yobobo Posted February 17, 2000 Share Posted February 17, 2000 Fionn and all the CM players out there. The program for the chess rating system is in place and working great. I would love to add CM to the line up. It would take me 1 hour to get it ready, so whats the hold up lol. Lets get it going You cant tweek the chess rating formula. Tweeking it blows the ratings to hell. But there is no need to tweek. Hmm i asked Fionn about this weeks ago, he only was interested in the my programs??? i guess he wants to do it all himself. Like i have said before this Tournament House is non profit, made by players for players. Again T.H. would be happy to add CM to the small list of games Rule changes can be made, anything is possible. http://tournamenthouse.com/ yobobo info@tournamenthouse.com Link to comment Share on other sites More sharing options...
Fionn Posted February 17, 2000 Share Posted February 17, 2000 Gees, now I have to respond to this post twice ;(. Yobobo, I was only interested in your programme because I do NOT think your current system is sufficient to rank CM games since it is based on a game with only 3 outcomes whereas CM has 8 or 9 possible outcomes. This needs to be reflected but isn't. Since I respected your desire to keep your ranking system as is I wrote you an email asking if you would share your programme so that I could have it modified to include the various items which I felt were important to deal with. You said no. So, rather than trying to cast aspersions on me and trying to make out that I want to "own all" etc why don't you realise that: a) it is QUITE alright for me to think your system is flawed for ranking CM games and players. it is QUITE allright for me to say so and choose NOT to co-operate with a ranking system I feel is inherently flawed. c) I asked for your programme and was quite clear that I wanted to modify it and post it to CMHQ and use it to create a ranking system there. IF you had agreed with me regarding the need to tweak the programme I'd have been happy to see it hosted on Mars. I don't really care BUT I'm not going to support a ranking programme I think is flawed. d) I find it MOST disheartening that you would come on here and cast aspersions on me which are quite baseless. When you said no, that you didn't want to share the programme I took your answer with dignity and left. NOW you pop up again and start accusing me of stuff which is blatantly false. hell, I could start giving out about your meaness of spirit in not sharing a programme which could form the basis of a CM-specific ranking system. Anyways, I won't say any more but I wanted to set the record straight about WHY I didn't support your venture (I don't support ANY venture I think is flawed) and also wanted it to be crystal clear that I don't appreciate your mis-presentation of events. Link to comment Share on other sites More sharing options...
Guest Ol' Blood & Guts Posted February 17, 2000 Share Posted February 17, 2000 Fionn, sounds like a lot of jobs I had. Trying to work for a company that I KNEW was flawed in their ideas and thinking. ------------------ "I want you to remember that...no bastard ever won a war...by dying for his country...He won it...by making the other poor dumb bastard...die for his country."--George S. Patton Link to comment Share on other sites More sharing options...
yobobo Posted February 17, 2000 Author Share Posted February 17, 2000 Well please tell the players and I how you would modify the chess system, as it is mathematically sound. It is not my system as i am not a mathematician, i did not make it up by pulling numbers out of my hat. By (tweaking) it you mean adding #s and deleting #s that you think should be, will only destroy the system. I asked if CM wanted to be ranked at the Tournament House, your email was a little rude. Stating that the chess system was flawed and that you Fionn would like to make it better and then you would let me have it to use, lol About me being mean, I was the one that asked if CM want to be ranked at the T.H. and as you can see im not in it for the money as not one banner add can be found on the site. Every game has different wins and losses Fionn. If it was a close game or an easy win, the outcome is the same. Again it is not my system. This rating formula has been around along time. I guess T.H. will not host CM ranking because Fionn says so and it is too bad because it would be fun. But please tell use Fionn how you will make the chess system better by tweaking it. yobobo Link to comment Share on other sites More sharing options...
Guest hunt52 Posted February 17, 2000 Share Posted February 17, 2000 Would it be possible to assign points (wins?) based on the score? Like a 90-10 game would net the victor nine wins and a loss and the loser a win and nine losses? I am not very familiar with chess ranking, but I would think something like this would work well. That said - I am against ranking in general. It makes commanders focus on winning instead of learning and enjoying themselves. I am sure that if a ranking system were predominantly displayed many new players would not play the more experienced vets. This would lead to a stagnation of tactics, and generally decrease the turnover in what works and what doesn't. I often find that new players try things that the older ones would simply scoff at, and every once in a while they are *really* clever. I think a ranking system discourages this kind of experimentation, sacrificing it to the god of the win column. I don't personally like this. If other people want a ranking system, fine. I just won't participate. - Bill Link to comment Share on other sites More sharing options...
