Ice's biggest problem has always, I feel, been in its type effectiveness distribution. You dance around this issue, but I think you neglected to mention the most important point: that the ratio of offensive-to-defensive type strength is the dominant factor in a type's viability. In this post, I will outline this argument and relate it to your motivating question regarding GameFreak's failure to successfully balance ice types. GameFreak appears to operate under the assumption that offensive coverage and defensive coverage roughly balance each other out. This can be easily demonstrated by analyzing the type relationship chart. As a result, steel is designed with many resistances (and few types that it is super-effective against), and as you point out, ice is designed with wonderful offensive coverage but poor resistances (weak to rock, fire, fighting, steel...). However, all things equal, for Pokemon of type a, a's defensive bonuses (resistances) are generally more valuable than offensive bonuses (super-effective relationships). This is true for several reasons: Offensive coverage can be attained even off-type. Obviously, you don't need to be of type a to use moves of type a. This is particularly true for Ice moves, which are often deployed off-type by Water Pokemon who have easy access to strong Ice moves and by miscellaneous (usually special) attackers looking to check the many ground/flying/dragon types or attain strong neutral coverage (how many times has bolt-beam coverage been ran off-type, for example). The marginal benefit of additional defensive bonuses is greater than that of additional offensive bonuses for type a, because a Pokemon may only have 2 defensive types, while they can have more offensive types, corresponding to the four (or more in some cases) moves they can use. The best counterargument to this point is the existence of STAB. However, the advent of several mechanics have lessened the importance of STAB and increased the importance of type coverage in the last several generations. These include, but are not limited to: The proliferation of strong dual-type Pokemon Mega evolutions (doubly so for those that change the Pokemon's type) Z-moves More raw power in the form of general power creep higher base stats better items more powerful attacking moves more powerful support/stat-boosting moves Yes, for type a, the average movepools, abilities, and stat distributions of the Pokemon instances of type a are important - but I will argue that these type relationships dominate the other factors: Most types include a wide range of Pokemon, many of which have no need to be viable or strong (such as lower evolution stages), so looking at average stats or movepools or abilities is unimportant. Notably, there are 44 Ice Type Pokemon - we need not see all 44 commonly used to be satisfied with its competitive use, so it is fallacious to attribute the type's weakness to flaws that only characterize a subset of the type, even more so when that subset would not be used for other, more compelling reasons. In the damage formula, realistic incremental changes to type have a greater impact than realistic incremental changes to the base power of moves or to the base stats of a Pokemon. For all trends identified in Pokemon of type a, there will be outliers in each trend. For example, you neglect to mention Cloyster, Mamoswine, or Rotom-Frost, all of which succeeded in ways different from those you outline. Thus, aggregate measures are less valuable in this context. All else equal, "glass cannon" types will always be deployed less than the robust, defensive types. Anecdotally, Steel-types are perennially used, while Ice continues to languish. Of course, simply buffing defensive stats or providing Pokemon with obscene abilities could serve a band-aid solution - changing Mewtwo's typing to Ice wouldn't take it out of Ubers, for example, and changing Snow Cloak to generate Doryuuzu-like weather sweepers wouldn't create an enjoyable meta. The best way to fix this is to change Ice's resistances, as you suggest. This can also be done by creating new abilities that add resistances or immunities (Think dry skin or levitate, respectively). An alternative strategy to improve the defensive capabilities of the type could be to focus on the theme of residual damage and healing that is associated with Hail/Ice archetypes. For example, Gen 4 Hail teams deploying Walrein to stall could have benefited from changing Ice Body to restore status conditions (similar to Hydration).