Support NeoGAF

[jj984jj]

Not looking at the accuracy as stricty as that, I think Square2005 did a good job this week. Gave a reasonable look at the week sales.

I'm sure first day sales helped him.

 

[apotema]

Guys please, we are starting the thread, can we please just use real data (Media Create, Famitsu, Dengeki)?

Later in the week we can let ioi smoke pot

 

[ioi]

And it starts ...

ioi's LTD = theoretical maximum sales since = shipped

In reality, they are at least slightly high.

ioi's weekly figures = guesses based on trends, past history, form fit functions, etc. = much less accurate than the LTDs

Media-Create

Spoiler

and sometimes Famitsu

are the "official" sources of GAF Japanese data, so arguing over ioi's guesses (for the weekly numbers) doesn't get us anywhere.

Well, slight correction as ever:

ioi's LTD = theoretical sales since launch < shipped

I don't always adjust data 'up', a number of games are adjusted 'down'.

You have to get past this idea of theoretical maximum and being slightly high. My figures are always a small amount less than shipped, in some cases more than others, in many cases higher than the trackers but in some cases lower. There are no hard and fast rules.

Guys please, we are starting the thread, can we please just use real data (Media Create, Famitsu, Dengeki)?

Later in the week we can let ioi smoke pot

The irony is that the 'real' data is actually less real than my data

 

[Square2005]

I'm sure first day sales helped him.

I never saw the first day sales...
I actually came up w/ those figures last Sunday when they were released.
BTW: thanks cvxfreak for the compliment...

 

[loosus]

Pokemon did well, but that does seem a bit lower than I expected for this week.

P.S.: For the betterment of mankind, please stop giving a shit about ioi's numbers. Please.

 

[ioi]

Not looking at the accuracy as strictly as that, I think Square2005 did a good job this week. Gave a reasonable look at the week sales.



http://www.neogaf.com/forum/showpost.php?p=4533102&postcount=344

I guess you really did readjust it.

I did, as I stated earlier today:

http://www.neogaf.com/forum/showpost.php?p=4551620&postcount=466

Do you agree with the MC numbers then, or were you just bluffing about having them?

If you have them, why not compare MC to MC, knowing your usual tendancies I'm disappointed you now think it's acceptable to compare data from two different tracking sources :-O

 

[cvxfreak]

I did, as I stated earlier today:

http://www.neogaf.com/forum/showpost.php?p=4551620&postcount=466

Do you agree with the MC numbers then, or were you just bluffing about having them?

I have it. 1,199,100. I'm not sure how the 28 rounds off to 100, but it's no big deal.

Since I don't have second week MC sales, I'm holding yours in high suspicion. Unless you admit you bought the data from them.

I'm sure first day sales helped him.

True, but still not bad at all.

 

[ioi]

I have it. 1,199,100. I'm not sure how the 28 rounds off to 100, but it's no big deal.

Well you obviously have innaccurate numbers then, it's a massive deal. Those 72 units make all the difference. Forget the fact that the number is ~20% too low anyway, it's the 72 we should all be worrying about :rolleyes

That, and the 1m+ MC week 3 for Pokemon G/S that famitsu have at 454k- how is that going to affect your comparison next week?

 

[enishi]

It happened back in the NES days, once that system really took off.

I remember it being bad enough for a while that even your typicall WalMart clerks were blaming it on memory chip manufacturing plants and such....

That's what happen when you sold nearly 700k cards in a single week, which follows a week that 1.8m cards were sold.

Not to mention that at least 300k are being sold every other week...

 

[Square2005]

Pokemon did well, but that does seem a bit lower than I expected for this week.

It'll hav legs and expect a huge surge during Christmas/New Years.
Check out Pokeman Emerald's weekly sales & it's rebound at the end of the year (also released in Sept):

643,987
160,155 (a 75% drop)
82,952
56,480
47,606
32,527
32,795
26,326
18,470
14,115
18,997 < Thanksgiving week :D 
26,538 < December begins
32,242
57,274
[B]84,557 < Christmas week[/B]
40,212 < New Years
30,041
14,611
7,824
[U]7,187[/U]
LTD by the end of Jan. 2005: 1,434,896 or 2.2 times launch week...

It stayed on the famitsu TOP30 up-thru July 05...

