• 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.

Media Create Sales 12/31 - 1/6 2008

Stumpokapow

listen to the mad man
JJConrad said:
Its the same argument worded differently.

Whether or not either ends up being true or false, it really isn't.

If we have two classes, each with twenty students, and the grades are
Class A: 100, 95, 90, 85, 80, 75, 70, 65, 60, 55, 50, 45, 40, 35, 30, 25, 20, 15, 10, 5
Class B: 100, 97, 92, 88, 84, 45, 20, 11, 8, 6, 5, 3, 3, 3, 2, 1, 0, 0, 0, 0

The statement "Class B has students who do as well as Class A's best" is true. The statement "Class B is as good as Class A" is false.

Again, whether or not the data ends up playing out in support of the idea that Wii third-party mid-level games are selling as well as PS2 third-party mid-level games, it's still entirely different than just comparing the top end.
 

sphinx

the piano man
PantherLotus said:
For the record, we ARE in the midst of an avalanche of announcements and releases. Just watch.

for real??

:(

my patience is wearing thin.

I don't think there has been any noticeable announcement since E3 2006, capcom's crazy fall 2007 announcements aside, of course.

Has there been an announcement about an exclusive game for PS3 or Wii done before their respective launches?? Resistance 2 is there... but since I don't care much about that game... >_<.
 

Xeke

Banned
Stumpokapow said:
Top <x> versus Top <x> is not really a good comparison. The argument about Wii third-party sales is not "there are no good third-party sales on the Wii", it's that "many or most third-party Wii games are bombing".

Can you show me data that just as many PS2 games didn't bomb?
 

Stumpokapow

listen to the mad man
Xeke said:
Can you show me data that just as many PS2 games didn't bomb?

I spent far too long either earlier in this thread or in last week's thread providing a plethora of early PS2 games of comparable quality to the Wii games included in the "bomba" category that sold substantially better. I also provided examples of software on the X360, PSP, and the first year of the DS. I'm not going through the effort again. You're welcome to go through Josh's data yourself and pull out a representative (in a subjective way, not a statistical way) sample of PS2 first-year third-party mid-grade sales.

If you want to take all the PS2's first year games, all the Wii's first year games, control for market size, control for game quality using either your personal heuristic or an aggregate of press reviews (I concede that the former is more likely given the lack of Japanese press review score aggregators that I am aware of), then do a multivariate regression on all the data and give me a graph, I'd be happy to accept that graph as a statistically valid determination of whether or not the claim that Wii third-party games are selling less than PS2 third-party games year one relative to their profile.

Of course, again, at this point I don't feel like even retreading this argument and I don't even care right now whether it's true or false. But comparing the top 10 of each is NOT a valid comparison. See my classroom example for the distinction. Normally I'd just say "median is a better measure than upper quartile" but given the fucking pathetic abuse of statistics on GAF, I'll stick to the analogy.
 

apujanata

Member
ethelred said:
Argh. People, please go take some stats classes before commenting here.

I always fall asleep during stats class, so you need to teach me (instead of just refreshing my memory using my college books).

Based on other people post, I think most of us has the same problem (not enough exposure to statistics).
 

JJConrad

Sucks at viral marketing
Stumpokapow said:
Top <x> versus Top <x> is not really a good comparison. The argument about Wii third-party sales is not "there are no good third-party sales on the Wii", it's that "many or most third-party Wii games are bombing".
Its the same argument just further down the scale. The facts of the matter are that both systems have games sell worse than they should have and games that have sold better than the should have. Trying to pin the cause of those sales on 3rd party status is naive.

Capcom and Sony did not have any major success stories on the PS2 in 2000, yet both had huge hits for the system in 2001. How were they turn their performances around even though Square, who was already very dominating on the platform, was releasing FFX at the time?
 

ethelred

Member
Dragona Akehi said:
What for? The first thing that they teach you there is that it's all made up anyway!

Not all of them. Just 83%.

Segata Sanshiro said:
i just want to climb up on the rooftops and yell

i miss my best friends dragona and ethelred

boop boop

I miss Betty Boop, too.

JJConrad said:
Its the same argument just further down the scale.

No, as Stumpy showed, it really isn't the same argument.

