Tuesday, July 14, 2020

Quest: Redundant


Dmiti got out of the truck and stretched the kinks out of his back. In his mid-sixties and not physically active, he knew the bouncing ride from Capiche to the Buffer-Zone was going to extract a price.

He fussily oversaw the unloading of the boxes of electrical equipment.

Sammy heard through the grapevine that equipment with wires was being unloaded. He figured that he should look into it.

Dmitri had come unannounced and Sammy was surprised to see him.

Dmitri said, “Tell these men where to put the equipment. Then you and I need to take a walk.”

Sammy directed the fighters to put the equipment into his workroom. Then Sammy picked up an AR and a vest with six magazines. That was in addition to the 9mm handgun that Spackle required officers to either wear at all times or have within arms reach, even when showering or sleeping. It would not do to be unarmed when the balloon went up.

Sammy chambered a round and safed the weapon. While not a master with the weapon, he could hit a five-gallon bucket at a hundred yards while shooing off-hand all day long.

“First, you did well to think of more sensors” Dmitri said. “That is something I should have thought about.”

“The other thing is that you cannot know what you don’t know” Dmitri continued. “This is not a criticism, but you only got the problem a quarter-right.”

Sammy started to say something.

Dmitri held up his hand to forestall the flood of explanations.

“You were absolutely right that we need more sensors and that we need them in the Buffer-Zone, and yes, west of the Buffer-Zone” Dmitri said.

“What you missed is that for technology to work, you need to think about the entire system.”

“What happens if you do get rolled over? How are you going to do your job and support the fighters in the Buffer-Zone and the irregulars west of the BZ?” Dmitri asked.

“I will just have to step-up my game” Sammy answered.

Dmitri cast a sidelong glance Sammy’s way. “I don’t think you understand how ugly this is going to be. Likely, you will have to destroy your equipment and evade capture. Then where we all be?” Dmitri asked.


Dmitri did not add that there was a very real chance that Sammy would be captured or dead if his position was rolled over.

Sammy started to speak again.

Dmitri shook his head.

“No, you are only human. You will hyper-focus, looking at the immediate problem and you will lose situational awareness. How many times have you been in a video game and forgot to eat?” Dmitri said. “War is the ultimate video game. Even if you remembered the extra sensors, it would be too much information for you to process, too many fighters in the field to feed the information to.”

“So what is the point of installing extra sensors?” Sammy asked bitterly. Dmitri as much as called him a failure.

“Think of it as a system. A chain is only as strong as its weakest link. If you triple the number of seismic sensors and than double that with an equal number of acoustical microphones, then you are saying you can process six times as much information” Dmitri said. “And that is not realistic.”

Sammy shrugged. There was nothing he could say.

“The answer, my young friend, is to find two more people who can do what you do and to set-up fall-back positions for you to work from in the event getting rolled-over becomes likely.” Dmitri said.

“You see, you aren’t the only bottleneck. We run into problems putting all of that information on screens. We run into problems with computer processing power. The answer is to find one more person to back you up and to provide them with redundant processing equipment.”

“I thought you said two more people?” Sammy said.

Dmitri smiled. “Once the bullets start flying, I plan to be one of the other operators. Do you have any fighters you have been training and show enough promise to be your back-up?”

*

“No!” General Spackle roared. He rarely raised his voice but this was an exception. “No way in hell will Dysen be Sammy’s back-up. She needs to be in Capiche where I know she will be safe.”

Dmitri let Spackle roar. He remembered Quinn from when he was a kid walked along the road collecting pop and beer cans to make enough money to buy an ice cream cone in town.

“She can do that job from Capiche” Dmitri said. “In fact, that is probably the best place for her to work from.”

Dysen perked up. It had seemed awefully unfair that she would not be able to help “her boys” when the shit hit the fan. It had not yet occurred to her that she would be taking calls for medical assistance and reports of casualties.

Dysen gave Quinn “that look”. He knew that he would not win. The best he could hope for was to lose gracefully.

He relented. “OK. But I insist that she go back to Capiche with you so she can begin training with her new equipment immediately.”

Ordinarily, Dysen’s pronounced speech impediment would have disqualified her for the job as data analyst and dispatcher but the Buffer-Zone had boiled down virtually all communication into four character alpha-numeric codes and GPS positions.

Alpha-numerics can be compressed into very short bursts and are harder to home in on and disrupt than verbal communication. They are also immune to background noise, a factor that was expected to be important.

Before Dmitri headed back to Capiche, he helped Sammy install three more monitors. One of the monitors was to capture seismic data from the sensors installed within the boundaries of the Buffer-Zone. The other two monitors were to display data from the devices Dmitri had designed and fabricated to capture and process air-born noise; devices he called “Pickelmans”.

Monday, July 13, 2020

Rules of Evidence: Part II



The Availability Heuristic
The availability heuristic merits special mention in Rules of Evidence.

Our minds are imperfect processors of data. A big part of that imperfection is how we structure our data storage and retrieval.

We string data together. Each bit of data has a pointer to several other bits of data. When we try to retrieve a particular piece of data, we shave the top of our stack and select the parcel that is closest to what we are looking for.

If that parcel is not the exact piece we want, we follow the pointers to the next level of data. We review them and if none of the pieces on the second level are exactly what we need, we select the parcel that we judge closest to what we want and follow the pointers associated with that parcel.

Some bits of data are boring. Some are commonplace. Some are "siloed" off from us.

Consider late-middle aged people with diabetes. A twenty-something rarely considers them as a population. Diabetes is a chronic, boring disease. There is nothing glamorous about diabetes. Having to diet for the rest of your life is a drag, it is work.

Diabetes is commonplace. Getting bit by a venomous, South American spider and dying is newsworthy and memorable because it is so unusual. Sidney dying of complications from diabetes in his mid-50s is neither.

In America, twenty-somethings don't hang out with fifty-year-olds. For most twenty-somethings, the most common causes of death for fifty-year-olds are as "available" as the common causes of death in Minsk, Belarus.

It works in the other direction, too. Causes of death, and death-rates for very young children are also not very "available" to twenty-somethings, the spear-carriers for Antifa and BLM.

Frankly, it is not just the twenty-somethings. For example, can you empathize or place yourself into the perspective of the 999,999,999 spermatozoa that did not win the race? No? I can't either.

Perspective
It becomes difficult to put risk into perspective when the person assessing risk is oblivious to the availability heuristic.

Suppose you were a twenty-something African-American who listens to the main-stream media.

Based on the stories you are marinaded in and the limitations on empathy, what would you judge the relative risks of an African-American female dying in a police no-knock raid versus dying in the first year of life?

Police no-knock:Infant mortality
1:1?
1:10?
1:100?
1:1000?
1:10,000?


Acta non verba


Never forget, Navajo Code talkers


Regardless of origin, we are in this together.

Rules of Evidence


Suppose you  go to a family party. Perhaps your nephew, Woodie, graduated from high school. It is a joyous occasion.

Your oldest sister is there with her new boyfriend, Biff. Biff appears to be in distress. You work in the medical profession. Maybe you are an EMT. Maybe you are a nurse.

Your sister tells you that Biff is just suffering some anxiety at meeting so many new people, that he gets panic attacks.

You look at Biff. He is rail thin. Every symptom points to a myocardio infarction, all the way to the bluish lips.

Do you ruin the party?

Slightly different scenario

You are still in the medical profession. You just moved into a new neighborhood. It is summer. One of your new neighbors is having a pool party.

You and your beautiful spouse mosey on over after the party is well underway.

Scanning the crowd with your professional eye, you see a young man who is way more impaired than the hour and potency of the punch indicates. You stroll over.

You see a medical alert necklace on the young man. His date assures you that he is just really drunk. Bending in close, you smell the distinctive, fruity aroma associated with ketoacidosis.

