Loading...
  OR  Zero-K Name:    Password:   

Forum index  > General discussion   >

Deliberately Studying Gamesense

6 posts, 384 views
Post comment
Filter:    Player:  
sort
USrankKnightshade
21 days ago
Deliberate practice is easy to define, and there are methods for training many hard-skills, but game sense is about as fuzzy as it gets for a soft skill. Despite that, we've all seen it happen. Certain players have an almost uncanny sense of timing, of know what their opponents are doing without scouting ot, etc.

Before I continue, Definitions I'm using:

* Deliberate practice: improving on a skill in a way that both isolates it and gives clear feedback

* Gamesense: the ability to determine from incomplete information both the current game state and the space of probable future and likely game states to a reasonable degree of accuracy.

It seems to me that part of it can be attributed to a combination of experience and game knowledge. E.g., if you present a player with a given problem, they typically only have a few solutions available. It's also possible to lead players into behavior (e.g. creating openings that are no such thing, by deliberately failing to counter something so that they invest further in it making later counters more effective, etc.) This would especially apply as you slow down to consider the game more.

However I find this an insufficient answer. Even if it's correct, not being able to deliberately develop the skill becomes a problem.

One thing I think isn't helpful is predictions in spec games, because for your predictions to be be correct, you have to blindly predict both sides without the ability to have game impact. Without this, your guess is just lucky, and there will be no clear feedback as to if that call was correct or luck.

I would appreciate the thoughts of others on this.
+1 / -0
AUrankUnicorn_cats
20 days ago
You touched on a part of this alrady, learning typical solutions to problems. Learning the typical esculation paths of diffrent factories and what they build to counter your factory (or your allies) lets you have a pretty good idea of what the enemy is going to make next. Which you can then plan around to eliminate or nullifiy that play.

As an example in 1v1 on Tarturusv7: you are spiders and your opponent is jump, the game state has proggresed to a point where they have an army of moderators, placeholders, and pyros. You have an army of venoms, fleas, and recluse. Becuase your recluse can skirmish their moderators and are protected from raids via venoms, your opponents current army combination will lose. Your opponet thus needs to make something new that can threaten your recluses (or something else like your base or venoms).

Your opponent has a lot of options; make a stinger (to buy time), firewalker, jugglenaught, fac switch or maybe an iris. Typicaly players go for the least expensive options if they need an answer soon, but if they belive they have the time (or econnomy, or were already building something big) they might go for something more expensive but harder to deal with. Players also typicaly use the options they have in their current factories if they would work decently rather then fac swap, unless fac swapping enables a lot of extra powerful options with which to smite you with e.g air. The game being earlyish on and on a moderate-low eco map thats small, the opponent probaly dosen't have the time or economy to make an expensive jugglenaught (this is game sense isnt it, how do we estimate eco and how fast the enemy will die?) They are much more likely to use a cheaper and effective option firewalker, knowing this you build an aegis and morph it to aspis a relativly cheap option that will protect your army.

You can use the cost of an Aspis as a bit of a mesuraing stick, by looking at how much of the map the other player controls you can get a good estimate of how much eco they have compared to you (Okay thats gamesense agian, how do we guesstamate eco?). If your's and their's economy is about equal we can assume that the firewalker costing a little more then the aspis should arive soon after your aspis is built, if it doesn't you should scout the enemy becuase we are out of the most likely and cheap options and should check what they are doing becuase the number of options is vast and the threat of a high cost solution is high, also I dont have a clue on what they could be doing.

Learning typical counters and esclations lets you guess at how your opponent is going to counter what you have, and thus let you counter those counter before they stop your army.
+2 / -0


SErankPLT_Godde
20 days ago
(edited 20 days ago)
I'm not sure it makes sense to practice gamesense. Pun intended.
To get a sense of where the game state is, you must first understand the game, what is possible in the game, then you must get a sense and try to predict what your opponent is doing and is capable of doing.

I don't like to try and guess what my opponent is doing.
Rather I prefer to either:
1. Scout the actual state of my opponents base, units and resources
or
2. I prefer to force my opponent to react to me so that I dictate the terms of the state of the game.

You can read more about that here:
How to Raid and Why You Always Should



Dealing with paranoia
How many scythes can your opponent have made without you noticing?
How many scythes has your opponent snuck through your lines?
How many scythes are standing right next your factory right now?

In some games, I will build extra Lotus turrets in my base for fear of Scythes even though the enemy never made any Scythes in that game.

It would be ridiculous to prepare for a Scythe in your base, the moment it could have been rushed out from the factory and reached your base.
It would be ridiculous to keep every mex and every con safe from an invisible threat, that could come from anywhere at any time.
One of the easiest way to deal with this is thinking in terms of opportunity costs.

A Scythe is worth about 2-4 raiders. If an unsuspected Scythe kills a con and prevents you from expanding, your raiders can likely also prevent the enemy from expanding or at least slow down their expansion enough to offset the loss of your con. Raiders also generally get favorable trades against Scythes, so if the Scythes reveal themselves too close to your raiders, you can run them down with your raiders and kill them costeffectively.

