• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

Official May 2007 NPD Prediction Thread

donny2112

Member
Sule said:
The prediction's should have stopped this morning or even better, yesterday, just to make sure. The thread was open long enough for everybody to have a chance to post their prediction's.

If people actually take the time to extrapolate U.S. sales from Canadian sales, let them. It won't make them win. We already have Japanese sales, so someone could always extrapolate from that, too. It won't make a substantial difference either way. The only way you could "cheat" to become #1 is to edit your prediction after the U.S. NPD results are posted. However, you'd also get ridiculed and banned from prediction threads after you were found out. :p

If you're actually worried about someone consistently beating you by basing their predictions off of analysts/Canada, then you really need to find a way of making better predictions. :lol

Regardless, the real blind guessing is the 2007 totals that we predicted in January, well before the January NPD came out. ;)

Edit:
Alkaliine said:
Wait, are you predictions allowed to be edited? Because I changed my mind a couple of days ago on a few figures. My numbers will still be accepted, right?

My understanding is that the parser starts from scratch each time it is run. With that in mind, your edits would be counted, as long as they are before 4 p.m. ET today.
 
donny2112 said:
If people actually take the time to extrapolate U.S. sales from Canadian sales, let them. It won't make them win. We already have Japanese sales, so someone could always extrapolate from that, too. It won't make a substantial difference either way. The only way you could "cheat" to become #1 is to edit your prediction after the U.S. NPD results are posted. However, you'd also get ridiculed and banned from prediction threads after you were found out. :p

If you're actually worried about someone consistently beating you by basing their predictions off of analysts/Canada, then you really need to find a way of making better predictions. :lol

Regardless, the real blind guessing is the 2007 totals that we predicted in January, well before the January NPD came out. ;)

I agree. If you're that worried and really serious about your prediction, WAIT until the usual predictions, projections and official numbers from other territories are in.
 
donny2112 said:
If people actually take the time to extrapolate U.S. sales from Canadian sales, let them. It won't make them win.

I think the Canadian sales can be a pretty good indicator of US sales once we have a larger sample and few formula's to apply.

If Canada is at all representative of the US, allowing for specific differences, then having a 7% sample is statistically huge. Presidential campaigns have been won or lost with less returns.

A few years ago, when Best Buy data for 3 districts was being leaked, we were able to predict hardware and key software titles with amazing accuracy
 
If these ratios follow a fixed trend, and counting 9 weeks between Feb and April and 4 weeks between April and May, and without being bothered to make any guesses at some curves and stuff (because Canada's correlation between with US each month seems to have no... correlation), here's May US NPD...

Quote:
DS: 362,111
Wii: 266,333
PSP: 202,000
PS2: 180,000
360: 140,500
PS3: 46,500

I so want that to be true. It would be a bloodbath.
 

Sule

Member
Donny2112 said:

Same here:

Sule said:
dammitmattt said:
That is exactly what he did in the Canadian thread. Took the April percentages and applied them to May to get these numbers. It's good as a measuring stick to compare against for shits and giggles, but dishonest to pass it off as your "guess."

Then you could argue that nobody really ''guesses''. Everybody has his or her own technique for ''guessing''. For example:

sonycowboy said:
My particular methodology is looking at the trending for the past several months and then looking at historical trending from April to May, finally accounting for the effect of big titles released.

Using trends from the past months, historical trends etc. Nobody is really guessing. Although I could agree with you the way he got his numbers (and the Canada x 10) are a lot less ''guessing'' than sonycowboy's and other techniques.

I'm not saying it's an illegitimate way of predicting but i do feel the thread should be closed earlier than has been the case for the last months. To make sure nobody gets their hands on leaks before anybody on the forum knows it and to minimize outside influences so that YOUR prediction represent the way YOU feel the HW sales etc. are going to turn out.

Again I think everybody has his or her own way of predicting, using Canada NPD should be allowed and the prediction threads are open long enough for everybody to notice them and make a prediction. Just close it a little earlier.

Have you guys set a deadline for next month? If you haven't let's discuss it here or start a new thread see what everybody has to say, do the same as with the new points system. The deadline for predictions and parser format should be in the OP (like you've already said) and should be highlighted so that everybody notices. Can you give them a different color maybe?
 

AirBrian

Member
sonycowboy said:
A few years ago, when Best Buy data for 3 districts was being leaked, we were able to predict hardware and key software titles with amazing accuracy
That was 3 years ago already? Wow, the time flies.
 

Parl

Member
RubberJohnny said:
I so want that to be true. It would be a bloodbath.

Unfortunately, it won't happen.

If Canada proves to be a useful tool when predicting US NPD, then I think predictions should stop before it comes in because we want it to be blind GAF instinct and as some would say, even though I hate the phrase, "seperate the men from the boys".
 

donny2112

Member
sonycowboy said:
If Canada is at all representative of the US, allowing for specific differences, then having a 7% sample is statistically huge.