 

[The_lascar]

11/ Brain Training (NDS, Nintendo)
12/ Talking Cooking Navi (NDS, Nintendo)
13/ English Training (NDS, Nintendo)
14/ Rhythm Tengoku (GBA, Nintendo)
15/ .hack// G.U. Vol. 2 (PS2, Bandai)
16/ Mario Kart DS (NDS, Nintendo)
17/ Tingle RPG (NDS, Nintendo)
18/ Romance of the 3 Kingdoms XI (Koei, PS2)
19/ Tamagotchi 2 (NDS, Bandai)
20/ Mario Slam Basketball (NDS, Nintendo)
21/ Tetris DS (NDS, Nintendo)
22/ Tenchu 360 (360, From Software)
23/ Kawaii Koinu DS (NDS, MTO)
24/ Kenshin (PS2, Banpresto)
25/ Monster Hunter Freedom - The Best - (PSP, Capcom)
26/ Dead Rising (360, Capcom)
27/ Big Brain Academy (NDS, Nintendo)
28/ Perfect Kanji Ability (NDS, TDK)
29/ Gintama DS (NDS, Bandai)
30/ Brain Scan (NDS, Sega)

 

[cvxfreak]

11/ Brain Training (NDS, Nintendo)

Brain Age is eternally stuck at #11. :lol

I don't think it'll dethrone DQ7, but AC/NSMB/BT2/DP are bound to.

 

[ioi]

It'll hav legs and expect a huge surge during Christmas/New Years.
Check out Pokeman Emerald's weekly sales & it's rebound at the end of the year (also released in Sept):

643,987
160,155
82,952
56,480
47,606
32,527
32,795
26,326
18,470
14,115
18,997
26,538 < December begins
32,242
57,274
[B]84,557 < Christmas week[/B]
40,212 < New Years
30,041
14,611
7,824
[U]7,187[/U]
LTD by the end of Jan. 2005: 1,434,896 or 2.2 times launch week...

It stayed on the famitsu TOP30 up-thru July 05...

Well I'd predict much better legs anyway than Emerald and 200k weeks over xmas so yeah. It's far from the end yet. 5m+ is likely, but I don't think it'll end much above NSMB / BT2 if at all

 

[saz]

1. (PS2, Bandai Namco) Dragon Ball Z Sparking Neo - 263,897 COMGNET 280K
2. (DS, Pokemon) Pokemon Diamond - 254,080 / 1,074,127 COMGNET 236K
3. (DS, Pokemon) Pokemon Pearl - 212,193 / 980,881 COMGNET 181K
4. (PSP, Bandao Namco) Gundam Battle Royale - 64,986 COMGNET 38K
5. (DS, Nintendo) New Super Mario Brothers - 40,114 / 3,233,793 COMGNET 44K
7. (PS2, SCE) Minna no Tennis - 32,962 / 416,507 COMGNET 37K
8. (DS, Square Enix) Final Fantasy III - 30,418 / 781,713 COMGNET 29K

Less than 5% difference : #8
Less than 10% difference : #1, #2
10% or more difference : #3, #4, #5, #7
Worst accuracy : 41%
Best accuracy : 3%

Overall accuracy : 11%

So about as good as ioi's numbers

 

[ioi]

Since I don't have second week MC sales, I'm holding yours in high suspicion. Unless you admit you bought the data from them.

I've just gone back and read this, so hang on, you're calling me a liar now? Even though I just gave you the first 4 weeks sales?

:lol

 

[kisakiproject]

11/ Brain Training (NDS, Nintendo)
12/ Talking Cooking Navi (NDS, Nintendo)
13/ English Training (NDS, Nintendo)
14/ Rhythm Tengoku (GBA, Nintendo)
15/ .hack// G.U. Vol. 2 (PS2, Bandai)
16/ Mario Kart DS (NDS, Nintendo)
17/ Tingle RPG (NDS, Nintendo)
18/ Romance of the 3 Kingdoms XI (Koei, PS2)
19/ Tamagotchi 2 (NDS, Bandai)
20/ Mario Slam Basketball (NDS, Nintendo)
21/ Tetris DS (NDS, Nintendo)
22/ Tenchu 360 (360, From Software)
23/ Kawaii Koinu DS (NDS, MTO)
24/ Kenshin (PS2, Banpresto)
25/ Monster Hunter Freedom - The Best - (PSP, Capcom)
26/ Dead Rising (360, Capcom)
27/ Big Brain Academy (NDS, Nintendo)
28/ Perfect Kanji Ability (NDS, TDK)
29/ Gintama DS (NDS, Bandai)
30/ Brain Scan (NDS, Sega)

Great start for Tenchu. Probably sold 8-10k, considering it was 5k day 1. Dead rising has some legs. Maybe 360 will stay above 2,000 this week. Pretty excited for famitsu 360s next monthly chart.

