|
||||
|
Title: State of the Challenge 2009 Post by Fritzlein on Mar 2nd, 2009, 11:49am Omar's comments in the 2009 Computer Championship thread seem like a good excuse to kick off the 2009 Arimaa Challenge thread with some speculation about the current man vs. machine balance of power. I propose that from now on we have a yearly "State of the Challenge" discussion. It will be a fun tradition to formalize what has been happening informally over the years. For example, I wish I could find 99of9's comments from 2004 about how he rated the chances of Omar paying out the $10,000 before 2020 to be significantly more than 50%. Those odds seem to have dropped yearly with Bomb ruling the roost, but have they bounced back significantly this year? Here's what Omar said: on 03/01/09 at 21:08:56, omar wrote:
I mostly agree with Omar's analysis. We have gotten overconfident from using Bomb as a nearly-stationary benchmark while we humans gradually improved. I'd call it four years though, not six; Bomb was definitely still improving in early 2005. After that Bomb's slight improvement came only from running on a faster chip, and even that improvement was minimized because Bomb wasn't parallelized. OpFor's incredible showing in the 2008 Postal Mixer was our first wakeup call. In hindsight I think we can rule out the hypothesis that OpFor's wins were due to simple overconfidence on our part. Ron Weasley had already sounded the alarm early in the tournament when several of OpFor's positions were still unclear, and which humans should still have won if paying attention was the only requirement. But there was something deeper going on; OpFor actually played quite well in "wild" positions with multiple traps contested. I definitely think that we will have to improve our strategic jugdment in multi-trap positions, and hone our tactics at trap control fights. Nevertheless, for the present I don't believe that humans will have to avoid certain types of positions in order to stay ahead of computers. It would be a severe disadvantage for us to have to play tame positions only, because then the struggle would mostly be about making the position wild or preventing it from becoming so. But the fact that a position is wild does not yet mean that a bot will win from either side. For example, a capture race that turns into a goal race definitely fits into the loose definition of tactical that Omar presents, but winning such a race against a bot is entirely possible, as blue22 has demonstrated repeatedly. If blue22 can chop to goal in about four turns in the northeast while a bot can chop to goal in about six turns in the southwest, it doesn't matter if the bot is a little better than him at racing; blue22 is still going to win the race. My point is that strategic judgment stands at the beginning of every tactical race. I don't think the bots can beat us with a mantra of "always complicate", because very often the move that makes the position wilder also concedes some disadvantage. For example, I suspect that part of the new strength of clueless is that it is eager to enter races rather than remaining tame (like Bomb) when the elephants are deadlocked. But clueless totally lacks the ability to distinguish a favorable race from an unfavorable one. Take my recent win over clueless in game 97957 (http://arimaa.com/arimaa/gameroom/comments.cgi?gid=97957). I deadlocked our elephants early with a smother, but the smother isn't worth much since it takes my elephant to hold in clueless' elephant. There's hardly any hope for me to rotate my elephant out of the smother, particularly since clueless is now smart enough not to advance rabbits into that situation. So what did clueless do? It activated its camel. As a general principle, this is smart for clueless to do. When elephants are deadlocked, you should get more free with your camel. But in this particular case the smart general principle was executed stupidly. The camel hostage that clueless offered to me was worth significantly more than the elephant blockade I had to give up to freeze the camel. In other words, clueless did exactly what Omar is suggesting bots need to do, by turning a strategic situation into a tactical one. Clueless didn't sit back and let me exploit my elephant smother, it quickly complicated the game. But the ensuing capture race is one I am going to win, even though clueless is better than me tactically. I'm going to win that race because my elephant is next to his camel, while his elephant is only next to my horses. The fact that I won a camel for a horse has little to do with great tactics on my part during the capture race; it was mostly about bad strategic judgment from clueless to offer me such a favorable race, and good strategic judgment from me to immediately switch gears and ditch the smother for something better. The 2009 versions of clueless, OpFor, and GnoBot have definitely raised the bar, and further such improvements in the years to come will make the Arimaa Challenge interesting again. We humans have to keep learning to stay ahead. I do not, however, believe we are at the point that we can beat bots only in quiet positions. Our learning does not have to be focused on how to prevent complications from ever occurring. On the contrary, I predict we still have years of being able to punish bots that complicate at all costs, by learning how to snap off their attacks and make them pay for being overly aggressive. |
