Proof that Open Source Kills Innovation

It's no secret that I am not a fan of open source. I have at various times stated that open source:

• Is no better than closed source in terms of quality
• Generally results in an inferior product
• Inefficient
• Fails 98% of the time (that's only after the first year, it fails even more after that)
• Not innovative
• Less secure
• Is killing the planet
• Is evil because it leads to increased complexity
• Harms innovation

It's this last statement that apparently flies in the face of the copyleft stalwarts because apparently the entire reason open source/copyleft software was invented was to address the mythical "harms" of copyright to innovation. Well, the empirical evidence is in, open source does in fact harm innovation. I don't have to rely on what amounts to mere opinions that claim the contrary, I have hard empirical, quantitative evidence and that evidence comes from the result of a game I invented to specifically test if copyright or copyleft was better for innovation. I called this game The Innovation Game and the concept for the game came out of a discussion here on LinkedIn. I won't belabor the origin here in this article, you can read about the history and the full rules but to recap:

1. At the start of the game, each player is given a die that they place in front of them on the number 1 and each player starts at zero points.
2. Players can perform one of two actions during a turn and must decide their action prior to any player announcing their action and without discussions. Players can take either of the following actions:

• Add feature: Add a feature to their software by incrementing their die by 1, adding point value to their die.
• Points roll: Earn points from their software by rolling the spare die an equal number of times to the value of their die and adding the total to their points.

At the end of the game, the player with the most points wins. You can play this game under two variations called copyright and copyleft. Under the copyright variation, players can only add features to their own die and only use their own die for points rolls. Under copyleft, players may add features to any player's die and use any player's die for points rolls.

I had the opportunity to moderate the first ever playing of this game at a game night session held by the company I work for. The only variation to the original rules was that in play testing we found that the number of turns was too many. So we halved the number of turns. Five professional software developers agreed to participate in the game and it is perhaps interesting to note that all of them are (or perhaps were) big proponents of open source software. They all came into the game blind so I explained the rules and the results, to me at least, were unsurprising.

At the end of the copyright version of the game, 30 total features were present meaning that each participant's die was on the number 6. Also, a total of 925 points were scored. At the end of the copyleft game, 17 total features were present, just over half as many in the copyright version. A total of 1,273 points were scored.

Now, again, the end results are unsurprising to me. Under the copyright version, players are incented to add features as quickly as possible in order to maximize their potential for points, something that two of the players figured out immediately. Under the copyleft version, once a single player's die gets increased, all players are incentivized to use that die for points roles and feature innovation effectively ceases.

The results clearly show that in the copyright version, the players that innovated the quickest made the most points while those that innovated the slowest earned the least points. The copyleft results are perhaps more interesting. Players that contributed no features earned the most points while those players that contributed the most features earned the least points.

Now, in the original concept for the game, the "points" represented dollars earned or saved by utilizing or selling the software. And here I would like to point out one of my many additional complaints about open source. You see when people freely contribute their time and energy for a non-profit, we call that volunteerism. But when people freely contribute their time and energy for a for-profit organization, we generally call that something else entirely. The degree to which certain open source companies and organizations profit from the work of others without fair compensation is obscene in my opinion. Oh well, I guess there's a software development sucker born every minute.

The most gratifying thing to come from this first, and likely last, playing of The Innovation Game is the discussion that came afterward. We all had an actual debate/commentary about the value of open source and innovation. It was pointed out that the points could represent innovative things done with the software. It's a fair point but one that I reject on the basis that the cost of software is miniscule in the grand scheme of things. Organizations that seek to innovate will spend exponentially more money on people, processes and systems versus software. But, it was nice to see an actual debate being held on the topic versus just blindly believing that open source software is a force for good the same way that the masses just blindly accept that there is this concept of a visual learner. Spoiler, the empirical evidence is that there is no such thing.

To date, I am not aware of a single, empirical bit of of quantitative evidence that indicates that open source helps innovation. I suspect that is because that evidence does not, and will never, exist.