Author |
Topic: "Playing Arimaa: How to Think" (Read 6228 times) |
|
megajester
Forum Guru
Istanbul, Turkey
Gender: 
Posts: 710
|
 |
"Playing Arimaa: How to Think"
« on: Apr 28th, 2011, 10:17am » |
 Quote  Modify
|
This is an idea that I ran past Omar a while back but never got around to implementing, and I'd like some input from the community.
I think the e-mail I sent Omar explains the premise fairly well:
When I started getting into Arimaa, I'm sure that one of the most important factors was that I read the wikibook very early on. Without the wikibook I think it's very hard for newcomers to get a grip on the game.
I'm trying to teach my wife to play Arimaa, bless her. Instead of silently playing against each other (which would be frustrating for her because all she would do is lose pieces without knowing why) we chat about our moves as we play them. What tends to happen is we both move our elephant and maybe a horse out a bit. She goes to make a move that would leave a piece hanging, and I show her why it's not a good idea. Then there's silence and she says "So what am I supposed to do now then?" It was similar with my dad.
In a way I think the Arimaa learning curve has a slight "spike" immediately after you've learnt the moves that you don't get with chess. Chess is a very tactical game, so new players can very quickly find something to do even if it's not strategically sound. A beginner starts by getting his pieces out, and then tries to capture stuff. He makes lots of mistakes and doesn't understand what they are, but at least the game is moving along. However in Arimaa once you've exhausted the tactical possibilities you're into the murky world of strategy, which is even murkier than in chess, and an uninformed beginner can be at a loss as to what to do next. This "murkiness" is of course why we're all here, so it's a good thing. But I think it would be nice if we could give beginners just a little boost to get them started.
(Something else that adds to the learning curve is the fact that Arimaa is a highly reversible game, by which I mean that positions can be reversed much more easily than in chess. This is another sticking point for beginners, because if neither player feels particularly adventurous the opening can theoretically go on forever, which is not the case with chess.)
At the time Omar agreed with me that a "Arimaa Crash Course" would be helpful. We talked about a sequence of videos. The first one could be "How to Play Arimaa" from the homepage, and the subsequent videos would help beginners out with getting to grips with basic tactics and deeper strategic concepts.
However after thinking some more I can see some potential downsides to this idea. A course of videos would be an awful lot of work to produce, and would likely just end up being a rehash of the information already available in the wikibook. In any case it would be too long and involved to give my wife what she's looking for when she asks me "So what do I do now?"
We just need a "bridge", a quick video, to take a beginner who only knows the basic rules and has no idea what he's doing, and make him feel like he's got the gist of Arimaa and can start playing some meaningful games.
I've only just got straight in my head exactly what I think this video might be like. Basically, it entails helping a beginner to formulate a thought process, by giving him a list of simple questions. The difficulty here is taking concepts (such as camel hostages) and boiling them down into simple questions you can ask yourself. The wikibook already explains concepts well enough. What I want is a few simple questions to give beginners that kernel, to point them in the right direction towards discovering those concepts.
The problem is I'm not sure exactly what these questions should be. So I'd like your help. What thought processes helped the veterans of Arimaa to find the concepts we all swear by now? What questions do you ask yourself when it's your move?
I'll provide a blueprint to get us started. Imagine something like the following.
Title: "How to Think in Arimaa"
"So you've learnt the rules. You don't just want to play Arimaa, you want to play Arimaa well. You want to feel like you're learning something every time you play.
"Well, if you want to get good, there are N simple questions you have to ask yourself on every move.
"Question 1: [Insert question here. The narrator goes into more detail about what the question means, and provides one or two examples from past games]. If the answer's no, it's time to ask...
"Question 2: ... "
...and so on and so forth.
I'll write down a few questions so you get the idea, and then you can correct me and suggest extra questions from there.
1. "Can I or my opponent get a rabbit to goal?" (Include examples of 1- and 2-move combinations.)
