User blog comment:Healingmantra/New attribute for supports/@comment-1330314-20161205044103

I very much like the core idea behind this. In a game that is now binding its classes more closely to specific stats, supports still have no stat that is bound to them, and there are clear negative consequences to that. Supports cannot afford to build AP in large amounts, for example, because any AP item cheap enough to work for them is likely to be abused by mages, and as a result, support items often just don't boost their kit power at all, and instead just give them powerful actives. These effects are fun, but end up homogenizing the support class, and make scaling on their abilities much more awkward, to the point where supports typically cannot function correctly in both solo and duo lanes at the same time.

As for the implementation, though, I'd perhaps do things a little differently. For one, I'd keep the AP scaling on utility abilities, and instead turn Spirit into a percent multiplier (e.g. "+10% utility"), so that supports could scale quadratically in their utility, just like tanks scale quadratically in their effective health. I would also avoid making the duration of any basic ability or CC scale with crowd control, for a number of reasons: many abilities rely on memorization of their duration so that you can react immediately after, and making that duration fluctuate could make that gameplay less clear (imagine of Zhonya's Hourglass had a ratio to its duration, for example, especially since many players already miscalculate it right now).

Another issue with duration scaling is uptime: abilities are generally balanced around a certain window of uptime, i.e. a rate at which they are in effect, and currently the only way to increase uptime is through cooldown reduction. Increasing uptime by enhancing duration runs the risk of leading to uptimes approaching or exceeding 100%, which is something you absolutely do not want on CC, especially hard CC. The solution to that, bar removing scaling, would be to nerf the base duration of those effects, which would stop them from reaching abusive uptime levels, but would run the risk of resulting in very unsatisfying base durations.

Regarding implementation of the stat, adding it to new items is one possibility, but ultimately you'd have to add a whole slew of Spirit items to accomodate support itemization around it. That's not impossible, but would require a ton of additional work. A potentially more economical solution could be to turn an existing stat into Spirit: health regen, for example, is currently a very niche stat that, despite being on a host of support items, isn't particularly appealing to many supports, and isn't even a stat that gets prioritized on tanks or fighters. Reworking it into Spirit, or an equivalent utility-boosting stat, would allow it to contribute to support itemization without too many problems, especially since tanks also tend to have a lot of personal shields and defense boosts that would benefit from the stat. There might have to be some adjustments to Tiamat and Titanic Hydra, but only because the stat is tacked onto that item without it serving a greater purpose.