 

[saz]

or alternatively just look on VG Charts where the thinking has been done for you

:lol

 

[Square2005]

Well I'd predict much better legs anyway than Emerald and 200k weeks over xmas so yeah. It's far from the end yet. 5m+ is likely, but I don't think it'll end much above NSMB / BT2 if at all

Ioi!! Can you quickly look up FFIII sales from 1990? Even if the weeks are doubled-over...

 

[cvxfreak]

It'll hav legs and expect a huge surge during Christmas/New Years.
Check out Pokeman Emerald's weekly sales & it's rebound at the end of the year (also released in Sept):

LTD by the end of Jan. 2005: 1,434,896 or 2.2 times launch week...
[/CODE]It stayed on the famitsu TOP30 up-thru July 05...

Emerald isn't gonna be a useful indicator for D/P. Save that for whatever the third version is.

I've just gone back and read this, so hang on, you're calling me a liar now? Even though I just gave you the first 4 weeks sales?

:lol

I'm asking you where you got the data. I've searched for Media Create's pre-2002 numbers for *years* in both English and Japanese and never found anything. I've never seen anyone refer to them before 2002, either.

 

[Square2005]

Emerald isn't gonna be a useful indicator for D/P. Save that for whatever the third version is.

I'm making a comparison b/c it was released in Sept. Yeah sales will be higher 'cos it's two pokermans but it should follow a similar path.

 

[ioi]

So about as good as ioi's numbers

Well let's take that point seriously for a second. Square 2005 has basically learned from experience over the last 3-4 weeks and gone from using a single scale factor across the board to a different one per game based on past trends. Each week he has been closer and closer.

Nobody asks why we use a scale factor of 280 for Dragonball Z and 870 for Pokemon? No other reason than in past experience this is what fits. If we had nothing but the last 3 weeks of COMGNET and MC data, and this week's COMGNET data (i.e we were never going to get MC data) then what Square2005 has done is use all available data to produce the best idea we can of how the games sold this week.

Much like I do (on a lesser scale, although going back to 1990 then not really) with the data on VG Charts.

Unfortunately what you are all doing is saying that I'm a fool, and we should just use the COMGNET raw data because it is an 'official source of data from a professional company', which would have given the following:

1 Dragon Ball Z Sparking NEO 1000
2 Pocket Monster Diamond 271
3 Pocket Monster Pearl 208
4 Gundam Battle Royale 181
5 Everybody's Tennis 102
6 New Super Mario Bros. 84
9 Final Fantasy III 45
10 .hack//G.U. Vol.2 &#21531;&#24819;&#12501;&#22768; 39

DBZ selling twice as much as Pokemon combined, 20 times as much as FFIII etc. WRONG in other words.

 

[Jonnyram]

I'm asking you where you got the data. I've for Media Create's pre-2002 numbers for *years* in both English and Japanese and never found anything. I've never seen anyone refer to them before 2002, either.

They've been tracking data since 1992 but didn't make it available freely until 2002. ioi must have paid for it, if he has it at all. It still makes no sense to use the shipped numbers so willingly.

COMG is a chain of game shops in Japan. They track their own data and noone else's. Using their data to estimate national sales is a non-starter. I don't know why anyone is ever doing it. Their figures are quoted on Japanese boards, but never ever considered as a serious grounds for estimates.

 

[Jasoncheng]

TOP30
NDS:20
PS2:5
XBOX360:2
PSP:2
GBA:1

 

[Square2005]

They've been tracking data since 1992 but didn't make it available freely until 2002. ioi must have paid for it, if he has it at all. It still makes no sense to use the shipped numbers so willingly.

