Switch is 1.24m down YoY as of October 9th and all indications are the gap is going to immediately start widening again after OLED launch week. We already know Japan is 23k down YoY for OLED's 2nd week, and 2020 continued to post 500k+ weeks all the way up to Black Friday.

To sell 29m+ Switch is going to have to sell 11.5m-12m in November and December alone. That would be the best Nov+Dec any console has ever had.

To say 29.66m+ to beat DS's record is likely is... very very optimistic.

Based on your post, it sounds like most people are doing a week by week tally of projected future sales, while I am doing a proportion based on previous years.

"To sell 29m+ Switch is going to have to sell 11.5m-12m in November and December alone. That would be the best Nov+Dec any console has ever had."
If Switch is going to have the best year ever, then it isn't unrealistic to say that it will have the best Nov + Dec ever.  Those are the two most important months out of the year.  Your argument is basically, "If Switch were going to break a record, then it would have to break a record."  This is circular reasoning.

The real question is, "Is Switch going to break a record?"  Last year it came extremely close, and the main thing that held it back is that Switch had a very disappointing holiday as compared to the rest of its sales that year.  With a new hardware model and a Pokemon game releasing this year, it is very likely to outsell 2020's holiday period by a good margin.

Obviously, none of us knows the future, nor do we really know how slight or extreme the chip shortage will be.  However, to say that Switch will be down this year is ... "very very pessimistic."

Sure it is similar to saying "if it's going to break a record it would have to break a record." The point is I don't see where you are getting any indication that Switch is suddenly going to jump from being 7.5% down for the year and suddenly be up 30% for the whole of Nov+Dec to beat that "likely" 29.66m.

 Here are the record Nov+Dec sales:

DS 2008 - ~11.6m
DS 2007 - ~10.5m
Switch 2019 - 9.85m
Wii 2008 - ~9m
Switch 2020 - 8.68m
PS4 2016 - 8.16m
360 2011 - 6.64m
3DS - 6.60m

Switch is currently down 1.24m, as indicated by Japans next week sales this gap is likely widening again. But even if we are generous and say the gap remains 1.24m at the end of October that would still mean sales for the year would be about 18.3m, meaning 11.36m sales needed from the last 2 months, an increase from 2020 of 30.8%, and that's being generous as we already know the next weeks data is down YoY in Japan.

This is the main problem. Switch did not have a poor holiday last year, it had a great holiday. "Disappointing holiday as compared to the rest of its sales that year" is true in terms of how much sales came from the holiday compared to the rest of the year in 2020. However this is due to the covid pandemic changing buying patterns, and your assumption that this year will be vastly different and suddenly 2021 will be very loaded towards the holiday for Nintendo is not an idea that I share.

So basically I think your idea that the we're going to see a similar holiday boost as years prior to 2020 isn't a correct one and final sales will be closer to your 2020 projection than your 2017-2019 projections in the OP.

I can see the Switch having a stronger holiday in 2021 than 2020, and I don't think it's impossible it does in fact beat 2020 overall, but best year ever for a console is fantasy in my eyes.

"The point is I don't see where you are getting any indication that Switch is suddenly going to jump from being 7.5% down for the year and suddenly be up 30% for the whole of Nov+Dec to beat that "likely" 29.66m."

The part you are missing is where these Switch systems come from.  Nintendo plans how much to manufacture, and they are going to plan based on the projections of several previous years.  2021 is actually quite different from 2020, because the sales boost in 2020 was sudden and COVID limited manufacturing.  The sales rate for 2021 has not had sudden surges or drops that would surprise Nintendo, while 2020 did.

So in practice this is what Nintendo is going to do.  They are going to trickle out systems for the next several weeks.  Then on Pokemon week millions of systems will be available worldwide.  There is a good chance there will be a "most systems sold during Pokemon launch" record being set here.  It doesn't even matter if fans like the game, because a ton of people are going to buy OLED that care nothing about Pokemon.  It's just that millions of these systems are being held back for this launch date.  There are also millions more systems being held back for Black Friday and for December.  October sales are kind of irrelevant compared to all of this, because they are based on shortages and not on limited demand.  The people who couldn't buy in October will just buy their system in November or December instead.

There is one way I can see that I am overestimating and that is if Nintendo does not get enough chips to manufacture enough systems.  There is nothing they can do about that no matter how much they plan ahead.  However, your argument is not based on how many chips you think they have.  Your argument is based on the idea that they can't break the record, because that would require them to break another record.  That is not good logic.  Whenever a major record is broken, like total annual sales, then you'd expect one or more other records to be broken too, like # of systems sold in Nov+Dec.  Yes, they might even break the Nov+Dec sales record by a good margin, because they've had plenty of opportunity to plan for this.  "They'd have to break a record to break that record" is not a good counter argument.

"your assumption that this year will be vastly different and suddenly 2021 will be very loaded towards the holiday for Nintendo is not an idea that I share."
Sales being loaded toward the holiday is the norm, and that is doubly true for Nintendo.  You might be right that it will be down this year, but that would make this year another exception.  I am predicting based on the norm.  You are predicting based on the exception.

My argument is that the potential for those kinds of astronomical sales isn't there. No matter how many consoles Nintendo ship they won't sell more than 29.66m this year.

Don't forget that shipments are likely a bit larger than that (probably 5.5-5.75) due to the OLED launch. It launching on the 8th means OLED shipments were happening within the previous quarter, I believe. 

Right around 24-25 million, so a modest drop. Could be a few million more, but I don't think it will quite match 2020.

In any case, I think it's better to wait for Nintendo's official announcement. Though it seems we're probably basing our predictions on VGC numbers.

Yeah, this thread is really asking for a prediction of VGChartz numbers.  Maybe there should be another thread for FY numbers.

Might be a good idea, though vgc numbers are always brought in line with FY numbers anyway. But yeah all predictions in this thread are based on the vgc numbers so if there is a big adjustment when the new report comes out next week, well then it's understandable if predictions change.