Fluke? If not, congrats to Geek News!

[Ian]

Only hours into April, and over 80 votes fot Geek News! That's amazing. Competition is already hot!

[robaustin]

I think this highlights a new flaw in the "monthly" voting system.

Since votes expire due to calendar month - once the "switch" is flipped - you see a whole new slew of podcastts that get the early vote. This will eventually be overturned as the older podcatss with the larger listener bases once again get their votes back up to their previous status.

I mentioned to Chris a more dynamic way of voting - where votes have a "freshness" of a certain number of days from when they were placed. This would create a much more dynamic and accurate count of what podcasts were gaining and losing momentum.

Think about it - if you have a "freshness" of 10 days on a vote - that is enough time for a weekly show to gain/lose based on its gaining/losing listeners. It's dynamic because 10 days - would be from whenever that vote is placed - so votes are constantly being placed and expiring - thus creating an ebb and flow effect, rather then a "let's see how many votes we can get THIS month" kind of effect.

Not trying to take away from Geek News' spotlight - just pointing out that it is too early to see if this is a lasting thing - or just a beginning of the month "bubble".

--*Rob

[allthewhile]

It's too early to tell who's gonna be at the top. give it a week or so.

[PaulCasting]

I really like the idea of the 10 day cycle. This would be a good way to have a running or changing number. Each vote is good for only 10 days. That means people need to have fresh votes coming in to stay on top.

Wouldn't be a bad idea to have the total for the previous month listed as well.

Good thinking

Paul
www.EastpointRadio.com

[Version3]

I like the votes so far showing up. I had thought for a moment about weighting old votes more than new votes (the 2 or 3 levels), but that would create too much opportunity for old podcasts to stay near the top forever.

I don't think the monthly reset is such a bad idea for the following reason: we are all trying to get new listeners, new and old. The only reason for the voting thing is exposure, so this way, everyone can work really hard once or twice at taking advantage of that exposure, without the person with the most listeners always being at the top, or the oldest podcast always being at the top. So for the first week every month, even the newest podcast CAN end up in the top 10 for a bit. Maybe they will snag a few new people before the Dawn and Drew show's huge fanbase gets their votes back on and pushes them back down. Ultimately and the end of the month, the listener base will still win... not the one time push voters.


That said, I'm happy for the opportunity to show up near the big shows. I think that it will help me with new people actually checking out my show, and in a few months, maybe our listener base will perpetuate itself as well.

[robaustin]

Yeah - but my point is - you'll show up for a couple of hours or maybe a day or two - then the big shows will push you back down.

I guess what I'm suggesting is a much more dynamic way that votes get used. This way - you need a constant stream of votes througout the whole month - AND year even...to keep a show in the top 10 or top 50.

This way - votes are reacting to podcasts actual content too. Suppose a podcast dies off during a given month - well then if you have freshness on votes - its status during the month will beging to die out as well. I mean it's better on a monthly basis rather then the "all time" votes were - but I still believe that a frehness voting system needing a constant stream of votes will be better for everyone.

Maybe I should just go and start my own podcast directory like this. I just don't have the programming skills to do it.

--*Rob

[yaz]

c'mon, we know who's gonna be on the top, because everyone wants to kiss their asses...

[theFerf]

I think this highlights a new flaw in the "monthly" voting system.

You caught the site as it was flipping over to a new month. The top 10 is generated by a voting script that runs every 15 minutes or so to recalculate it. This is to speed up the server so when users access the xml / opml top 10 and 50, its not determining the tops every time.

There will be a short time lag at the beginning of every month like that where the roll over is off, but if you look now, it is accurate.

[Version3]

Yeah - but my point is - you'll show up for a couple of hours or maybe a day or two - then the big shows will push you back down.
This way - votes are reacting to podcasts actual content too. Suppose a podcast dies off during a given month - well then if you have freshness on votes - its status during the month will beging to die out as well. I mean it's better on a monthly basis rather then the "all time" votes were - but I still believe that a frehness voting system needing a constant stream of votes will be better for everyone.


I see your point, however dedicated listeners to big shows would still dominate the top spots, just because more listeners always results in more resources. If they stay on top all of the time, the newbie checking out the top ten will increase the big shows chances of staying there. It wouldn't be hard for D&D to stay at the top with this system, it only takes one person to post over on their board to design a vote schedule to put their 480/month votes to better use than my 50.

I'd rather show up near the top for a few hours or a day or two than not at all. That's all I'm saying.

[jeffoest]

The voting system is very interesting. The analogy that comes to my mind is competition for local bands in a particular market. The local bands that get more bookings tend to have a lot of 'friends/fans' that can guarantee a good take for the bar/club. That means more bookings for that band and corresponding more visibility and 'marekt share'.

Does that always mean that the most 'talented' local bands get the gigs? No. It's an imperfect system. Just like this one.

BUT I do believe that over time, the bands that have talent DO end up with audiences (as long as they somehow get hooked up to a fan base - and a connector/maven or two - if you've read "The Long Tail") and over time a band that really isn't all that good loses their buzz even though they may offer their friends a good night out. But again, it's not perfect.

So goes our current voting system on the Alley. It's not perfect - podcasts with more 'friends/fans' will get more visibility and that visibility will lead to even more visilibity. But over time, I believe, popular well-liked podcasts will be the ones most featured in the top 50.

However, I would offer a couple of tweaks that I think would improve things:

1) extend the count from monthly to something larger. if you take a top podcast with say, 200 votes. one could assume that 75% (just guessing) of that number represents a very loyal fan base. They will probably vote every voting period. A longer voting period distributes that loyal fan base over a longer time and consequently lowers the impact a bit as compared to new voters/new listeners. I think that will help keep the list fresh as new podcasting audiences come to our market. Yea - possibly a show like DND will be number 1 in that system for six months, but I'm ok with that - they have earned it by establishing a personal relationship with their listeners which is what all of us should be doing. And we may be surprised - in a short number of months, new listeners will probably dominate existing listeners and voters - I think the list will be more dynamic that under the current system.

2) in addition to the top 50 list I would like to see Podcast Alley (Ferf!) feature more podcasts more often on the front page. It is in the best interest of the community to steer traffic away from the top 10 into other content every now and then to help 'lift the whole ship up'. Soemthing akin to what you do with the Featured Podcasts now but in my opinion, it needs to be more frequent.

[Version3]

I was thinking the same thing on the features... something like Weekly. Just depends on how much more work it is to put together.

[theFerf]

consider it done. i will work on doing a feature more often, maybe 2 a week or something like that for now. obviously it would be cool to do more, but i have to get the back end all worked out first.

[jeffoest]

Ferf, I'm sure that if there are any bandwidth issues (i mean in terms of work for you vs. available time - not literally bandwidth!), members of this forum would probably be happy to help in any way.

[PupuStudios]

I would have to disagree. I think the PK and J Show should be featured indefinitely.

[jeffoest]

Well...... except for them ;-)