JJConrad said:
Capcom and Sony did not have any major success stories on the PS2 in 2000, yet both had huge hits for the system in 2001. How were they turn their performances around even though Square, who was already very dominating on the platform, was releasing FFX at the time?

How do you define "very dominating?"
 

Neo C.

Member
Stumpokapow said:
Whether or not either ends up being true or false, it really isn't.

If we have two classes, each with twenty students, and the grades are
Class A: 100, 95, 90, 85, 80, 75, 70, 65, 60, 55, 50, 45, 40, 35, 30, 25, 20, 15, 10, 5
Class B: 100, 97, 92, 88, 84, 45, 20, 11, 8, 6, 5, 3, 3, 3, 2, 1, 0, 0, 0, 0

The statement "Class B has students who do as well as Class A's best" is true. The statement "Class B is as good as Class A" is false.
There's a difference between grades and sales. You can have high average grades or low average grades, there is no limits except the grade range. When we talk about sales, we must consider the size of the userbase, the money and the time the average consumer might spend.

For example my limit is about ten games yearly. Let us say I buy 8 first-party games this year. In that case, I simply only have the money and time for 2 third party game.

Of course, the Tie-ratio isn't stable and vary greatly between the plattforms. Still, I think the absolute software sales and the Tie-ratio of the Wii are acceptable, It's just the first party games are extremely successful.
 

creamsugar

Member
Famitsu

DSL&#12288;103000
PSP&#12288;84000
PS2&#12288;15000
PS3&#12288;34000
Wii&#12288;82000
360&#12288;5500

1.Wii Fit&#12288;93000&#65288;1005000&#65289;
2.Mario Party DS&#12288;52000&#65288;1493000&#65289;
3.Wii Sports&#12288;41000&#65288;2633000&#65289;
4.Wii Play&#12288;26000&#65288;2120000&#65289;
5.Layton 2&#12288;23000&#65288;697000&#65289;
6.FF4&#12288;20000&#65288;547000&#65289;
7.MHP2&#12288;20000&#65288;1551000&#65289;
8.Mario Sonic&#12288;20000&#65288;457000&#65289;
9.Mario Kart DS&#12288;19000&#65288;2783000&#65289;
10.Rune Factory 2&#12288;18000&#65288;74000&#65289;
 

Yoshi

Headmaster of Console Warrior Jugendstrafanstalt
:( @no Mario Galaxy, so we will most likely not see it reaching a million and therefore have to wait till end of year to see whether it reached this mark...
 
creamsugar said:
Snipped figures

Wow - quite a slump. ISTR this being pretty typical, though - can any other long-time SalesAge GAFers confirm?

:( @no Mario Galaxy, so we will most likely not see it reaching a million and therefore have to wait till end of year to see whether it reached this mark...

Super Mario Galaxy: 18000&#65288;827000 LTD&#65289;

It's going to make it eventually, even if this week marks the point at which sales start to really fall off.
 

max-pain

Member
creamsugar said:
Famitsu

DSL&#12288;103000
PSP&#12288;84000
PS2&#12288;15000
PS3&#12288;34000
Wii&#12288;82000
360&#12288;5500

1.Wii Fit&#12288;93000&#65288;1005000&#65289;
2.Mario Party DS&#12288;52000&#65288;1493000&#65289;
3.Wii Sports&#12288;41000&#65288;2633000&#65289;
4.Wii Play&#12288;26000&#65288;2120000&#65289;
5.Layton 2&#12288;23000&#65288;697000&#65289;
6.FF4&#12288;20000&#65288;547000&#65289;
7.MHP2&#12288;20000&#65288;1551000&#65289;
8.Mario Sonic&#12288;20000&#65288;457000&#65289;
9.Mario Kart DS&#12288;19000&#65288;2783000&#65289;
10.Rune Factory 2&#12288;18000&#65288;74000&#65289;

WTF:lol
 

creamsugar

Member
New title
Katekyoo Hitman Reborn! Dream Hyper Battle! Wii&#12288;6900
ARMORED CORE 4 360&#12288;2000
ARMORED CORE 4 PS3&#12288;1400
&#38646;&#24335;&#12288;600


