This is the final part of our series dedicated to constructing frameworks for decision-making at different stages of the game. Today we are going to concentrate on the Late Game. Not every game goes late and in most cases, the outcome of the match is decided much earlier. Because of this, many inexperienced players really struggle with making correct choices at this stage. We will not be able to solve the problem outright — this mostly requires in-game experience, but we can help with identifying things to pay extra attention to.
Most of the information from the previous posts in the series still applies.
There is a reason most professional teams don’t attempt to go high ground without Aegis, even when they have a sizable Net Worth lead. Once one team has full control of the map, there is usually little reason to be hasty and potentially squander the lead.
For the leading team, playing cautiously when there are no safe objectives on the map is of the utmost importance. No matter how much of a Net Worth lead one team has, even their most farmed hero will most likely die if they are left in a 1v5 situation. Too often teams get too greedy and try to play all three lanes at once.
This is an easily punishable mistake, especially around Roshan timings: a support death can and will tip the balance of power on the map for a period. Even if it is a support with 2k net worth, while the leading team has 10k+ advantage.
Never forget about positioning. The more of an advantage you have, the looser your formation can be, since you will probably have more time to react, but the usual mantra of frontliners being on the front and safe supports being deeper and further away from the enemy is still as important.
Having map control usually doesn’t mean you control 100% of the map, unless you are absolutely crushing the game. Settle for two thirds of it: play two lanes, out pushing the necessary ones, and keep relatively close together. Most importantly, around the Roshan timer you need to start prioritizing the lanes closest to the Rosh pit, especially in the later stages of the game, when heroes are more likely to have Boots of Travel.
If, theoretically, the best way for the leading team to keep and increase their lead is to play two lanes, the losing team still has some room to play. Pay close attention to the heroes shown on the map by the enemy: if you see two main cores in two different lanes, you can generally safely occupy the third if you need to farm up.
Economy-wise it is an un-ideal situation, but as discussed in the previous blog post, it is usually preferable to suicidal attacks on a much stronger opponent in an open field. However, the Roshan control becomes much harder, if the enemy is dictating the part of the map you are allowed to play in.
This is where your pushing heroes really need to shine. Heroes who deal fast and efficient structural damage are important not only when you are trying to breach the enemy base: they are important in terms of macro when you need to focus the enemy attention on a part of the map. If you have heroes, who are a threat to structures, ignoring them means the enemy loses some map control and will have to give away some gold, whenever they are attempting team maneuvers, like, for example, taking Roshan.
Worst case scenario, you will trade Roshan for an objective, not allowing the Net Worth gap to get too big. Best case: the enemy will start trickling in, making for easier targets. Be aware of the worst case, however: when the enemy fakes attempting Roshan and is responding in full force. This is what makes vision so important, and this is what will allow you to retreat safely, delaying the end of the game.
Finally, a quick note on map control: while Top and Middle lane towers are the most important structures when it comes to map control, since they are the closest to Roshan, Bottom Lane barracks are usually the best to take in the early game, simply because it takes a lot more time to make rotations from the Rosh Pit to the Bottom Lane to outpush. Keep it in mind, when trying to breach High Ground. For Dire, the Bottom Lane is typically the best lane to push in naturally, since it opens the enemy jungle, so if you have an opportunity to do so in the early and mid game, try to take at least the ranged barracks in it: macro-wise it is usually worth it.
The last thing to keep in mind is how your team distributes the gold and farm. Position one carry getting extra gold to change one of his bigger items for a slight upgrade to stats is usually inferior to a support player getting a crucial item.
This doesn’t mean you have to give away farm as a carry if you need an MKB against Phantom Assassin, where the item you are going for is absolutely crucial for your damage output in a teamfight. But it does mean you have to give away farm if you are selling your early game Battlefury to get something like Mjollnir: the difference in effectiveness can be noticeable, but it is not going to be as noticeable as something like Aghanim’s on Disruptor or Grimstroke, in a game versus Terrorblade.
Smart economy distribution is very important, especially in the later stages of the game. At this point, the game is less about the efficiency of the items, but rather about the items being a direct answer to enemy threats. In the vast majority of cases, extra damage on an already high-DPS hero is going to be inferior to the added utility of a possibly less expensive item on a support. Think where the next 5k gold on your team is going to have the most impact, and act accordingly.
This concludes our series on Analyzing Game States. Dota is an incredibly complex game and we probably haven’t even touched 20% of the things top tier professional players need to pay attention to, but we hope that we at least helped newer players to better organize their understanding of the strategic side of the game.
What else do you think players should pay attention to in their games? Did we miss something important? Let us know in the comment section below.