If we consistently get Canadian data, maybe we can start putting it in as a predictor. JoeMolotov and Fuzzy are covering that angle this month. Now, I'm not saying that looking at Canadian sales can't give a general feel how the U.S. sales will look, but looking at past months and trends gives you basically the same advantage. How you massage the data from that point to come up with an estimate plays a huge part in what actual ranking you take, though.

Parl said:
If Canada proves to be a useful tool when predicting US NPD, then I think predictions should stop before it comes in because we want it to be blind GAF instinct and as some would say, even though I hate the phrase, "seperate the men from the boys".

"separate"

I agree that if a particular system consistently gives amazing results, we should stop the predictions before that data comes out. Sort of like how card counting in casinos is frowned upon. I just don't feel that using Canadian data/Analyst estimates will consistently give those "amazing results."
 

Parl

Member
donny2112 said:
"separate"

I agree that if a particular system consistently gives amazing results, we should stop the predictions before that data comes out. Sort of like how card counting in casinos is frowned upon. I just don't feel that using Canadian data/Analyst estimates will consistently give those "amazing results."

Yeah, me too, but if we start to notice a correlation, then I think we should pull the plug within a year.

Ah, "separate". I have spelling problems when it comes to double/single letters and vowels, especially with my region's lazy dialect. My memory system is worked together in a non-typical fashion where things that logically link are remembered really well, but other stuff, not so well, like character's names in fiction unless that person "looks like a ____" somehow.
 
Alkaliine said:
Wait, are you predictions allowed to be edited? Because I changed my mind a couple of days ago on a few figures. My numbers will still be accepted, right?

Yes, donny was correct in stating that the program downloads the thread anew each time. I've left it running on a timer to do so this afternoon at 4 pm Eastern.
 
Seperate, seprate, definately, and my personal favorate proabally, are all dialect/accent problems. :p Even worse, I sometimes wonder how how non-rhotic dialects like Bostonian or Austrailian (spelled how someone with a Midland accent, also known as American Broadcast, would think it might be spelled) manage to spell words like card, bird, and juggernaut.

I hope the previous paragraph didn't hurt anyone.
 
sonycowboy said:
I think the Canadian sales can be a pretty good indicator of US sales once we have a larger sample and few formula's to apply.

If Canada is at all representative of the US, allowing for specific differences, then having a 7% sample is statistically huge. Presidential campaigns have been won or lost with less returns.

A few years ago, when Best Buy data for 3 districts was being leaked, we were able to predict hardware and key software titles with amazing accuracy

So the short version is that Nintendo wins again. O_O But I don't think that Wii > Nintendo DS.
 

Sule

Member
GhaleonEB said:
Find a clock. The one on your computer will work. Count down to the time in the OP that the data is usually released. :p

thisisneogafdude.gif :p

Not everybody lives on the east coast. We would even have to look up timezones etc. Plus you just know there are going to be people asking ''how long till NPD'' ten times. Might as well have an official countdown clock in the OP.
 
hamzik said:
When are the numbers supposed to come in?


Last month, it was at 1:08 AM on friday...

And two months ago, it was at 12h57 AM on friday too...

Hope that sonycowboy can maintain this improvement trend :D
 

Sule

Member
Brandon F said:
...Japan-Time...

54n4a5t.jpg
 

kottila

Member
GhaleonEB said:
Can someone explain the whole Japan-Time meme? I missed out on the origin and have no idea what it even means.

After reading the first 1000 posts of the SSBB-thread, you will understand
 
Merovingian said:
Is it today?
maybe you should check the OP for the information you're looking for. it's strange and weird, I know, but it might just work.

I don't do predictions, but I do like reading the threads. :lol so entertaining.
 

Sule

Member
Cosmonaut X said:
There's something that tickles me about someone posting a GIF for an overused GAF meme to complain about an overused GAF meme... :)

I'll make it up to you in the NPD thread with a brand new GIF by me, based on another, relatively new, GAF meme... :)
 

Evlar

Banned
bmf said:
Seperate, seprate, definately, and my personal favorate proabally, are all dialect/accent problems. :p Even worse, I sometimes wonder how how non-rhotic dialects like Bostonian or Austrailian (spelled how someone with a Midland accent, also known as American Broadcast, would think it might be spelled) manage to spell words like card, bird, and juggernaut.

I hope the previous paragraph didn't hurt anyone.
I had a college professor from Boston. He would occasionally write out the names of Greek letters as "delter", "beter", "alpher", etc.
 
The Sphinx said:
I had a college professor from Boston. He would occasionally write out the names of Greek letters as "delter", "beter", "alpher", etc.
So, kind of a "If you're not sure, add an 'R' and hope for the best" philosophy?
 
Top Bottom