2. "Can I or my opponent threaten any pieces?" (Include examples of 1- and 2-move combinations.)
Deep Strategies
3. "Which is the strongest free piece?" (Explain the concept, why it's a good idea to have the strongest free piece, and give a couple of examples of how hostages and frames can help you to get it.)
4. "Who has control of each of the traps?" (Explain trap control etc.)
5. ... ?
|
|
 IP Logged |
|
|
|
ocmiente
Forum Guru
Arimaa player #3996
Gender:
Posts: 194
|
|
Re: "Playing Arimaa: How to Think"
« Reply #1 on: Apr 28th, 2011, 1:43pm » |
Quote Modify
|
I like the idea of a checklist when making a move. In fact, I made a few different versions of my own during the World League last year to try to avoid doing stupid stuff - and it worked sometimes. For next year's championship, I'm adding 'is the opponent's horse hanging?' question near the top since I missed that in two of my games.
I would like to think that I go through something like the following list each move, applying each question to myself and my opponent (e.g. do I have an immediate goal? does my opponent have an immediate goal?) before and after my planned move, and hopefully before actually making the move. In practice, I often don't go through the list and end up paying for it.
1. immediate goal, or does it look like a forced goal?
2. capture?
3. frame (or how can I/my opponent break one)?
4. hostage?
5. threaten capture?
6. blockade?
7. swarm?
8. rotate out piece?
9. block? (phalanx or other)?
10. rabbits advanced appropriately?
11. balance, and how are the pieces working together?
12. long term plan?
One of the problems I have with this is that there is so much to think about, it has been difficult to use such a list every time when under time pressure. On the other hand, I would be very interested to know if other people make such lists and what is on them.
|
|
IP Logged |
|
|
|
Swynndla
Forum Guru
Arimaa player #1821
Posts: 235
|
|
Re: "Playing Arimaa: How to Think"
« Reply #2 on: Apr 29th, 2011, 12:08am » |
Quote Modify
|
It seems to me (just my opinion) that those lists would be good for an intermediate player who has already past the stage of reading the wiki (and therefore wouldn't be good as a bridge between learning the rules and the wiki).
I'd like to see something like a game where both elephants are dragging pieces to their own traps, and then one side gets the upper hand by controlling an enemy trap, but then blunders & loses when the the other side gets a rabbit through & wins (something like that)?
|
| « Last Edit: Apr 29th, 2011, 12:08am by Swynndla » |
IP Logged |
|
|
|
ginrunner
Forum Guru
Arimaa player #5449
Gender:
Posts: 163
|
|
Re: "Playing Arimaa: How to Think"
« Reply #3 on: Apr 29th, 2011, 12:18am » |
Quote Modify
|
I watched the commented games by C&G and Fritz
|
|
IP Logged |
|
|
|
megajester
Forum Guru
Istanbul, Turkey
Gender:
Posts: 710
|
|
Re: "Playing Arimaa: How to Think"
« Reply #4 on: Apr 29th, 2011, 12:38am » |
Quote Modify
|
on Apr 28th, 2011, 1:43pm, ocmiente wrote:I like the idea of a checklist when making a move. In fact, I made a few different versions of my own during the World League last year to try to avoid doing stupid stuff - and it worked sometimes. For next year's championship, I'm adding 'is the opponent's horse hanging?' question near the top since I missed that in two of my games.
I would like to think that I go through something like the following list each move, applying each question to myself and my opponent (e.g. do I have an immediate goal? does my opponent have an immediate goal?) before and after my planned move, and hopefully before actually making the move. In practice, I often don't go through the list and end up paying for it.
1. immediate goal, or does it look like a forced goal?
2. capture?
3. frame (or how can I/my opponent break one)?
4. hostage?
5. threaten capture?
6. blockade?
7. swarm?
8. rotate out piece?
9. block? (phalanx or other)?
10. rabbits advanced appropriately?
11. balance, and how are the pieces working together?
