1 = 0.0099 so just 1 in 100. You can't just magically make a combined factor for the very thing you are trying to calculate and also include it! :)

[@_invencible_]

yeah and what does "The probability that an interracial couple would be in car" even mean?

[@joshbrent4950]

The picture of Betty White made me sad.

[@lincolnheron4122]

This video didn’t age well for Betty white

[@vikumwijekoon3166]

Wait making up probabilities, isnt that perjury? Shouldnt the prosecutor be punished for literally making up the numbers and lying to the court?

[@GodOfWood421]

I think the thing that ticks me off the most is that the get away driver, that happened to be a black man, got a worse sentence than the attacker, the white woman.

[@mikeymoughtin6573]

using a pic of betty white in this video is pretty poor timing

[@jacobkodad3065]

Rest in Peice Betty White, aka Juanita Brooks

[@shockingshane9700]

So the male getaway driver got a much heavier sentence than the female person that actually assaulted and robbed the person... God I love the criminal justice system... They don't even try to hide it. They are wrong on almost every account.

[@robertthomas5906]

Women seem to never get an equiv. sentence. Often it's a fraction of what a man would get. No so called equality. He also failed to mention Jim Crowe was by the democrats. Governor Wallace, Senator Byrd, Joe Biden, all Jim Crowe era politicians and all democrats.

[@JohnSmith-kb4re]

almost like the nation whose existence is built upon corruption and hypocrisy is corrupt and hypocritical.

[@Transformers2Fan1]

This reminds me of an example my professor once gave in a Stats class: He once got asked to weigh in on if students had cheated. Since it is technically possible to blindly guess 4-option multiple choices correctly, no matter how many questions there were. It's also possible I get smited by a meteorite 5 seconds after posting this. It's 4^(-x) where x=# of questions and it gets really tiny, really fast. Administration went "So it's possible? Okay, they didn't cheat." (they totally did)

[@aothanhhuy2814]

The most infuriating thing here is,let say that they are right,SO WHAT,THEY HAVE ONLY PROVE THAT THERE IS A LOW POSSIBLITY THAT THEY ARE HERE,IT DOESNT EVEN TOUCH A THING ABOUT WHY THEY ARE HERE,IT DOESNT PROVE ANYTHING.