No, but it seems though that you know very little about aggregated polling.
Really? I've had a pretty keen interest in polling methodology, poll aggregation methodology, and the relative accuracy of those differing methodologies in varied scenarios for quite some time.
Let's see how this plays out.
The graphs you posted is [are] the averaged polling of the most recent poll of each individual poll, for Biden.
No, they are not. You give the appearance of being unaware that poll aggregators each have their own statistical methodology for aggregating polls and producing trend line graphs. The methodology that you describe above is the simplest and is comparable to what frequently cited (and often preferred in conservative media) Real Clear Politics (RCP) uses for example. The main issue that I have with RCP is that they utilize too small of a sample size, relative to available public polling data, and
may tend to cherrypick at times. Still, during election season, I will frequently look at and compare their numbers to other aggregators as their methodology isn't one that's invalid or should be ignored; I just think that it tends to be a somewhat less accurate method overall.
538, which is also frequently cited (
perhaps the most frequently cited over the past several years), is the aggregator to which I linked and from whom the graphs were taken. They use a
much more complex methodology than what you describe. First, they use every available public poll which meets widely recognized and accepted basic scientific standards for methodology and only a small fraction of polls are excluded on that basis. Second, each poll is
weighted prior to inclusion based on several different and carefully analyzed and structured criteria. This is done for multiple reasons, not the least of which is that by choosing to use only minimal screening of polls, the accuracy of their aggregation could be skewed by polls with lower quality methodology, statistical bias trends, a track record of lower accuracy, smaller sample size, older data, among other considerations of statistical accuracy. It goes deeper than this and their methodology is transparent and highlighted with links on their site. Actually, had you looked, or were remotely interested in their methodology, there is a link highlighted in large blue type incorporated in the title above the trend line graph on the page that I linked which says "How this works". There is also additional detail available on their site on each of the weighting criteria.
538's individual poll weighting can vary dramatically based on several factors: sample size, pollster rating for bias and accuracy, how recent the poll is, polling method, frequency of polling from one source, and more. Your statement that they use the "most recent poll of each individual poll" is also incorrect. There may be multiple polls from one pollster included from different time periods; the weighting, however, will be reduced on a sliding downward scale over time. The differences in polling methodology between say RCP and 538 generally does not produce significant differences when looking at aggregate national polling like presidential approval. As of this moment RCP's approval rating for Biden is currently 44.1% and for 538 it is 44.6%. It's really in the much more difficult state polling where you'll tend to see larger differences and it's the exceptional accuracy exhibited by 538 in
some of the past presidential and senate elections that has been their claim to fame for their more complex, highly analyzed, and weighted methodology. I say
some because the accuracy of their poll aggregation to actual results in a given election, particularly at the state level, will depend on the quality and accuracy of the many polling services and the volatility and polarization exhibited in individual elections. It's likely the most important factor for presidential forecasts, however, is the number of swing states in play and how narrow the margin is in each of those.
Because their methodology is complex relative to what other aggregators use and is readily available on their site, I won't go into further detail. I will say, it bears little to no resemblance to your description.
Whereas the graph for Trump, and Obama, has been interpolated already since we know what all of his numbers are.
Interpolated? No, they're the same methodology described above and are exactly how the approval aggregation appeared for each day in a term during their administrations.
That's the purpose of the comparative graphs. It's the same methodology aggregated and displayed
separately for each day in their respective terms. They are
aggregated day by day, I don't know where you're coming up with interpolation. 538 graphs the data using a
local polynomial regression which by design is somewhat coarse to better show movement in the polls, but the method is exactly the same for each aggregated point, on each graph, for each president. It's a mystery what difference you believe there is when you write "since we know what all of his numbers are"; as the
only numbers they use
are the numbers they know, regardless of whether it's a current or past presidential office holder. Again, they are simply aggregating,
day by day, the president's
daily aggregated approval compared to the
same days in a predecessor's term.
That part of the equation is pretty simple and direct.
Quinnipiac's poll is the most recent update. So, that average is relying on just one recent update and the rest outdated by a couple of weeks, at best.
No, Quinnipiac's poll is
not the most recent poll. There are several that are more recent. In fact, it is the
oldest of all the October polls. Despite that fact, it is currently the highest weighted poll with a 1.85 weighting as they have a high pollster rating grade and a reasonable sample size.
Now perhaps Quinnipiac is an outlier and Biden is actually doing better then [than] Trump, but we will not know until the others perform new polls.
Quinnipiac
is an outlier and a bit dated relative to several
more recent polls, but still has the highest weighting for the reasons previously described. As for the rest of your statement, refer to the posts above.
There is your basic middle school math lesson of the day. I will be testing you on it next week. 
Yeah, thanks. Learned a lot.
