SDA site and individual speedrun traffic stats

Edit history:

Thinkshooter: 2009-03-02 08:52:36 am

2009-03-02 08:45:48 am

Runs: FCP, FC3, XWA, HD1-2, KotOR1-2

Hello. I somehow always thought SDA was as popular with everyone else on the Internets as with me, and consequently that the ad traffic was able to support a couple of (at least) part-time site admins. I just learned that this site is run as a spare time hobby instead. Double respect to the crew!!!

Are there any website traffic stats available? How about the individual speedrun downloads? Are the videos downloaded in the hundreds? Dozens? Thousands?

Good thing there are forum statistics. Do the speedrun downloads correlate to the forum topic views, e.g. is a game with 40000 topic views 400 times more likely to be downloaded than another with 100 views? Are the downloads from archive.org also counted?

Thanks.

Thread title:

nate

2009-03-02 09:48:14 am

several thousand people visit the site every day.

we don't have reliable statistics on anything involving run downloads, though. reasons include download accelerators and archive keeping even less accurate counts. rule of thumb is the more popular the game, the more popular the run. that makes no sense to us here in the forum, but the site gets so much traffic that the majority of people downloading choose based on whether they've played the game, not on how cool the run is. the recently released sorceress run may be the biggest run this year.

having said that, every run gets downloads.

the site costs over $200 a month to run in server bills alone. we make less than $100 a month off the ads. this is decreasing every month. donations from people like you and me are keeping sda running now. not to worry, though; i'm perfectly capable of funding sda entirely myself if need be. it just bothers me a little, because i would like to think that sda doesn't need me in any way.

DJGrenola

2009-03-02 09:49:18 am

guffaw

I can vouch for the fact that the site is incredibly hard work to run, so I'm sure the administration (in whose number I no longer count myself) are grateful for your remarks.

In general, site policy has been not to release information about run download popularity, as it promotes a rich-get-richer state of affairs.

Thinkshooter

2009-03-03 08:20:30 am

Runs: FCP, FC3, XWA, HD1-2, KotOR1-2

Thanks guys!

Quote from DJGrenola:

Site policy has been not to release information about run download popularity, as it promotes a rich-get-richer state of affairs.

Totally understand. That would be too high a price to pay.
Though part of me still kinda wishes that submitters would be able to find out roughly how many people watched their runs.

nate

2009-03-03 09:37:10 am

i agree. unfortunately, the two seem to be mutually exclusive to some degree.

tycho102

2009-03-03 06:43:20 pm

Some people think they can outsmart me.

Quote from nate:

the site costs over $200 a month to run in server bills alone. we make less than $100 a month off the ads.

I'm willing to mirror the site media files.

You can keep the forum database all to yourself, however. You can keep it, and you can nuzzle it tightly to your breast as if it were a genetically-challenged child.

nate

2009-03-03 07:33:36 pm

should note that we don't actually pay for bandwidth (sda would be totally unsustainable then). however your offer of a mirror is welcome. we just need to figure out a way for you to update which files you're hosting so the link can show up in demo.pl. i'm sure we can figure something out.

of course an easy way is to just host torrents. Smiley

Enhasa

2009-03-03 09:32:14 pm

everybody wanna tell you the meaning of music

I think if we tried to make money off SDA, we could totally do that. It's just that we don't have much of any priority for that.

Quote from nate:

that makes no sense to us here in the forum, but the site gets so much traffic that the majority of people downloading choose based on whether they've played the game, not on how cool the run is.

Huh? I know there are lots of regulars who will watch anything, but it seems completely obvious that most people (including me) aren't that way. An ok run of a game I've played is much more interesting to me than a supercool run of something I've never played that I won't even understand or appreciate. Even for those that doesn't apply to, isn't it natural people will go for games they recognize? It's not about how good a run is in a vacuum. When it comes to personal enjoyment, particularly for casual watchers, a 10-minute SMB1 would have a much better hit rate than a 10-minute Demon's Crest.