12. long term plan?
One of the problems I have with this is that there is so much to think about, it has been difficult to use such a list every time when under time pressure. On the other hand, I would be very interested to know if other people make such lists and what is on them.
|
|
Thanks ocmiente that's really helpful.
As a player who's fairly advanced, all these questions reflect the understanding you already have. What I'm trying to do is boil it all down to the questions that help a beginner to reach that understanding. I want to be able to give a beginner maybe 3-5 easy-to-remember questions he can ask himself with every move, that will hold true throughout his Arimaa career as he discovers the deeper concepts. A easy way for beginner to answer the question "So what do I do now?"
So for example, your point 1 fits with my question 1 "Can I or my opponent get a rabbit to goal?" In the video perhaps we could hint at your point 10, that long term you need to think about making goal threats.
My question 2 "Can I or my opponent threaten any pieces?" would cover your points 2 and 5.
My question 3 "Which is the strongest free piece?" could be used to introduce your points 3, 4, 6 and 8. Even if we don't mention them all in the video, telling a beginner to always ask himself the question "Which is the strongest free piece?" will help him towards discovering those concepts.
My question 4 "Who has control of each of the traps?" would similarly cover 7, 9 and 11.
And so on. Do you see what I mean?
|
|
IP Logged |
|
|
|
UruramTururam
Forum Guru
Arimaa player #2537
Gender:
Posts: 319
|
|
Re: "Playing Arimaa: How to Think"
« Reply #5 on: Apr 29th, 2011, 2:22am » |
Quote Modify
|
Heh, the checklist is great. I've printed it out! 
|
|
IP Logged |
Caffa et bucella per attactionem corporum venit ad stomachum meum.
BGG Arimaa badges - get your own one!
|
|
|
ocmiente
Forum Guru
Arimaa player #3996
Gender:
Posts: 194
|
|
Re: "Playing Arimaa: How to Think"
« Reply #6 on: Apr 29th, 2011, 12:26pm » |
Quote Modify
|
on Apr 29th, 2011, 12:38am, megajester wrote:| ... What I'm trying to do is boil it all down to the questions that help a beginner to reach that understanding. I want to be able to give a beginner maybe 3-5 easy-to-remember questions he can ask himself with every move, that will hold true throughout his Arimaa career as he discovers the deeper concepts. A easy way for beginner to answer the question "So what do I do now?" |
|
Yes, I think I understand what you mean completely. When I first played Arimaa, I gave up on it for a while because I couldn't grasp a lot of the basic concepts - and I wasn't inclined to ask a lot of questions on the site to get the basic information needed. It wasn't until Fritz' book came out that I was able to begin to understand the game.
Unfortunately, telling someone to read a book to learn how to play an abstract game is a lot to ask of most people who just want a casual experience.
So, I hope you're able to succeed in condensing the basic concepts down to just a few things. Something like that would help a lot.
With respect to my list, I left out at least one major thing: Frozen?
|
|
IP Logged |
|
|
|
omar
Forum Guru
Arimaa player #2
Gender:
Posts: 1003
|
|
Re: "Playing Arimaa: How to Think"
« Reply #7 on: Apr 29th, 2011, 1:40pm » |
Quote Modify
|
Great idea Joel. A checklist is a good way to get better at Arimaa. Initially it might seem a bit tidious, but later on it becomes second nature.
Explaining the rules of Arimaa to a new player is like explaining how the steering wheel, brake and accelerator in a car work to a new driver. They need to know that, but it doesn't tell them how to drive. When you are first learning to drive you need to consciously think about stuff like staying in your lane; using your turn indicator; looking out for traffic signs, etc. But after years of driving it becomes second nature and you probably can't even enumerate all the things that you do automatically while driving.
So with Arimaa also I think just lots of experience is the only way to let your unconscious mind automate playing well, but when starting out having a list to consciously review will certainly help.
