Sales Tracking Accuracy: Famitsu vs. Media Create (W/E May 11th) - News

by VGChartz Staff , posted on 16 May 2008 / 2,847 Views

How accurate are sales trackers? The answer is not really very easy to determine, since there are rarely any standards to benchmark against. NPD has a monopoly in the US, so the only comparison can be made with manufacturer sales to retail and a ballpark estimate of retail stock. Chart-Track and GfK typically have local monopolies in Europe, and rarely release public sales figures besides. Japan, however, has a unique situation, with two professional sales trackers releasing weekly sales figures to the public. Both Famitsu (Enterbrain) and Media Create cover about 2/3 of the market, so how much can that remaining 1/3 affect the accuracy of the figures? Let's take a look at the latest week of sales data to get an idea.

First, a look at the hardware sales data from each tracker:

	Famitsu	Media Cr.	% Diff
PSP	73k	90k	21%
Wii	71k	67k	6%
DS	48k	51k	6%
PS3	8.5k	8k	6%
PS2	8k	7.5k	6%
360	1.7k	1.3k	27%

What do we see here? Well, what stands out the most is the large 21% difference in PSP sales between the two trackers. Combined with the 6% difference in Wii sales, look at how the sales narrative changes. In Famitsu, Wii has nearly caught up to PSP again. In Media Create, PSP sales are still 34% above Wii sales. 360 sales are also 27% different, though the small base values mean that the large percentage is only 400 units -- emphasizing the difficulty in accurately tracking small numbers. The other hardware numbers are all within 6%, but none are any less than that amount.

Now for software sales from the past week:

	Famitsu	Media Cr.	% Diff
Mario Kart Wii (Wii)	119k	107k	11%
Monster Hunter Portable 2nd G (PSP)	95k	82k	15%
Wii Fit (Wii)	64k	53k	19%
Link's Crossbow Training (Wii)	50k	49k	2%
Meccha! Taiko Drum Master DS (DS)	34k	34k	0%
Pokemon Ranger Batonage! (DS)	28k	27k	4%
Deca Sports (Wii)	13k	22k	55%
Wii Sports (Wii)	21k	21k	0%
Boura wa Kaseki Holder (DS)	14k	17k	20%
Super Smash Bros Brawl (Wii)	11k	16k	38%
Mario Kart DS (DS)	26k	16k	50%
Wii Play (Wii)	18k	16k	12%

The top three titles are all more than 10% different, with Famitsu having higher sales for all three. Four of the next five titles agree very closely. Yet even within the top 10 there are some very large differences as the sales volume decreases, most notably the 55% disparity in Deca Sports, 40% in Brawl, and 50% in Mario Kart DS.

Just for completeness, let's look at the sales rank for each of these titles for each tracker:

	Famitsu	Media Cr.
Mario Kart Wii (Wii)	1st	1st
Monster Hunter Portable 2nd G (PSP)	2nd	2nd
Wii Fit (Wii)	3rd	3rd
Link's Crossbow Training (Wii)	4th	4th
Meccha! Taiko Drum Master DS (DS)	5th	5th
Pokemon Ranger Batonage! (DS)	6th	6th
Deca Sports (Wii)	11th	7th
Wii Sports (Wii)	8th	8th
Boura wa Kaseki Holder (DS)	10th	9th
Super Smash Bros Brawl (Wii)	16th	10th
Mario Kart DS (DS)	7th	12th
Wii Play (Wii)	9th	11th

The top six are the same across the board, and the top 10 contains eight of the same titles for each. However at the bottom end of the top 10, the large differences in the raw numbers shake up the order of the titles a bit, with some titles off by as many as six ranks and each tracker having two titles not present in the other's top 10.

What can we conclude from this? The first is that even professional trackers can be off by large amounts at times, despite being in close agreement most of the time. The second is that the differences have a greater impact when the sales volume is low, as despite 10-20% differences in the top three software titles, the order was still consistent. The third is that even assuming that the "true" value happens to be directly between the Famitsu and Media Create numbers, each tracker is still off by as much as 30% on some of the numbers, despite a majority of market coverage for each -- and even that's when only including the top 10 titles from each list.

While this piece isn't meant to compare VGC numbers, as a side note the VGC values for each piece of hardware and software listed were somewhere in between the Famitsu and Media Create numbers.

Contact VGC at kjackson@vgchartz.com

30 Comments

kirby007 (on 04 June 2011)

i lol

kirby007 (on 04 June 2011)

i lol

kirby007 (on 04 June 2011)

i lol

kirby007 (on 04 June 2011)

i lol

kirby007 (on 04 June 2011)

i lol

kirby007 (on 04 June 2011)

i lol

kirby007 (on 04 June 2011)

i lol

kirby007 (on 04 June 2011)

i lol

kirby007 (on 04 June 2011)

i lol

mmolichking3 (on 23 February 2009)

greepoman (on 20 May 2008)

WOW you compared a whole 1 WEEK!!!!!! That must've taken a lot of time. So many fluctuations, etc. can happen in a whole week...how on earth can you try to draw conclusions from it????????????????????????

Try looking at comparisons for say the last 52 weeks..you'll see the % difference it's tiny.