As an aside, I really really detest the opposing view. It's the #1 thing I can't stand about tasvideos philosophy, which is to watch a run/tas of a game they haven't played and then because they don't enjoy the run, conclude that the game itself is shit. I promise, this is not even considered strange on that site.

Anyway, most people have no idea what an optimized run looks like, and others do but don't care. Think about the story run requests we get. We would be much more popular if we had longplays or fullruns than speedruns. This is why I'm so glad you are in charge of a/v and the bar is set where you personally want it and not mine. I honestly think video quality is our main selling point. I don't mind though, they keep the site popular enough for the technical gameplay people.

I think if we tried to make money off SDA, we could totally do that. It's just that we don't have much of any priority for that.

[quote="nate"]
that makes no sense to us here in the forum, but the site gets so much traffic that the majority of people downloading choose based on whether they've played the game, not on how cool the run is.[/quote]
Huh? I know there are lots of regulars who will watch anything, but it seems completely obvious that most people (including me) aren't that way. An ok run of a game I've played is much more interesting to me than a supercool run of something I've never played that I won't even understand or appreciate. Even for those that doesn't apply to, isn't it natural people will go for games they recognize? It's not about how good a run is in a vacuum. When it comes to personal enjoyment, particularly for casual watchers, a 10-minute SMB1 would have a much better hit rate than a 10-minute Demon's Crest.

As an aside, I really really detest the opposing view. It's the #1 thing I can't stand about tasvideos philosophy, which is to watch a run/tas of a game they haven't played and then because they don't enjoy the run, conclude that the game itself is shit. I promise, this is not even considered strange on that site.

Anyway, most people have no idea what an optimized run looks like, and others do but don't care. Think about the story run requests we get. We would be much more popular if we had longplays or fullruns than speedruns. This is why I'm so glad you are in charge of a/v and the bar is set where you personally want it and not mine. I honestly think video quality is our main selling point. I don't mind though, they keep the site popular enough for the technical gameplay people.

nate

2009-03-03 10:34:25 pm

what you say is probably true, of course. but i wonder how many people are as lucid about these things as you are.

Quote from Enhasa:

Think about the story run requests we get. We would be much more popular if we had longplays or fullruns than speedruns. This is why I'm so glad you are in charge of a/v and the bar is set where you personally want it and not mine. I honestly think video quality is our main selling point. I don't mind though, they keep the site popular enough for the technical gameplay people.

i think at the end of the day i just get a kick out of distorting reality for people. helping them find out what the limits really are. people think there can be no free high quality video on the web? bam. they think no one does hard work for free? bam. they think they can't direct download popular videos? bam. but surely they can't watch it in the browser too? bam. it's the same for me as it always was. breaking everything just because i can - just to see if i can.

they call it "realization of potential," and it has two meanings: once you've seen what's possible, you can't go back - you've gained the potential to follow through.

what you say is probably true, of course. but i wonder how many people are as lucid about these things as you are.

[quote="Enhasa"]
Think about the story run requests we get. We would be much more popular if we had longplays or fullruns than speedruns. This is why I'm so glad you are in charge of a/v and the bar is set where you personally want it and not mine. I honestly think video quality is our main selling point. I don't mind though, they keep the site popular enough for the technical gameplay people.
[/quote]

i think at the end of the day i just get a kick out of distorting reality for people. helping them find out what the limits really are. people think there can be no free high quality video on the web? bam. they think no one does hard work for free? bam. they think they can't direct download popular videos? bam. but surely they can't watch it in the browser too? bam. it's the same for me as it always was. breaking everything just because i can - just to see if i can.

they call it "realization of potential," and it has two meanings: once you've seen what's possible, you can't go back - you've gained the potential to follow through.

Siyko

2009-03-03 10:48:13 pm

I used to be athiest until I realized I was God.

you guys are nuts

xXIkuto

2009-03-03 11:48:57 pm

being nuts is only temporary. when nate & others accomplish what they want (world domination) it will become normal to break limits and constantly improve.