His date assures you it is the smell of the pineapple juice in the punch.

Do you ruin the party?

We run into these situations all the time. If we were Biff or the "drunk" date, we hope the medical person decides to ruin the party.

When do we act on evidence? When do we let it play out without acting?

Rules of Evidence
Humans are very imperfect processors of data.

While the rules or hierarchy of evidence might vary by profession, there are some commonalities.

Following common Rules-of-Evidence tells you when it is time to ruin the party.

Hear-say evidence is always low quality
Your older sister said "Biff gets panic attacks." That is hear-say unless your sister is a psychiatrist or a Ph.D. psychologist and Biff is her patient. And if Biff is her patient, there are rules against her dating him because emotion is pollutes the ability to make meaningful diagnoses and design effective therapies.

You hear the shortness of breath. Biff tells you the pain he is feeling. You can see the color of his lips.

You decide to ruin the party because the evidence you collected first-hand almost always out-weighs second-hand information.

Regarding the pool party. A lot rides on what the medical alert necklace says. I bet it says the "drunk" is diabetic. If you can collect that last bit of information, then you make the party a memorable one rather than a tragic one.

We tend to accept second-hand information when it validates our fears or hopes or sense-of-specialness.

Being the first to scoop information is a very common way of feeding that sense of being special. "Hey, look. I am connected. I was the first to know." Facebook would collapse overnight if members waited until the third of forth report before chiming in.

Bill of Rights
The Sixth Amendment to the US Constitution enunciates our God given right to face those who accuse us.

Anonymous accusers can claim more "knowledge" than they actually have. They can be "false witnesses" and claim to witness events that they heard about second or third or fourth hand. If you were to tot up all the people who claim to have been on the raid that bagged Osama bin Laden, you would have needed a dozen 747s to deliver them all to the compound. Most of them were not there.

Another problem with anonymous witnesses is that we cannot evaluate any "moral hazard" that might bias the witness's testimony. Suppose the witness is an ex-romantic partner and they parted on a sour note. Suppose the witness is a business competitor who stands to benefit if the accused is taken out-of-play?

What if I am my own anonymous witness? Would you expect an unbiased testimony regarding my character? If you knew I was providing my own character reference than you would rightfully discount the testimony.

Pattern Seeking
We are imperfect processors of data. Our brains tend to fail in predictable ways.

For instance, our brains are pattern-seeking. As soon as our brain finds a pattern that comes close to matching the requirements, it stops looking.

Your sister wants to stay at the party. Biff does not like crowds. Your sister's dog pants during thunderstorms. "Biff is having a panic attack". Problem solved.

The "drunks" date doesn't want to make a scene. Being drunk is temporary. Having diabetes is a life-time thing. She WANTS him to be drunk. It fits all the symptoms she is capable of assimilating. Problem solved.

Vivid vs. Boring
Our brains (except for Boris who is half Vulcan and half Cray Super-computer) can process very limited amounts of data at a time. Quick: What is 12,345 + 43,210?

We take out a piece of paper, write down the problem and then ask our self "What is 5+0?" and then chew our way through the problem one step at a time.

We compensate for our limited mental capacity by recursing, that is, following a simple procedure (like sorting) over-and-over until we reach an acceptable solution.

Think of it as combing snarls out of your hair. You take a small hank of hair. You start near the end. You comb out the end 1/2". Then the end 1". Then the end 1-1/2" and so on.

The problem is that the information that is most readily accessible to our mind is the vivid information that we connect with emotionally.

EMOTION is the native operating system of our brain. LOGIC is a foreign system that is emulated in the native system. LOGIC is always slower.

Because our brains process EMOTIONAL data more quickly that LOGICAL data, and because our brains stop working when we reach an OK solution, the EMOTION driven 7/8ths of the brain nearly always finds an OK solution sooner than the LOGICAL side.

