A heartbreaking scene took place at a local coffee shop. This locale is the type where, if you see a sign for a certain septuagenarian, your heart leaps for joy for the contrarian. In the same way, if you see a sign for a certain other septuagenarian in a rural area, you’re pleased with the contrarian there. The scene took place in such an area of geographical dogma.

I sat there reading Liberties (https://libertiesjournal.com/) when a familiar felllow frequenter and octogenarian entered. This coffee shop greets both him and me by name when we enter because we’re “regulars”.

This cane-walking, senescent man is a character. He’s interesting. I’ve heard him burst out in random hums as we, separated by 5 decades but together in the demographic of loner coffee drinkers, sit at this locale. He once answered the requisite, “how are you?” With, “I’ll tell you in a second!” I identify as an old man and I love watching this eccentric character.

In this scene, he walks in. I’m giddy with expectation as to what wisdom, goofiness, or quip will come.

After he orders all the cookies and a small coffee, I don’t think it matters what or how he orders now, it’s all fascinating. He said something about how many cookies were left and he might as well take them all. I have stopped reading and have turned my whole body in his direction as to not miss anything. The sweet lady behind the counter asks what he used to do.

“Well, at 18, I was drafted into the Korean War for ten years”. The meaning of this was lost on everyone. I was devastated. Everytime the topic of drafts comes up, every male I know has forgetten that they signed some card at 18. And every female I know is utterly nonplussed by the action. Our lady behind the counter was nonplussed. Our old man went on, “Then I went to seminary for 5 years, took 5 years of Greek, took 5 years of Hebrew. Discovered after 5 years I had more questions than answers, and figured I was unfit to be a minister.” What gold! That’s the candor, wit, and wisdom I was expecting! I’m glad I’m listening.

At this moment, some young schmuck, who must wait in corners of rooms just waiting to be called upon said, “Thank you for your service.” Thank you for your service. Larry David excellently panned this phrase in an episode of Curb Your Enthusiasm. I despise this ridiculous rite in our civic religion of statism. Fuck off. Our old man had a better response, “Well it ruined my life.”

The sweet lady behind the counter let an “aw” escape. She still doesn’t know every 18 year old male signs away his life on a draft card. In the men’s rights movement, they refer to this as “male disposability”. Meanwhile our schmuck who came running to shout a platitude is back at his seat. How is no one fascinated by this guy!? How is no one pissed off!?

I think our old man senses that no one gets it. He mentions how this country has been in a constant state of war since WW2. Then he makes a mistake that no one should make in this geographic dogmatic place. He says he’s sad Trump lost because war lessened for a moment.

Now that’s interesting. If you’re sensitive to drafts like I am or your life has been ruined by drafts, then you may have an anti-war priority. You may also have a priority for kindness. As a statist, you must weigh your priorities and give some septuagenarian nuclear codes. Glad that’s not on my conscience. Nuance is dead. Our sweet lady behind the counter did not find this interesting, said she didn’t know about all that. Our old man asks about a more comfortable chair at the “executive table”. I can’t get over this guy. He’s hilarious. All the comfy coffee seats were full, he had to sit at a sort of meeting table. He called it the executive table. He’s so interesting and everyone around him is dogmatic and boring.

We Believe

In this house, we believe:

in putting banal signs in our yard.

minorities are our pets and are solely defined by how we* treat them.

men’s rights don’t matter so much. We need to keep lowering expectation on women because that’s sexist, I mean sexy.

no human is illegal unless it’s convenient to ignore that fact.

science is real except for HBD of course.

in tolerance for things we agree with.

diversity is important for inmutable traits like hair color et cetera. Different ideas are intolerable. Did I mention we’re not racist?

*I know this sounds self-absorbed and white-supremicist, but I swear we’re anti-racist.

Started Amazon Flex

My last post was written in February of 2018. I went to the Winter Olympics in Pyeongchang, I visited San Diego June of 2018, forgot about the post, questioned my life choices, and started driving for Amazon Flex.

Amazon Flex is the Uber of packages. Using an app, you accept routes within the hour and deliver packages. There are Amazon Prime routes with your normal packages, Prime Now with 2-hour delivery, and Whole Foods with 2-hour delivery.