I still can't believe you're basing your conclusions on 1 week.....wow...that's just bad.

johnsobas (on 17 May 2008)

to be fair here the error isn't near as bad as the error between NPD and VGC this month, and the numbers on most of the software are much, much, much smaller making it hard to make as accurate of an estimate.
This is just for 1 week, if you average it out over a month the difference is most certainly even smaller.

Dolla Dolla (on 17 May 2008)

I think it would be interesting to compare the monthly Famitsu numbers to monthly Media Create numbers to see how different they become. Nonetheless, interesting.

Esmoreit (on 17 May 2008)

Could this be a basis for allowing other tracking services within the US market? Thus basicly revoking NPD's monopoly in favor of a more diversified market-analysis?

hunter_alien (on 17 May 2008)

Well there is a pretty big difference between the two . Well Il rather listen to VGC :P

bardicverse (on 16 May 2008)

The official Japan sales are here:

Wii - more than God can count to
PSP - 84 trillion
DS - Eleventy billion
PS3 - two truckloads
XB360 - 4

;)

Final-Fan (on 16 May 2008)

I wonder what their overlap is -- that is, if they combined forces what would their market coverage be?

shams (on 16 May 2008)

Its really interesting - in the end "no one" is right, but the more sources of data you get - the better a comparison you can make.

I always liked the idea of averaging Famitsu & MediaCreate - it would also be great to get both organisations together, and comparing data - to find out where the data "holes" are.

CrazzyMan (on 16 May 2008)

Well, in both cases, you get estimates of sales, that are closest to real sales. =)
So, people shoudnt look at these numbers like to a facts. Its just some data to consider.

Anyway, with more time, more data comes, and with shippments numbers from publishers the ability to get better estimate appears. =)

dolemit3 (on 16 May 2008)

Could the op (or someone else) provide the reference as to how many percent the two firms track? The best I could find is the absolute number of retails for Enterbrain:
[quote]Drawing from our extensive network within the Japanese game market, our sources include:

3,200 participating retail stores in Japan

31,000 other franchise and convenience stores in Japan
[/quote]
[url]http://www.enterbrain.co.jp/en/c_outline/mark.html[/url]

TWRoO (on 16 May 2008)

^See I wasn't using it to say VGC is more accurate than NPD. For two tracking services to be 20%+ apart when both track accurately 60%+ of the market means that NPD can easily be 10% + out when they also track about 60%.

Of course VGC is usually further than 10% with at least 1 of the 6 hardwares each month, but I was stating to stop people using NPD as "the truth" which people do quite a lot on here.

Yes they are a better estimate than VGC, no they are not 100% correct as many people seem to assume (admitedly it is much less here on VGC but on other forums etc they never mention NPDs margin of error)

GfK likely have a much smaller margin of error given that they track 90% of the market (likely diferent between countries, but in general) which is why ioi takes much more notice when VGC is off GfK, but people complain more when there are differences with NPD (perhaps except a few like diomedes who complain equally)

erikers (on 16 May 2008)

TWRoO says:
People should consider this when they class NPD as "the truth" when they track a similar (slightly smaller) amount of the market as these two do.

Uh...the differences between Famitsu and Media create are incredibly close with only one being 17k off. It's not like they were hundreds of thousands off like VGC was from NPD this month for umpteenth time. That's not even a good comparison to try and make.

When you project 2 difference consoles at over 300k for the month and they both do under 190k you're not even in the ballpark. It just shows that every time there's a month with a huge game, price cut or holiday month (in other words a month where your tracking ability would be on showcase) the numbers end up being off more than usual.

People don't class NPD as the "truth", just every company in the gaming industry who uses their numbers in their press releases so you might want to use their numbers as a better estimate, just a common sense thought. You don't see microsoft saying they sold 309k consoles instead of 188k this month do you?

famousringo (on 16 May 2008)

Lies!

Statistics are all lies and there's no way that useful conclusions could possibly be drawn from them!

Seneque (on 16 May 2008)

Nice read. I wanted to write something like that since a long time, but couldn't get it out without talking about gaussians, standard derivation, variance and stuff like that.
If it can get some people to realize the difficulty of tracking, it will be a big step forward.

superchunk (on 16 May 2008)

Great article. Helps prove that VGC is doing an excellent job.

The Ghost of RubangB (on 16 May 2008)

Who are we supposed to believe DKII??? It's maddening!

NintendoMonopoly (on 16 May 2008)

Well in that, case I feel honored to be número uno .

Aidman (on 16 May 2008)

@NintendoMonopoly if you're post isn't the first Nintendo is still Number One and this is more than enough for us lol.

NintendoMonopoly (on 16 May 2008)

Darn! I was so close to being the 1st to post.Interesting stuff. You should have put VGC #'s up anyways. It's surprising how the #'s can be so far off when they both cover so much of the market. If it wasn't for vgchartz I wouldn't know who to believe. Definitely a good read. thx for the info.

TWRoO (on 16 May 2008)

People should consider this when they class NPD as "the truth" when they track a similar (slightly smaller) amount of the market as these two do.

Existing User Log In

New User Registration

Alien: Rogue Incursion Announced for PS VR2, SteamVR, and Quest 3

Switch 2 to Reportedly Have Magnetic Joy-Cons

Resident Evil Creator Originally Planned to Leave Tango Gameworks 8 Years Before Doing So

GALIX: New Horizons Announced for PS5 and PC

Sales Tracking Accuracy: Famitsu vs. Media Create (W/E May 11th) - News

More Articles

Popular Stories