Inaccurate Criminal Justice System

The American judge Alex Kozinski highlights 12 assumptions in the criminal justice system that we either do not know are true or have evidence to suggest are not true.

(please read the whole PDF for plenty of citations and exposition)

1. Eyewitnesses are highly reliable.
research shows that eyewitness identifications are highly unreliable…Yet, courts have been slow in allowing defendants to present expert evidence on the fallibility of eyewitnesses; many courts still don’t allow it. Few, if any, courts instruct juries on the pitfalls of eyewitness identification or caution them to be skeptical of eyewitness testimony.

2. Fingerprint evidence is foolproof.
When tested by rigorous scientific methods, fingerprint examiners turn out to have a significant error rate.

3. Forensic evidence of other types are scientifically proven and infallible.
Spectrographic voice identification error rates are as high as 63%, depending on the type of voice sample tested. Handwriting error rates average around 40% and sometimes approach 100%. False-positive error rates for bite marks run as high as 64%. Those for microscopic hair comparisons are about 12% (using results of mitochondrial DNA testing as the criterion).

4. DNA evidence is infallible.
… DNA evidence is only as good as the weakest link in the chain.

5. Human memories are reliable.
The mind not only distorts and embellishes memories, but a variety of external factors can affect how memories are retrieved and described. In an early study by cognitive psychologist Elizabeth Loftus, people were shown videos of car accidents and then questioned about what they saw.26 The group asked how fast the cars were going when they “smashed” into each other estimated 6.5 mph faster than the group asked how fast the cars were going when they “hit” each other.27 A week later, almost a third of those who were asked about the “smash” recalled seeing broken glass, even though there was none.

6. Confessions are infallible because innocent people do not confess.
Innocent people do confess with surprising regularity. Harsh interrogation tactics, a variant of Stockholm syndrome, the desire to end the ordeal, emotional and financial exhaustion, family considerations and the youth or feeble-mindedness of the suspect can result in remarkably detailed confessions that are later shown to be utterly false.

7. Juries follow instructions.
This is a presumption-actually more of a guess-that we’ve elevated to a rule of law.35 It is, of course, necessary that we do so because it links the jury’s fact-finding process to the law. In fact, however, we know very little about what juries actually do when they decide cases.36 Do they consider the instructions at all? Do they consider all of the instructions or focus on only some? Do they understand the instructions or are they confused? We don’t really know

8. Prosecutors play fair.
The U.S. Justice Department, for example, takes the position that exculpatory evidence must be produced only if it is material.

9. The prosecution is at a disadvantage because of reasonable doubt.
To the extent this psychological research is applicable to trials, it tends to refute the notion that the prosecution pulls the heavy oar in criminal cases. We believe that it does because we assume juries go about deciding cases by accurately remembering all the testimony and weighing each piece of evidence in a linear fashion, selecting which to believe based on assessment of its credibility or plausibility. The reality may be quite different. It may be that jurors start forming a mental picture of the events in question as soon as they first hear about them from the prosecution witnesses.

10. Police are objective. 
And not just small-town police in Podunk or Timbuktu. Just the other day, “[t]he Justice Department and FBI [] formally acknowledged that nearly every examiner in an elite FBI forensic unit gave flawed testimony in almost all [of the 268] trials in which they offered evidence against criminal defendants over more than a two-decade period before 2000.”49 Do they offer a class at Quantico called “Fudging Your Results To Get A Conviction” or “Lying On TheStand 101”? How can you trust the professionalism and objectivity of police anywhere after an admission like that?

11. Guilty pleas are proof of guilt.
… this fails to take into account the trend of bringing multiple counts for a single incident—thereby vastly increasing the risk of a life-shattering sentence in case of conviction56—as well as the creativity of prosecutors in hatching up criminal cases where no crime exists57 and the overcriminalization of virtually every aspect of American life.

12. Long sentences deter crime.
As with much else in the law, the connection between punishment and deterrence remains mysterious. We make our decisions based on faith

Embedded Link

There are a dozen or so common algorithms computers use to sort data

Here is a visualization and audio-ization of 15 of them.

A description of how an evolved neural network learned to play a level of Super Mario…

A description of how an evolved neural network learned to play a level of Super Mario World.

Fruit, vetetable, or….achene?

Fruit, vetetable, or….achene?

I'm not making any moral or other type of statement about this data other than…


I'm not making any moral or other type of statement about this data other than to say it's interesting to see.

Evolutionary circuit design

This is a super fascinating article peering into the work of Adrian Thompson at the University of Sussex.  

Dr. Thompson is using natural selection to create novel microchip designs.

One interesting insight involves one evolved design that didn't seem like it should work at all:

It seems that evolution had not merely selected the best code for the task, it had also advocated those programs which took advantage of the electromagnetic quirks of that specific microchip environment. The five separate logic cells were clearly crucial to the chip's operation, but they were interacting with the main circuitry through some unorthodox method– most likely via the subtle magnetic fields that are created when electrons flow through circuitry, an effect known as magnetic flux. There was also evidence that the circuit was not relying solely on the transistors' absolute ON and OFF positions like a typical chip; it was capitalizing upon analogue shades of gray along with the digital black and white.

On the Origin of Circuits

Letter of recommendation for John Nash to Princeton University


(For those unaware, John Nash is the subject of the movie A Beautiful Mind)

If you're attempting to replicate scientific research, you're awesome

If you're attempting to replicate scientific research, you're awesome

Vox has an article up about the recent fraud by Michael LaCour in his "study" featuring gay canvassers.

It contains this tidbit about David Broockman who exposed the fraud:

Broockman was consistently told by friends and advisers to keep quiet about his concerns lest he earn a reputation as a troublemaker, or — perhaps worse — someone who merely replicates and investigates others’ research rather than plant a flag of his own.

I have to say that someone who "merely" replicates studies is super awesome in my eyes.  This probably stems from my interest in meta-science…science about science…but someone out there replicating research and making science better is my hero.

How the biggest fraud in political science nearly got missed
Why power dynamics and poorly designed incentive structures in science could have caused a recently retracted Science article to stay on the record.

It's pretty frustrating that an introverts preferred state of being is often…

It's pretty frustrating that an introverts preferred state of being is often interpreted as snobbishness or some other negative statement about the relationship between the introverted person and others.

Additionally, it's frustrating when an introvert explains that this isn't the case and it's taken as if there's something wrong with the introvert.