Returning to our original example, the developers of Puzzle Quest actually should have considered cheating – but in favor of the player. The game code could ensure that fortunate drops only happen for the human and never for the AI. The ultimate balance of the game could still be maintained by tweaking the power of the AI's equipment and spells, changes that appear fair because they are explained explicitly to the player. The overall experience would thus be improved by the removal of these negative outliers that only serve to stir up suspicion.
When the question is one of fairness, the player is always right.
I say: Bullshit.
A few comments: (1) I consider this to be utterly antithetical to the old-school D&D sensibility. We say: Here are some simple, equitable rules for both sides; the DM will adjudicate them neutrally; it is up to you to figure out a way to survive/succeed, based on your playing skills. You may either win or lose. (2) Is this not similar to the requirement that all big-budget Hollywood movies have a happy (and predictable) ending? Perhaps the actual claim is that for maximal sales, play-act as though the player is always right. And perhaps as hobbyists (instead of a corporatist stance), we are more free from this restriction. (3) As I say to my colleagues and math students: “Random numbers will mess with your head”. I encourage developers of games to confront this head-on; use the game as a learning tool for dealing with randomness, and unexpected setbacks, and our built-in intuition leading us astray. Do not surrender to just feeding back to players their own advance expectations all over again. We must use games as training for the real world, not for life in the Matrix.
Halfway through the GD article, another example is brought up favorably: the technique of “rubber-banding” cars in racing games (that is, giving whoever is behind an automatic speed-boost). This strikes a resounding chord in me, because my first game-industry job out of school was as a programmer for Papyrus Racing Games in Boston (a branch of Sierra/Vivendi at the time), where we made racing sims like Nascar Racing, IndyCar Racing, and Grand Prix Legends.
The emphasis at Papyrus was on racing simulations, and I loved working there in its heyday. Many of us went to racing school for practical experience. Indeed, we routinely received complaints from players, publishers, and reviewers in the vein: “The game is too hard. The cars are too difficult to control. It's hard work to keep from spinning. If you have one collision you cannot win the race.” And we as a company would say: "Tough." That's exactly what real racing is like; the cars are powerful, easy to spin, and there is no comeback from an engine-shattering wreck. There is no save game. Sometimes the greatest challenge in racing is the concentration and endurance necessary to withstand a 2-hour race without damaging your vehicle in heavy traffic. If you want to truly learn racing, this is your game. If you want a cartoon that strokes your ego, it is not. Would we have ever implemented “rubber-banding” in a Papyrus racing sim? Hell, no.
And let me say this: In all the games that I've played heavily (D&D, Nascar Racing, Starcraft, Poker, etc.) the most interesting part – practically the only part I find of interest any more – is this: What does a player do under an enormous, unfair setback? Do they crumble and give up? Do they yell and complain after a bad beat (as in poker)? Or do they have the fortitude to gather themselves up, get creative under adversity, and fight back? Can we learn to be cool under pressure?
If we make our games coddle players – if we start cheating in favor of the players, so that they never have unfair setbacks – what we will actually cheat them of is this critical opportunity to grow and learn about ourselves through gameplay.