Uber and Lyft have gone by the wayside because Amazon Flex is four times better statistically.

Let me show you.

I have a spreadsheet where I enter daily data for time, miles, routes, expenses, stops, packages, earnings, tips, et cetera. Everyday, I meticulously track what I’m doing so that the spreadsheet calculates what and how I’m performing. In my previous posts, I showed how Tuesday was the worse day for ridesharing; well, my pivot table says Wednesday is the worst day for packages. The usual caveats and confounders should be noted here like “in my market”, and “for my schedule” and so on.

Over the past 404 days, I have worked 307 days compiled of 1460 hours for Amazon Flex, or 60 days worth of delivering packages. I have averaged $27.36 per hour and $1.78 per mile in revenue.

After my first 70 days of Amazon Flex and delivering 81 Amazon Prime routes, I quickly realized driving to the airport for an Amazon Prime route was not efficient. I started phasing out Amazon Prime, Lyft, and Uber, and I started prioritizing Prime Now and Whole Foods routes. I have done those routes exclusively for 237 days. The Prime Now and Whole Foods routes include tipping!


With the above table that is dynamic with whatever data I enter daily, I was able to prioritize Whole Foods routes. I also used a dynamic pivot table with highlighting rules to show me performance over day of the week.


As you can see not only is Wednesday the worst performance day, but also Wednesday was the day I did the least number of routes so perhaps a self-fulfilling prophecy; however, if I count how many days I worked each day of the week, Wednesday is right in the midst, so I don’t think it is serial correletaion.

Again, a lot of caveats. I didn’t do Amazon Flex before 11am. Perhaps morning routes would change these stats. The few morning routes I did, it seemed efficiency was way down, but I have no data to back this up. While my hours and miles include my commute; they don’t include time of day. If I could extract a CSV from the app data on the phone that contains what time of day these routes occured, I might be able to glean something.

All that being said, let’s incorporate costs and see how much money I’m making doing Amazon Flex.


Now that’s a good model right there.

Compared to ridesharing with Lyft and Uber, I am better off by a factor of 4 in regards to hours and a factor of 1.7 in regards to miles.

What is fascinating is that my constant term has remained relatively the same between ridesharing and Amazon Flex-ing. That constant term represents my car. When I step into my car to work at the beginning of the day, I am $30 in the hole for expenses. I have to pay my way out of gas, insurance, registration, cleaning, etc before I can start making a profit.

Added Uber

-Written February 2018-

I have now rideshared for 643 days total. After the previous post and on the 408th day, I started driving for Uber in addition to Lyft. The added efficiency is remarkable. While waiting for a ride, not only have I doubled the chance of receiving a ride request, but also increased my discriminatory power. Why accept the ride 5 miles away when there is a request less than a mile away?

Methodology remains the same. yada, yada, yada…

One problem this time around is separating the time spent ridesharing that day between the two platforms. For instance, I may have ride-shared 5 hours on some day, but the time spent online on both platforms does not add up to 5 hours. The minutes multitasking both platforms will add up to 6 hours and 2 minutes precisely. I say precisely because the month I tracked how long I was actually ridesharing versus how long the apps said I was online resulted in a ratio of .83 to 1. I am reluctant to add this ratio to my total hours spent ridesharing, because I don’t really know the accurate number and I want the model to be on the lowest possible end.

In the same vein, I don’t separate from personal tire-use and ridesharing tire-use. In the same vein, I’ve added more costs to the model this time around. I have added the following costs: brakes, air filters, yearly state-mandated inspection, and windshield wipers.

For this model, I am also removing the mentoring money and the time spent mentoring. Lyft did away with mentoring a year ago and the quality has decreased as a result. It seems customers were not willing to pay for the added interview step for drivers.

For the 643 days I rideshared, I made $4.95 per hour and $0.43 per mile, with a constant term of -$33.25.

These 643 days of observation include the 408 days of Lyft exclusive data with Uber data being zeros. If I run the model on 235 days of Lyft and Uber data, the model loses significance. I would think it would be the other way around. That being said, with a p-value of less than .1, the 235 days of ridesharing both Lyft and Uber, I made $2.34 per hour and $0.68 per mile, with a constant term of -$36.56. The per hour term lost significance. The model seems to be pulling more from the per hour variable than the per mile variable and to be giving miles more credit. Weird.