Guest Zulu1 Posted February 17, 2000 Share Posted February 17, 2000 I went and had a look at the rating system and I have to agree with Fionn. It is fine for chess where there is only 3 possible outcomes ie: win, lose, or draw. It will also work for CC2 etc where the outcomes are win or lose. But as Fionn stated, CM has 8 or 9 possible outcomes. These need to be taken into account. ie. you should get a bigger jump in rankings for getting a Total Victory against a higher ranked opponent than squeeking out a Marginal Victory against the same opponent. Fionn asked a legitimate question and you dumped on him. Also, why did you have to publically post this other than deliberately sniping at someone. IMHO you should have handled this privately via email. You owe Fionn a public apology. Link to comment Share on other sites More sharing options...
Guest Charles Posted February 17, 2000 Share Posted February 17, 2000 Damn! Forum ate my post!! Fionn, in my original post I stated I thought you might have jumped on this guy's' case a little too hard. It seems to me that he truly wants to help and thinks he has a good system which can be used to rank CM players. Seems to me there is room for some compromise and cooperation here. Although I am a little unclear whether Yobobo actually wrote the program we are talking about, I would like to ask you how would you feel about giving someone your program when the person you are giving it to indicates he will make whatever changes he feels necessary, with little or no input from you? If I had developed such a program, I would be somewhat leery of giving it away to a stranger, even a stranger with a good reputation like yourself. Of course, if the program isn't the guys work, then of course he can't give it to you without the author's consent. I know I may not have all the information here, but it seems to me that this guy wants to help and thinks he has a good system to rank CM tournament players. I say let's give this guy a chance. Work with him to see if any changes are really necessary to make it work for CM. If it works, great! If not, try something else. ------------------ Not THE Charles from BTS Link to comment Share on other sites More sharing options...
tss Posted February 17, 2000 Share Posted February 17, 2000 Would it be possible to assign points (wins?) based on the score? Like a 90-10 game would net the victor nine wins and a loss and the loser a win and nine losses? I'm not an expert on chess ratings so I don't know how that approach would affect the scores. However, if it doesn't cause any problems I'd suggest using a scale of assigning 20 wins per game so that the score could be calculated with the accuracy of 5 points. For example: score 45-55 leads to 9 and 11 wins (plus 11 and 9 losses). - Tommi Link to comment Share on other sites More sharing options...
yobobo Posted February 17, 2000 Author Share Posted February 17, 2000 Reply to Zulu1, You wrote But as Fionn stated, CM has 8 or 9 possible outcomes. These need to be taken into account. ie. you should get a bigger jump in rankings for getting a Total Victory against a higher ranked opponent than squeeking out a Marginal Victory against the same opponent. If you beat a higher ranked opponent you do gain more points than beating a lower ranked opponent. But as for type of victory there is only one outcome. I do undertand what you are saying about the different types of wins. But as in all games a win is a win. I do apologize for sounding a bit pissed with Fionn, I am just trying to help. Link to comment Share on other sites More sharing options...
Berlichtingen Posted February 17, 2000 Share Posted February 17, 2000 <BLOCKQUOTE>quote:</font><HR>That said - I am against ranking in general. It makes commanders focus on winning instead of learning and enjoying themselves.<HR></BLOCKQUOTE> Coming from a military background, I have to mention that a commander's focus should be on winning On the otherhand, you learn the most from losing. Link to comment Share on other sites More sharing options...
Guest hunt52 Posted February 17, 2000 Share Posted February 17, 2000 Actually, in CM a win is not a win... There are several scales of victory from tactical (?) to Total. If you beat your opponnent by one unit killed and neither of you has enought troops left to clean toilets much less fight you should not get the same boost in rank as you would if you crushed your enemy, drove him before you, and heard the lamentations of his woman (to paraphrase... ) I think this variation in victory levels is what Fionn is talking about. Also - tss - I think you are right. 20 would be better than 10. (Ideally, 100, but this is probably unrealistic...) - Bill Link to comment Share on other sites More sharing options...
tss Posted February 17, 2000 Share Posted February 17, 2000 But as in all games a win is a win. I have to disagree here. There are many games, even classical board games, that assign different scores to different kinds of wins. For example, in Backgammon (which is one of the oldest games known) the score is doubled if you manage to finish before your opponent has entered the final phase of the game (I don't know the correct English terms). I think that any rating system for CM has to take into account the different victory levels. - Tommi Link to comment Share on other sites More sharing options...
yobobo Posted February 17, 2000 Author Share Posted February 17, 2000 CC2 has 4 types of wins, Minor Victory, Major Voctory, Deciseve Victory, and OverWhelming Victroy. We use this type of rating for knock out tournament play. For tourneys, players play both sides of the same map, once as allies once as german. After both games have been completed, point are totaled and the winner announced. Link to comment Share on other sites More sharing options...
