Thursday, February 14, 2013

Decision Making in Dota 2: When (not) to Build Hand of Midas

Hand of Midas is an item that I frequently see used incorrectly.  In my opinion the reason that it gets used incorrectly so often is that players do not understand the costs associated with building a Midas, instead focusing on the benefit of the item, namely increased gold and experience acquisition rate.  Here I hope to shed some light on the real costs of building Midas, and offer some scenarios in which building Midas is justified.

What is Hand of Midas?

Hand of Midas is an item which has two main benefits.  Firstly, if grants +30 attack speed.  Secondly, and most notably, it has an active ability that allows the bearer to instantly kill a non-player unit and receive 190 reliable gold, and a bonus 2.5x normal experience as a reward, on a 100 second cooldown.  It is created by combining Gloves of Haste with a 1400 gold recipe.

Why is Hand of Midas Valuable?

When completed, Hand of Midas will allow a hero to accelerate their rate of both gold and experience farming, enabling them to farm expensive items and gain levels more quickly, and eventually will grant them a power advantage over other heroes who have not benefited from this acceleration.  It also allows the user to instantly kill any player controlled creeps, making it more valuable when facing heroes like Chen or Enchantress.

What costs are associated with building a Hand of Midas?

Starting with the obvious, Hand of Midas costs 1900 gold to build, which is extremely expensive for an item that grants such a small amount of stats.  It is important to realize that you are forgoing any other combination of items that you could have built with that 1900 gold in order to build Midas.  Here is a list of items that that 1900 gold could have been used on instead:

Phase Boots - 1350 gold
Power Treads - 1400 gold
Perseverance - 1750 gold
Helm of the Dominator - 1850 gold
Mask of Madness - 1900 gold
Yasha - 2050 gold

Every single one of these items offers stronger stats, and will allow you to have a better chance of scoring hero kills, AND will allow you to kill creeps faster which means your farm rate is increased by purchasing them.  Not only are you forgoing the gold and stats, you are also forgoing mid game power and presence, which will allow the other team better opportunities to take your towers and win teamfights.  This is the true cost of Midas, you are giving up mid game power in order to have a stronger late game.

When is Hand of Midas the correct item choice?

When choosing whether or not to build Midas you want to ask yourself a few questions.  Most importantly would be: Can my team handle an underpowered mid game without losing too much advantage (towers, gold, experience) in order to allow me enough time to gain benefit from Midas?  You need to look at the heroes on your team and the opponents' team, and understand what parts of the game your team will have strength, and which parts your team will struggle with.  In my eyes there are 2 main scenarios in which Midas is a good idea:

1. The enemy team is obviously aiming to have the game go late, and have heroes are able to provide enough space for their late game hero(es) to farm up, while also defending against any attempts we might make to end early.

2. My team is 4 skilled players who have incredibly strong mid game heroes that will be able to play the game 4v5 until late game when I start to gain benefits from my Midas.

Outside of these two scenarios, I do not believe that Midas worth what you are sacrificing.  As I mentioned earlier though, when facing heroes that rely strongly on creep domination (Chen, Enchantress), the value of a Midas goes up.  At the same time, when you are playing a hero that strongly relies on attack speed as a stat (e.g. Faceless Void), the cost of building a Midas is lessened slightly.

Conclusion

Midas is not a license to print money.  It is an item that disadvantages your team greatly throughout the mid game in order to grant advantage during the late game.  It is a niche item that should only be built in very specific scenarios.  Please use Midas responsibly.

No comments: