Go to Top

The Balance of Vision: Obs, Sentries, and Smokes

A great visionary of early DotA once said, “This game is complicated at first, but in the end it’s very simple. It’s just a main objective: you have to fill up.” At first glance, a hero becomes more powerful from obtaining two resources: gold and experience. Gold translates into items, experience turns into levels, and eventually you steamroll the opposing noobs with a double naked rapier. But there are other metrics to measure the strength of a team at any given time. A gold and experience lead does not ensure victory. Spell cooldowns, buyback status, Aegis timers, inventory space, gold stash, and dozens of other factors affect the advantage. And Dota is all about the advantage. Advance when you have it, survive when you don’t. The team with the advantage dictates the flow of the game. Unless TMNT is running the “turtle time” game plan from level one, everyone wants the advantage.

Perhaps the most crucial advantage to seek after gold and experience is positioning. Good positioning results from manipulating an infinite number of moving parts. World champion team Invictus Gaming does not have perfect positioning, and they never will. We play an amazing game with an unreachable skill cap dangled far above our heads by Zeus himself. What a little Olympus mountain troll.

There is one aspect of positioning that is so blatantly important, and that is vision. Without vision, everything goes to shit. Most positioning tactics are useless while blind, or at least incur extreme elevated risk. Vision is a resource in itself, and such an obvious one that everyone knows about it. The lowest level pubs feature plenty of observer ward purchases. Do new players use them well? I saw this ward while playing with a friend last night…

perfect observer ward placement

Perfect observer ward placement

I didn’t even flame. Maybe I’ve matured. The point is that vision mechanics can be extremely basic. Beginners try to place an observer ward where it will provide maximum (and useful) vision. Experts treat vision as a whole game of its own, reaching levels of complexity that only the most experienced support players can fathom. The vision game is different in every match. Many heroes have the ability to improve their own vision, or to decrease the enemy’s. But regardless of heroes, the majority of the vision game in high level play is fought with three sacred items: the observer ward, the sentry ward, and the smoke of deceit.

Vision is important; we’ve established that. So naturally the balance of vision is also important, as it has a major effect on the balance of the entire game. The balance of vision involves the cost, duration, and rarity of observer wards, sentries, and smokes.

The balance of vision involves the cost, duration, and rarity of observer wards, sentries, and smokes.

Every part of this balance can be reduced to numbers. Numbers change. Numbers have changed. The vision game items have been balanced and re-balanced for years, just like every other item. And the current balance is… awful.

Now before I am deemed a heretic and sacrificed before the lord IceFrog himself, let me explain. Balance is an art, especially in a game with the complexity of Dota. Every item can be purchased by both teams, so balance is actually relative balance to other items, with the primary goal of creating ideal game play. To put it simply, if observer wards cost 50 gold with a 10 minute duration and no cooldown, they would be “imbalanced” in the sense that both teams would heavily favor the item in every game, resulting in a decreased value of the item’s counter, sentry wards. Ideal game play (for the most part) includes all items being useful in some scenarios while maintaining a value relatively similar to other items. This balance affects how the items are used as well as how often they are purchased. For the sake of exciting game play, the three vision items are currently unbalanced.

Observer Ward
Cost: 150g – Charges: 2 – Duration: 6m – Max stock: 2 – Stock cooldown: 6m

Observer WardObserver wards are ultimately defensive. Yes, they can be used offensively to spot easy ganks or snipe a courier, but for the most part their usage decreases action. Maintaining rune and Roshan vision can be argued as offensive and defensive, but the most common usage is to prevent ganks. This is not new information. The stock limit and cooldown are relatively new features to observer wards when considering the entire history of DotA: Allstars. These restrictions were imposed to set limits on the usage of observer wards. This balance change suggests that ideal game play includes observer wards, but not too many. Observer wards add an extra layer of strategic insight and preparation to the game, but in bulk remove a layer by minimizing fog of war and simplifying the vision game. In short, observer wards are boring. They prevent ganks. We want ganks. Ganks are kills. Ganks are action. Ganks are exciting.

Sentry Ward
Cost: 200g – Charges: 2 – Duration: 3m – Max stock: ∞ – Stock cooldown: 0

Sentry WardSentry wards are ultimately offensive. They counter observer wards, creating opportunities for ganks. They remove the enemy’s vision, forcing them to play blind and, as previously stated, take risks. It’s a risk to farm without vision, and more importantly, it’s a risk to place another observer ward. Risks are exciting. By definition, risks cause exposure to danger. Risks are also often accompanied by large reward. That means kills. Sentry wards have no max stock and no cooldown — ideal game play voted “yes” on unlimited sentries. In fact, sentry wards used to have a larger vision range similar to observers. Vision removed? Offensive. Sentries can also be used for killing invisible heroes. More kills. On the other hand, sentries can be used as protection against invisible heroes, including smoked heroes. Less kills. This game is not simple.