The net effect is similar to compound interest. The EMOTIONAL side gets amplified 1.10^n while the LOGICAL side is attenuated 0.9^n where "n" is the number of iterations needed to completely comb out the tangled hair.

Because of EMOTION's ability to dominate decisions and because it is almost impossible to be objective about EMOTION, EMOTION is only used as a tie-breaker.

If you are hiring somebody, sort through the candidates using boring, LOGIC and weed out every candidate who is unlikely to be a good fit. Then, use interviews and EMOTION to pick from the much smaller pool of fully qualified candidates. At that point, you could throw a dart and be sure you would hit a good one.

Looping back to the two vignettes at the beginning of the post
The medical person assessing Biff and the "drunk" is trained to give ZERO credence to everything except the hard evidence.

The fact that others made bad diagnoses before you arrived. The fact they will get angry if you blow-off their assessment. The fact that your wife is giving you the stink-eye as you quarter your way toward the "drunk" like an English Pointer quartering a stubble field searching for a covey of quail...they have zero bearing on the medical facts.

Isn't that the foundation of civilization? We have many people who are trained to be color-blind to emotion and unswerved by empathy so they can do their jobs in the best possible way.

Getting to cases
The twenty-something who objected to "old people" saying "America is the greatest" isn't getting any sense of "special" by accepting that viewpoint. She is not the first to be in-the-know. From the standpoint of a recent college graduate who invested a significant percent of her life-to-date, there is little satisfaction in affirming something that has been "known-and-provable" since 1918.

The same tendency that leads people-with-means to buy a Mercedes when everybody else is driving a BMW or a Maybach when everybody else is driving a Mercedes drives the twenty-something to tout Scotland over the US, for instance.

In one way, it is a good thing that she does not blindly accept Aggie writing, "Hey, I have been in twenty countries in my lifetime and none of them are as good as their tourist propaganda claims. In fact, none of them can hold a candle to the good, old USA"

Another point in her favor is that she accepted first-hand testimony from fellow college students who came from places like Scotland, The Netherlands, Ghana and the like.

On the other hand she clearly accepted the second (or third of fourth-hand) testimony of a T/A who was projecting "I am in the know, the US sucks".

Tangential story
I worked with an engineer I will call Mike. Mike was from Romeo, Michigan. Mike was at a party at a prestigious University in Michigan shortly after he graduated from college.

At the time, the dominant contractor in the Detroit Metro area had a Mediterranean last name. Because the last name was long, his company went by his first and middle name. For the purposes of this story, let's say his company's name was "Don Mario".

The contractor lived in Romeo.

Talk at the party turned to corruption in the construction business. One of the drunk frat boys opined that Don Mario probably won most of his business via kick-backs, bribes and intimidation.

Our intrepid friend from Romeo saw an opportunity to show himself to be a man-of-the-world. "Oh, hell. yah! I am from Romeo and I know ALL ABOUT HIM. You don't know nothing. The stories I could tell you would curl your hair."

There was a young lady at the party. It was a girl he vaguely remembered from high school. She graduated a year behind him. She was cute. She tried to inject "You can't know that. That is not true. Don Mario has to deliver performance to keep getting contracts...."

Our hero hooted her down.

The next morning, after he sobered up and the worst of his hang-over left him he had an epiphany. He grabbed his yearbook just to be sure. The young lady had a very long, very Italian last name; something like Carraracannaberra.

Mike had this cold lump in his gut. He had just spent an appreciable portion of the previous night absolutely disrespecting a multi-millionaire (like $100M) in front of his beloved daughter.

If he had not been so desperate to impress the frat-boys (boys he would never see again), but rather had defended her father using the simplest rules-of-evidence...something Don Mario's daughter had attempted...Mike could have been waking up in her bed and have an inside-track at being Don Mario's son-in-law.

But the need to fit in, the need to be "special" within that ephemeral group was his undoing. He let his emotions drive his bus. And he was an engineer.