Actually I remember the-magicbox posting media-create sales going back to 1999...
I never could find where that data was coming from.

 

[polg]

I don't think there are enough stop posting gifs to express how annoying this "ioi vs square2005 vs theworld" thing is...

 

[cvxfreak]

Actually I remember the-magicbox posting media-create sales going back to 1999...
I never could find where that data was coming from.

I think those were actually Dengeki/Media Works. Jonny explains it nicely. MC just magically pops up in 2002 because that's when some of the data became public. Back when I was on GameFAQs they used to be referred to as Bloomberg numbers because they would be the first to post them.

So that would contradict ioi stating a long time ago that MC used to post their top 50's publically (or he meant Dengeki).

 

[Square2005]

I don't think there are enough stop posting gifs to express how annoying this "ioi vs square2005 vs theworld" thing is...

Wrong thread...
So far it's been cvxfreak vs. ioi

I actually agree w/ some of ioi's methods (it's similar to what I used to do) but I think there is no need to scale-up data anymore. Seems pretty accurate as of late (whether famitsu or m-c).

 

[ioi]

They've been tracking data since 1992 but didn't make it available freely until 2002. ioi must have paid for it, if he has it at all. It still makes no sense to use the shipped numbers so willingly.

Well it shows what you guys know.

Ever heard of a website called gamejapan? They used to post the FULL Media Create top 50s right back from 99 to about 02 when it stopped and all we got was the top 10s posted on the MC Site. Having said that, I do have all data going back to mid 90s from a number of different sources, but just to show that your comment is yet again incorrect.

We really need to sort this shipped thing out.

Let's use Pokemon G/S. Nintendo shipped 7.21m in Japan? How much did they sell?

1m? Unlikely
3m? Nah
5m? What about the other 2?
6m? At least
7m? Most likely

Probably something around ~7.1 - 7.2m, with upto 100,000 copies MAX remaining unsold for whatever reason.

Is that fair?

So when we have famitsu polling ~40% of retailers and coming up with a figure of 5.6m then isn't it reasonable to think they've undertracked the game?

On such an obvious example as this with such large numbers involved I can't see how you can argue other than to just blindly suggest to me that there are 1.5m copies of Pokemon GS sitting in shops :lol

I think those were actually Dengeki/Media Works. Jonny explains it nicely. MC just magically pops up in 2002 because that's when some of the data became public. Back when I was on GameFAQs they used to be referred to as Bloomberg numbers because they would be the first to post them.

So that would contradict ioi stating a long time ago that MC used to post their top 50's publically (or he meant Dengeki).

No, definitely Media Create. Dengeki also posted top 50s for a long time.


And Square2005, I agree that data doesn't generally need adjusting these days. Coverage by trackers is now very high and I RARELY DO ADJUST DATA FOR 2005+ other than to average between MC / Fam / Dengeki, this conversation has stemmed from my disapproval at using raw, unconsidered data from 7+ years ago to compare to today. You can't do it.

 

[Square2005]

I think those were actually Dengeki/Media Works.

Could be and I copied down the new data every week back then (yeah i've been into VG sales for awhile). Still hav some of the data too...

 

[cvxfreak]

The Pokemon GS debate is over and we've said our cues.

Why not share the MC data with everyone? That's what I'm most interested it.

 

[Doom_Bringer]

too much handheld shit to be honest. Also are PC games ever included in these charts or what?

 

[Square2005]

Hey I just found this from years ago in my old notepads (before excel)
11/15-11/21/1999:
I labeled it GameJapan (?)
1. Poke Silver 601,594new
2. Poke Gold 597,334new
3. Chrono Cross 352,300new
6. Derby Stallion 99 14,505/770,531...

...It goes on to 30.

 

[cvxfreak]

Hey I just found this from years ago in my old notepads (before excel)
11/15-11/21/1999:
I labeled it GameJapan (?)
1. Poke Silver 601,594new
2. Poke Gold 597,334new
3. Chrono Cross 352,300new
6. Derby Stallion 99 14,505/770,531...

...It goes on to 30.

Share it with us!!!!!! Upload those Notepads ASAP!

 

[Square2005]

The Pokemon GS debate is over and we've said our cues.

Why not share the MC data with everyone? That's what I'm most interested it.

And the FFIII famitsu sales from 1990!!

