Patch 7.00 changed the landscape of hero builds with the introduction of talent trees. Replacing the attribute bonus upgrade, players now have free upgrades at levels 10, 15, 20, and 25. Not only do these buffs smooth over some of the frustrations of playing many heroes, it opened up interesting, new builds by making them viable.
What the talents do is replace the attribute bonus upgrade and add more weight to those intermediate levels between and after leveling skills. A level 20 upgrade is no longer just a +2 attribute bonus, but it can be something like Lich’s +150 damage upgrade. Unfortunately, many talents appear to be placeholders, with as much flavor as the old +2 stats bфonus, such as a minor movement speed upgrade, +5 to all stats, or +6 to one stat. It’s a natural consequence of adding 8 talents to every hero. Not every talent can be a major upgrade.
But there are those talents that feel more significant than others, even if they are minor buffs. At level 10, Ogre Magi’s +100 cast range changes how the hero feels. It’s a mini Aether Lens that can stack with Aether Lens, and it’s an upgrade that can’t be easily substituted by another item, as with talents that add to main stats.
Talents That Open Up Builds
This isn’t just about the level 25 talents that gives heroes an Aghanim Scepter-like upgrade to their skills, but also the talents beneath that tier that can fundamentally change the experience of playing a hero. An extra +20 damage at level 10 for Ember Spirit doesn’t feel like much when compared to a 15% Spell amplification that capitalizes on Ember’s mid game magic damage peak. Before this talent tree, Ember Spirit players usually leveraged his magic damage in the mid game to transition into the late game with physical damage items like Battlefury and Daedalus. But his spell amplification talent, along with 15% cooldown reduction at level 20 and +2s Searing Chains at level 25, opened up a new magic damage build.
Then there’s Jakiro, who by level 20 turns into an artillery hero with both a +400 attack range and +120 cast range upgrade. The alternative is +250 health and +150 Gold/Min. The impact for those upgrades aren’t felt immediately, but rationalized later. You won’t feel the impact of a health upgrade until you narrowly escape a fight. With +Gold/Min, it’s difficult to balance the opportunity cost, because the effect of the talent is spread out over time. Twenty minutes in the game, did that extra gold/min help, or would a cast range upgrade help you net kills and gold that you otherwise wouldn’t have?
The Allure Of Gold Talents
Ogre Magi players favor +Gold/Min, but it comes at a cost
To compare, Midas gives 114 GPM, outside its XP benefits. Most gold talent for supports, like Rubick, Ogre Magi, and Ancient Apparition, start at +60 GPM at level 10. It’s a non-trivial amount, but players are often overrating its impact compared to the other choices. The simplest comparison is comparing win rate for both talents.
The majority of Ogre-Magi players (52.4%) choose gold over +100 cast range, but those players suffer a 1.2% win rate decrease. In a more drastic case, 80% of Rubick players choose +60 Gold/Min over +50 damage, with no significant difference in win rate.
Passive gold gain takes time to accumulate, except that in the late game, there’s more space for supports to farm. The contrast with gold/min talents is that there are many ways to get more gold on the map, while there are few options to compensate for the alternate talent. Disruptor players have the same choice of +GPM or +Cast Range at level 10. Aether Lens is the only other way they can achieve that benefit. Enigma at level 15 gets the option of either +120 GPM or 15% Cooldown Reduction (+2% win rate). Again, there’s only one other item in the game, Octarine Core, that can reduce cooldowns.
Greed isn’t always good. Midas’ strength comes more from its XP boost than its gold gain. And players who opt to buy Midas, rather than use that 2050 gold for something else, are either sacrificing their early and mid game or had already lost it, in hopes to find an advantage later. The same goes for players who gravitate towards GPM talents.
Ancient Apparition players overwhelmingly favor his gold talent
AA players who pick his GPM talent over 8% spell amplification currently have a -1.0% win rate. More than most supports, AA’s impact spikes with Aghanim’s Scepter. GPM can accelerate that item build, but spell amplification is a rare mechanic. 8% spell damage is more than what Aether Lens gives, and it’s equivalent to the bonus 128 Intelligence would give.
Stop Dying or Die More
Across the board, players who choose the talent that reduces respawn time have a lower win rate. AA players pick the respawn talent 70.3% of the time and lose 8.5% more. The same goes for Enigma players—70.6% pick rate with a 4.5% win rate decrease. One case could be that the players who want lower respawn times are the kind of players who tend to die more frequently. Another likely case is that since the talent is more beneficial the more you die, it means you’re probably in a losing situation. Still, the result is that the win rate is lower. The talent either currently doesn’t do enough (-30 seconds at level 15 is a 58% reduction in respawn time) or it’s fundamentally faulty.
For many heroes, the alternate choice is one that increases survivability. Beastmaster gets strength. Enigma gets +300 health. Clockwerk gets magic damage resistance. Again, like picking cast range over GPM, the choice here benefits you both now and later.
Patch 7.01 tweaked a multitude of numbers, while buffing the respawn time talent for four heroes. The patch is still young, and talents offer a way for Valve to focus their balance changes. The new talent trees present us with a choice, but the right one isn’t always clear.