11.SMG&#12288;18000&#65288;827000&#65289;
12.MP8&#12288;16000&#65288;1138000&#65289;
13.DQ4&#12288;16000&#65288;1139000&#65289;
15.SO1&#12288;14000&#65288;175000&#65289;
16.WEPS2&#12288;13000&#65288;595000&#65289;
17.New SMB&#12288;13000&#65288;5051000&#65289;
18.ACWW&#12288;12000&#65288;4554000&#65289;
19.COD4PS3&#12288;11000&#65288;50000&#65289;
20.Layton&#12288;11000&#65288;803000&#65289;
21.Pokemon dungeon &#12288;11000&#65288;1313000&#65289;
24.Brain Age 2&#12288;9100&#65288;4816000&#65289;
25.Golf PSP2&#12288;9000&#65288;194000&#65289;
26.&#23478;&#35336;&#12288;8300&#65288;396000&#65289;
27.REUC&#12288;8200&#65288;242000&#65289;
28.Pokemon&#12288;7900&#65288;5425000&#65289;
30.&#40845;2&#12288;7700&#65288;95000&#65289;
 
Wii Fit refuses to give into the post-holiday slump.

AC4360&#12288;2000
AC4PS3&#12288;1400

Is that Armored Core 4? If so that's some pretty bad numbers.
 

Stumpokapow

listen to the mad man
Neo C. said:
For example my limit is about ten games yearly. Let us say I buy 8 first-party games this year. In that case, I simply only have the money and time for 2 third party game.

Normative analysis. "Gamers ought to buy fewer third party games on the Wii because they ought to be spending more money on first party games". Might be true, but in the mean time, positive/descriptive analysis tells us something that's far more useful for third parties "Third parties software is not doing well".

JJConrad said:
Its the same argument just further down the scale. The facts of the matter are that both systems have games sell worse than they should have and games that have sold better than the should have. Trying to pin the cause of those sales on 3rd party status is naive.

Okay I'll just leave it at "median comparisons and median comparisons with outliers removed are both statistically more valid and empirically more useful comparisons than an arbitrary top-x comparison, which is even less useful or valid than the incredibly useless top-quartile comparison".

And I'll say for the third time in as many pages that I don't even care about the actual argument about Wii third-party sales anymore because I'm so frustrated about it. I just genuinely think that whatever comparison you want to use, an arbitrary top-x comparison between systems is a garbage piece of analysis and not fit to draw any conclusions other than "Each system has games that have sold well", which is as trivial of a conclusion as anyone has ever found.

If you want to actually do proper legwork here, do a linear regression or a multivariate regression and do a one-tailed t-test or any other null hypothesis rejection test and test the hypothesis. I will warn you before you waste your time that the chance of committing a sampling error is quite high given that the PS2 data is substantially poorer than the Wii data in breadth but the Wii data does not have any updates from the yearly top 500, and as such the strength of the Wii data we have is quite weak for the most part.
 

jgwhiteus

Member
What was Sega's biggest game before M&S? It seems like the Wii version alone is a virtual lock to pass 2M WW (I think someone calculated it was already past 1.2M not accounting for the US December NPD's and a few other territories), maybe even 3M and beyond...

And Famitsu's figures are still 50K-100K off MC's and Nintendo's for WiiFit - both said WiiFit passed 1M the first week of January, and Famitsu has it passing the 1M mark this past week.
 

test_account

XP-39C²
BishopLamont said:
Is that Armored Core 4? If so that's some pretty bad numbers.

AC4 is over a year old though. I wonder why it suddently turned up on the charts again. Maybe its some best price thingie to "prepare" people for the the next AC4 game that is out in 2 months or so :)
 
ARMORED CORE 4 360&#12288;2000
ARMORED CORE 4 PS3&#12288;1400
They are budget re-releases

15.SO1&#12288;14000&#65288;175000&#65289;
Slowly getting near its first shipment (200k). Doesn't look like it'll beat SE target (250k AKA what the original did)

19.COD4PS3&#12288;11000&#65288;50000&#65289;
Heh, Media Create and Dengeki had the PS3 edition over the 360 one in their first weeks but Famitsu didn't. I guess they are correcting it a little. Good numbers, looks like it'll do 100k between the 2 platforms easily.

25.Golf PSP2&#12288;9000&#65288;194000&#65289;
Nice to see Mingol P2 legs, as usual with the series. Around half the LTD of the first one full-priced release (which is at something like 660k counting the budget release).

