+1 for TWRoO

Thank you.

Also if anyone knows when Nintendo was properly at 1.8 million per month I would appreciate the info... at the moment I am swaying toward November being the start of 1.8mil, but I am still unsure.

Similarly any info on when Nintendo announced 1.5 million per month would be apreciated.

Nicely done TWRoO!

Very well done indeed.  

I do however, think your productions rates ramp up slightly too slowly.  The production was reported to be just shy of 1.5m in late May '07, and I remember there was talk about the 1.8m as early as July but later it came out that there was a parts shortage that pushed it back into Aug.  I'm having a hard time locating the article for the 1.8m in August part but I'm fairly sure that Reggie has said they hit 1.8m production during the summer which would be Aug at the latest.  I also remember hearing a lot about 1.8m well before September but that doesn't mean a lot by itself. I do think the 1.5m production needs to be pushed back into Jun at least it actually puts it in line with the rest of the production increases right before a new quarter.

I do have to admit I'm confused why you said you don't think there was significant stockpiling in the other thread.  By your unaltered numbers there were 3m Wii's stockpiled going into Nov '07 and I would definitely call that significant. Some of that is normal production but about 1.2m of it isn't. Similarly at the end of Feb there is about 2.8m stockpiled by your numbers with 1.8m of that being normal production and around 1m being stockpile.  I guess it depends on the definition of significant but for me 1m qualifies.  And if you were to move your production timetable back a month for 1.5 and/or 1.8 it would add another 300-600k to those stockpiled numbers.

In any case, I want to be clear that I'm not trying to diminish the work you've done as I think you've done a good job with it.  I just don't fully agree with your conclusion for the reasons mentioned above. With that said I agree with about 95% of it.

I back you up Sqrl. I also don't have a link, but I am sure 1.8 Million since end of august/begin of september.

I think you may have not totally understood what I wrote...(looks at your claim of 2.8 million stockpiled in Feb) The table shows a running counter of what has been produced, but in terms of shipments I assumed a month interval for the time it takes to ship, which happens to tie in ok with the numbers... so you should in fact look to the produced number at the end of January to see what has been shipped by the end of February.... thus I there is no "stockpile" as production is about 600k above the sales figures VGC has.(supply)

regarding the stockpile you mentioned in November, I didn't really look at monthly to monthly numbers as I would have had to estimate shipments by using VGC sales... so I meant little stockpiling (apart from Nintendo's usual quarterly holdover) was done from fiscal Q2 into Q3.

-----

For your estimations of when each production ramp-up happened, yours seems to make the least sense even if the dates ring a bell (they may only be ringing bells because other posters have mentioned them here though)
If indeed the increase to 1.5 was 2 months earlier, AND the 1.8 increase was also a month earlier, then that adds 900k onto all production totals after August, which means that they had produced 20.1 million by the end of November, and therefore would have had no need to air-freight consoles from Decembers produce, and would have had no trouble supplying early 2008.
Also by the end of January they will have produced almost 24 million already, whichhappens to be a hell of a lot more than were sold.

If that model were to be true, then it must take much longer than 4 weeks to ship normal stock... 6-8 weeks would tie-in with Q3 2007 and would involve Nintendo air-shipping Novembers stock so that it got there in December [instead of January]
However that then doesn't fit with the previous fiscal year, because if it takes longer than 6 weeks to ship stock, then Nintendo couldn't have shipped the 5.84 million they managed by the end of March last year (as they had only produced 5 million by the end of January, so a max would be 5.5)

Ahh, Damnit, I think I just figured it out.

Right,
-using the same months you mentioned Sqrl,
-and under the assumption that normal shipping takes 6 weeks,
-also remembring that the production isn't an exact value per month, and neither is it an average, but will fluctuate (and knowing Nintendo any figures they give us are lowballed) and may average slightly higher per month than what they say.

So although adding up the months comes to only 5.5 million produced (with 6 weeks to spare before the end of March) the discrepancies from vague launch numbers (2 million) and from slightly fluctuating (probably trending upward) supply Nintendo squeezed an extra 340k out before March ended and thus shipped 5.84 million.

For the next quarter... 8.4 million had been produced by the end of April... add in half of May (so that we are 6 weeks from the end of June) which is 750k, and you get 9.15 million... so again from the slight discrepancies 120k were squeezed out to make the 9.27 million shipped.

For fiscal Q2.... 12.9 million were produced at the end of July, add half of August (900k) and it comes to 13.8 million, which is 630k over 13.17 million that were shipped.
If we take that 630k and assume it was stockpiled.....

for fiscal Q3.... 18.3 million were produced at the end of October, add half of November (900k) and then add the 600k stockpiled from Q2, and it comes to 19.8 million, so the remaining 330k must have come from the air shipping (which would be the produce from the later weeks of November being flown in just in time for Christmas)

finally fiscal Q4 (incomplete).... there were 21.9 million produced at the end of December, add half of Januray (900k) and it comes to 22.8 million, which is presumably what had been shipped by the end of February (sales were 22.2 mil)

Whew! I think that explains it close enough. I think I will make a new thread with a revised table and explanation, and just link to this thread if anyone wishes to see the confusion I had to go through before getting to this.
Thanks Sqrl.
I'll make the new thread tomorrow, for now I will just update the table and post it below this.