Cloaks are arguably the most annoying match-up in this sense, because you can never be quite sure, if the enemy has made Scythes, Imps, Phantoms, or have a bunch of Reavers walking down on your factory in the cover of a Conjurers cloak field.
Even if none or those units are used or any of those tactics are employed, consider them in your opportunity cost analysis.
Avoiding an Imp hitting a bunch of your raiders can simply be countered by spreading out your raiders when you advance, but that comes in the cost of slower movement across the map and not having the concentrated fire that you want when you engage the enemy.


You can considered the opportunity costs of your opponent as well.
If you send your raiders to the enemy base early in the game, and notice that the enemy is expanding slowly, or don't have enough units to defend their expansion, you might suspect that there is something going on.
Maybe the enemy is sending Scythes into your base or maybe the enemy has a group of raiders sneaking along to the edge of the map to surprise you and do crippling damage.

I also agree with what AUrankUnicorn_cats wrote.
+4 / -0
USrankKnightshade
20 days ago
Interesting, I didn't realize there was a Medium article out there to read, this is excellent thank you. SErankPLT_Godde Do you believe these rules remain in effect for team games where (I feel) scouting becomes more difficult and the combination of more competent opponents and a smaller slice of the income pie would presumably make it harder to control the options your opponent has.
+1 / -0


SErankPLT_Godde
20 days ago
(edited 20 days ago)
quote:
One thing I think isn't helpful is predictions in spec games, because for your predictions to be be correct, you have to blindly predict both sides without the ability to have game impact. Without this, your guess is just lucky, and there will be no clear feedback as to if that call was correct or luck.

I think predictions in spec games does help your gamesense. Not all predictions have to rely on predicting both sides.
For example, once defensive buildings have been set up, they are static and don't move. It is then up to attacking player to exploit the weaknesses of the defense and if you want to predict how the attacker performs the attack, you have to understand the state of their mind and if they are likely to exploit the gaps or weakpoints or not.

Predicting commander snipes in 1v1 is especially fun in my opinion.
This is a more dynamic case where the players are basically taking turns predicting eachother, while you, as a spectator, tries to predict both players predictions.
An unupgraded commander can lose to just 3 Scorchers or 6 Glaives and that trade is basically always worth it for the raider player to take.
However, the commander have the ability to build 1 to 3 Lotus turrets before the raiders reach the commander when spotted on radar. However, if the commander is building Lotus turrets between mexes out in nowhere, that defense can easily be circumvented and might just be wasted resources for the defender if the raiders choose to go for a different target. The attacker is typically also working outside radar coverage so they might not even know where the commander is or how many nearby enemy units there are to support the commander. If the commander player is bold, they can just assume that enemy doesn't know or doesn't dare to go for the commander kill and avoid building defenses, basically bluffing.
For the attacker, the commander explosion can also really make it really tense. The attacker might know that there are nearby units, like riots, but if the commander is just far away enough from the other units and defenses, you might be able to snipe the commander quickly with lots of raiders, but you also want to avoid losing too many raiders to the commander explosion and surrounding units, because if you lose 20 raiders to kill one commander, they enemy might just counter-raid and you lose all your expansion, making you lose the game afterwards.
+2 / -0


SErankPLT_Godde
20 days ago
quote:
Do you believe these rules remain in effect for team games where (I feel) scouting becomes more difficult and the combination of more competent opponents and a smaller slice of the income pie would presumably make it harder to control the options your opponent has.

When it comes to teamgames, you do not just have to predict your opponent but also what your allies will do.
However, you always have the ability to zoom out to see the whole map, track when your allies scout, see the frontlines and judge what parts of the front are likely to hold, advance or break.
Now if you have actual teamplay, then teammates should be delegated roles so that they can focus on microing their part of the front. Preferably you need a player to make call-outs and have an overview of the whole battlefield, while others are focusing on their part, but individual players can also make call-outs like; "Attack here", "Push", "Defend", "Help", "Skuttle", for teamplay.
An air player will generally have a good overview as their role is to watch the whole battlefield to look for targets of opportunity to bomb.
Having flying scouts can also be a force multiplier for the whole team as it makes each individual player able to judge the composition of the enemy force, and be prepared if the enemy is massing units behind the frontline to try and overwhelm a part of it.

With experience and knowledge comes gamesense. It will be easier and faster to judge the state of the frontline, and if parts of the frontline are likely to collapse, as you get better at the game and understand what the players are doing.

If you have a noob ally that is monospamming snipers, Phantom, which is easily countered by raiders and spam, it might still work if their opponent doesn't know how to deal snipers, or lack cheap spam or raiders to send at them.
You might be tempted to tell them to make Glaives or Reavers to counter spam, but they might lack the capability to multitask and keep the Glaives and Reavers close to the snipers without losing them.
They might be performing over their rating focusing on just the sniper monospam, without having to divert their attention. Maybe it is easier to convince them to hide behind other friendly units or send units to protect them yourself.

I like to say that League of Legends and most MOBAs are particularly toxic with random teammates.
Because in order to play well, you need teamplay with fast communication and players are expected to fulfill their role according their teammates best interest and understanding of the game.
What happens if the teammates disagree on how to play a role?
Well toxicity is likely happen and blaming eachother when people don't play how they expect them to play.

A lot of balance changes in Zero-K has been done to accommodate teamplay with randoms.
A free factory is one thing. Energy requirements for strider hub is another. And factory plates is a third.
+2 / -0