Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

How spammers are gaming the podcast charts

chartable.com

Post reply on HN
Print view View original

1–10 of 32 posts

Re: How spammers are gaming the podcast charts

#5

Post by CPLX »

I think this article should be retitled to remove the word "How" at the beginning, since it doesn't even really speculate how one would go about doing this. It's interesting nonetheless.

Any ideas? Assume that someone decided to harness one of the many click-farms to do automated podcast subscriptions?

I suppose it's also possible that Apple did an ill-advised algorithm change or something that changed the weights very heavily, and the spammers who had been quietly doing their thing for months suddenly all got promoted.

Re: How spammers are gaming the podcast charts

#6

Post by smock »

arprocter wrote: post #3

The algorithm seems to be weighted to boost new subscriptions - looking at the Chartable top US-all there is one at #20 which recently launched (has 6 episodes) which is above This American Life at #28

While Apple doesn't comment on how the rankings are calculated, the common knowledge is indeed that its heavily biased towards recent subscriptions.

Re: How spammers are gaming the podcast charts

#7

Post by smock »

CPLX wrote: post #5

I think this article should be retitled to remove the word "How" at the beginning, since it doesn't even really speculate how one would go about doing this. It's interesting nonetheless. Any ideas? Assume that someone decided to harness one of the many click-farms to do automated podcast subscriptions? I suppose it's also possible that Apple did an ill-advised algorithm change or something that changed the weights ve…

We can't say for sure how the charts are being gamed, but there is strong evidence that its related to a pay per play scheme to pump subscription numbers in a small time window. The implication of the graph visualization is that the same iTunes users are subscribing to all of the suspect shows, and all on one storefront (the U.S.).

Re: How spammers are gaming the podcast charts

#8

Post by ballenf »

A podcaster (and Youtuber) who I really enjoy is CGP Grey. He has been proposing allowing voluntary removal of view numbers from Youtube and similar platforms.

I think Apple should fix or remove their rankings but, please, don't add any more metrics.

The podcast world is currently mostly immune from so much of the clickbaity trash on Youtube (that arguably isn't even YT's fault).

I admit it's very counterintuitive or almost subversive these days to suggest that internet points be kept secret.

Edit: Here's a CGP Grey video where he talks indirectly about how he sees the internet fracturing his (our?) ability to pay attention and focus. https://youtu.be/wf2VxeIm1no

Re: How spammers are gaming the podcast charts

#9

Post by JohnJamesRambo »

smock wrote: post #7
CPLX wrote: post #5

I think this article should be retitled to remove the word "How" at the beginning, since it doesn't even really speculate how one would go about doing this. It's interesting nonetheless. Any ideas? Assume that someone decided to harness one of the many click-farms to do automated podcast subscriptions? I suppose it's also possible that Apple did an ill-advised algorithm change or something that changed the weights ve…

We can't say for sure how the charts are being gamed, but there is strong evidence that its related to a pay per play scheme to pump subscription numbers in a small time window. The implication of the graph visualization is that the same iTunes users are subscribing to all of the suspect shows, and all on one storefront (the U.S.).

I'm reminded of the images from China where one person is in front of rows and rows of phones clicking on stuff and that is their job. Would that do the job here?

Re: How spammers are gaming the podcast charts

#10

Post by smock »

JohnJamesRambo wrote: post #9
smock wrote: post #7

Earlier quoted context omitted.

We can't say for sure how the charts are being gamed, but there is strong evidence that its related to a pay per play scheme to pump subscription numbers in a small time window. The implication of the graph visualization is that the same iTunes users are subscribing to all of the suspect shows, and all on one storefront (the U.S.).

I'm reminded of the images from China where one person is in front of rows and rows of phones clicking on stuff and that is their job. Would that do the job here?

Apple doesn't publicly comment on how the chart rankings are calculated, which is why we can't draw strong conclusions, but our understanding is that this would do the job. Since the somewhat catastrophic failure of the charts a week ago its clear that Apple did update their algorithms, but its not clear how (and some of the same players identified in the blog are still ranking highly).
Post reply on HN