Pinch Media records a lot of data as part of our reporting on the AppStore, but there’s a lot we deliberately don’t publish. For instance, we don’t report on reviews. Reviews are currently the easiest-to-manipulate aspect of the AppStore. A person doesn’t have to use the application to review it, and the number of reviews are low enough that an unscrupulous developers can ’stuff the ballot box’ – similar to asking all of your friends to vote for your items on Digg.

For applications that collect e-mail addresses, I/m sure we’ll see the sort of aggressive review manipulation that comes to all review-based systems in time. Here’s a few of the tricks we commonly had to deal with back when I worked for a comparison-shopping engine:

a) asking users for reviews in corporate promotional e-mails,
b) conducting surveys on user satisfaction, and then asking only the happiest users to provide reviews,
c) tying reviews to incentives – only providing the incentives when the review had been written.

In short, online reviews are an extremely weak thing to base your rankings on. Which is why I’m happy to report that Apple doesn’t seem to do this. We took AppStore applications’ review scores (out of 5) and ranks (where they appeared on the ranked ‘top 100′ lists) and checked to see if the variables were correlated – that is, if higher ranks meant higher reviews, or vice versa. The result: there’s no significant relationship between reviews and AppStore ranking. (For the statisticians out there, the correlation coefficient was around -0.10.)

For AppStore rankings, reviews simply don’t matter. Considering the number of ways they can be manipulated, this is probably a good thing.

Responding to developer complaints, Apple has fixed the immediate alphabetization problem with its AppStore. For the most part, numbers, spaces, and punctuation at the start of names are simply ignored – the application name ‘2 Across’, for instance, appears in the alphabetical order list as if it were simply ‘Across’. There appears to be some exceptions – I suspect a comma still is considered, since ‘20,000 Leagues under the Sea’ currently occupies first place on the alphabetical order list.

While this doesn’t prevent unscrupulous developers from naming their applications something like ‘aaaaaacrobats’, the quick fix shows both Apple’s commitment to a level playing field and their desire for an aesthetically-appealing customer experience in the AppStore. A good sign for the future.

Right now the AppStore is brand new, and everyone’s trying to figure out how to use it correctly. We’re all experimenting, and I’m sure Apple is too – trying to figure out the policies that’ll result in maximum growth of the AppStore, maximum growth of the developer community, and maximum respect for the user experience. I suspect here the iPhone application ecosystem and the Facebook application ecosystem share some similarities – minimal policies at first, tightened over time to prevent the inevitably-emerging abuses.

This morning I noticed at least two behaviors that are clearly ‘commons dilemmas’ – when one individual takes advantage of them, they make rational, economic sense for that individual, but when everybody does it, the end result is a classic tragedy of the commons. In other words, short-term interests conflict with both long-term interests and the common good.

First, it’s possible to name (or rename) an application whatever you want, and the ‘all’ and ‘free’ lists are sorted alphabetically. It’s always better to be at the start of a list than the end of a list, which is probably why one company recently prefixed ‘#1 Rated’ to the start of their games. Clever, but this sort of behavior, once adopted by everybody, inevitably results in names like ‘####!!!!111AAA A A A Solitaire’. Remember paper copies of the Yellow Pages? Exactly the same behavior. In addition, claims in the title that may or may not be true make me hesitate. Would a brand new game called ‘Most Popular Solitaire’ or ‘Most Downloaded Sudoku’ be acceptable?

Second, updating your application refreshes the ‘released’ time, which is interesting, because the ‘all’ and ‘free’ lists have a ’sort by release date’ option. This morning both the Mobile News Network and New York Times applications were near the top of these lists. Perfectly fine if the application’s actually being updated – which, of course, the Mobile News Network and New York Times applications were. But if the sort-bump and placement on the ‘new’ list results in more downloads, tweaking your application for the sake of updating becomes tempting for the unscrupulous. There’s no direct financial incentive to preserve the bandwidth and respect the not-always-unlimited data plans of the current installed userbase, because they’ve already paid. I know I’d be more comfortable if ‘released’ and ‘updated’ were two separate dates, with separate filters (’recently released’ vs. ‘recently updated’).

Does anyone else see any ‘commons dilemmas’ or other questionable practices in the AppStore? I’m sure Apple’s going to have to deal with all of these issues in time – along with review-stuffing and companies purchasing their own applications to get on the ‘hot’ list (hey, you get 70% of your money back) and a host of other things not yet anticipated.

For those of you looking for numbers – there’s now 713 applications in the AppStore this morning (around 8:30 AM PDT), 61 more than the last time I checked. If my count is right, 553 (78%) are paid, 160 (22%) are free. The numbers are skewed somewhat by AppEngine’s $0.99 books, of which there are currently 114, 16% of the total applications in the store. Why do I suspect Apple will be adding a books category shortly?

Since I was asked a couple of times today what types of applications are generally free and what types of applications are generally charged for, I decided to put together some quick stats, based on data collected from the AppStore this morning:

The results are somewhat obvious – news and social networking are disproportionately free, since a) it’s difficult to charge for content that’s freely available elsewhere and b) social networks grow in value with the number of participants. Entertainment and games are disproportionately paid, reflecting a belief (correctly, I think) that people will pay money to have fun. Since the AppStore’s applications are disproportionately entertainment and games (helped along by a lot of $0.99 e-books), the AppStore’s applications are predominantly paid. The most common price for an application in the ‘games’ category is still $9.99, although the second-most common price is $1.99.

It’s hard to draw generalizations about the other categories, in the middle – the number of applications per category is still pretty small – but I suspect that applications that rely on user-generated content are disproportionately free, and those that rely on developer-and-designer-provided content disproportionately cost money.

Apple’s been busily adding applications to the AppStore overnight – we’re up to 652 of them as of 9:00 AM EDT this morning, exactly 100 more than yesterday. (Interesting coincidence – I wonder if Apple’s going to release 100 a day for a while?) Of these applications, 502 (77%) are paid and 150 (33%) are free. The most common price point is still free, although $0.99 (with 124 apps) is now pulling away from $9.99 (with 98 apps). The next most popular pricing point remains $4.99, with 71 apps. Here’s Pinch Media’s iPhone Application Pricing Distribution chart, updated with fresh data:

A few more observations on pricing:

• the $69.99 ForeFlight preflight planner is no longer the most expensive application in the AppStore – it’s been replaced by a time-tracking application called LionClock Plus, which goes for $79.99. (Sorry, Tyson, I know you were enjoying the attention.)
• from Pinch Media headquarters, anyway, it looks like the on-phone download figures have been removed – I’m seeing ‘0 downloads’ for every app. This is unfortunate, because developers were using them to estimate sales, and it would’ve been useful data for price point optimization – but with media outlets starting to backward engineer revenues from it, I’m not so surprised it’s gone.
• a select group of developers still know almost precisely how many downloads and sales they’re getting – the developers who installed Pinch Media’s Pinch Analytics code before submitting their applications to Apple. It’s running right now on several AppStore applications, reporting back first-time users, sales estimates, session data – everything these developers need to optimize and improve their applications. (And as of yesterday, it can do even more.) Many developers we talked to prior to AppStore launch were worried about Apple’s approval process – and believe me, we understand – but we’re happy to report our code checked out fine.

Pinch Media will be releasing more data later in the day – subscribe to our blog so you don’t miss out. (Anything in particular you’d like to see?) And enjoy your 3G iPhones! I’ll be picking my own up just as soon as I can.