I can p-hack and make the model significant, but the results change to $2.12 per hour and $0.79 per mile. That per mile variable is more than ridesharing charges per mile! That makes no sense. I don’t know what is happening.

If I remove all costs and make the model about revenue instead of profit, I made $6.56 per hour and $0.42 per mile. That confirms that all the costs are being pulled from the per hour variable. Biscuit crumbs!

Driving for Lyft

Hi, I’m a Lyft Driver and Mentor.

I have driven for Lyft for exactly a year now and I want to share some findings.

Before I get into the data, let me say, in general, I don’t think it’s an economical job. It is time and car consuming. As a supplemental and temporary job, driving for Lyft is great. Hence the name ridesharing, it is best for sharing a ride for which you are already engaged. When the getting of rides is good, it is quick, fun money. When it’s slow, it’s painful. There seems to be diminishing returns.

If you have the time and the capital, it can pull you out of a jam. Lyft saved me from demise and I am forever thankful. If you are willing to sacrifice future car use for present cash, then go for it. I suppose I just summed up capitalism right there. Do you spend now or save for later? Your situation, capital, cash on hand, and city could change the economics of ridesharing.

I made $36,968.15 ($30,720.18 after costs outlined below). I gave 2,037 rides. I mentored 515 new drivers. I got $2,308 in tips. I drove 35,000 miles.

Based on the Lyft service fee charged to passengers and the Lyft fees from rides, Lyft has made $6,087.25 from my services. As a mentor, driver, ambassador, and promoter of Lyft, I would guess I’ve added about $20,000 value to Lyft.

I have spent almost $400 on cleaning/vacuuming my car for passengers. That averages to 20 cents per ride.

For the past 366 days (it is a leap year), I have worked Lyft 312 days.

For those days, I have been online and ready for a ride an average of 5 hours and 41 minutes per day. I am a Lyft Mentor which means I communicate, train, and recommend new drivers to Lyft. So with Mentoring and Driving, I have worked 7 hours and 50 minutes a day. I worked an average of 54 hours a week. I worked 46 Mondays, Tuesdays, Wednesdays, and Thursdays, 45 Fridays and Saturdays, and 38 Sundays.

I have kept a spreadsheet of lot of data. My miles driven and hours spent includes travelling to and from home. It includes being stood up, cancelled on, and burned out. I debated including money and time spent eating while on the road and money spent on phone bill. Before Lyft, I was a restaurant connoisseur so I believe it is a lifestyle choice and not a cost imposed by working Lyft, so I have not included my Big Macs, Cherry Limeades, or Crunchwrap Supremes. In the same way, my cell plan, while being extensively used by Lyft, is not a direct result or cost from my Lyft work.

For the following models and regressions and data, keep in mind that I have taken into consideration a lot of costs. For instance, my tires, oil changes, and gas are more expensive than the average. Grip, 5W-40 synthetic, and 93 octane are expensive, but my car saves me from having a personality or being genuine. Also, I have not separated my tires or oil changes between personal and work use so these results are on the lowest possible end. I have separated between work and personal gas and miles.

For the linear regression, I’m using ordinary least squares because the variant least squares are the only ones I know. I’m very academic. What this means is I’m assuming a lot. I’m assuming my factors are unbiased and not themselves correlated together (which isn’t true). I’m assuming I’ve used my head to determine the best variables. I’m assuming linear variables. Perhaps the wear on my tires changes exponentially when there’s rain on the ground and I’m shifting to 2nd gear to do donuts. Whatever. I’m also assuming my error or residuals are evenly distributed and not skewing my data and sending mixed signals.

Let us look at what I’m making. I want to see what my profit (x) is per mile, per minute, and per mentor session. The slopes (β) of these variables will tell me the rates.

x = β(total distance) + β(total hours) + β(mentor sessions) + ε

As a function of income, my variables are trip money, mentor session money, tips, cancellations minus Lyft fee, gas, oil, tires, insurance, registration, and cleaning

f(x) = trip money + mentor monies + tips + cancellation – Lyft fee – gas – oil – tires – insurance – registration – cleaning

Running this, I can see how much I’m making per mile, per hour, and per mentor session in pure profit after all my income and costs.