Here is a file that I had started with tips for beginners. It's a bit more than a checklist, but you might find parts of it useful.
http://arimaa.com/arimaa/learn/tips.html
It's last modified date is Feb 14, 2004  I don't know why I didn't finish this and link it somewhere.
|
| « Last Edit: Apr 29th, 2011, 1:44pm by omar » |
IP Logged |
|
|
|
Swynndla
Forum Guru
Arimaa player #1821
Posts: 235
|
|
Re: "Playing Arimaa: How to Think"
« Reply #8 on: Apr 29th, 2011, 5:14pm » |
Quote Modify
|
I taught my boys how to play arimaa, and I think it was important that I didn't say too often "don't do that because this" and "don't do that because of the other", otherwise they'd sit there and say "well what can I do?", where I'd prefer them to try all sorts of stuff and find out a certain amount from experience (although I point out what they might have done differently after they've moved if they missed a really good move). So at their level, advancing rabbits early works (when they play each other), as it often leads to a win, and I'm absolutely fine with that.
It's like playing chess - I'm not that good, but I certainly have fun trying all sorts of things when I play others at my level. I also know at the same time that experts would say my moves are wrong (and that I should be doing safe, solid moves), but when I play opponents at my level, it doesn't matter and it's a lot of fun (and therefore not wrong).
So when beginners play arimaa against other beginners, the long-term strategy of slowly pulling an h-file rabbit shouldn't even be mentioned (for example).
I remember when chessandgo first started playing arimaa, and when I played him I was trying to "school" him (if you can believe that), and I was saying that him advancing his E, M & H at the same time early in the game wasn't a good idea. I was the one that got schooled! 
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
Re: "Playing Arimaa: How to Think"
« Reply #9 on: Apr 29th, 2011, 5:53pm » |
Quote Modify
|
My hunch is that there are a number of different overlapping frustrations that beginners have. It is very easy as a teacher to be over-zealous in trying to communicate everything that they need to know. I doubt that the best answer to the question, "So what should I be trying to do?" is to give a checklist of everything important to think about. Information overload isn't the best cure for confusion.
The key might be to understand the specific nature of the question a little better. Why is a beginner having a problem generating a move? Certainly part of what I think about on every move is whether I can take a piece or score a goal or protect my hanging piece or defend a goal. However, if any of those situations were present, the beginner would probably be asking, "How can I get that rabbit to goal?" or, "Can I capture that piece?", i.e. they would be wondering how to achieve an identifiable objective instead of wondering what the immediate objective is.
At one point in my writing about Arimaa, I went to great lengths to explain the "two threats" doctrine, i.e. that no matter what you try to achieve in Arimaa, the opposing elephant can stop it, so you need to be thinking, even as you make one threat, about a potential for a second one. The beauty of thinking in terms of "two threats" is that it heads off a second round of beginner frustration that will almost inevitably follow from a careless answer to the first question. If you tell a beginner that what they should be doing in the position at hand is to pull a rabbit, or to try to get a horse frame, or to swarm an opposing trap, etc., then the beginner will gamely try to do what you told them to do, find that the opposing elephant has stopped it, and then be left wondering, "Why did I do that? Did it help me in any way?". Quite possibly it didn't help them at all. For example they pulled a rabbit that turned into a goal threat, or framed a horse when the frame can be broken to create a swarm, or launched a swarm that got their own elephant pinned. None of the strategic objectives we toss around casually are worthwhile in themselves; all of them can backfire.
What a beginner needs to understand after mastering basic tactics, and is indeed capable of understanding after mastering basic tactics, is that the opposing elephant can defend against any one threat, and the great majority of strategy is an attempt to work around that limitation. I don't frame a horse for the sake of the frame; I make it with the intent to rotate out my elephant to make a second threat the opposing elephant can't defend, and if the second half isn't feasible the first half is a bad idea.