Grim prognosis
  • Stories are more vivid than exposition. That is why I rely heavily upon stories to make a point or relay information. People listen to, and remember stories.
  • Face-to-face is more vivid and interactive than video.
  • Video is more interactive than pure audio.
  • Voice is more interactive than the written word.

Vivid data overwhelms even logical minds unless concrete barriers are positioned to partition it away from us.

Voices and stories that honor the knowledge gained by generations of painful experience are being silenced.

Voices and stories and video that are dripping with emotion and fantastical solutions are being promoted on the internet.

Good judgement comes from painful experiences.

Painful experiences come from bad judgement.

The inability of younger people to accept the authority of history and Tradition dooms them (and ultimately us) to avoidable, painful experiences.

And the younger people will wail "I just didn't KNOW!" while their parents will say "We tried to tell you but you would not listen."


Quest: Vimy Ridge redux


Sammy was embarrassed. Not just personally, but more importantly, professionally.

Sammy was a nerd.

As a young man, before Ebola, he read classic Science Fiction. One of the books in the pantheon was Gordon Dickson’s book Dorsai.

Dickson pointed out that battles in the vastness of space could only happen by mutual consent. The distances were so great and even massive space battleships were like motes of dust on the far side of the Pacific Ocean.

General Spackle was fond of comparing his forces to the forces of World War I. In some ways it was a valid comparison. The technologies were not that far apart.

Where Spackle was way off base involved the number of fighters per mile of front.

One battle in WWI that Spackle demanded that his Lieutenants study was the Battle of Vimy Ridge. A force of Canadian soldiers attacked the German force holding Vimy Ridge. That battle took place from April 9 through April 12, 1917. So considering the differences in climate, it was an almost exact analog for the battle looming at the end of April in Michigan.

The ridge was approximately 4.5 miles long and offered favorable elevation for the defenders. The difficulty in excavating trenches, on the other hand, mitigated against the defenders.

The defenders were under-manned (estimated to be 70%-to-80% of the T-o-O) and resupply was fifteen miles to the rear. The Buffer-Zone was also under-manned according to the original Tables of Organization and resupply was thirty miles to the rear.

The parallels continued. Spackle was willing to let the attackers earn acres of ground as long as it placed his fighters in advantageous positions around the salient. That is, Spackle encouraged his defense to bend as long as the attackers paid a heavy price in blood and as long as the attackers pushed themselves into a “fire sack”. That was a strategy similar to the German defense plan.

Where the two scenarios diverged wildly was that the Germans had 8000 defenders per mile of front and the attackers threw five times that number per mile at the defenders.

And those were just the primary fighters. Reserves were in addition to that number.

The Buffer-Zone, in contrast, had about five hundred fighters defending thirteen miles and many of them were biased toward the north end where the interstate was.

Essentially, Spackle had forty fighters per mile defending the southern eight miles of front. And they were not stacked on the line. Rather, they were deployed-in-depth to support the bend-but-dont-break philosophy Spackle was promoting.

It was a recipe for disaster if the attackers achieved surprise and poured a significant amount of their force along a second axis of attack in the sparsely defended southern two-thirds.

The operative verb is “achieved surprise”.

Sammy knew that sensors were cheap. They were little more than a beer can or “Red Solo” cup with a magnet and coil. Amplifiers were cheap. Batteries and small solar panels were cheap. Software is replicated at almost zero cost.

The expense in sensors is in the installation.

The sensors east of the Buffer Zone had been installed by Sammy, himself. It had been done in the dark of night under arduous conditions.

Comparatively speaking, it would be a walk in the park to install lines of sensors within the Buffer-Zone. It could be done in the light and fighters could be trained to do it.

He started to transmit an order to Dmitri for sensors to populate three picket-lines within the Buffer-Zone. One along the eastern boundary. One along the western boundary. And one along the spine of the land between the West Branch (ironically on the east side of the Buffer-Zone) and Doan Creek.