||||
|
Title: Re: 2009 Arimaa Challenge Post by omar on Mar 3rd, 2009, 1:04pm I really like the idea of having a yearly "State of the Challenge" discussion. Maybe we should rename the thread similarly so that the current name can be used for discussion of the actual challenge match games, announcement of defenders, etc. Also makes searching for these discussions easier. I am certainly not a good tactical player and perhaps I am using my own weakness as a basis for predicting that human players will want to steer the game into pro strategy positions. Players like Fritzlein and chessandgo have an amazing ability to handle tactical positions. Perhaps as the bots become stronger such players will continue to learn from the bots and stay ahead. However, players like me, who hardly calculate more than a ply or two on just a few selected moves will get left behind as the bots continue to get better tactically. Fo me being able to steer the game into pro strategy positions will become essential as the bots get better. |
||||
|
Title: Re: 2009 Arimaa Challenge Post by Fritzlein on Mar 3rd, 2009, 4:53pm on 03/03/09 at 13:04:46, omar wrote:
Let's wait to rename the thread until at least one other person has chimed in with an opinion on the State of the Challenge. ::) Quote:
What if CeeJay had said in 2005 that the only way to stay ahead of the bots is to find a trick like bait-and-tackle? Sure, bait-and-tackle might have been the only way he could reliably beat Bomb, but that didn't mean it was the only avenue for humanity to explore. And sure enough, lots of players who hadn't even discovered Arimaa at that point are now able to reliably beat Bomb without the bait-and-tackle. I think your pro-strategy/pro-tactics distinction is not only short-changing the subtlety of Arimaa, it is also short-changing your ability to learn strategies from other players. You shouldn't suppose that the next method for staying ahead of the bots someone else comes up with will be beyond your ability to execute. I suspect that over the next few years you are more likely to broaden your arsenal of strategies you use against bots than you are to shrink your arsenal into trying to force the game into a small handful of position types. |
||||
|
Title: Re: 2009 Arimaa Challenge Post by Janzert on Mar 3rd, 2009, 8:24pm Last year at this time I thought computers had about a 10% chance of winning. But with the performance of OpFor in the postal mixer and then the great leaps Clueless and Gnobot have made I'm hovering up around 50% now. Janzert |
||||
|
Title: Re: 2009 Arimaa Challenge Post by 99of9 on Mar 3rd, 2009, 9:16pm on 03/02/09 at 11:49:51, Fritzlein wrote:
Maybe you're referring to the forum post linked below. Your memory has been kind to my predictions! Re-reading this thread makes entertaining reading, there are now-humorous prophecies from a number of people. http://arimaa.com/arimaa/forum/cgi/YaBB.cgi?board=talk;action=display;num=1128966698;start= Nevertheless, I'm still an optimist, and certainly consider the 2020 chances nearer 50% than 10%. |
||||
|
Title: Re: 2009 Arimaa Challenge Post by Fritzlein on Mar 4th, 2009, 6:38am Thanks for the link, 99of9. It is indeed amusing to go back and read our old predictions from time to time. I wanted to expand on something that occurred to me in chat with aaaa yesterday. Suppose that we had a good way to measure true skill at Arimaa, something like the game room ratings but more accurate. Suppose that we plotted the gap in true skill between the top human and the top bot as a function of time. What would we expect the graph to look like? The image floating in my mind is something like the graph of f(x) = log(x+1) - x/10. The gap started near zero, and has risen dramatically, but at some point our advantage will level out, and after that it will get smaller every year, almost linearly, until we lose to computers. After that computers will get so much better than us it won't even be funny. I wonder if the same sort of image isn't influencing other people's predictions as well. If you believe that there will be one turning point in the graph, then predicting the Challenge is a matter of reading when the turning point has come, and guessing how sharply the U-turn will happen. Omar's comments seem to be as good as saying that the tide has already turned, and from here on out humans are on the downward slope. I believe in the one-turning-point model as a rough approximation. However, it is not easy to predict the turning point within that model because the actual graph is not smooth when you look at it close up. Both human learning and bot improvement happen in a series of small