In my annotation of my 2005 Postal Mixer game against Omar, I laid great stress on the two-threats doctrine. Indeed, I was so successful communicating the idea to IdahoEv, a relative beginner at the time, that he concluded Arimaa was pretty simple after all, i.e. that there wasn't very much to Arimaa strategy, and wanted to complicate the rules of Arimaa to make it less simple to comprehend. That's the sort of feedback that lets me know I have been clear as a teacher!
As I think now about how I teach and talk about Arimaa, I realize that I have been assuming that the two-threats doctrine is second nature to everyone. Like Omar says, it's like turning on the blinker before making a lane change; it is automatic. But it is totally futile trying to explain swarms or rabbit pulls to someone who doesn't get the importance of the two-threats doctrine.
I submit, therefore, that this is the core of answering megajester's concern about the bump in the learning curve. Suppose a beginner finds himself knowing basic capture and basic goal, but not knowing any strategy. How to start to bridge that gap? Perhaps the very next concept is to show how elephant defense means that no progress can happen without making two threats. Then when you start talking about hostages, frames, etc., it is in the context of creating two threats, i.e. the learner will know why these other strategic markers are important. Hopefully, if we explain it clearly enough, they will not think Arimaa is frustratingly opaque, but rather join with IdahoEv in saying that Arimaa strategy is fundamentally very simple.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
Re: "Playing Arimaa: How to Think"
« Reply #10 on: Apr 29th, 2011, 6:09pm » |
Quote Modify
|
on Apr 29th, 2011, 5:14pm, Swynndla wrote:| I taught my boys how to play arimaa, and I think it was important that I didn't say too often "don't do that because this" and "don't do that because of the other", otherwise they'd sit there and say "well what can I do?" |
|
Ah, yes, I didn't consider to what extent fear of making a bad move might be the root of the problem. You definitely don't want to paralyze the people you are teaching with criticism.
Quote:| It's like playing chess - I'm not that good, but I certainly have fun trying all sorts of things when I play others at my level. I also know at the same time that experts would say my moves are wrong (and that I should be doing safe, solid moves), but when I play opponents at my level, it doesn't matter and it's a lot of fun (and therefore not wrong). |
|
Yes, I quite agree with this attitude. It is relatively bad pedagogy to reject someone's move out of hand, because they might take it to mean that they shouldn't be trying to accomplish what they are trying to accomplish. At worst the teacher should praise the intent of the move while expressing doubt about the efficacy of the move. That way the learner will know that he at least has a grasp on what strategy is about. He can then say, "OK, I know what I am trying to do, but am not going about it the best way."
I was answering a different question, namely how to help a beginner who can't think of anything to do next. I believe that this confusion is a real problem that needs to be addressed. The defensive nature of Arimaa makes it more likely that beginners will have some frustration like this. But also we teachers of Arimaa should be very aware that the complaint, "I can't think of anything to do," could very well be code for, "I can't think of anything to do that you won't criticize, so I am afraid to move."
|
|
IP Logged |
|
|
|
megajester
Forum Guru
Istanbul, Turkey
Gender:
Posts: 710
|
|
Re: "Playing Arimaa: How to Think"
« Reply #11 on: Apr 30th, 2011, 5:36am » |
Quote Modify
|
on Apr 29th, 2011, 5:53pm, Fritzlein wrote:
What a beginner needs to understand after mastering basic tactics, and is indeed capable of understanding after mastering basic tactics, is that the opposing elephant can defend against any one threat, and the great majority of strategy is an attempt to work around that limitation. I don't frame a horse for the sake of the frame; I make it with the intent to rotate out my elephant to make a second threat the opposing elephant can't defend, and if the second half isn't feasible the first half is a bad idea.
[...]
