One of the easiest ways for a game designer to damage their games is to pay too little attention to the methods by which they design the mechanics of the game. The flaw that comes up the most when I see it is a failure to correctly scale games’ mechanical structures in relationship to each other. Often, developers get lazy and use exponential or even linear scaling in their games and don’t realize the impact that their actions have on the difficulty and balance of their work.
The first issue with scale is the fact that you have to consider technical limitations. Both in digital and tabletop works, you’ll still run into numbers that are just too big; 379 HP is a lot to track mid-battle, and while it’s definitely possible for players of a pen-and-paper game to deal with that, it’ll be very annoying and is likely to come down to estimation and could suffer from mathematical errors-fine in the concept of a casual storytelling-oriented game, but something that could cause major issues in competitive or regimented play. In modern video games, you really don’t even have to worry too much about this; unless you intentionally define values to a single byte or something like that you are unlikely to run into issues with scaling outside the possible bounds of a computer’s memory. These technical limitations really shouldn’t limit your design process, but it is important to consider the players’ limits in tracking values that are visible to (and managed by) them.
However, a more plausible issue with scale is the fact that it’s used to cover shoddy design. In a Final Fantasy game, for example, characters may start out with double-digit health point counts, and by the end of the game increase their health a couple orders of magnitude. This is not an issue given the linear design of the game, but applied to something like Fallout it would be fatal. A more balanced system would perhaps use smaller numbers, scaled by thresholds relative to the power of characters; that way a powerful character could still do a lot of damage without forcing the numbers used by the system into increasingly larger values. Dungeons and Dragons, for all its merits, has a major scaling issue where an end-game character is simply going to succeed at things that a beginning character would have to try hard at, even if they’re relatively difficult; were a character to specialize in Balance, for instance, they’d be able to Take 10 and apathetically walk across a surface a mere two inches wide; they could even risk a roll and fail only in extreme circumstances.
Another difficulty with scaling games is that it can lead to circumstances where success or failure are just not possible; in many cases this won’t adversely impact a game, but there are many cases in which this can ruin the challenge or desirability of an experience. For instance, one of the gripes frequently leveled at the single-player portions of modern military shooters is that they don’t let anything happen without the player’s direct actions, and on the other hand they only allow the player to participate in the game in very specific ways, discouraging them with incredible difficulty (or outright killing their character) unless they follow the game designer’s intents. In this case, this is usually done intentionally through level design and scripted events, but accidental scenarios can occur in just as many games that create insurmountable challenges. For instance, a shooting gallery may provide ten targets, ask the player to hit all the targets, and give them seven bullets. Barring the ability to penetrate multiple targets, the player is guaranteed to fail. While this may seem a little absurd, it’s a tool used by many game designers to direct the flow of players, but it’s also one that causes an incredible amount of frustration and anxiety.
A footnote in scale issues is that there are also narrative scale issues in certain things. These are most common in tabletop games, and occur from either intentional genre-based design or from negligence with potential results. For instance, it’s perfectly plausible to have a character with 9 Speed go 2500 km/h in a superhero game, but if you were to do that in a WWII war-game you’d likely have some issues with scale. These are usually prevented by forcing any such exponential design principles into closed systems or transitioning to linear improvement past a certain point, but I’ve seen a few examples of this that can lead to issues with cheesing the system to produce gimmick characters. Dungeons and Dragons is guilty of more than a few of these on account of stacking spell effects and the interactions between presumably separate class effects.
Fortunately, there are some ways to manage scale more effectively. On the tabletop, we see open-ended systems like West End Games’ d6 or Shadowrun that rely on a probability-centric return system that curves outcomes so that high-skill characters will still perform better than others, without running into problems with negating the impacts of character skill and making players feel like they are being ripped off for having invested four dice instead of three. In video games, logarithmic expansion can serve much the same purpose, allowing increased rewards for encouraged behavior but preventing the benefit in a competitive setting from becoming too intense. Likewise, a closed system solves the incredible scale problems of a linear or exponential setting by limiting the maximum scale of gameplay; several editions of Shadowrun feature these in various forms, and my own ABACUS system is built around them, and they’re just as common in video games. Likewise, allowing an expansion of the caps can slow down scale bloating, meaning that it’s impossible for players to cheese the system as efficiently or quickly.
In short, scaling in games can be used intelligently to provide new challenges and direct players, as well as giving them opportunity for expansion, but it also has many pitfalls, especially with unchecked exponential or linear scaling. It’s possible to prevent these issues handily by imposing caps or evaluating the situations in which the scaling occurs to determine a more appropriate outcome, or using a logarithmic curve that softly caps out the maximum power of characters and mechanics without totally negating the effect of prolonged investment.