He wrote an article about games. The crux of his article is that games are two hard. Games punish players and they should not do so.
In essence, he's putting forth a notion of design: that games should avoid punishments. That a game is "better" if it does not punish the player, if the player is easily allowed to progress through the game and see more stuff.
The fundamental flaw with his notion is this: it's not objective. It's subjective. It doesn't work as a credible theory of design because of this. His argument isn't that games that use strong Risks in the Challenge dimension are broken; it is merely that some people won't play them.
A proper Theory of Game Design is not intended to say what the most people will play. Ultimately that is a subjective notion, one that changes over time based on what the population wants. An objective Theory of Game Design's purpose is to aid a game designer in being able to tell if a game is functional, as well as describe how functional it is.
My section on pathologies does not detail subjective problems. Something is pathological if it is objectively broken. There is some subjectivity to it, as the Situation Model does model a degree of subjectivity in its representation of the player's attributes. The Situation Model only goes so far as to allow the game designer to assign a generic value to a player's attributes in an area. A kind of, "You must have this much reason to pass this section."
Given that, is it possible to have a game who's Risk dimension is too great? For a particular player, and generalized over a number of players of similar mindset, yes.
As an example, the Risk for being on one's last life in World 8-4 of Super Mario Bros is enormous. Fail, and it's back to the beginning for you, where at best, you'll have to trek through 7 levels before getting back. As children, this risk was somewhat acceptable; children tend to have more time and tenacity on their hands. Plus, children get games when their parents are good and ready to give them to them.
Even so, Risk can be mitigated by another factor: how much fun it is to play. Super Mario Bros. game mechanics are rather enjoyable. While the early levels do get boring after a while, even a seasoned expert gets a thrill from some of the World 8 levels. If a game is fun enough, Risk can be averted.
However, there is one problem. While Chris summarized his position as wanting less Risk (using different terms, of course), that's not what his article actually described. The description of why he doesn't complete games did not deal with Risk; instead, it dealt with Challenge as a whole.
For one reason and another, I was never able to complete them because at some point while playing, I would hit a "blocker." (A blocker is a term that game designers use to call a part of the game that stops the player's forward progression because of a complex puzzle, or arbitrary twist in the game.) When I hit a blocker, I give it a couple of tries, decide that this seems like a great place to take a break for the night, and lo and behold, the game begins to collect dust and I never play it again. That's the story for most games on my shelf.
That is not an aversion to Risk; that is an indictment of the entire domain of Challenge.
Do videogames need Challenge in order to be videogames? No. The definition of videogames, as I present it, does not explicitly or implicitly require Challenge. It only requires interactivity and the intent to provide enjoyment.
However, do games, whether in videogame form or otherwise, need Challenge? Absolutely; Challenge is at the core of games.
Look at the difference between Chess and Tic-Tac-Toe. Why is Chess played at high skill levels, while Tic-Tac-Toe is considered a child's game? Because Chess provides a much greater degree of Challenge. Chess played against a tough opponent requires a great deal of Knowledge, Reason, and Comprehension in order to be successful. By contrast, Tic-Tac-Toe against the smartest player of the game in the world will always result in a tie.
In terms of non-videogame games, Challenge is ultimately all there is. Enjoyment is gained from little else; the base interactivity of non-videogame games tends to be moving pieces around a board. Table-top RPGs and high-end miniatures games with grand backstories can buck this trend, but for the most part, non-videogame games derive their enjoyment entirely from the Challenge that their mechanics generate.
In terms of gaming videogames, Challenge is still a very strong aspect. The Situation Model demands it. Though it models interactivity, because most of the dimensions of Challenge are based on factors in the Situation Model, Challenge is an important component of gaming videogames.
After all, what good is gaining Knowledge if the game does not confront you with a Situation in which that Knowledge is required? What good is Experience if the game does not test it? And the only way to test such things is by creating a Situation with Challenge in the appropriate dimensions.
It's like rats in a maze, really. It was discovered that rats running through a maze actually learned how to get around simply by being there. When they started putting food in various locations, the rats were able to find the food much faster than rats who were new to the maze. But nothing indicated to the researchers that the rats learned the maze until they started placing goals in it.
Only when something is Challenged are you certain that it is really there.
However, unlike most non-videogame games, videogame games can create enjoyment from non-Challenging features. The act of interaction, for example, can be enjoyable, as can the videogame's setting or other non-game factors. Even so, much enjoyment can be had by having the player surmount a Challenge.
The difficult, of course, is Pacing. That is, providing the right Challenge at the right place at the right time. As mentioned, some people like sharper pacing curves than others. Despite the fact that Myst sold many units, very few of those buyers actually liked the game. It had a very sharp pacing curve, one that was too sharp for many.
So, what Chris is asking game developers to do is lower the pacing curve. This begs a question though: is there a right pacing curve for a game? A single correct solution for a specific videogame?
No. Ultimately, everyone likes their pacing in a certain way. Some players get bored with games that are too easy. Other players get frustrated with games that get too hard too fast.
What Chris ultimately wants is for game developers to cater to the low-pacing crowd. That's his call, but it need not affect your decisions as a game developer. Higher-paced games are still valid and viable games, from an objective standpoint.
On a personal level, I can't really bring myself to care for the low-pacing crowd. It simply isn't that interesting of a task to design games for them. Their enjoyment tends to stem from the base interaction or other facets of the videogame's construction than from anything involving the game rules themselves.
This musing went kind of wide of its initial purpose, but maybe it's an interesting read none-the-less.