26.Household Diary&#12288;8300&#65288;396000&#65289;
30.Ryu Ga Gotoku 2 (The Best!)&#12288;7700&#65288;95000&#65289;
RGG2 budget release doing good, nearing 700k between the original release and this one. RGG is at 800k. Stellar numbers for both, wonder how much PS3 can do for the third one.
 

liuelson

Member
Stumpokapow said:
Whether or not either ends up being true or false, it really isn't.

If we have two classes, each with twenty students, and the grades are
Class A: 100, 95, 90, 85, 80, 75, 70, 65, 60, 55, 50, 45, 40, 35, 30, 25, 20, 15, 10, 5
Class B: 100, 97, 92, 88, 84, 45, 20, 11, 8, 6, 5, 3, 3, 3, 2, 1, 0, 0, 0, 0

The statement "Class B has students who do as well as Class A's best" is true. The statement "Class B is as good as Class A" is false.

Again, whether or not the data ends up playing out in support of the idea that Wii third-party mid-level games are selling as well as PS2 third-party mid-level games, it's still entirely different than just comparing the top end.

Unfortunately, the problem with comparing generations is that, on a macro level, you are comparing only 2 things. To extend your classroom analogy, there may be a lot of reasons why the middle and lower performing students in Class B are doing worse than the middle and lower performing students in Class A, none of them involving the relative merits of Teacher B v. Teacher A. The time scale also matters - Teacher B's students are today; Teacher A's students are from the past. That kind of comparison is inherently fraught with statistical difficulties.

Edit: "statistical difficulties" not in terms of actual analysis, but in terms of ability to draw meaningful conclusions.
 
jgwhiteus said:
What was Sega's biggest game before M&S? It seems like the Wii version alone is a virtual lock to pass 2M WW (I think someone calculated it was already past 1.2M not accounting for the US December NPD's and a few other territories), maybe even 3M and beyond...

And Famitsu's figures are still 50K-100K off MC's and Nintendo's for WiiFit - both said WiiFit passed 1M the first week of January, and Famitsu has it passing the 1M mark this past week.
Sonic the Hedgehog 2, I think. It's hard to get any decent numbers from ye olde days, but it was packed in with the Genesis during the prime of its life and I do seem to recall SEGA trumpeting an "over 5 million sold" on it at one point.
 
Stumpokapow said:
Okay I'll just leave it at "median comparisons and median comparisons with outliers removed are both statistically more valid and empirically more useful comparisons than an arbitrary top-x comparison, which is even less useful or valid than the incredibly useless top-quartile comparison".
Given that most of our data comes from weekly Top 30 lists, we rarely see how the lesser titles are actually doing, though. The top is the only place where the data is anywhere near full.
 

farnham

Banned
jgwhiteus said:
What was Sega's biggest game before M&S? It seems like the Wii version alone is a virtual lock to pass 2M WW (I think someone calculated it was already past 1.2M not accounting for the US December NPD's and a few other territories), maybe even 3M and beyond...

And Famitsu's figures are still 50K-100K off MC's and Nintendo's for WiiFit - both said WiiFit passed 1M the first week of January, and Famitsu has it passing the 1M mark this past week.
Virtua Fighter 2 sold one million
Oshare Love and Berry sold one million
 

Stumpokapow

listen to the mad man
jgwhiteus said:
What was Sega's biggest game before M&S? It seems like the Wii version alone is a virtual lock to pass 2M WW (I think someone calculated it was already past 1.2M not accounting for the US December NPD's and a few other territories), maybe even 3M and beyond...

Love and Berry was Sega's biggest game in Japan EVER. Still is. Worldwide, a number of Sega's previous-gen titles broke a million. This gen, Crackdown might have broken a million worldwide but I'm not sure because despite loving 360 games, I don't follow their sales much except for NPD.

And Famitsu's figures are still 50K-100K off MC's and Nintendo's for WiiFit - both said WiiFit passed 1M the first week of January, and Famitsu has it passing the 1M mark this past week.

Wii Fit is literally the edge case for tracking since it's being carried and sold by a lot of mainline stores that normally have a small video game presence. Given that it's highly unlikely that either MC or Famitsu have separate extrapolation models for special case software (Fuzzy assures me NPD doesn't and have only one extrapolation model for each platform), it's likely that between tracking coverage and extrapolation models, the difference is accounted for.

