Theory: Gold Values

In your travels through Runeterra you may at some point have encountered strange phrases such as "item efficiency", "cost-efficiency", or "gold value". Theorycrafting gurus expound on these strange, mystifying terms, claiming that they indicate one item or mastery is secretly overpowered. There's a method to the madness, and madness to the methods.

The basic concept of an attribute's gold value is straightforward. All you need to do is take a basic item and divide its gold cost by its attribute. Voila, now you know that 1 Attack Damage is worth 20 gold, and 1 Ability Power is worth 21.75 gold. Simple, obvious, and useful.

Theorycrafters use these values to determine what items are cost-efficient. An item is cost-efficient when the sum of the value of its attributes is equal to or greater than the cost. It's fairly straightforward math once you have the gold values for all attributes.

The problem, referred to previously as "madness", is that we don't have gold values for all attributes. A significant number of attributes don't have any basic items on which to base their gold values. Cooldown Reduction, Life Steal, Spell Vamp, Penetration, Movement Speed and others are only found through upgrading items and other special cases. As a result, the only data points we have to work with are the combine/upgrade costs of items.

Unfortunately, combine costs have essentially nothing to do with the value of the stats gained. You need only look as far as Trinity Force for this to become self-evident. Its combine cost is a ridiculously small 3 gold, yet it gains 5 Ability Power, 10 Attack Damage, 12% Attack Speed, 50 Health and more. The value of the stats I listed alone is 880 gold, not to mention the nebulous value of the Movement Speed and item passives. Upon closer examination, all item combine costs are just as devoid of meaning.

This presents a problem for theorycrafters as the gold values of roughly half of all attributes cannot be determined. Many theorycrafters try anyway, but the results are mere conjecture. Do we base the value of Cooldown Reduction off of Kindlegem, putting the value of 1% Cooldown Reduction at 32.23 gold? Or do we base it off of Glacial Shroud, giving us -6 gold? With Fiendish Codex and Stinger giving us values of 16.75 gold and 45 gold per 1% Cooldown Reduction respectively, it should be immediately obvious that there's simply no way to assign a meaningful gold value to these sorts of attributes.

Some theorycrafters lack the conviction to admit it, but it is essentially impossible for us to have a complete picture of cost-efficiency. Without an incontrovertible method of determining gold values for advanced attributes most items can't be fully assessed. The best we can do is see how close an item comes to being cost-efficient before making a personal judgement as to whether the remaining attributes are worth the difference.

Theory: Donating to Junglers

It is currently normal, even expected, that junglers will donate their Wraiths and even the rest of their jungle to the lanes as the game goes on. This is because it's generally assumed that the lanes, not the jungler, will carry the game later. However, it's mathematically possible, and even practical, for a jungler to achieve or exceed solo lane levels of farm by reversing this relationship.

The math is pretty simple. A jungle clear of the small camps is worth 154g and 441exp before scaling and can be done approximately once a minute. Two creep waves, one minute's worth, are worth 228g and 529exp before scaling (and without siege minions). Donating as few as four creeps to a passing jungler is enough for them to surpass both the gold and exp potential of a solo lane who nails every last-hit. Even if we take into account scaling this remains true throughout a game.

First and foremost this illustrates the importance of coordinating recalls with your jungler. Even if you are only going to miss a handful of creeps while you are gone, having the jungler there to collect them boosts a jungler significantly. More often it's the case that many creeps will be missed, and giving them all to the jungler is enormous. Junglers like Nocturne, Udyr, or Shyvana, who thrive on farm, go from decent to powerful if they can cover for lanes.

But simply covering lanes isn't the point of discussing this topic. Rather, the point is to suggest that instead of having a farm-dependent champion top or mid and a tanky/support jungler feeding them jungle farm, it's entirely possible, even safer, to have a tanky/support champion top or mid feeding a farm-dependent jungler lane farm.

There are many tanky/support champions perfectly capable of holding their own in a lane, but who don't make quite as good use of the farm as a carry. Having them roam or recall frequently to allow their jungler to quickly clear a creep wave before returning to the jungle has the potential to provide the carry with more farm than any solo lane save those rare cases mid where it's safe to clear both Wraith camps repeatedly.

The big advantage to this strategy is that it's very hard to counter directly. So long as the relevant jungle entrances are warded or the nearby lanes keep tabs on their opponents, the jungler is perfectly safe even if their jungle is warded. Only an extremely concerted counter-jungling effort could threaten the farming jungler, and only at great risk. It's really the easiest way to guarantee a given champion will be farmed.

The disadvantage is that it's much easier to track a roaming mid or top lane than it is a roaming jungler. With the jungler tied up in farming, and conspicuously mid or top when their ally is absent, there will generally be fewer ganks or at least far more predictable ones. This can be offset in various ways, but in general the lanes will have to be significantly more self-sufficient than normal.

At the same time, having Smite on a carry is awkward. It's not impossible to work around, and there is the potential to have some shenanigans with Smite on your mid champion, but it is a difficult adjustment that teams will have to deal with.

Ultimately, however, this is the best alternative to the current meta I'm aware of. As it stands junglers are almost always tanky, while carries in the lanes take their farm. Many underused champions might see the light of day again if we reverse this relationship.

Quote: On Theorycraft

"You can discover more about a champion in an hour of play than in a year of theorycraft."

-Plato