Humble Bundle 2 - Non technical discussion

[Frak]

Some reason may include: We have less games, so we spend more for whats available, the Linux crowd is on average higher disposable income, we really want more games so we want to support the people who make Linux games.

Not the case at all. What we have here is a case of mathematical fallacy. The average shows that Linux users spend more, which is not really the case at all, and I'll give an example to show why:

Let's say one Windows user gives $1,000, another five Windows users give $50, and thirty Windows users give $1.

((1000 * 1) + (50 * 5) + (30 * 1)) / 36 = An average of $35.55 per person. A total of $1,280 all together.

Let's say one Linux user gives $500, another five Linux users give $50, and ten users give $1.

((500 * 1) + (50 * 5) + (1 * 10)) / 16 = An average of $47.50. A total of $760 all together.

It looks like the Linux users are giving more, when the fact is, the Windows users have given roughly 60% more than the Linux users.

Because of the low population of Linux users, one even moderately generous donator can heavily inflate the average donation but still keep the overall total low. We need to keep a look at the overall donation and not the averages, as the average leads the conclusion that Linux users give a lot, while the truth is the population is still too low for developers to care. It's the simple case of userbase beating independent investment. Averages are meaningless.

[EpicEraser]

Not the case at all. What we have here is a case of mathematical fallacy. The average shows that Linux users spend more, which is not really the case at all, and I'll give an example to show why:

Let's say one Windows user gives $1,000, another five Windows users give $50, and thirty Windows users give $1.

((1000 * 1) + (50 * 5) + (30 * 1)) / 36 = An average of $35.55 per person. A total of $1,280 all together.

Let's say one Linux user gives $500, another five Linux users give $50, and ten users give $1.

((500 * 1) + (50 * 5) + (1 * 10)) / 16 = An average of $47.50. A total of $760 all together.

It looks like the Linux users are giving more, when the fact is, the Windows users have given roughly 60% more than the Linux users.

Because of the low population of Linux users, one even moderately generous donator can heavily inflate the average donation but still keep the overall total low. We need to keep a look at the overall donation and not the averages, as the average leads the conclusion that Linux users give a lot, while the truth is the population is still too low for developers to care. It's the simple case of userbase beating independent investment. Averages are meaningless.

I don't understand what you are trying to say...
Averages are not meaningless. They simply do not give information about the actual distribution, but why would this imply that the reasons given do not explain these averages? If an average is higher, this means that on average, Linux users spend more. There may be more people spending more, or just some people spending a lot more, but using only this data, I would certainly try to find explanations for this number.

Furthermore, Linux users do not only spend a lot on average. They also contribute a large amount to the whole revenue, as can be read from the graph on the website. Linux users contribute about 23%, whereas OSX users contribute less (in absolute values), making the contribution from Windows users a bit over half the total amount. Because of the numbers of users of the various operating systems, it is clear that Linux users provide a great contribution. Again, this may be caused by these outliers, but because of the lack of further data, that explanation, to me, is as good as the ones given by Frak.

As a last point, the outliers you described do not actually play such an important role.
The total amount is more than 774 thousand dollars, and the largest ten contributions are only ten thousand dollars in total.

[taotree]

This was definitely a fantastic purchase. I had a few problems with Revenge of the Titans, but most of the issues are already in their bug tracking or solved through other means.

Just a side note, Puppy has released an update of Revenge of the Titans already. To upgrade, download the latest demo directly from their site.

How do you get that update? You can't download from their site. It says right on the download page that humble bundlers can't register it if you download it there--you'll just end up with the demo.

But I tried downloading again from the humble bundle page and got the same 1.71 version.

[Spr0k3t]

How do you get that update? You can't download from their site. It says right on the download page that humble bundlers can't register it if you download it there--you'll just end up with the demo.

But I tried downloading again from the humble bundle page and got the same 1.71 version.