Consider this an EXCEPTIONALLY simplified version of the math at hand.
For example, a department store chain whose marketshare is 2% of the Japanese video game market, will likely have a >2% share of Wii Fit sales (let's say 4% for shits and giggles). So when it comes time to extrapolate and you have 50k representing what you think is 50% of the market but is in reality 60% of the market, you extrapolate 100k when in reality the sales are 83k. Make sense?

Extrapolation and sampling differences will explain all differences between trackers, but Wii Fit is a special edge case because of its nature. Besides the whole "mainstream/adult" aspect of it, it's also an unusual form factor which probably means that department stores are more well equipped to stock it than Kazuki's Video Game Dungeon.
 

Stumpokapow

listen to the mad man
liuelson said:
To extend your classroom analogy, there may be a lot of reasons why the middle and lower performing students in Class B are doing worse than the middle and lower performing students in Class A, none of them involving the relative merits of Teacher B v. Teacher A. The time scale also matters - Teacher B's students are today; Teacher A's students are from the past. That kind of comparison is inherently fraught with statistical difficulties.

Well, obviously besides the fact that I was merely intending to use this analogy (which is crafted rather than real-world) to show the difference between a top-x/top-% comparison between two samples and the more apt regression or median comparisons, I also don't really accept this criticism because it centers on the idea that all analysis must be normative in nature rather than positive/descriptive.

It's true that this sort of comparison masks the fact that the students may be doing better or worse for reasons entirely different than their teachers, but it's also true that the hypothesis "Class A is doing better than Class B" doesn't care why it may or may not be the case, just that it is or isn't.

I do get your point though. It also sucks, because a lot of the posters here, even the very very very smart and knowledgeable ones, don't have a good background in statistics. Hell, even mine just consists of the handful of courses I had to complete for my math minor. So you get a situation where some of our posters are exceptionally bright high school students or college lib arts students who are great at laymen's analysis but not up to speed on stats. I'm not thinking of any people in particular (in fact, I know very little about who is doing what and who has what background) but it's obvious from the lack of actual math and statistics in the general GAF analysis that it's the case.

I still remember the time some poster made a graph of hardware sales and then showed differentiations and integrations of the graphs to describe the phenomenon of "closing the gap [in sales]". That's when I realized that a clear limitation of an open community like GAF is that no matter how high of a quality the community, some people weren't going to get those graphs because they didn't have a calculus background.

JoshuaJSlone said:
Given that most of our data comes from weekly Top 30 lists, we rarely see how the lesser titles are actually doing, though. The top is the only place where the data is anywhere near full.

I certainly agree with this and I acknowledge it later in the post or another post when I mention the risks of doing a regression or median analysis. But if we need to choose between accepting that our data is not good enough to immediately test a hypothesis or using a clearly tortured test that really doesn't even represent the core of the hypothesis, I think we should choose the former.

I really do believe that the top 500 will answer the key question from a descriptive point of view; with off-chart growth added, do some of the mutually agree upon Wii bombas manage to sneak up in the numbers?

I think the actual "why" of the sales will likely be answered with time as we see whether or not Nintendo adjusts their strategy, third parties adjust theirs, the quality rises or falls, the tie ratio rises or falls, etc.
 

schuelma

Wastes hours checking old Famitsu software data, but that's why we love him.
Interesting- Wii Fit doesn't appear to be cannnibalizing Wii Sports sales much at all. Wii Sports continues to sell to about half of the new Wii owners for the week.
 

liuelson

Member
JoshuaJSlone said:
Given that most of our data comes from weekly Top 30 lists, we rarely see how the lesser titles are actually doing, though. The top is the only place where the data is anywhere near full.

Most of this "argument" is also not very meaningful statistically. Why even make a distinction between 1st party and 3rd party sales? They compete for the same yen. Statistical comparisons between current gen and previous gen, between different manufacturers and their different corporate strategies / philosophies, using an arbitrarily defined segment of the market (3rd party sales only)...Let's not pretend that we're doing anything other than having fun with numbers.