There, that is what prodution probably looked like, and if you want to work out what had been shipped by a specific month, go up the "cumulative" column by 1 and a half month.
so for the shipment number at the end of May... look up to between March and April, which means the shipped number was about 7.8 million.

Having said that, I seem to have found some massive errors in the past data of VGC... I checked the sales to see what they were at the end of May and they are at 7.9 million... Now although Wii supply has been low i'm quite sure it wasn't negative.
I thought at first I might have discovered a flaw in the production table, but I checked the end of each quarter to see what VGC has compared to Nintendo's fiscals.

End of December 2006

VGC sales - 2.96 million
Shipment -  3.19 million

supposed supply = 230k (this is phenomenally low, but is believable as few could find Wiis then, even in Japan)

End of March 2007 

VGC sales -  5.77million
Shipment -  5.84million

supposed supply = 70k (this is just odd)

End of June 2007 

VGC sales -  8.93million
Shipment -  9.27million

supposed supply = 340k (This is very low for the middle of summer when Japan seemed to have ok supply, Europe wasn't copius but it wasn't impossible to find one, and America was poor as usual)

End of September 2007 

VGC sales -  12.26million
Shipment -  13.17million

supposed supply = 910k (this is actually a much more believable number, as supply had recently had a big boost)

End of December 2008

VGC sales -  19.33million
Shipment -  20.13million

supposed supply = 800k (this ones seems ok, I suspect Japan may have had half of that supply for itself as there was likely a good shipment for the new year (plus it wasn't sold out beforehand either)

Here is some evidence for the 1.8m being much earlier:

http://www.videogamesblogger.com/2007/11/09/despite-18-million-nintendo-wii-manufactured-monthly-there-will-be-wii-shortages-for-holiday-shoppers-says-nintendo-president-in-video-interview.htm

"Nintendo, the world’s largest maker of handheld game players, plans to keep monthly production at 1.8 million Wii consoles, after boosting the output during the three months ended June 30, 2007."

This supports what folks may remember which is that they said in November that they were keeping production at 1.8m, that is what spurred all the 1.8m articles in November. The press largely pays no attention during the rest of the year and the production information close to the holidays when everyone wants to know about Wii production obviously got a lot of coverage.

The two articles I've linked seem to indicate to me that some time during December '06 Nintendo decided to increase production and around late February was the first increase. Now, I provided a link above that says they hit around 1.5m production by the end of May and this new article indicates that they continued increasing production until they hit 1.8m at the end of June.

I have to admit I am skeptical that they had production at 1.8m in June as it really doesn't fit with what is observable. I think that the answer lies somewhere between the the timeline I mentioned above and what TWRoO listed originally. I'm going to take a crack at this and I think I will end up a lot closer to TWRoO than what those articles indicate:

First of all I want to say that the production numbers I'm using in each month are the result of what I speculate was the average production rate.

For Feb I think the increases came too late to have a real impact. Then over the course of March through May they increased to just shy of 1.5m like the report I linked above indicated, next they leveled off at 1.5m for June and July to assess the situation a bit due to their notoriously cautious nature. Finally realizing it was warranted they increased it over the course of August and reached the 1.8m rate by the end of August (which is why the month averages out to around 1.65m. I think the large increase over Aug was accomplished by preparing for it during June and July by getting their proverbial "Ducks in a row" so that if they decided to go ahead they could make the adjustment quickly.

We really don't disagree all that much on the final numbers just a bit about when it all happened which is completely fine since there is plenty of room for interpretation seeing as we don't have the exact numbers.

^ah excellent, we are very close... I didn't want to go into trying to average out increases myself without any proper sources... I haven't read the link you posted yet but I guess from your table they did not have a lengthy period at 1.2 million like I had, but had slight increases leading up to 1.5 million. (though it works out very similar)

by the way how do you make those tables like that? is it from a seperate program and pasted in or did you use the editing features from the table function on VGC?


---
Also as I added in my last post, researching this using VGC seems to have brought up something wrong with the past numbers, mainly with the end of March numbers, but also end of June seems to have VGC too close to the shipments too.

 @tables,

I paste them in from my Open Office spreadsheet, although I think I could make them with the VGC tables as well they would be a real pain in the ass that way.

@averages,

The averages I used are basically educated guesses. I started by filling in the production rates which we had information about and then worked out from there.  Next I went back and estimated the average based on what I thought the rate was at the beginning of the month and the end of the month.  I think this is actually where our difference in numbers comes from TBH, and its insignificant enough that I'm really not worried about it.

@VGC data issues,

I'm going to wait and see what the FY report looks like before I worry to much.  I know that supply was extremely tight during those early quarters so some of the low 200k type numbers don't really surprise me as much as they normally would.  The 70k seems to be the real potential issue and it falls during a period where I think it is just as likely that we're off as it is that VGC is off.  The production rates leading up to and during March '07 aren't that clear imo.