I do agree that the old "X day weighted unit sales average" is still a big component of this calculation, otherwise the entire store would have been turned on its head today instead of the very specific behavior we have seen. They could be simply adding an extra weighted modifier to the calculation: average daily sales for the *entire life* of the application. It could be that the high review/rating numbers are just an indicator of how well that particular app sold. Thus we see old apps surfacing now with amazingly large numbers of reviews. If this sticks around for a few more days I think we'll get a lot more data points from people, and may be able to nail down what is really going on.
Bounce On is currently around 25 in adventure, it only moved 1 rank since yesterday and has been sitting between 20-30 for the past few months(pretty consistent sales). I've noticed a few apps around it has either risen or fallen. Rolando(older game) got a big jump up while Spider(newer game) went a few down. Sunday sales for Bounce On was 503 units. Bounce On Lite moved up 10 ranks(in US) and 28 in Germany(wow!). Sunday sales for Bounce On Lite was 5376 units. I would be very interested in knowing unit sales of other apps. It makes sense if the ranks now take a longer slice of time to calculate the time. What is interesting is that Bounce On Lite probably had 80% of its downloads in February last year when it was #2, so if it drops quickly after this month that would imply its looking at the past 365 days(assuming nothing else affects its rank). I've always wondered what exactly the "popularity" sorting was in any sub category for games? If you go to games->adventure and change the sorting from release date to popularity it is different then the "top app" list to the left. My guess is that popularity is entirely based upon units downloaded(maybe including upgrades even?)... or I wonder if Apple monitors usage of apps.
Anyone knows what happened to the app Unblock me? It was in the top 100 games yesterday and today it seemed that it disappeared from the appstore.
prediction I predict: 50% of appstore developers will be delighted with this change 50% of appstore developers will be horrified with this change
I know you're joking, but still I don't think it's quite that clear-cut. *If* this new ranking system remains and *if* I am correct in my guesses on it's calculation it opens things up to rank manipulation through review/rating counts. That will piss off 99% of developers once the unscrupulous 1% of them start abusing it
I'll tell you what though ... this whole thing gives me stomach acid. It's not that Trenches dropped a couple spots based on a new algorithm ... we didn't move THAT much. It's the fact that devs must have the ability to anticipate chart movements and build marketing campaigns around them. If Apple wants the AppStore to really mature, they need to create a balance between new apps hitting and long-time successful apps remaining. Frankly, I preferred the old, old way. Every new update got on the "New Release list". It benefited those devs that were committed to long-term support. In the end, Apple should figure out a system for ranking that doesn't include chicken feet and eye of newt. For some of us (meaning 99%), this is a business and we need to be able to learn, know and anticipate the "market". I'm going to go take some Pepto.
LOL! And yes, I agree that these changes are unnerving for developers trying to run a serious business...
Sorry to jump in late here, but our past sales data has indicated to us that it's more like a 3 day moving average (or at least, was). We've tested that by advertising heavily for just a single day and then watching the effect of the ranking change. It takes about a day and a half for the jump in the charts to max out and then another day and a half for the app to fall back to its normal position. This is an excellent thread!! ___________ Co-Founder and Lead Developer, Fairlady Media Inc. Developers of: Stinky Socks, Spazzle, Spazzle II , Whizzit games for Preschoolers and Puzzle Sleuth Twitter: jrtb
Out of curiosity, does anybody remember when Apple changed the App Store so that free downloads didn't add towards your paid ranking? Did Apple give a public announcement stating that? (Not a rhetorical question.) If they did, this is probably just a glitch because a change this massive would probably have been announced on the developer portal. If not, this could very well be a new algorithm for rankings.
They didn't announce that. I believe it was within the first month or two of the App Store. They really don't announce any of these kinds of changes. No announcement for: rate on delete, star rating reset on updates, updates don't appear on release list anymore. We just get to discover it after it happens, and hope we don't go out of business.
Well, our new rank position for Spazzle Free in the Family Games category seems to have held from yesterday (we just dropped one place). I'm curious to know how many dl's we got, but the iTunes report is delayed... For those of you who had big rank changes yesterday: did they hold up today?
TanZen jumped 40 spots yesterday, up to #25 in Puzzles. It fell several spots, and then went back to #25 again overnight. I don't think this is a glitch anymore.
Yeah, it's been over 24 hours now, and you can go refer to any of the apps and games I have brought up in this thread or via twitter and see that everything has remained in place. I think this ranking algorithm is here to stay for a while. Whatever the calculation is now, it contains some of the original calculation for moving average of the past few days sales, and a new component that appears to be looking at the total amount of ratings or reviews that app has received. It's possible this is actually using lifetime sales/downloads as the modifier, as the effect would be similar... apps with tons of reviews and ratings were likely downloaded MANY more times than others. Of course, Rogue Touch with it's 900 ratings cannot compete with iMob Point Pack apps that have 65,000 ratings.
Still stuck on the rating thing? I honestly think it's a glitch, if it's not -- it's most likely just a larger weighted average window.
What do you think the effect of going 'free for a day' would be in this possible new system? Loads of downloads, loads of (bad?) reviews... but does it continue to boost your rank when you're no longer free?
it cant be. Save Kitty was selling around 10-20 copies a day for.... 8 months, maybe more (except for christmas of course). this wouldnt get it to top 100 arcade, would it ? BUT it has 30.000 reviews. something strange is happening, and seems like games that were free at some point of their lifetime suddenly received a huge boost in rankings
Thank you! You can lead the developer of "water your body" to water, but you can't make him drink, huh?