leaps rather than in a smooth ascent. Plotted over a long enough period the increase in human playing strength and the increase in bot playing strength will look like smooth curves, but in the short run the graphs will be stair-stepped. Looking at bot strength plotted over the past four years we have at first a nearly flat graph of Bomb2005 being the benchmark and getting better only from running on a faster chip, followed by a spike in the last few months as clueless (and OpFor and GnoBot?) surged past Bomb. I admit that we probably got overconfident due to the flat part of the bot-strength graph, but by the same token it would be unreasonable panic to take the recent rate of improvement as the new normal. If you look how much bots have improved in the past four years rather than in the past four months, the graph is going to have a very different slope, and your long-term predictions will come out very differently. Because of the stair-step nature of bot improvements, I'm still claiming we haven't reached the turning point in the man-machine gap. Of course my prediction also assumes that human playing strength has not yet plateaued, but I already went into that in my previous posts in this thread. The human strength curve is probably even harder to predict accurately than the bot strength curve, so we'll just have to wait and see. In another four years 99of9 can link back to this post and we can have another good laugh. :) |
||||
|
Title: Re: 2009 Arimaa Challenge Post by omar on Mar 4th, 2009, 6:49am on 03/03/09 at 16:53:39, Fritzlein wrote:
Not at all. I can definitely learn new methods and strategies and in fact will need to because the bots become better at avoiding the ones that used to work. For example its harder to get clueless to fall for bait-and-tackle. Also I don't mean to classify all positions as either pro-tactic or pro-strategy. It's really a spectrum; I was just referring to the extremes. |
||||
|
Title: Re: 2009 Arimaa Challenge Post by Fritzlein on Mar 4th, 2009, 8:04am But do you think that the tide has turned and the human advantage will get smaller from this year forward? |
||||
|
Title: Re: 2009 Arimaa Challenge Post by Fritzlein on Mar 4th, 2009, 6:41pm OK, I need to backpedal a little. The game clueless vs GnoBot (http://arimaa.com/arimaa/gameroom/comments.cgi?gid=99023) in Round 3 of the Computer Championship was mind-blowing and scary. I think I might have lost that to either bot from either side. So maybe, yes, there are types of positions I'm going to have to steer clear of. |
||||
|
Title: Re: 2009 Arimaa Challenge Post by 99of9 on Mar 4th, 2009, 7:35pm Some specific estimates and predictions: 40% chance of the challenge falling by 2013 60% chance by 2020 In the year it falls, surprise will be an essential element of the bot's win. That is, after losing, humans will find areas of weakness that will enable them to win reasonably often. In the year it falls, the strongest human opponent will hang a piece during one of his/her losses. By 2013, at least one defender will have lost 0-3. If the challenge has not fallen by the 2018 challenge, developers will then join forces in a final push for the last two years. |
||||
|
Title: Re: 2009 Arimaa Challenge Post by Fritzlein on Mar 5th, 2009, 5:58am Nice predictions, 99of9. It takes guts to get specific; you inspire me to respond in kind. By mid-2011 (that is to say within two years of Arimaa's commercial release) there will be some new player (that is to say someone who doesn't even know how to play Arimaa at the time I am writing this) who is 200 Elo points better than chessandgo and I are today. S/he will achieve a game room rating of 2700 playing against only humans. Equivalently, in case there is more club play and less on-line play at that time, s/he will be able to win three of four tournament games against chessandgo and me, unless the two of us have similarly improved to keep up. By 2020 there will be at least one human player with a playing strength of 3000 on the current rating scale. |
||||
|
Title: Re: 2009 Arimaa Challenge Post by Adanac on Mar 5th, 2009, 11:50am on 03/05/09 at 05:58:21, Fritzlein wrote:
My predictions are similar to Fritzlein's. We have a tiny, tiny Arimaa player pool, with only 18 people on the entire planet willing to participate in the 2009 World Championship. Once the game gets marketed and our player pools grows exponentially we’ll have dozens of players with 2600+ ratings. And perhaps a few of those will reach the 3000 level. I’ll predict that 3 players will have 3000+ ability by 2015 (measured by today’s standards). The bot developers have a huge advantage at this moment in time because unlike chess or go where the ratio of active players to developers is perhaps 100000:1 or more, in Arimaa the ratio might be 10:1 (?). Once we recruit a deeper pool of human players, the gap between humans and bots will grow much larger. I’ll predict that the best Arimaa bot will become permanently better than the best human in 2028. I will estimate there’s a 10% chance that a bot developer will win the Arimaa Challenge within the next 5 years. Even if a bot developer pulls it off, so many super-geniuses will become hooked on Arimaa in the mid-2010s that humans will once again reign supreme and many years will elapse before bots close the gap again. |
