These next two weeks will be interesting. I anticipate all states to have a spike as travel and socialization is up with the holidays, but I expect the reported cases to be way down as the holidays will disrupt testing and data updates. Also, there is a possibility that people will avoid testing as they plan visiting family.
That being said, last week’s prediction accuracy was 81.21%, raising my total accuracy from 70.14% to 78.91%. Wooooooo!
So my plan for this week is to just quickly analyze the data and trends then find what 5/7ths of that number may be (roughly 71.43%). The thought process being that there will be no testing on the holidays. Between those two numbers we should see the actual numbers fall in that range.
Then again, maybe we will see higher than normal numbers from previous weeks increased travel???
TRAVIS COUNTY (AUSTIN, TEXAS)
Travis County ended up being higher than predicted, coming in at 3051 cases. A 36.7% increase from 12/7s 2232 cases. The spread rate went from 1.03 to 1.37.
This is easily from the increase in travel and socialization. Not to mention “Covid Fatigue”. So I was right about seeing an increase, but I shot too low.
FORMULA FORECASTS
4 Week: 1.16 Spread Rate = 3528 Cases
3 Week: 1.25 Spread Rate = 3807 Cases
2 Week: 1.20 Spread Rate = 3659 Cases
14 Day Average: 377.36 Avg = 2641.5 Cases
7 Days Average: 436 Avg = 3051 Cases
MY PREDICTION
The Travis County Spread Rate has been creeping back up. See the chart above. My closest prediction model last week was the 2WASR (2 week average) with a prediction of 1.19, 2652 cases. That same model is now predicting a 1.20 spread rate or 3659 cases.
The Holiday Prediction (only counting 71.43% of the full prediction) is 2613 cases. That’s a wide gap, almost 1000 case gap. But considering the unpredicatableness of everything…..it is what it is.
So, Travis County Prediction for 12/21: 2613-3659. Maybe call it 3000 to be safe…
PASSAIC COUNTY (NEW JERSEY)
Passaic hit lower numbers than I predicted with 2122 cases. This is the second week in a row that the cases and spread rates have declined. (I predicted a small spike this week). Spread Rates dropped as follows:
11/23- 1.27
11/30- 1.05
12/7- 0.89
12/14-0.82
I see a trend! And according to the graph below, it looks like the spread rate is dropping ever so slowly from the Halloween Massacre.
FORMULA FORECASTS
4 Week: 1.01 Spread Rate = 2137 Cases
3 Week: 0.92 Spread Rate = 1949 Cases
2 Week: 0.85 Spread Rate = 1808 Cases
14 Day Average: 336.71 Avg = 2357 Cases
7 Days Average: 303 Avg = 2122 Cases
MY PREDICTION
The closest prediction formula for the actual case count last week was, again, the 2WASR with a prediction of 2510. But even that was nearly 400 too high. I now believe the new lockdown orders in the area may actually be doing something. Despite the fact that it is hurting many business owners that I know, which is hard to watch.
With this information and knowledge of local politics, we can assume the spread rate will continue to fall. But then there are all the Holiday gatherings. So the two might offset each other. Meaning the cases won’t spike, but they won’t exactly fall off either. For those reasons I will go with the 3WASR of 0.92 or 1949 cases.
The Holiday Prediction of that number is: 1403. But that’s not an exact science. (Who am I kidding? None of this is an exact science.) So let me add to that number. Why? Because I wanna!
My Prediction: 1500 cases (if there are no cases recorded on the Holidays)
SUFFOLK COUNTY (BOSTON, MASSACHUSSETTS)
Suffolk County numbers were also down compared to predictions. New cases for the week were 4023, compared to the previous weeks 3985. The spread rate only increased 0.95%. But according to the graph below, the Spread Rate has been staying in a specific range where the average Spread Rate from the past 35 weeks is 1.04.
FORMULA FORECASTS
4 Week: 1.23 Spread Rate = 4968 Cases
3 Week: 1.33 Spread Rate = 5355 Cases
2 Week: 1.03 Spread Rate = 4135 Cases
14 Day Average: 614.93 Avg = 4304.5 Cases
7 Days Average: 666.86 Avg = 4668 Cases ** This average hasn’t changed much since last week. Its increased, but not by much.
MY PREDICTION
The 7 day average hasn’t changed by much. But when compared to a week of cases, it comes out to a Spread Rate of 1.16, higher than last week. Given the Holidays, I’d say this number is possible.
Now for the Holiday Prediction formula….beep…boop….beep….boop……3361 Cases.
Sure lets go with that: 3361 Cases
RHODE ISLAND (Yea, like the whole state)
Admittedly I am not doing well with Rhode Island. You guys are all over the place. I predicted 8400, seeing as the previous two weeks were in the 8500 to 9000 range and you were going into another lockdown. Your actual was 6293….so good for you for getting that Spread Rate down! (0.74 for those actually playing along at home)
FORMULA FORECASTS
4 Week: 1.02 Spread Rate = 6423 Cases
3 Week: 1.07 Spread Rate = 6708 Cases
2 Week: 0.85 Spread Rate = 5338 Cases
14 Day Average: 1059.86 Avg = 7419 Cases ** This number has been declining since 12/13 where it hit its highest of 1259.
7 Days Average: 899 Avg = 6293 Cases **This number has been declining since 12/10 where it hit its highest of 1328
MY PREDICTION
Looking at the above graph, we can see the Spread Rate is declining since its spike in September. But I suspect the Holidays will stunt the decline. So I will be sticking with the 4WASR prediction of 1.02 or 6423 Cases.
With Holiday Prediction math……4624 Cases