Then a voice from the past reached out to him. He could not remember where he had heard or read it, but it was crystal-clear “Failure to plan for failures is to passively accept all consequences that spring from that failure”

Sammy doubled the order of seismic sensors.

There as absolutely no reason to not deploy picket-lines of sensors WEST of the Buffer-Zone, between the Buffer-Zone and Capiche.

The worst-case scenario, from a failed mission standpoint, would be if the defenders failed to engage the attackers and they punched through the defenses and then had freedom-of-movement in the rear of the Buffer-Zone. The Buffer-Zone would lose logistical support from Capiche and nobody would know where the attackers were or where they would attack next.

Then, as an afterthought to his afterthought, Sammy added a gross-lot of acoustical mics to is order. He also included a very short note that captured his thinking. Anything Dmitri could do to pre-program the equipment for the mission would be much appreciated.

Sunday, July 12, 2020

Persimmons

Persimmon seedlings. The seedlings in this photo receive about four hours of light shade a day, followed by heavy shade.
Insanity has been defined as doing the same thing over-and-over again and being surprised when the outcome does not improve.

I moved about thirty persimmon seedlings into a row in the garden about two weeks ago. They promptly fried. I had about four survivors.

Southern Belle is a project girl. She was more than happy to help me move some more.

We moved them last night. I did the shovel work. S.B. did the bending over work. We planted them in fours, about 6" apart in each direction.

All told, we moved about 80 seedlings.

This time, I covered them with a thin, fluffy layer of wheat straw.

If you embiggen and look closely, you will see two seedlings hiding down there.

My hope is the seedlings will grow through the straw even as the mat compacts.

The picture in my head is that they will receive much less sun loading, almost zero wind (which can dry them out in a New Yawk minute) and that the mulch will hold the humidity from the ground close to the leaves.

It will take a few weeks to know if I am successful. By then, the roots will have integrated into the soil and the seedlings should be growing with joyful abandon.

Parentage
I have persimmons on two ends of my property. The south end has seedlings and grafts. Some of the seedlings are male.

On the north end of my property I have all grafted trees and two males. One male is "Szukis" which is a switch-hitter. That is, it is a male that produces some fruit.

The difficulty in growing seedling persimmons and hoping for good fruit is that it is impossible to know if a given male is likely to produce "spitters" or "smile-and-swallow" fruit. "Szukis" produces good fruit and has a track record of a high percentage of good progeny, especially when paired with Juhl or similar persimmons.

The other male branch is one of Jerry Lehman's selections. He was kind enough to sell me persimmon scions. I told him I had a half-azzed interest in breeding persimmons and asked if he would be willing to sell me a male that he thought had potential. The exact experimental number of the male that he sent me is lost in antiquity, but the branch is still alive and grafted into a tree that is 90% Morris Burton.

These seedlings are almost certainly to be Morris Burton pollinated by Jerry Lehman's male.

The story behind Morris Burton is that a pig farmer named Morris Burton noticed a trail his hogs had beaten to one, particular, wild persimmon tree.

He tasted the fruit and decided it was good enough to propagate.

The first few years that Jim Claypoole was hybridizing, he used Morris Burton heavily (it is a parent of Dollywood) but then he drifted away from it.

Morris Burton fruit are smaller than most cultivated American persimmons. It is slightly more likely to produce progeny that have cosmetically ugly specks of tannin.

And then Lehman noticed that a few of the Morris Burton offspring could be eaten before they were dead-soft. That is, they were so low in tannin distributed at the cellular level that they could be eaten like tiny apples.

Suddenly, Morris Burton pollinated by one of Lehman's "elite" males sounded like a wonderful rootstock for grafting.

Suckers
The thing about American persimmons is that they sucker profusely. You might plant a grafted persimmon but in time, your tree will be overwhelmed by a thicket of suckers that spring up from the rootstock.

Most of these persimmons are likely to end up in plantings to attract deer. I want the rootstock to at least have the potential to produce good-to-excellent fruit for human consumption.