Pootrain

2009-03-04 04:10:55 pm

train kept rollin

Do you not even get the amount of plays the embedded flash movie gets? I know this wouldnt be an accurate representation either but, you know.... Just curious.

nate

2009-03-04 06:25:32 pm

don't think i can distinguish those from people downloading the mq.

Emptyeye

2009-03-05 08:53:10 pm

Talk to the Hand

Quote from Thinkshooter:

Totally understand. That would be too high a price to pay.
Though part of me still kinda wishes that submitters would be able to find out roughly how many people watched their runs.

I've honestly wondered this myself. I'd love a way that the author of a run could, in isolation (In other words, ONLY they can see the stats, and ONLY for their own runs, no one else's. This is to try and avoid it devolving into a popularity contest), see how many downloads their runs get. That said, I totally understand if it isn't a priority/possible.

That said, anecdotally, quite a few people watch every run--my piddly site has gotten a traffic spike when each of my last two runs have come out (Those being the all-time classic Astyanax, and the even more fondly remembered Chuck Rock ;)).

nate

2009-03-05 09:03:00 pm

ironically, the best data we have is comparisons of one run to another. a hard count for a given run doesn't exist, unfortunately. blame http.

Thinkshooter

2009-03-06 01:55:51 am

Runs: FCP, FC3, XWA, HD1-2, KotOR1-2

Quote from nate:

ironically, the best data we have is comparisons of one run to another. a hard count for a given run doesn't exist, unfortunately. blame http.

I know what to do. Create a mock speedrun with just 1 (one) download. Not that hard, right?
According to the above, a comparison to this mock speedrun would give us the exact count each other run was downloaded.

nate

2009-03-06 09:08:27 am

close, but no cigar. would depend on how you downloaded it how you made the other numbers look. and there are many different ways of downloading a run.

Thinkshooter

2009-03-07 01:57:50 pm

Runs: FCP, FC3, XWA, HD1-2, KotOR1-2

OK I was partly being funny but I hoped someone would see the possibility in my suggestion.

I think that a carefully selected speedrun with a *known* number of downloads would give a pretty exact base for comparison by which other speedrun downloads could be measured. It's probably not impossible to gather the necessary number of trustable people who would correctly report the number of successful downloads. That would give us an "actual viewings" to "number reported by webserver" ratio. Couldn't a ratio such as this be used as a good approximation in the case of other runs?

I understand it's an *estimate* since there are variations in how speedruns are watched but it would certainly be close enough for my taste.

nate

2009-03-07 05:20:04 pm

getting a large number of people to report numbers, so far so good. but how do you tell those people to download? how can you guarantee that how they are downloading is representative of how everyone is downloading? the way corporations do this is they randomly survey visitors to their site. but it's safe to say no one on the administration cares enough to put that much work into implementing that.

there are other problems, too - like variations in how people download things based on what they're downloading. so for example a 6 gig file might be downloaded differently than a 1 meg file because you are going to resume the 6 gig file more, maybe use a download manager more, etc. and that would have to be known.

JaggerG

2009-03-25 02:36:25 pm

Jumping Turtle

Quote from Siyko:

you guys are nuts

CONFOUND IT ALL I love it though....

Seriously, many of my friends have the attitude of "That's pretty cool, but you don't really gain anything from it, so there's no point." They just have trouble comprehending that not getting money doesn't necessarily mean there's no point. It's like searching for the limits is just something you do as a child and you need to mature out of that "phase" once you reach a certain arbitrary age. People still collect stamps, they keep mint-condition action figures, the keep track of professional sports teams, they read books, and still for some reason believe that playing video games long after the point of simple frustration is for some reason beneath them. Yes, it is a hobby, no, I'm not gonna quit, and certainly not anytime soon.

Also one of my friends thinks you guys are getting rich just because you have an established site. :^P