Israel's supporters claim that the United Nations has quietly revised Gaza's death toll. According to the figures published on the UN website on 6 May, more than 9,500 women died and more than 14,500 children died.
Report on May 6
In the May 8 report, the number dropped significantly, with only 4,969 women and 7,797 children dying, a 50 percent drop
Report on May 8
As a result, Israel's supporters began to use the change as evidence to accuse the United Nations of manipulating Gaza's death figures. Of course, Gaza's supporters are either really stupid, really bad, or both. In fact, the United Nations did not make any changes to the death data, but only added a classification - identification.
The so-called identification refers to the presence of a clear identity ID and time of death information on the death record. This segment had only 24,686 people on April 30 (with a lag in identification). The latter 4,969 women and 7,797 children who died were also only identifiable death lists, and half of the data could not be identified — often because the bodies were found unrecognizable and unrecognizable or unclaimed by relatives, which is common in war environments.
It's a very clear picture, but the Jews, including the Israeli Foreign Ministry, are claiming that the United Nations has tampered with the data, presumably in the spirit of being able to deceive one by one.
The Israeli Foreign Ministry said the United Nations revised the death data
In fact, the change in the UN statistics came from the Gaza health department, which on May 4 said that 21,720 victims had complete data, while 11,371 victims had incomplete data (at least one field was missing in fields including identification ID, age, date of death, etc.).
Gaza health department bulletin on May 4
Jews have been working hard to discredit the Gaza death figures
I wrote in "Hell on Earth, Gaza Death Breaks 30,000, Is the Civilian Death Data Released by Hamas Reliable?" cited the Lancet study to show that death data in Gaza's health sector are of high quality. But Zionists have been working hard to discredit the data, the most deceptive of which is a March 6, 2024 post by Abraham Wyner, a professor of statistics and data science at the Wharton School of the University of Pennsylvania, titled "How the Gaza Ministry of Health Falsified the Casualty Figures," in which his statistical analysis of the casualties reported by the Gaza Ministry of Health "highly suggests that the process of reporting these numbers is irrelevant or loose in relation to reality."
In his article there is such a chart:
Chart in "How the Ministry of Health in Gaza fakes casualty figures".
The coefficient between the data in the chart is R2 = 0.999, which Wynner describes as revealing "an extremely regular increase in the number of casualties during this period" and concludes that "this regularity is almost certainly not true". This is a clear bully of Israeli supporters for a lack of mathematical knowledge. Because his icon is a daily cumulative death data. Since the cumulative number is much larger than the increase, the cumulative number is always a linear regression.
The previous chart dataset is presented with per-death data instead
If you use the number of deaths per day, R2 quickly becomes 0.233, which becomes less regular, and Abraham Wynner only used the 15-day number set, so how to put him in the whole set is even more obvious
The graph above is the cumulative number of deaths, the chart below is the daily number of deaths, and within the red line is the data cited by Wynner
As you can see from the graph above, even if the variance in the number of deaths per day is large, it has almost no effect on the R2 value of the cumulative number, which should be the knowledge of a freshman year in statistics, and it is hard to imagine a professor of statistics and data science making such a low-level mistake.
Other conclusions in "How the Gaza Ministry of Health Falsified Casualty Figures" have the same problem, such as his statement that "more female casualties were reported on the days when men were killed", and when viewed in the whole data set, the conclusion is almost the opposite
The red dot is the Wynner reference dataset, and the blue is the full set
For another example, he said that the deaths of women and children were not correlated, with an R2 of 0.00787, but in fact, the R2 of the complete set of data was 0.3244, showing a weak correlation.
Red is Wynner's dataset and blue is the complete set.
There is also Wynner, who ignores the demographics of Gaza and mistakenly believes that 70 per cent of the casualties in the indiscriminate bombing will not result in women and children. Simple mathematics contradicts this: if children make up 50% and adults are evenly distributed, women and children will make up about 75% of the population. This doesn't even require the statistics level of the freshman year, and the upper elementary school can be competent.
Abraham Wynner's article fully embodies that "hooligans are not terrible, but they are afraid that hooligans are cultured"
But the good thing is that it's not just hooligans who are educated