I am a person of lists. Each Monday morning I write a todo list for the week which is broken into sections for work, personal, and hobby goals for that week. Well, last week was my birthday, and not just any birthday, last week I turned 50 and what follows is my todo list to accomplish by June 3, 2029:
1. End the decade happily married to Shelli with 4 healthy kids (or more)?!
2. Buy a vacation property I can rent out when not in use. Somewhere Shelli and I love. Something around $1m. Big enough for several kids to stay with us there. Hawaii? Pensacola? Grand Lake?
3. Get a new job (preferably within Cisco). Especially a travel job after the girls graduate NN. Travel with Shelli. Don’t be scared!
4. Mentor someone (not a family member)
5. Start/join a men’s coffee group.
6. Own a performance Tesla (0-60 <= 2.4)
7. End the decade with 3x my net worth of June 2020 in total net worth.
8. Help Austin find a career opportunity
9. Help Evan find a career opportunity
10. Learn the song meaning behind the lyrics of as many songs as I can.
I have created a proximity tracing tool using data exported from Prime Infrastructure. The tool is currently located at http://ciscoProximityTracer.com Please check it out and give me feedback.
To use the tool, the user inputs a “client sessions” report from prime infrastructure. The report has to have the fields of AP name, Session Duration, and RSSI added in, using the exact order shown below.
You can generate for any length of time, I’m currently using either 1 or 7 days when I generate reports. Sorting the output data does not matter. When finished, export to .csv and use that file for the input of the tool.
To use the tool input your file and a MAC address of interest (this would be the person with COIVD19 that you want to track who they came next to). The other 3 fields (max RSSI diff, time resolution, and time offset are optional, defaults are given)
Push the Trace contact button, and the script fires away. It comes back with something like the following:
Simple and effective.
For the week of May 16 MAC 84:85:06:BC:F6:C8 was in contact (<3dBm) with 84:85:06:bf:fc:46 for 1 hour 40 minutes at the AP named MEMO-MAIN-129, and so on.
If you want to play with the tool you can use the PI logfile named “agahome_may28.csv”. and search for the spreader MAC of e4:b2:fb:87:8f:62. That’s my iPhone. Try using different RSSI resolutions and time values.
The tool is located at http://ciscoProximityTracer.com
here is an “a” that loops 3 times then stops.
Over the weekend I have been thinking about reference points for the current COVID19 pandemic and how that has compared to other periods in modern history. I’m talking here about the pure mortality rates from COIVD19, not the economic impacts.
We have all heard horrific things like convention centers turning into morgues in NYC and Italy, and I wanted to flesh out how true that would be on the surface. My thought was to find how many people die on an average non-COIVD day, so any given day in 2019, for example. Here are the numbers I came up with presented without commentary. Commentary to follow the images.
First, an explanation of these four tables and then commentary.
Table 1 shows the causes of death in the state of New York. Currently in 2020 the CIA estimates the mortality rate of the USA to be 830 per 100,000 — meaning that in the year 2020 if you take a group of 100,000 Americans 830 will die at some point in the year (the 830 was a pre-COIVD rate). So using the state population of New York of 20 million, you can see the number of New Yorkers that die in any given day to heart disease (69) or road injury (10). Total daily NY non-COIVD deaths equal 455. On Easter the NY COVID deaths was 755. Meaning COIVD caused a 150% jump in the daily deaths.
What other event caused around a 150% jump in daily deaths? How about Germany during WW2. Using the same logic, in Germany in the early 1940s you should have had about 3,500 Germans dying each day, but the WW2 deaths were another 5,000 daily on top of that.
So in that light New York’s current outbreak is similar in daily magnitude to Germany’s experience in WW2 (not good).
Now, take another place that is not a current epicenter like Italy and NYC have been — take Texas. In table 3 we see daily expected deaths in Texas are 659, and there were 11 COIVD deaths on Easter 2020 in Texas, so the total went from about 660 to about 670, or a 2% increase (unlike NY’s 150% increase). What else has about a 2% increase? The USA’s daily death total in WW2. In the US each day in 1940 about 8,000 people died, and an additional 500 died from the war for about a 2% increase.
Who has the best gingerbread house? Three entries this year:
Team 1: Austin and Addison
Team 2: Neville and Evan
Team 3: Shelli and Emerson
$10 prize for the winner!
Today is a big day in the life of my college football playoff predictor site (playoffpredictor.com). Today is the second CFP committee ranking for the 2019 season, which means it is the first prediction week for the computer model.
what is in store for tonight? According to the model we will have Ohio State, LSU, and Clemson in three of the four spots. No surprises there. One surprise that the playoffpredictor says that differs with the AP committee poll – Georgia, not Alabama is in the fourth slot.
Personally, I think they will put Oregon in that slot – I think they will consider a last-second lost to Auburn on a neutral field much superior to a loss to South Carolina on Georgia’s home field. The problem with the first Prediction of the season is that there is not much bias information, those biases tend to smooth out as the season goes on.
If I were a voting member of the committee I would advocate for exactly what the computer says, which is Minnesota and Wisconsin in spots three and four. No Clemson, no Alabama, no Georgia. Minnesota is obviously unbeaten, but I just don’t see the committee changing on a dime from Voting them 17 to voting them number three. I hope it happens, but I’m not holding my breath. As far as Wisconsin? Well they were destroyed by Ohio State, but Ohio State looks fantastic. Other than that just a one point loss to a decent Illinois team. That certainly just as good or better than anybody else’s one loss who has some quality wins to go along with it. Alabama has nothing in the terms of quality wins. Their best win is Texas A&M, followed by Tennessee, southern Mississippi, and Duke. Yes, Alabama’s second best win was to a team that also lost to an FCS level team this year at home. Ouch.
Stay tune for 7 PM tonight, when we see if the first prediction is 75% correct or 100% correct.
5 years of college football data are in the books and I have enough data now to look at the playoffPredictor biases and make some determinations about habitually overrated and underrated teams that the playoff committee loves or snubs.
A little primer if you need it — each week of the college football season the computer assigns a rating and ranking to each top 25 team. During weeks 9-15 the playoff committee also assigns each team a ranking. Each week we can compare the committee rankings to the computer rankings and make an objective determination about over-ranking or under-ranking.
Using final season average rating biases, here is what we have after 5 years.
Conclusion? The perennial over-ranked team are also the teams that most often make the college football playoff. 3 out of 5 years for Alabama, Clemson, Washington and Baylor. 4 out of 5 years for Oklahoma and Mississippi State.
Interestingly, Ohio State (the only team besides Alabama, Clemson and Oklahoma to make multiple playoff appearances) has zero seasons over-rated or under-rated by the committee.
But there you have it. Conculsive proof that the rich in college football get richer, not because they are better, but because us humans are biased to think of the bluebloods as better.