Install the downloaded 1.71 from the HumbleBundle then download the demo from the main site (www.puppygames.net). This is detailed in the Support section as follows:

Our games check for updates when starting and will notify you of any new versions we release. To get the new version, simply download and install the latest demo version from our front page.

Just noticed this: http://www.puppygames.net/blog/

[Frak]

I don't understand what you are trying to say...
Averages are not meaningless. They simply do not give information about the actual distribution, but why would this imply that the reasons given do not explain these averages? If an average is higher, this means that on average, Linux users spend more. There may be more people spending more, or just some people spending a lot more, but using only this data, I would certainly try to find explanations for this number.

Furthermore, Linux users do not only spend a lot on average. They also contribute a large amount to the whole revenue, as can be read from the graph on the website. Linux users contribute about 23%, whereas OSX users contribute less (in absolute values), making the contribution from Windows users a bit over half the total amount. Because of the numbers of users of the various operating systems, it is clear that Linux users provide a great contribution. Again, this may be caused by these outliers, but because of the lack of further data, that explanation, to me, is as good as the ones given by Frak.

As a last point, the outliers you described do not actually play such an important role.
The total amount is more than 774 thousand dollars, and the largest ten contributions are only ten thousand dollars in total.

What I'm trying to say is, there are a lot of users jumping on the bandwagon that Linux users are carrying the HIB on their back. This is not the case. In fact, I'd say that the average with the lowest amount are the ones giving the most, as the divisor is quite high. Now, as a corollary playing off some basic mathematical laws, a higher divisor will result in a lower average.

What I'm trying to put into peoples heads is that averages do not mean what people think they mean. People are trying to use this number as a justification for people to invest in Linux development while any competent financier will simply look away from said average and indulge in the pie as a whole. Linux developers have been known before to hide true numbers behind expressions such as the paying average and look away from the true profit margin, and this is that case once again.

[cchhrriiss121212]

the truth is the population is still too low for developers to care.

The rest of your post made sense, but what is this statement based on? By "care", I assume you refer to the ability to recoup a financial investment?
As demonstrated by yourself, the payment data is made hard to evaluate due to the "pay what you like" scheme, but it can still be useful with a little effort.

If we interpret the data (total revenue from each platform/average payment from each platform) we see the following user population breakdown:
Windows: 75000 (71%)
Mac: 18700 (18%)
Linux: 11500 (11%)

Are over 10000 users not substantial enough to make money from? (I don't mean this rhetorically, I have no idea what it costs to develop a game, or how many man hours it takes to port one to a new platform, if anyone has estimates I would like to hear them).

[Arlanthir]

Are over 10000 users not substantial enough to make money from? (I don't mean this rhetorically, I have no idea what it costs to develop a game, or how many man hours it takes to port one to a new platform, if anyone has estimates I would like to hear them).

Perhaps more interesting is the fact that the total revenue generated by Linux users is on par with the revenue from Mac users.

If developers often deem Mac as a now worthy platform for ports, I'd assume this is an indication that Linux can be as worthy too (at least in this case, obviously).

Also, I may be wrong, but if you port something for Mac you're likely using OpenGL already, thus making the port for Linux even easier. Please enlighten me if I'm wrong

(Steam comes to mind!)

[madjr]

dont make assumption guys, here is the real data:



http://somethingmild.blogspot.com/20...after-all.html


well at least for the first one, the second one is looking even better for us.

[amauk]

These bundles will do wonders for Linux gaming (and possibly Linux adoption as well)

Someone posted a link to a thread on the Unity3D forums a few days ago, where it's users (game developers) were complaining that they were being left out from the HIB because the game engine didn't support Linux

That 25%-odd slice is significant, and they're starting to realise that not supporting Linux is hurting their pockets

Take any game engine that doesn't support Linux, and I bet it's users are saying the same things

[Scooter_X]

If developers often deem Mac as a now worthy platform for ports, I'd assume this is an indication that Linux can be as worthy too .


(Steam comes to mind!)

Those have been my thoughts too as I watch the pie chart.