I submit, therefore, that this is the core of answering megajester's concern about the bump in the learning curve. Suppose a beginner finds himself knowing basic capture and basic goal, but not knowing any strategy. How to start to bridge that gap? Perhaps the very next concept is to show how elephant defense means that no progress can happen without making two threats. Then when you start talking about hostages, frames, etc., it is in the context of creating two threats, i.e. the learner will know why these other strategic markers are important. Hopefully, if we explain it clearly enough, they will not think Arimaa is frustratingly opaque, but rather join with IdahoEv in saying that Arimaa strategy is fundamentally very simple. |
|
Bingo!
There's no other word for it.
I know you don't believe in giving beginners a long list of questions to learn, but I still feel like we can condense it all down into three questions.
Question 1: "Can I or my opponent get a rabbit to goal?"
Question 2: "Can I threaten any of my opponents pieces?"
Question 3: "What are my two threats?"
Of course the meat and potatoes are in question 3. As Fritz said, we have to make sure we explain it clearly. You could have a couple of examples, such as camel hostage+goal threat, and horse frame+swarm. You wouldn't go into detail about all the individual concepts. But still at the end of the video you'd be able to tell a beginner that he has all the basic tools he needs to discover the game.
What do you say, are we ready to start drafting a script?
on Apr 29th, 2011, 1:40pm, omar wrote:
Explaining the rules of Arimaa to a new player is like explaining how the steering wheel, brake and accelerator in a car work to a new driver. They need to know that, but it doesn't tell them how to drive. When you are first learning to drive you need to consciously think about stuff like staying in your lane; using your turn indicator; looking out for traffic signs, etc. But after years of driving it becomes second nature and you probably can't even enumerate all the things that you do automatically while driving.
|
|
I like this analogy. That tips page is also very nice. Perhaps this could be turned into a third video, "Tips and Tricks"?
|
| « Last Edit: Apr 30th, 2011, 5:39am by megajester » |
IP Logged |
|
|
|
Hippo
Forum Guru
Arimaa player #4450
Gender:
Posts: 883
|
|
Re: "Playing Arimaa: How to Think"
« Reply #12 on: Apr 30th, 2011, 10:43am » |
Quote Modify
|
I would add one more tought ... sometimes it is not question of where do I have second threat, but who has stronger threat. Will I have time to apply my threat or I will be forced to start defending stronger threat.
Interesting question is how many steps do I need to apply the threat and how many steps opponent needs. If number of defense steps required plus number of attacking steps required is higher than 4 player could not do both (interference could spoil this computation).
... in my 2010WC game with naveed, I have sacrified cat for free not thinking that way ... naveed had 4 step threats and must do 1 defense step. I have sacrified cat to be captured in 3 steps ...
So ability to create strong enough threat even when 1 step defense is possible often prevets 4 step threats to be applied. This should be taken into account when swinging wings with phant is considered.
OK ... this probably does not help with easy questions to ask ... so rather avoid racy positions till you become familiar with "threat step optimisations".
|
| « Last Edit: May 3rd, 2011, 6:09pm by Hippo » |
IP Logged |
|
|
|
megajester
Forum Guru
Istanbul, Turkey
Gender:
Posts: 710
|
|
Re: "Playing Arimaa: How to Think"
« Reply #13 on: Apr 30th, 2011, 2:23pm » |
Quote Modify
|
OK... so maybe we could generalise question 3 slightly, eg. "What are the second threats?" meaning both yours and your opponent's. Just like with the first two questions.
|
|
IP Logged |
|
|
|
megajester
Forum Guru
Istanbul, Turkey
Gender:
Posts: 710
|
|
Re: "Playing Arimaa: How to Think"
« Reply #14 on: May 2nd, 2011, 6:09am » |
Quote Modify
|
on Apr 30th, 2011, 2:23pm, megajester wrote:| OK... so maybe we could generalise question 3 slightly, eg. "What are the second threats?" meaning both yours and your opponent's. Just like with the first two questions. |
|
Or might it be clearer for a beginner if we say "How might we tie down each other's elephant?"
|
|
IP Logged |
|
|
|
|
Home
Help
Search
Members
Login
Register
Reply
Notify of replies
Send Topic
Print