 

[ioi]

Hey I just found this from years ago in my old notepads (before excel)
11/15-11/21/1999:
I labeled it GameJapan (?)
1. Poke Silver 601,594new
2. Poke Gold 597,334new
3. Chrono Cross 352,300new
6. Derby Stallion 99 14,505/770,531...

...It goes on to 30.

Still dyslexic I see Square2005, your NPD stuff is riddled with numerical errors

Share it with us!!!!!! Upload those Notepads ASAP!

Haha, it's all on vgcharts already

 

[cvxfreak]

Still dyslexic I see Square2005, your NPD stuff is riddled with numerical errors




Haha, it's all on vgcharts already

I'm not interested in VGCharts' numbers. I'm interested in untampered Media Create charts.

 

[Square2005]

Still dyslexic I see Square2005, your NPD stuff is riddled with numerical errors

ioi, you've got quite a few errors on your site and in your famitsu data, none of us are perfect...

 

[ioi]

I'm not interested in VGCharts' numbers. I'm interested in untampered Media Create charts.

Why?

 

[Square2005]

I'm not interested in VGCharts' numbers. I'm interested in untampered Media Create charts.

Back then I only collected the HW sales from the magic-box and only a few weeks of GameJapan and only the games I was intersted in.
The next entry I hav is 12/13/99 and only:
1. Gran turismo 2 815,030new

 

[duderon]

Jesus christ ioi, why do you shit up Media Create threads? Why not make your own goddamn thread when the media create thread comes out?

 

[jiggle]

spamming of your own website should be against the rules.

 

[cvxfreak]

Why?

...because I can just go to VGCharts and see the stuff for myself. Untampered MC is much more interesting. But clearly you're not interested in sharing them, so I'll stop asking.

Back then I only collected the HW sales from the magic-box and only a few weeks of GameJapan and only the games I was intersted in.
The next entry I hav is 12/13/99 and only:
1. Gran turismo 2 815,030new

Interesting... but check your PMs anyway.

 

[Square2005]

Then I hav Christmas week (not sure if it's doubled-over):
and only DC:
12/20/99-1/2/2000?
1. Shenmue - 304,781/380,472
17. Sea*man - 32,699/365,867
30. Space Channel 5 - 20,813/89,267

Looks extremely high BTW...

 

[ioi]

Untampered MC is much more interesting.

Why?

Then I hav Christmas week (not sure if it's doubled-over):
and only DC:
12/20/99-1/2/2000?
1. Shenmue - 304,781/380,472
17. Sea*man - 32,699/365,867
30. Space Channel 5 - 20,813/89,267

Looks extremely high BTW...

It was very high back then and quite erratic. I've said all this before that MC went through a stage of massively overestimating some titles (Mario Tennis 64 on 1.35m compared to 1.10m shipped) but settled down into 2001.

So you don't have Donkey Kong, Mario Party 2 or GT2 for that week Square 2005? Or even Pokemon G/S (lying in 2nd/3rd, combined it destroyed Shenmue)

 

[cvxfreak]

Why?

Because I've never seen them.

 

[jedimike]

Great start for Tenchu. Probably sold 8-10k, considering it was 5k day 1. Dead rising has some legs. Maybe 360 will stay above 2,000 this week. Pretty excited for famitsu 360s next monthly chart.

Me too... I'm more curious because DCharlie said he couldn't find a used 360 anywhere.

 

[ioi]

Because I've never seen them.

Funny that you and everyone else are in such a great position to judge the validity of the data on vgcharts then, when I have all the data and have spent the last 7-8 years analysing it in detail to get things right.

I'm still waiting for one valid reason why the figures on vgcharts are so 'questionable'?

 

[duderon]

TOP30
NDS:20
PS2:5
XBOX360:2
PSP:2
GBA:1

Don't call it a comeback, for 360!

 

[Drakken]

ioi......SHUT THE HECK UP.

 

[cvxfreak]

Funny that you and everyone else are in such a great position to judge the validity of the data on vgcharts then, when I have all the data and have spent the last 7-8 years analysing it in detail to get things right.

I'm still waiting for one valid reason why the figures on vgcharts are so 'questionable'?

In history, if a secondary source including statistics doesn't show its primary sources (and all of it, in detail), then it can be questioned. Simple as that.