Connect with us
Click me!


Math prediction — 1.2M by year-end




(Author’s Note. As an AB graduate in Mathematics, I conferred with a colleague, who is a retired ADB consultant, to validate my mathematical formula. I am presenting a simple layman-friendly formula, which college math students can easily understand. This is pure cold mathematics — no politics, no emotions, no speculations. I am open to corrections from readers. My projections are intended not to sow fear, but to present the mathematical reality, data that can help us prepare for the worst in advance, to enhance our foresight, in case there is a sudden rapid surge. A false positive picture may catch us off-guard.)

Sec. Harry Roque, I hope it is okay to give a slight correction to the figure you gave earlier to media. The UP OCTA prediction is not wrong, just slightly off the mark — 40,000 confirmed cases by the end of June, from 24,000 three weeks ago. In your media interview, an actual figure of 36,438 was given as of Monday, 29 June. But this is not the end of June. The end of June is reflected on the statistics taken at the end of 30 June and/or released on the morning of 1 July, which is 38,000 not 36,400.

The UP OCTA was off by a mere five percent of their 40,000 prediction. That is still a good prediction. They deserve to be praised, not condemned. May I comment that they had no intent to put the government in a bad light. They are just doing their duties as scientists dedicated to serve the nation to contain the pandemic. Thank you, UP OCTA.

I used a simple mathematical formula to project a figure at the end of 2020, pure math, not considering other factors or variables, like continuing shifts from strict to moderate quarantine, sudden spikes left and right, availability of test kits and speed of testing, speed of contact tracing (which is delayed by 15,000 as of 3 July (Inquirer), etc.

My prediction for July of 67,526 is higher by about 11 percent than UP OCTA’s prediction of 60,000. I am explaining here in layman terms the exact mathematical process I used, shown below, so anyone can correct my formula. My mathematical method is simple and straightforward, not sophisticated, not using complex algorithms.

Exact Mathematical Formulation

Assumptions (given): 24,000 positive, 10 JUNE, increased to 38,000 in 30 June (three weeks)

Actual Historical Values


F2 = 38,000, ENDING VALUE

Step 1.

Determine percent rate of increase for the three-week period = R3.

The formula is — END VALUE F2 (38,000) MINUS BEGINNING VALUE F1 (24,000) DIVIDED BY BEGINNING VALUE F1 (24,000)

R3 = (F2 – FI) / F1 = (38,000 – 24,000) / 24,000

R3 = 0.583 rate increase for three weeks.

Step 2.

Determine rate increase R4 for four weeks based on rate increase R3 for three weeks by ratio.

R4 = R3 X (4/3) = 0.583 X (4/3) = 0.777

R4 = 0.777 rate increase for four weeks

Step 3.

Compute the predicted value for July (FJuly) using R4. The formula is –FJuly = F2 X (1+R4) = 38,000 X (1+0.777)

FJuly = 67,526, predicted COVID positive for July (rounded off to two decimal places)

Note that the UP OCTA figure is 60,000, 11 percent lower.

Step 4.

Do the same to compute for the rest of the other months until December 2020.

FAug = FJuly X (1+R4) = 67,526 X 1.777 = 119,993.70, August

FSep = FAug X 1.777 = 119,993.70 X 1.777 = 213,228.81, September

FOct = FSep X 1.777 = 213,228.81 X 1.777 = 378,907.59, October

FNov = FOct X 1.777 = 378,907.59 X 1.777 = 673,318.79, November

FDec = FNov X 1.777 = 673,318.79 X 1.777 = 1,196,487.49, December

My colleague and I agreed that, because the sampling period of only three weeks may be too short, there should be an allowance for error. Also, possible unpredictable erratic spikes within the time period should be taken in consideration. Because all these defy quantification, and historical data is not readily available, we make a QUALITATIVE judgement to give a RANGE rather than an exact value — 1,100,000 to 1,300,000 confirmed cases by December or year end, give or take 100,000 from the exact value of 1,200,000.

Recent spikes have alarmed everyone, and there is a tendency to be emotional — fear or panic — which may cloud our ability to be on top of the situation. Given this cold mathematical prediction devoid of politics or emotions, it is perhaps better to avoid premature relaxing of quarantine, and keep our guard up for the worst, without fear or panic and with determination and political will, avoiding finger-pointing and employing gentle constructive criticism, not harsh condemnation. The government listens. The greatest test for the Filipino is to work together in an extreme crisis.

Reactions to

Click to comment