||||
|
Title: Re: 2009 Arimaa Challenge Post by Fritzlein on Mar 5th, 2009, 12:51pm on 03/04/09 at 19:35:08, 99of9 wrote:
These prediction percentages translate into a declining per-year chance of the Challenge falling. By Bayes' Law, you are implying that if the Challenge hasn't fallen in 2009-2013, there is a 33% chance that it will fall in 2014-2020. Assuming that the per-year chance of the Challenge falling is constant in the first epoch and again constant in the second epoch, then you are saying there is a 9.7% chance for bots to win each year from 2009-2013, and a 5.6% chance for bots to win each year 2014-2020. Do you really feel that the per-year chance of the Challenge falling will decline over time? My hunch is that the per-year odds for the bots will decline for a few more years, but then get start to get better again each year as we approach 2020, because human learning will plateau at some point, although it hasn't yet. Adanac's scenario that bots might win in the short run and then be left in the dust by humanity the following decade is yet a third option that seems far from either my intuition or 99of9's. This thread is turning into a good State of the Challenge thread after all. I have re-titled my first post to give the thread a new name. |
||||
|
Title: Re: State of the Challenge 2009 Post by mistre on Mar 5th, 2009, 6:06pm Doesn't the percent chance of the challenge being met depend on the human opposition? Obviously if we have our three current and most active highest rated players (Fritzlein, Chessandgo, and Adanac) defending then we have a much better chance then if someone like me or a lower rated player is called to defend (I played one match last year). If we can keep the top 3 players active and interested in always defending (considering also that there may be some new players out there that could get close to a 3000 rating, at least 2700 does not seem out of the question once the user base expands) then I think the chances are pretty good that we can make it to 2020 without losing the challenge. On the other hand, if Omar insists on spacing out the challengers and letting some lower rated players assist in defending, then I swing the favor back to the bots. I have yet to play the latest versions of Clueless or Gnobot and wonder if I would even be 50-50 with them. |
||||
|
Title: Re: State of the Challenge 2009 Post by Fritzlein on Mar 5th, 2009, 6:47pm on 03/05/09 at 18:06:04, mistre wrote:
Spreading out the Challenge defense to lots of players is not a requirement; it is a gesture of confidence similar to offering a piece handicap when you expect to win anyway. My hunch is that if the Challenge starts to seem in jeopardy, not only will the handicaps disappear, the names of the defenders will get a lot more predictable. |
||||
|
Title: Re: State of the Challenge 2009 Post by Oystein on Mar 6th, 2009, 6:10am I predict a breakthrough will occur and the challenge will fall around 2015. |
||||
|
Title: Re: 2009 Arimaa Challenge Post by Fritzlein on Mar 6th, 2009, 10:27am on 03/04/09 at 19:35:08, 99of9 wrote:
I forgot to respond to the element of surprise. That's an interesting point because Kasparov was definitely surprised by Deep Blue's ability to make progress in a closed position, so surprised that he (A) resigned in disgust when he could have salvaged a draw, and (B) accused the Deep Blue team of cheating by having a human give advice. On the other hand, although we know Kasparov was surprised, we don't know whether he could have recovered enough to win a longer match. I expect he could have come back in 1997, but subsequent man-vs-machine matches showed that Deep Blue's win was no fluke. By the time software can win the Arimaa Challenge, it should be mature enough to have no obviously exploitable flaws, and therefore will not benefit much from surprise. If some bot can win with a surprise trick, then some other bot will shortly thereafter be able to permanently win with no surprises. In the grand scheme of things it doesn't matter much whether the Challenge falls a year earlier or a year later if the trend is relentless and clear, as it is in chess. To an individual developer, however, it makes a world of difference whether he wins it this year or someone else wins it next year. When it looks like two or three bots each have a reasonable chance to win the Challenge, the collaborative information sharing among developers that has ruled so far will probably give way to secrecy. Also, if I'm still playing Arimaa at that time and I am still good enough to be a Challenge defender, I will probably become completely partisan to human players, and therefore stop pointing out to developers when I spot a bot weakness. Human players might start keeping secrets up their sleeves as well. |
