Recently, I read something interesting about some differences between the ways that Novice chess players and Master chess players perceive and analyze game states on a chess board. I wonder if Shadow Era is a sufficiently complex enough game that many of the same concepts can be carried over. This article is my attempt to do that.
Experiments have been done where Novice players and Master players are exposed to “snapshots” of different chess games, for 5 seconds. The players are then asked to re-create the game state that they observed, from memory. The result was that, overwhelmingly, the Masters were able to very accurately re-create the correct game state (all the pieces in their correct positions), much more so than the Novice players. But it gets much deeper than that, and much more interesting.
When asked to re-create the chess boards, Novice players tended to make mistakes involving individual piece placements, whereas Masters had re-created their boards with entire GROUPS of pieces in the wrong positions. (For example, a cluster of four pieces were actually off by 1 position left, and 2 up...but their distance from each other on the board did not change; the whole group was moved as a whole.)
It showed that the Masters perceive the board in "chunks", whereas Novices see the board as made up of many localized pieces.
In the field of AI, there was an intuition that a computer could be programmed to beat all of the chess Masters. They assumed that all they would have to do is create a software that could "look ahead" so much further than humans can, make decision trees, consider all possible game paths, and then select the best move to win. The result - so far they haven't even come close.
The chess Masters, it was discovered, weren't really "looking ahead" any more than the Novice players were. Instead, any given change on a board state could cause Triggers, causing the player to scan for the best known sequences for how to play from there...
Novice players perceive the board as made up of distinct [chess] pieces, or in our case, individual cards. When Novice players are asked to describe a given board state, they may say something like this: “Player 1 is playing Majiya. Majiya
The Master, on the other hand, perceives the board in “chunks”, or larger groups of pieces and generalized concepts. A professional can describe that same board state as something like this: “Gwen dealt with the rush pretty well but needs to draw more cards at some point. Gwen currently has board control.” Notice that the Master is able to summarize what is happening on the board without referring to a single card by name. The individual "pieces" are ignored, and instead the master is attuned to the patterns involved.
Moving on, in the First Experiment, all the players were given “snapshots” from actual games that had been played. In a Second Experiment, there was a twist – none of the “snapshots” came from actual games; they were generated randomly. Pieces were placed onto the “snapshot” arbitrarily. This time, the Masters scored about the same as the Novice Players with regard to recalling the positions of the pieces on the board.
Over time, players become aware of the patterns of the game they are playing. Players begin to notice a pattern from T1 to T2 to T3… or T6-T7-T8. Players then take these patterns and will “chunk” those moves together. Meaning, players who become aware of the patterns of the early game can start to identify the strongest opening sequences, and those sequences they can now intuitively replicate in every game they play. T2 Puwen
The idea is that a Novice player is looking at each card as a separate “piece”. When it comes time to make a decision, they evaluate each and every piece, making sort of “decision trees” in their head, whether they know it or not. For example, when it comes time to sacrifice, the Novice player will look at each card and try to “look ahead” in the game to find out which cards will be useful in later turns, and which cards are expendable. The Master, however, does not see bad moves. It’s like the bad moves don’t even register for them, unless they make an effort to spot them out. The Master has an affinity for the patterns of the game, and when they look at their options, they intuitively scan the strongest possible chunks of plays, drawing from their memory and experience.
Human memory is best suited to recall 6-7 digits at a time. If you want to memorize 60 digits of Pi, you would have a very hard time by trying to memorize all 60 at once. It is recommended to learn them 6-7 digits at a time. Once you memorize that much, it gets “chunked” into one “thing.” Then you learn the next 6-7 digits, and that gets “chunked” into another thing. Then you can take seven chunks/things, and chunk those together. Each “chunk” has a trigger that efficiently allows you to recall the sequence that comes after it. (The trigger is the first number in the sequence, and it makes you recall the Pattern that follows.)
So, I sit down to play. The cards are being dealt, and they are coming in, one by one. There are certain cards that do not really register for me, I seem to think nothing of them. Other cards are Triggers. I see an Honored Dead
Puwen
Another type of Trigger Card is one that is a problematic card against your deck, triggering the sequence for how to deal with it. For example, Armored Sandworm
So, we learned that is possible to perceive a system on a higher/chunked level, and completely ignore the lower level/separate cards. For example in a sentence, you can grasp the meaning of the words without ever referring to the meaning of each individual letter in the word. (As you read this very sentence, you - as a Master of language - are most likely reading Words, which are chunks of letters, but no particular individual letter strikes you as being important to understanding the sentence. You can extract the meaning of this sentence without referring to: The writer used the letter T, the letter h...).
In the same way, one can perceive the board as clusters of patterns (opening sequences, retaking-the-board sequences, matchup sequences, sequences that transition you from the early to mid game, etc.) So that a professional can describe the board state without ever naming a single card on the board. "X has board control; Y has card advantage" - these statements are meaningful, and yet they ignore the lower level completely. And generally the masters were far superior in this regard, but about equal when it came to lower level descriptions.
So, what does this all mean? To me, this concept suggets two possible paths to becoming a Master of Shadow Era:
1) Look at every card and analyze it. Make a decision tree for each card in your hand and use reason to try to determine what is the best play.
2) Play lots of games.
The concepts we have been considering seem to suggest that, in a game such as Shadow Era, option #2 is where it's at. Option 1 deals with innate skills that some people have - the ability to use reason, the ability to make good decisions, the ability to follow decision trees to greater and greater lengths, etc. Now, these are critical skills to have, to be sure. However, my guess is that those skills alone cannot differentiate a Novice from a Master. There may be someone who is new to Shadow Era, but who has great skills in strategic thinking and reasoning; and a Master who has been playing Shadow Era for a long time will not necessarily have any greater reasoning power than a Novice.
On the other hand, the differentiating factor is that the Master has played lots of games, whereas the Novice is new to the game. In this sense, the Novice can only see the board from the lower level perspective - they are looking at every card, perhaps even turning the physical cards over to read what they do in-game, etc. The Master has access to a higher level of thinking, however - the Master knows what patterns of play will generally come up, and can exploit this. The Master can tell you what is going on, or who is going to win, without referencing a single card on the board. The Novice player simply will not have this advantage until they play lots of games.
So, my conclusion is that for a game like Shadow Era, the best way to get better is to play lots of games. Be aware of the patterns of play. Get a feel for the game; its speed, its stresses and challenges, its nuances. Not really a novel concept, but I personally enjoyed thinking about these things, and I hope you did as well!
- 1ndeed
Reply With Quote
Originally Posted by Atomzed
Bookmarks