If you run a search to say find me the polynomial that matches the difference between these two plots (the lines on the graph) the best well then that's what you're going to get, the polynomial with the highest correlation.

If you're doing this on a single graph you might just be fooling yourself. They do it on other graphs and that is more suspect because they should not match so well though in the video they don't show their working out clearly enough to fully verify it.

That said the same as in Lyndell's video, even for a single graph this is curious.

If you look at the key or the poly nominal it's sort of a wave. It goes up and down. If you look at the registrations it's like a wavy wave. You see the large wave with the big ups and downs then the higher frequency wave with small ups and downs.

The polynomial shown is just the difference. The delta series. For that to exhibit so little suboscillation compared to the parent is indeed very strange. There's a big drop in variance that seems a bit implausible to me.

I would expect the correlation to match a polynomial increasingly more as you add degrees and for it to be quite high but things like less than a thousandth when the parent graph oscillates a few percent is a bit too tight.

If you run a search to say find me the polynomial that matches the difference between these two plots (the lines on the graph) the best well then that's what you're going to get, the polynomial with the highest correlation.

If you're doing this on a single graph you might just be fooling yourself. They do it on other graphs and that is more suspect because they should not match so well though in the video they don't show their working out clearly enough to fully verify it.

That said the same as in Lyndell's video, even for a single graph this is curious.

If you look at the key or the poly nominal it's sort of a wave. It goes up and down. If you look at the registrations it's like a wavy wave. You see the large wave with the big ups and downs then the higher frequency wave with small ups and downs.

The polynomial shown is just the difference. The delta series. For that to exhibit so little suboscillation compared to the parent is indeed very strange. There's a big drop in variance that seems a bit implausible to me.

I would expect the correlation to match a polynomial increasingly more as you add degrees and for it to be quite high but things like less than a thousandth when the parent graph oscillates a few percent is a bit too tight.

If you run a search to say find me the polynomial that matches the difference between these two plots (the lines on the graph) the best well then that's what you're going to get, the polynomial with the highest correlation.

If you're doing this on a single graph you might just be fooling yourself. They do it on other graphs and that is more suspect because they should not match so well though in the video they don't show their working out clearly enough to fully verify it.

That said the same as in Lyndell's video, even for a single graph this is curious.

If you look at the key or the poly nominal it's sort of a wave. It goes up and down. If you look at the registrations it's like a wavy wave. You see the large wave with the big ups and downs then the higher frequency wave with small ups and downs.

The polynomial shown is just the difference. The delta series. For that to exhibit so little suboscillation compared to the parent is indeed very strange. There's a big drop in variance that seems a bit implausible to me.

I would expect the correlation to match a polynomial increasingly more as you at degrees and for it to be quite high but things like less than a thousandth when the parent graph oscillates a few percent is a bit too tight.

If you run a search to say find me the polynomial that matches the difference between these two plots (the lines on the graph) the best well then that's what you're going to get, the polynomial with the highest correlation.

If you're doing this on a single graph you might just be fooling yourself. They do it on other graphs and that is more suspect because they should not match so well though in the video they don't show their working out clearly enough to fully verify it.

That said the same as in Lyndell's video, even for a single graph this is curious.

If you look at the key or the poly nominal it's sort of a wave. It goes up and down. If you look at the registrations it's like a wavy wave. You see the large wave with the big ups and downs then the higher frequency wave with small ups and downs.

The polynomial shown is just the difference. The delta series. For that to exhibit so little suboscillation compared to the parent is indeed very strange. There's a big drop in variance that seems a bit implausible to me.