Edit:

Stumpokapow said:

I pretty much agree with everything in your last post. :)
 
Sony should stop lurking NeoGAF...just after I ask why they don't launch a new PSP color in a week with a major software, they say there is one coming with MHP2G release. Come on :/

Oh, and SEGA has done well protecting Valkyrie of the Battlefield, taking it out of the REALLY crowded March, and putting in on 27th April, which is just before Golden Week. Nice improvement for PS3 Golden Week already, Madden last year, Valkyrie this one :lol
 

Stumpokapow

listen to the mad man
liuelson said:
Most of this "argument" is also not very meaningful statistically. Why even make a distinction between 1st party and 3rd party sales? They compete for the same yen. Statistical comparisons between current gen and previous gen, between different manufacturers and their different corporate strategies / philosophies, using an arbitrarily defined segment of the market (3rd party sales only)...Let's not pretend that we're doing anything other than having fun with numbers.

I think on a descriptive level there is a meaningful distinction between the two because if you get a hypothetical console (not arguing the merits of this example wrt any real-world consoles/handhelds) where one company does exceptionally well and other companies do exceptionally poorly, those other companies are going to have to take into account the fact that overall statistical analysis of the console will reveal numbers skewed up by the one successful company.

I guess one could get into the argument that it might be frivolous to arbitrarily claim 3rd party as the meaningful unit of analysis rather than individual companies and obviously in an ideal data-perfect world we'd be able to draw much more meaningful conclusions by analyzing by genre, budget, marketing, pedigree, quality vis-a-vis review aggregation, company, CERO rating, etc. In the mean time, lacking the data to granulate any further than "3rd party" and a few of the larger individual companies, I feel it's a decent metric.

Certainly true. But I think the normative v. descriptive issue is the reason for the arguing. People are interpreting the descriptive analysis ("Nintendo bombas") as a normative evaluation. Even using the word "bombas" has normative connotations.

Something tells me that sales-age threads wouldn't get half the replies they do if we published all out thoughts in LaTeX and used only academically rigorous language :lol

I think, too, there's still two camps in sales-age; those who are interested in numbers (whether or not they also have their own console preferences), and those are interested ONLY in seeing through a console war agenda. The latter group is by definition incapable of positive analysis and will see things only in terms of "Agrees with what I think ought to happen" and "Disagrees with what I think ought to happen" and the way that the discourse flows in the forum format, if a Console Warrior accuses a Numbers Guy of being a Console Warrior, other Numbers Guys might begin to treat that Numbers Guy as being a Console Warrior, which causes a shift in the overall perception of the original analysis that sparked the argument.

Edit: I think we've probably eeked out most of the useful discussion we can have on this issue, or at least I'm starting to get exhausted :D What's your academic background out of curiosity? I've seen your posts on adoption rate curves so I'm assuming it's either econ or business with an econ/math focus?
 

jarrod

Banned
apujanata said:
Do you have the WW # of Dead Rising and Lost Planet ? I don't have the number, so I could not comment on your statement. If those two games can do the same (WW) or half (50%) of Capcom's successful new IP # on PS2 (like Devil May Cry), then all is good. If is is significantly less than 20% of DMC, then it (the plan) is not good.

If you don't like DMC, you can change it to some other new IP that was released during PS2 first two years, like Dead Rising and Lost Planet.
Direct from Capcom's IR page... so as of 09.30.2007...

Dead Rising (360) 1.300.000
Lost Planet: Extreme Condition (360) 1.400.000

Onimusha: Warlords (PS2) 2.020.000
Devil May Cry (PS2) 2.160.000

Resident Evil (PS) 2.750.000
Dino Crisis (PS) 2.400.000


farnham said:
Sega had bigger success on Nintendo platforms then on their platforms... of course only speaking about sales
Not at all... SEGA had Nintendo like domination of their own consoles actually, and moved far more units of their own games. Same story with PCE or NeoGeo, the first Japanese console where 3rd parties really ruled the roost was PlayStation. Before that, the only thing really comparable were computer formats like PC88 or MSX.


Stumpokapow said:
Love and Berry was Sega's biggest game in Japan EVER. Still is. Worldwide, a number of Sega's previous-gen titles broke a million.
Nope, VF2 is still SEGA's best seller ever in Japan. VF1 should be in 3rd place (around 900k iirc).

Segata's right on Sonic 2 being the best ever, ever worldwide at 5m+, but it was also a pack in for part of it's lifetime. So was Sonic 1 actually.
 
Top Bottom