For the 312 days I worked for Lyft, I made $4.49 per hour, $0.21 per mile, and $31.84 per mentor session with a constant term of -$7.30. (Regression 1)

I thought looking at each individual day might lend some insight. Both the regressions and pivot table do not seem to say anything. A comparison of profits per weekday:

Mon Tue Wed Thur Fri Sat Sun
C -$2.1* -$2.56* -$5.72* -$14.22* -$15.74* $4.6* -$14.6*
Hours $3.32 $4.36 3.04* $4.57 $3.28* $9.17 $6.93
Distance $0.27 $0.14 $0.29 $0.25 $0.36 -$0.05* $0.12*
Mentor $29.69 $32.78 $29.51 $34.01 $33.51 $29.43 $34.18

*Statistically insignificant.

A pivot table comparing days of the week:

Monday Tuesday Wednesday Thursday Friday Saturday Sunday
Avg. Mentor Session 1.58 1.57 1.76 1.52 1.41 1.65 1.19
Total Mentor Session 68.00 66.00 74.00 64.00 58.00 71.00 43.00
Avg. Total Hours 7.24 7.60 8.33 7.72 8.40 8.62 6.89
Avg. Earnings 107.39 108.98 121.03 109.73 123.10 141.09 101.54
Avg. Earnings per hour 14.83 14.34 14.53 14.21 14.66 16.37 14.73
Avg. Total Distance 97.76 104.88 115.01 98.73 123.86 153.84 108.36

Variables and regressions:

Date – the date mm/dd/yyyy
Weekday – where 1 is Monday and 7 is Sunday. I know Sunday is the first day of the week but Monday is the first day of pay period
Rides – # of rides
Gas – my gas per day
Gallons – gallons of said gas
Cancellation – under certain criterion, I make $5 per cancellation
Driver mode – how long I was online and ride ready
Total hours – driver mode plus mentor sessions and other time used working for Lyft that I estimated after each day
Trip time – hours on a ride
Trip money – money made on trip
Trip fee – Lyft takes fee for providing platform, support, insurance, and a host of other amenities
Tips – extra monies for smiling and putting groceries in passenger’s homes
Mentor session – how many mentor sessions I did that day
Mentor monies – how much I made doing the mentor session
Acceptance rate – percentage of how many rides I completed versus ignored/cancelled
Trip distance – miles on a trip
Total distance – total miles driven that day
Insurance – car insurance per day
Registration – car registration per day
Oil – oil change cost per day
Tires – tire use per day (determined by miles driven)
Cleaning – cleaning fees for keeping my car nice for the dirty passengers.

Regression 1:

Sample: 1 312        
Included observations: 312        
Variable Coefficient Std. Error t-Statistic Prob.
C -7.303133 3.495254 -2.089443 0.0375
Total Hours 4.492697 0.558361 8.046217 0
Total Distance 0.211364 0.028294 7.470388 0
Mentor Session 31.84723 0.894924 35.58652 0
R-squared 0.824132 F-statistic 481.1031
Sample: 1 312 IF Monday        
Included observations: 46        
Variable Coefficient Std. Error t-Statistic Prob.
C -2.104648 8.992235 -0.234052 0.8161
Total Hours 3.320041 1.360093 2.441039 0.0189
Total Distance 0.268772 0.077631 3.462171 0.0012
Mentor Session 29.69494 2.45127 12.11411 0
R-squared 0.806808 F-statistic 58.46661  
Sample: 1 312 IF Tuesday        
Included observations: 46        
Variable Coefficient Std. Error t-Statistic Prob.
C -2.564712 8.44325 -0.303759 0.7628
Total Hours 4.358777 1.166166 3.737698 0.0006
Total Distance 0.144683 0.066297 2.182339 0.0347
Mentor Session 32.77808 2.027303 16.16832 0
R-squared 0.865487 F-statistic 90.07954  
Sample: 1 312 IF Wednesday        
Included observations: 46        
Variable Coefficient Std. Error t-Statistic Prob.
C -5.715844 10.72408 -0.532992 0.5968
Total Hours 3.035765 1.507945 2.01318 0.0505
Total Distance 0.286691 0.078762 3.639972 0.0007
Mentor Session 29.51292 2.487967 11.86226 0
R-squared 0.816557 F-statistic 62.31818  
Sample: 1 312 IF Thursday        
Included observations: 46        
Variable Coefficient Std. Error t-Statistic Prob.
C -14.21771 8.236708 -1.72614 0.0917
Total Hours 4.574548 1.283521 3.564062 0.0009
Total Distance 0.246761 0.067743 3.642595 0.0007
Mentor Session 34.01033 2.068539 16.44172 0
R-squared 0.88572 F-statistic 108.5059  
Sample: 1 312 IF Friday        
Included observations: 45        
Variable Coefficient Std. Error t-Statistic Prob.
C -15.74197 9.362551 -1.681376 0.1003
Total Hours 3.275764 1.648024 1.987692 0.0536
Total Distance 0.355421 0.089613 3.966191 0.0003
Mentor Session 33.50861 2.513927 13.32919 0
R-squared 0.846305 F-statistic 75.25398  
Sample: 1 312 IF Saturday        
Included observations: 45        
Variable Coefficient Std. Error t-Statistic Prob.
C 4.602541 11.21363 0.410442 0.6836
Total Hours 9.168054 2.3874 3.840183 0.0004
Total Distance -0.046804 0.108577 -0.431069 0.6687
Mentor Session 29.43256 2.689396 10.94393 0
R-squared 0.759721 F-statistic 43.21166  
Sample: 1 312 IF Sunday        
Included observations: 38        
Variable Coefficient Std. Error t-Statistic Prob.
C -14.60224 8.821457 -1.65531 0.1071
Total Hours 6.927102 1.714551 4.040184 0.0003
Total Distance 0.124461 0.083474 1.491006 0.1452
Mentor Session 34.18317 2.540284 13.45644 0
R-squared 0.849268 F-statistic 63.85534


My driver rating is a 4.92 out of 5. For 13 weeks I maintained a 4.95+ rating. I am not happy about my 4.92 currently.

I have been flagged for safety, friendlieness, and cleanliness six times. I have been flagged for navigation nine times.

93.9% of my rides were 5 stars. 10.58% of those 5 star ratings bothered to leave me a comment.

I received 5 negative comments. Here they are:

“I NEVER RODE WITH YOU SO WHY AM I CHARGED? I couldn’t find you nor was there a number that was answering. No no no.”
“He literally stayed in the most congested lanes of trafic.”
“Driver did a good job however the quote was more than $10-$12 driver took the long way”
“Arrival time. Took fairly long to come. But very friendly guy. Enjoyed the ride.”
“Nice guy”

Here are my 163 other comments:

“Pleasure meeting you! Thank you!”
“Super nice and friendly! Very safe driver.”
“Awesome guy very friendly”
“Great convo. Wonderful driver !”
“Very friendly”
“Great conversation”
“Very personable and nice/friendly. Knee”
“Nice person”
“Personable driver very comfortable environment to ride in”
“Awesome driver and awesome service!!!”
“Super friendly you’re the best :D”
“Thanks man”
“V cool. V friendly.”
“Good to talk to”
“Such a delightful ride!”
“The SAAB!”
“James is great. He’s very personable. We discussed happiness and he was very insightful.”
“Great ride”
“Awesome dude.”
“Cool guy!!!!! Personable, and inviting!”
“awesome driver, very personable!”
“Awesome driver”
“Great driver !”
“James was a delight! Great riding experience.”
“Fast and friendly”
“Great Driver ! Love Him (:’”
“We got him as a driver once again. To the airport and then back from the airport. He is a great driver. Very friendly and quickly and safely drove us back. Thanks again James”
“Great Guy”
“Very cordial. Very good driver.”
“James was very easy to talk to and helpful.”
“Very comfortable to ride with and drives very safe!!! Also very helpful….”
“Awesome and kind !!!!!”
“Wow!!! James was the friendliest nicest lyft driver I have ever had. What an awesome experience. Thanks James!!!!!!!”
“Super nice guy. Thanks!”
“Amazing jokes and such”
“A quiet but substantial conversationalist1”
“Really nice and fun to talk to”
“Best ever”
“Very good driver and good person”
“Really friendly and super sweet.”
“Good conversation.”
“Very whole hearted guy. Great to ride with! Definitely recommend”
“One lyft driver to another thanks for getting my kids home safe”
“Great convo! Derrick Anderson is his fav Pathners player so he is really funny!”
“Thanks for welcoming us to Charlotte!”
“Nice Manuel car nice drive friendly”
“Just a nice ride”
“Awesome driver and better dude”
“He was great!!”
“Very personable guy”
“He was Awesome!”
“Very helpful and friendly!!”
“What a great guy. He had oreos!”
“Safe driving”
“Great driver, very friendly.”
“Local. Knows the streets. Nice guy.”
“The Saab!”
“James is awesome! He’s been my driver a few times. Always in good spirits!”
“He was polite”
“He is awesome!!!”
“James was very friendly , Great lyft driver !!”
“Amazing driver”
“He was real nice and very pleasant. Great driver?”
“Great driving skills!”
“Loved her personality.. excellence service!!!”
“He was very delightful”
“He’s so wonderful and hip!”
“Nice guy!!!”
“Very clean car. Very nice.”
“Best driver I have had yet. Incredibly knowledgable and a very positive person to be around ”
“Very friendly and great navigation”
“Awesome driver and awesome person!!!”
“Great guy”
“Awesome spirit”
“James is awesome and so friendly”
“Nice guy”
“So polite and friendly!!”
“He was on time, helpful, courteous, and very personable.”
“The driver was so nice he help me to put my luggage.”
“Cool guy”
“Awesome guy!”
“So friendly and passionate about the company!”
“He was super nice!”
“Awesome ride, very personable guy and fantastic driver. If you like a good stick shift this guy can whip it.”
“Really nice and funny”
“Awesome !”
“Quick and courteous!”
“Great assistance”
“Awesome guy”
“Cool guy!!!!!”
“Very friendly and upbeat.”
“Great service”
“Super nice guy!”
“Awesome conversationalist. Love that he grew up in Charlotte!”
“AWESOME PERSONALITY!! You’ve got a wonderful driver on your hands so hold him close !!”
“Friendly.. helpful.. REALLY NICE CAR!!.. ID RIDE AGAIN”
“Super funny, great driver!”
“Love your car and your hair!!! Best driver ever!!”
“great customer service and very pleasant”
“All around baller”
“So cool and funny”
“Very nice guy”
“Great communication – got me to the airport fast! Highly recommend.”
“Very personable! Great ride!”
“His speed and flexibility”
“Easy going and kind”
“Personable, thoughtful, and knowledgable. Enjoyed the conversation!”
“Very friendly”
“Great ride and great convo”
“So very nice!”
“Great guy wish I had more to tip”
“Highly recommend James! Hope to see him again on another ride!”
“The conversation, good upbringing, a real gentleman.”
“He’s awesome!”
“Super friendly!”
“Great service”
“Good driver”
“Quick pick up, great convo! Thanks!”
“Great conversation- super nice”
“Repeat awesome driver; hope I get him over and over! Love “L”!”
“So conversational and a good, safe driver!!”
“Great ride!”
“Nice guy. Very pleasant.”
“Very safe good conversation and i love his hair btw”
“Excellent customer service! Great Guy!!”
“Friendly knew correct route”
“Another great Lyft driver!”
“Super nice guy! Made me forget I was supposed to be playing pokemon go on this lyft ride!”
“James was great! Friendly and easy to talk to. Thanks for the ride to the airport!”
“Nice driver”
“Nice guy and great car”
“Very friendly and spoke highly of lyft … We will defiantly use again! Thanx!”
“Good company”
“Cool dude.”
“loved the pink mustache, nice touch! ”
“Great guy!!”
“Great ride !!”
“Friendly and funny!”
“Best driver I’ve had thus far.”
“Very nice!”
“He was safe and efficient.”
“Great guy, very pleasant ride!”
“Thanks!! Definitely going to text you with the number from your car for future rides (if okay). My name is Haley!!”
“Very nice”
“Best Lyft yet! Thanks so much!”
“Super nice guy”
“We felt really safe”
“Wonderful conversation!! Awesome car! :)”
“Very sociable, would love to get him again”
“Have a great evening, James ;)”
“He was good”