Okay, here's the locked FAQ concerning how I figured out the XBLA Sales charts, how it's done and what I use.
Q. How do you go about finding tracking information for XBLA titles?
A. I use www.mygamercard.net to track the information. They have the most complete, and best gamercard tracking on the internet.
Q. Does anyone else track sales like this?
A. Not with as many numbers, not to my knowledge, atleast. Major Nelson (http://www.majornelson.com) gives a weekly top-10 of XBLA sales, but no numbers or LTD information. Also, joaomcgd at NeoGAF has a similar xml-enabled system, but has much less of a userbase vs. MGC (he has under 1,000 users. I'm a part of his tracking too, btw). His is at http://joaomgcd.planetaclix.pt/GafGamersLeaderboard.xml . His numbers, ironically, fall in line with what I use on some titles.
Q. How is this accurate? What about all of those demo downloads? What about pack-ins?
A. I believe this is rather accurate due to the huge sample size. MGC tracks almost 20% of every single Xbox that's ever been on the internet, and about 12-14% of the entire X360 userbase. This is about 3x the sample size that VGC uses for their sales data. At such a large percentage that is very un-biased (as it comes from too large of a group, from around the world), I think this is as accurate as we're going to get unless MS starts giving out sales info.
A #2. Demo downloads are taken out of the equasion: I track only valid gamercards with achievements. You cannot unlock achievements unless you've downloaded the full paid-for game, or pack-in title (ie, Arcade, Unplugged, 48hr download). Pack-ins are counted in sales, as there's no way to differentiate between who has a XBLA Unplugged copy, Arcade Bundle, or otherwise.
Q. If it's done via achievements, does that mean a game with "hard" achievements might recieve less sales, defined by your charts?
A. Unfortunately, yes, currently. I don't believe it's a huge margin on many games, but I do feel if there are certain games that's core component stays outside of something that just "happens" during normal gameplay (ie, a game that's meant for Multiplayer mostly, or entirely, has Single Player achievements). I hope to correct this in the future, but for now, it's there. But again: from what I've seen versus any other tracking service, and Major Nelson, there are only a few titles this might effect. Even then, I don't think it's huge.
Q. If the achievements throw off sales, why not use the total user's base, instead of achievements?
A. Not all XBLA titles are created equal. Some games are pure crap, and users know it, and do not download the full game. A great example would be Ecco: The Dolphin or Street Trace: NYC. Neither game has fared very well on the demo to attach ratio (under 5% for both). This is totally opposed to a game like Geometry Wars, or Castlevainia where the attach ratio is over 30%. That's a much bigger difference than hard achievements, I believe.
Q. What exact kind of multipliers do you use on a game for your methodology?
A. Currently, the multiplier used in older games is around 3.7x that of the current card base. New games are around 4.5-5.0 vs. the actual base of unlocked achievements. This is due to the fact that, on average, the newer the game, the less likely achievements have been unlocked, and therefore, counted.
Q. Why that number?
A. Using a small-er sample of the few games I have, I felt this was a safe number with a ~10% margin of error, most likely on the side of caution. If a user can let me know of a more accurate multiplier for a given game that has easy-to-see leaderboard charts, I'll gladly use it. The game I based this multiplier off of is Aegis Wings. Currently, the multiplier for cards w/ achievements to ranked users is 4.06 : 1.00. The ratio of estimated XBLA users to MGC cards is 4:86 : 1.00 using (VGC X360 sales / 60% attach ratio for XBLA), which is what MS has released as the XBLA attach ratio for consoles. Could it be off? Absolutely. However, I don't see it being lower than 3.5 or above 4.2 under most any circumstance.
Q. Does this agree with most sales charts?
A. Currently, yes. Most charts mirror my charts, minus the pack-ins and free titles. The charts are usually full of cheapo (ie free, or pack-ins), but the legitimate titles are almost always there, mirroring any MS-sanctioned chart.
Q. Does this focus more on accuracy for newer games, or older?
A. Typically older. The issue with card tracking is that the older a game gets, the larger, and therefore easier to track, the base is. Don't expect uber-accuracy with ultra-new titles, unless they sell very well like Puzzle Quest. In the case of PQ, my numbers seem to agree kinda close with joaomcgd's....Cept mine are slightly higher.
Q. I think your methodology is too high/low, or what your doing isn't working right.
A. Then feel free to correct me, or show me how to do it better. I want this to be as accurate as possible, and need help from anyone that would like to. I am sure that my numbers aren't 100% perfect, and for those that can lend an idea, feel free to PM me.
Q. Are the notes/exceptions calculated into the overall revenue of a game? If so, why not?
A. Notes and exceptions are generally not taken into account in terms of revenue. Merely because I'm not a rocket scientist, and haven't done this forever. I have been upfront with what games are obviously suspect, and hope that it helps those questioning the charts, to have some idea of where titles might not of profited as much as I show.
As for the reasoning, aside from the not a rocket scientist part, each title is getting moneyhatted (paid for) some way by MS. So at some point, they are making revenue from some crazy deal....But obviously not as much.
List of notes/exceptions for various games with Pack-ins, and such:
() = Indicates an absolutely free game. Yaris, Aegis Wings, and Totemball.
* = Indicates a game that is normally full price, but recieved a 48hr "free download". Carcassonne (US/PAL), Texas Hold'Em Debut, and Hexic 2 (Korea)
** = Indicates a game has recieved a temporary price drop. Thus far for each game, it's been 50% for 1 day. Zuma, Small Arms, Dig Dug, Gauntlet, Frogger, and Bankshot Billards 2.
^ = Xbox 360 Arcade Pack-in Title. Boom Boom Rocket, Luxor 2, PacMan: CE, Feeding Frenzy, Uno
^^ = Recieved a permanant price drop for Xbox Live Arcade Hits program, effective 12/5/07. Bankshot Billiards 2 and Lumnies! Live were reduced from 1200>800 Microsoft points. Marble Blast Ultra, Small Arms, and Zuma were reduced from 800>400 Microsoft points. Revenue for each game was calculated using old prices for week ending 12/8/07, but then reduced to proper pricing from then on.
*** = A fully playable game in Xbox Live Arcade Unplugged. Bejewled 2, Geometry Wars: Evolved, Hardwood Backgammon, Outpost Kaloki X, Wik and the Fable of Souls.
^^^ = Included as a free download in the Xbox Live Vision Camera. Uno, TotemBall, Robotron 2084. All of which are given codes for free downloads (however, TotemBall cannot be played w/o the camera. Uno and Robotron 2084 can).
Currently, that's it for exceptions and notes. That's alot to cover, but it's very important to note, as these can skew the data to higher-than-real numbers.
Back from the dead, I'm afraid.