Smoke of Deceit
Cost: 100g – Charges: 1 – Duration: 40s – Max stock: 3 – Stock cooldown: 12m

Smoke of DeceitSmoke is a very new item, and admittedly took me a long time to gain comfort with. Introduced just over two years ago in 6.70c (I am writing this article while 6.77c is the newest version), smoke is also an offensive item. Woo! Go smoke! Another direct counter to observer wards, smoke creates opportunities to sneak past areas of vision to exploit the enemy’s false sense of security. In other words, “You have vision, but it’s bad vision. Time to smoke up.” Smoke has a larger max stock than observer wards, but only comes with one charge and has a very long cooldown. In current game play, the max stock and cooldown are almost a non-factor for smoke, as it is extremely rare to attempt using more than three smokes in 12 minutes.

Now, I’m going to assume we’re balancing for competitive ideal game play, as opposed to casual. In public matchmaking, everything is actually pretty awesome. Sometimes teams have observer wards, sometimes they don’t. When they do, they rarely get countered with sentries. The false sense of security provided by observer wards is closer to true. On the rare occasion that a smoke is used, it works. So here’s my matchmaking tip of the day: buy early smokes and gank. And don’t stop — they never learn.

There is a huge emphasis on early control, implemented primarily with trilanes and observer wards.

Back to competitive play. The opinion of current unbalance stems from the conception that current game play is not ideal. Compared to past iterations, it’s boring. There is a huge emphasis on early control, implemented primarily with trilanes and… observer wards. When a team chooses to trilane, they’re choosing to win that lane. The only instance when this is not true is with an offlane (offensive) trilane vs. another trilane, at which point the offensive trilane starts at a disadvantage without easy access to pulling and no tower to hug. The goal is not necessarily to win the trilane, but rather to disrupt the enemy’s trilane which they chose for the “guaranteed” win. This is a risky move because it assumes that the two 1v1 lanes will have a favorable outcome. Because of this, the majority of matches feature a 1v3 lane, a 1v1 lane, and a 3v1 lane with the trilanes typically farming the jungle as well.

How do you beat a trilane? You gank. If three heroes are devoted to a side lane (usually the safe lane, the furthest away from the rest of the action as it is separated by a jungle), the remaining two solo heroes are prime targets for ganks. They know this. They use observer wards to prevent this. A successful gank requires prior sentry warding or a smoke, both of which cost more than the observer ward they are meant to counter. Ganking becomes an investment, and all it takes is a missing call from teammates or decent map awareness to back up and avoid the gank. The ganking hero spent more gold and wasted more time away from their lane/jungle (more gold and experience) than the retreating solo hero. The culprit: the cost of observer wards relative to the cost of sentries and smoke.

6.75 - September 30 2012
* Observer ward cost reduced from 200 to 150.

In the words of Liquid`ixmike88, “welp”. Before this buff, all three vision items had the same cost per charge (100g). Paired with the closely following 6.76 cost reduction of Boots of Speed (500g -> 450g), these changes were considered support buffs to allow ward buyers to acquire bigger items faster. But that’s not what happened. Instead, the observer ward buff acted as a farming buff. Cheaper protection for the carry results in a better risk/reward ratio. Instead of farming bigger items, supports now focus on stacking and pulling, using their extra gold to buy…

Instead of farming bigger items, supports now focus on stacking and pulling…

Smoke. More obs, more smoke. If observer wards weren’t already being constantly bought out before, they certainly are now. The cooldown is the limiting factor, not the cost. The fastest way to perform a gank through an observer ward is a smoke. Teams know this, and they expect smoke ganks. Playing passively falls into the category of defensive, boring play. Team A tries a gank. Team B suspects it because heroes are missing. Team A fails the gank. Everyone misses out on farm, no kills result, and the game takes longer to finish. Heroes can go missing without trying to gank, but a similar result occurs. No kills and someone stops farming to protect themselves. But the best teams know they need to farm to win. Sometimes it’s greedy farm, and sometimes it’s smart. Regardless, the only sure way to farm safely is to protect yourself with…

Sentries. More gold spent on consumables, less gold spent on power items. The worst part? Protection against smokes has become one of the primary uses for sentries, and it also happens to be one of the only defensive uses for sentries. zzz…

We have now entered the vicious circle of unbalanced vision.

Observers < Smoke < Sentries < Observers

This is not Dota. This is not ideal game play.

The fix? Any of the three vision items could be changed. I think the most reliable solution would address the changes that caused this problem: the introduction of smoke and the cost reduction of observers. Smoke is often used in predictable situations, but part of the problem comes from large team smokes. When four or five heroes are missing, a smoke has to be assumed and everyone has to get back. No action. But of course smokes can lead to successful ganks, and that’s what we want. I don’t think anyone wants to see smokes removed entirely, but a change could be in order. A friend of mine by the name of Soothsayer suggested that smoke could only work on up to three heroes in the area of effect. If you have another idea, post it in the comments of your favorite community site or tweet it to me @Spitwad. But let’s start with the easy one and bump obs back up to 200g — they’ll still be bought out every game.

By the way, that visionary was wrong.

About Aaron Stern

Aaron oversees the day-to-day operations and production elements at NEO Dota. After a mediocre career playing competitive Dota from 2004-2009, he turned his efforts towards developing the Dota community, and has since been a journalist, league administrator, and broadcast coordinator for organizations such as GotFrag, CEVO, and TGBF. He is best known for creating and maintaining NADotA.com.