||||
|
Title: Re: 2009 Arimaa Challenge Post by 99of9 on Mar 6th, 2009, 5:08pm on 03/06/09 at 10:27:35, Fritzlein wrote:
What I mean is, I expect that there *will* be easily exploitable flaws, but they won't be obvious until humans have had time to play a few games after the challenge. Once they are found, most humans will be able to execute the exploit. |
||||
|
Title: Re: 2009 Arimaa Challenge Post by omar on Mar 6th, 2009, 5:19pm on 03/04/09 at 08:04:22, Fritzlein wrote:
Your question sounds so tempting to answer, but when I think about all the factors that come into play it is mind boggling and it will be incredibly hard to predict what that curve will look like. My short answer would be no, because there is still a lot of room for humans to get better. But its fun to think about the factors that are involved. For the bots we can certainly expect continued improvements in the immediate future, but the rate of improvement might change at various times. This is because the source of increased performance in bots are either software advances or hardware advances. Today Bomb will be playing a decisive game and could get knocked out of the tournament if it loses. I think this may get Fotland to improve Bomb to at least make it multi-threaded and perhaps add some knowledge about the game. Bomb sort of reminds me of the hare in 'The Tortoise and the Hare' fable :-) But the active developers will surely want to keep the lead and so the international bot race should really start to heat up. I think there are still a lot of tweaks that the developers can make before it starts getting to the point of questionable improvements. Perhaps in about 5 years the evaluation code for the bots will get to the point where it is just too many factors for a human to tweak properly. At that time the bots will hopefully hit the end of the conventional tree search software improvement ramp and have to start looking for a paradigm shift (a.k.a. the breakthrough). Interestingly enough there is an editorial in the current issue of the ICGA journal title "GO FOR A BREAKTHROUGH" that talks about how the Monte-Carlo search method has emerged as a significant breakthrough in Go. Also mentions how David Fotland made a new version of his program based on this and jumped to the head of the pack again. The hare woke up, but it wasn't too late :-) In addition to the software improvement ramp the bots are currently also riding the hardware improvement ramp. However I think the slope of this ramp has decreased in the past 6 years. Prior to 2002 the processors were getting faster each year due to both faster clock cycles and greater transistor density (i.e. more transistors per chip). Lately the clock cycles have not been increasing as much so the gains are now mostly due to increasing number of transistors per chip and reorganizing the chips into multiple CPU cores. We are nearing a time when another major paradigm shift is needed within the next 10 years to keep the ever increasing computation/dollar ramp going. Maybe we will shift from an era of transistors to an era of memristors http://en.wikipedia.org/wiki/Memristor. What actually happens in this area will have an effect on the bot performance curve. For humans also I think we can certainly expect continued improvements in the immediate future. Here also the rate of improvement might change at various times. The key sources that fuel the level of human performance I think are: the total number of active players, the recorded knowledge about the game and the ability for players to make a living from the game. Without these even a person who could potentially reach an astronomical Arimaa rating would hit various obstacles that would hinder or limit their progress. A large playing population increases the chance that some of the very best human players for the game are actually playing the game. It also increases the number of high quality games being played everyday. This helps to push the ability of the top players beyond the current level. Just having a large playing population is not enough. The new players need ways to quickly come up to speed without having to reinvent everything. As we document and share the knowledge that has been gained about the game, it will allow new talented players to get to the cutting edge of the game more quickly. But once they get there even these talented players with all the current knowledge about the game available to them will hit a different kind of limit. The limit of time itself. It is not that they don't have the potential to get better, but rather that they can't devote their full time and attention to getting better. Also it just takes time to learn, practice and get better; there is no short cut. To allow players to get past this barrier Arimaa will need to become a game where people can make a living playing it. Fortunately there is a lot of room to improve before we get to that point. Beyond this I think we will hit the limit of human thinking capacity which will be insurmountable without drugs or other