So I shot a little too high in my predictions last week. Based on all the historical data for holiday spikes and then the double holiday at the end of the year, I thought there would be a three week case spike. But as you will see below, last week saw a case drop in all locations. Potentially a sign of Covid spread decline due to the lack of social gathering holidays. The next ‘big’ holiday is Valentine’s Day, and if you do that with more than one other person….you might be from Utah.
Ooooooh, I should track Utah case spikes after Valentine’s day…..
Overall weeks Accuracy: 79.21%
TRAVIS COUNTY (AUSTIN, TEXAS)
Prediction: 5840
Actual: 4181
Accuracy: 71.59%
Spread Rate for 1/18: 0.85
Austin dropped by 731 cases last week, roughly 15% from last week. If this continues then that would make the week of 1/11 the highest one week increase ever for Austin. Hopefully we continue to see a steady decline in case increases, anything with a spread rate lower than 1.00 is good.
If next week has a spread rate of 0.85 again then we would have 3559 cases. The spread rate would have to stay at 0.85 for 51 weeks before Austin saw 0 cases a week….Shit, that number hurts.
Side note, the last time we saw less than 1000 cases a week was the week of 11/2 with 948 cases. Hopefully we can get back down to those numbers soon.
FORMULA FORECASTS
4 Week: 1.10 Spread Rate = 4619 Cases
3 Week: 1.07 Spread Rate = 4480 Cases
2 Week: 1.05 Spread Rate = 4403 Cases
14 Day Average: 649 Avg = 4546 Cases
7 Days Average: 597 Avg = 4181 Cases
Note: The 7 Days Average dropped by 104 in 7 days.
MY PREDICTION
All of those Weekly forecasts are above 1.00, I think those numbers are a little high. As we know, the social season is over, now comes the depressive “There’s Nothing to Celebrate, Maybe I’ll Take Up Drinking Alone At Home” season.
If the 7 Day Average Drops by 104 again, the case count for next week would be 3453.
If the Spread Rate stays at 0.85, the case count for next week would be 3554.
I might be overly hopeful, but I’d like to think these numbers are close to what might actually happen.
Prediction: 3600
PASSAIC COUNTY (NEW JERSEY)
Prediction: 2000
Actual: 1842
Accuracy: 92.10%
Spread Rate for 1/18: 0.91
Passaic was my closest forecast, but they still hit lower than expected. Last week was huge drop off from the counties highest of 4866 on January 4th. Last week was 2.64 times lower, but still only 174 cases lower than the week of 1/18.
Passaic county would have to stay at a Spread Rate of 0.91 for 79 weeks before there were 0 new cases.
FORMULA FORECASTS
4 Week: 1.39 Spread Rate = 2564 Cases
3 Week: 1.58 Spread Rate = 2908 Cases
2 Week: 0.66 Spread Rate = 1223 Cases
14 Day Average: 276 Avg = 1929 Cases
7 Days Average: 263 Avg = 1842 Cases
These numbers are all over the place.
Note: The 7 Day Average dropped by 25 in 7 days.
MY PREDICTION
If the Spread Rate stays at 0.91, next week cases will be 1683.
If the 7 Day Average drops by 25 again, next week cases will be 1667.
But Passaic hasn’t been consistently below a Spread Rate of 1.00 since June and that was only for 5 weeks. So it’s hard to tell what’s going to happen. I think I’ll stick with my gut and say that the cases will continue to go down. I’ll use my Day Average to help guide me.
Prediction: 1600
SUFFOLK COUNTY (BOSTON, MASSACHUSSETTS)
Prediction: 4600
Actual: 3401
Accuracy: 73.93%
Spread Rate for 1/18: 0.81
Bahston….looks like you finally stopped sending people out to get your Dunkies and decided to brew your own coffee at home. Probably for the best. Dunkin, it’s nothing personal, you know I love you and I miss you.
This is the lowest case increase in 7 weeks. The lowest prior was 1966 during the week of Thanksgiving, the following week had a spike reaching 3810.
FORMULA FORECASTS
4 Week: 1.02 Spread Rate = 3484 Cases
3 Week: 0.99 Spread Rate = 3373 Cases
2 Week: 0.80 Spread Rate = 2737 Cases
14 Day Average: 544 Avg = 3811 Cases
7 Days Average: 486 Avg = 3401 Cases
Note: The 7 Day Average dropped 117 from last week.
MY PREDICTION
If the Spread Rate stays at 0.81, next week’s cases will be 2741.
If the 7 Day Average drops by 117 again, next week’s cases will be 2582.
I suspect the Spread Rate will be below 1.00, however, given that Bostons Spread Rate has consistently been above 1.00 for months, I don’t think the Spread Rate will stay at 0.81. So I’ll split the difference.
The 3WASR (0.99) and 2WASR (0.81) average out to 3055 cases.
Prediction: 3000
RHODE ISLAND (Yea, like the whole state)
Prediction: 6600
Actual: 5229
Accuracy: 79.23%
Spread Rate for 1/18: 0.61
Data Update Notice: When I looked at the Rhode Island numbers, I noticed that there was reported cases on 1/16 & 1/17. When I published the update last Monday, these cases were not in the spread sheet I generate my formulas from. However, when I look at the spread sheet I used for the California vs Florida breakdown, those dates have numbers. My guess is there was a late update sometime last week. My experience with this project is that Rhode Island has a hard time actually submitting their data on time. If I’m being completely honest, this lost data probably made my accuracy better….
The case count for 1/11 was actually 8610. Making this past week a 3381 case drop.
FORMULA FORECASTS
4 Week: 1.05 Spread Rate = 5497 Cases
3 Week: 0.94 Spread Rate = 4905 Cases
2 Week: 0.91 Spread Rate = 4738 Cases
14 Day Average: 989 Avg = 6919 Cases
7 Days Average: 747 Avg = 5229 Cases
Note: The 7 Day Average dropped by 264.
MY PREDICTION
If the Spread Rate stays at 0.61 (unlikely) then next week’s cases would be 3176. (Last time the cases were that low was 10/26, and that was at the beginning of a spike from Halloween that never really ended for the state.)
If the7 Day Average drops by 264, next week’s cases would be 1848. Even more unlikely.
I could be hitting too high here, but I think the spread rate will be closer to 0.90. I say this because 0.61 was a huge drop and is fueled by the previous week being the state’s second highest weekly increase ever. I think the spread rate will even out before steadily dropping again.
Prediction: 4700