All "official" charts are totally biased. The only way to make a non-biased chart is to take the data and make your own chart.

That is somewhat problematic however, as I think the official data has been fucked with. The total death data on Dec. 14th was 2,854,838. By Dec. 31st it was reported as 2,913,144. By Jan. 6th 2021 it somehow jumped up to 3,187,086. As of Mar. 31, 2021 its up to 3,358,854.

Acording to the CDC there is a delay in data:

At NCHS, about 80% of deaths are automatically processed and coded within seconds, but 20% of deaths need to manually coded, or coded by a person... ...[for deaths due to covid they say] mortality data [could be] approximately 27% complete within 2 weeks, 54% complete within 4 weeks, and at least 75% complete within 8 weeks of when the death occurred

So for all deaths non-covid they say the death reports will be 80% done instantly, and 20% done within a week or so. The covid will lag (for some reason) but lets do some math.

According to the CDC data in the last 8 weeks of 2020 the total Covid count was 132,175 in the US. Even if we assume that zero of those made into the reporting by the Dec. 31st data above (2,913,144) and 20% of the other deaths were on a one week lag (6548) that should be a total possible death total in 2020 of 3,051,867. Now, almost certainly some of those are double reported (because I assumed zero covid deaths got reported), but it gives an upper limit.

The official number is 3,358,854 - 3,051,867 = 306,986 deaths that are unaccounted for. Where did they come from?

According to this report from Ireland (among other fuckery) the official data seems to have been backfilled from flu deaths in Jan 2020 to increase the numbers in April 2020. Jan. 2020 in the US also had a lower than average death rate (Although no where CLOSE to late 2020 flu season numbers, where we apparently cured the flu). Lower than average is not conclusive, but it is interesting. This flu season had the lowest death rate of all time... ALL TIME. And not just by a little bit, but by orders of magnitude. Did we backfill 2021 deaths into 2020 deaths to fudge the numbers? Did someone just create death reports de novo? I don't know.

More interesting things...

According to this article from Nov. 22, 2020 the data showed a drop in the previous leading death types. Now the data shows the opposite (from the last link above).

You can download the official data here.(click "Export").

The data is all kinds of crazy. Looking at the last week in December for example, apparently CA accounted for more than 1/8th of all deaths in the US with a jump of almost twice as many deaths than the average week. The entire last few weeks across the board are completely insane, and so improbable from a data perspective I would have thought a nuclear bomb had gone off in L.A. or something.

Another interesting thing is the "Covid with other", and "Covid alone" death data. It was previously reported that the vast majority of deaths were "comorbid". Now the data is just confusing, as if they don't want you to understand it. It says that in the US the total Covid deaths are 365,117, and the Covid alone deaths are 333,363. That's a comorbidity rate of 8.7%. Yet when I looked at the data in December it was the opposite, with a comorbidity rate of about 93%. I assume they are just trying to make it confusing, but who knows.

Here, the CDC (2019) says the death rate is 869.7 per 100,000: https://www.cdc.gov/nchs/fastats/deaths.htm But here, the CDC says the provisional death rate for 2020 is 715.2 per 100,000 and it's going up to 828.7 because of COVID: https://www.cdc.gov/mmwr/volumes/70/wr/mm7014e1.htm This doesn't begin to pass the smell test. From 870 to 715 in a year? They decreased the rate in order to assign the difference from the real rate to COVID.

Covid saved lives . I can tell you that .

https://data.cdc.gov/NCHS/Monthly-Counts-of-Deaths-by-Select-Causes-2020-202/9dzk-mvmi/data

So you can change what data set you want to use and make whatever chart style you like. I have it data set at deaths by month in 2020.