unnatural aids. Once we get to that level then for sure the human advantage will get smaller each year until the bots eventually pass us up by just riding the improvements in hardware. Hopefully that won't happen before 2020 with a breakthrough in software. But who knows how high this limit of human thinking capacity is. In terms of elo ratings I hope its about 3500. So if the bots are rated 2000 now and pack on 100 elo points per year, it would take 15 years for them to get to this level. But how fast will it take for humans to reach this level. My guess is probably at least 15 years with a large active playing population and our current best ratings being around 2500. The Arimaa challenge is not just a challenge for the bots programmers, but also a challenge for the human players to try and stay ahead. It will be interesting to see how it plays out. |
||||
|
Title: Re: 2009 Arimaa Challenge Post by Fritzlein on Mar 6th, 2009, 8:06pm on 03/06/09 at 17:19:25, omar wrote:
These are the predictions of a true optimist: Both humans and computer programs will reach astronomical heights of Arimaa skill! I'm not the optimist I once was, but I still always hope the optimists are right in their predictions. |
||||
|
Title: Re: State of the Challenge 2009 Post by omar on Mar 8th, 2009, 7:03am I've always thought you were more optimistic than me :-) |
||||
|
Title: Re: State of the Challenge 2009 Post by Fritzlein on Mar 8th, 2009, 7:18pm on 03/08/09 at 07:03:46, omar wrote:
In general perhaps, but you far outstrip my optimism for the pace of improvement of Arimaa players, bot and human alike. And in your expectations for the sale of Arimaa sets and books, your optimism is off the charts. :) |
||||
|
Title: Re: State of the Challenge 2009 Post by gatsby on Mar 9th, 2009, 3:50am Keeping in mind that hopefully arimaa is harder for computers to play than chess, I expect a large arimaa playing pool to be a guarantee that humans will remain unsurpassed by computers, since the legion of chess bot-developers have only achieved a minimum edge for bots (the high rating difference doesn't reflect the true because of anti-computer strategies). |
||||
|
Title: Re: State of the Challenge 2009 Post by jdb on Mar 14th, 2009, 7:04am First some thoughts, Humans have a clear edge when the position is not tactical. Even at the current level of human skill, this advantage will remain for several years. When the position is complicated, but the board is still relatively full, humans also have an edge. However, this advantage will not last as long. When the position is complicated and the board is empty enough that there are goal threats, humans and bots are about equal. By next year bots will have the edge in this area. These types of positions are all about calculating variations, and thats what bots are good at. If the current level of human skill remains about the same, the challenge will be over in 5 years. The known arimaa strategies can all be taught to a bot without too much difficulty. After that, it is just the case of investing the time to actually do it. If someone new comes along and plays arimaa at a 3000 level, then the challenge will be safe. This will require new strategic discoveries. |
||||
|
Title: Re: State of the Challenge 2009 Post by Fritzlein on Mar 16th, 2009, 3:30pm on 03/02/09 at 11:49:51, Fritzlein wrote:
Thank you naveed, for making my forecast seems prescient with this game (http://arimaa.com/arimaa/gameroom/comments.cgi?gid=100093). Apparently the day has not yet arrived when we have to avoid slugging it out with bots. Of course, I'll probably be too chicken to do this myself in the Challenge games, but that is my problem, not humanity's. |
||||
|
Title: Re: State of the Challenge 2009 Post by 99of9 on Mar 17th, 2009, 5:04am Well the humans are really smashing the bots so far this year, only one loss out of the first sixteen games! |
||||
|
Title: Re: State of the Challenge 2009 Post by Fritzlein on Mar 17th, 2009, 8:23am True, true. I am thrilled by humanity's current run. But since this thread is for long-term forecasts, I may just make myself look silly by getting caught up in the present. :P |
||||
|
Title: Re: 2009 Arimaa Challenge Post by Fritzlein on Apr 18th, 2009, 2:09pm on 03/04/09 at 19:35:08, 99of9 wrote:
Wow, it didn't take long for this to look prophetic. I hung a dog (http://arimaa.com/arimaa/gameroom/flShowGame.cgi?gid=103223&s=w&client=1) already! But I guess this isn't the year it falls, and I'm not the strongest defender, so your prediction is not yet decided. :P |
||||
|
Title: Re: State of the Challenge 2009 Post by 722caasi on Apr 20th, 2009, 4:22pm I think predicting what will happen is fundamentally flawed. There is no way to tell. An enormously powerful computer could come out, arimaa might become more popular, etc. |
||||
|
Title: Re: State of the Challenge 2009 Post by Belteshazzar on Feb 4th, 2019, 8:58pm on 03/06/09 at 06:10:56, Oystein wrote:
Right on. |
||||
|
Arimaa Forum » Powered by YaBB 1 Gold - SP 1.3.1! YaBB © 2000-2003. All Rights Reserved. |