Colin Posted February 17, 2000 Share Posted February 17, 2000 Wow this is a really long conversation for some reason.........this will be an interesting thread although I think Steve should close it just to keep our insanity in check. Also another update or even a demo would be useful in quelling these (soon-to-be) riots. ------------------ Visit my webpage! http://cm4mac.tripod.com Link to comment Share on other sites More sharing options...
Guest hunt52 Posted February 18, 2000 Share Posted February 18, 2000 Ok how about this then: An algorithm for ranking in CM: dP = [ P(opponnent) / P(player) * S(player) ] - [ P(opponnent) / P(player) * S(opponnent) ] P = P + dP * correction factor (to determine inflation if desired) Let's say everyone starts out with ten points. Then I really hose Scott in CE (74 to 26) My score would be: 10 + { [ 10 / 10 * 74 ] - [10 / 10 * 26] * say .1 } = 14.8 and Scott's would be 5.2 (sorry Scott... ) The next time we play he pulls a Bil on me and the score is 100 - 0 him. He steals all my points and I now have a score of 0 to his 20. Anyway - how this would work is you steal points from opposing commanders. You could engineer in an inflation to the gaining points which would make the total number of points in the game go up very easily. This would encourage taking on the big guns to get the big points! - Just an idea - Bill Link to comment Share on other sites More sharing options...
Guest hunt52 Posted February 18, 2000 Share Posted February 18, 2000 Sorry - I wrote that while doing calculus homework... umm... S(player) is the players score in the game P(player) is the players score in the ranking system. Really you would have to do something to make sure people never ran out of points or beating up on them would be no fun! Also you should maybe allow a maximum of 15 or 20% points to change hands in one game... - Bill (again ) Link to comment Share on other sites More sharing options...
Guest hunt52 Posted February 18, 2000 Share Posted February 18, 2000 If anyone is interested I can code that up probably in 1 or 2 hours. Only know C though (and javascript sort of)... - Bill (too lazy to use the edit post thing) Link to comment Share on other sites More sharing options...
Fionn Posted February 18, 2000 Share Posted February 18, 2000 Well, I've developed a mathematical formula which not only takes into account the size of the win but also takes into account the disparity in strengths between the forces involved. This formula thus means that you can now play someone in an UNBALANCED scenario and lose yet still increase your ranking based on how well you did vs how well an average player would have done. I asked yobobo for his programme under the clear pretext that it would save me a lot of time in creating a ranking programme at CMHQ since I'd only have to tweak his programme. He said no so I'm now starting from scratch. All this means is that the CM community will have to wait for a longer time period to have a truly applicable rating system. Lastly I'll just say that my only interest here is to get the best ratings system possible created for CM. I'm not going to make money from this so my ONLY motivation is to get the best rating system possible. Since I'm not worried about making money from this I'm free to look around for the BEST system and, when it isn't in existence I am free to create it. While I wasn't going to say so publicly one of the utilities I wish to develop (and was asking for programming help with) IS a ranking system based on TGN's web servers. Link to comment Share on other sites More sharing options...
Eridani Posted February 18, 2000 Share Posted February 18, 2000 Yobobo- I can definately see that you are trying to help. However, you do show a clear lack of experience to CM. Your statement that "as in any game a win is just a win" is most definately false in context for this game. There are MANY ways a game can end... I agree that the simplest way to gauge the "level" of the victory is the final score, although that is still a little rough around the edges. There is a very real difference between a pyrric (sp) victory of 51-49 where the battle ended simple because there was no combat-capable units (but still a tactical victory for one side mind you) and a sweep of 100-0 where one side is simply squashed. This is something which a 3 way system cannot model. I'm seriously beginning to think that we should set up an Email list or dedicated board to debate a new CM rating system... A lot of good ideas from all parties, we should put together one standardized algorithm. We're not saying (at least most of us I think) that your help would not be appreciated... Especially if you have Java or CGI script programming abilities, we are saying that ur current system is not sufficiant, and it would take considerable adaptation to make it suit CM. thanks -EridanMan Link to comment Share on other sites More sharing options...
Guest hunt52 Posted February 18, 2000 Share Posted February 18, 2000 Fionn, What language are you going to be writing it in? Maybe I can help some... I can definitely help code up the HTML interface if you need that. - Bill Link to comment Share on other sites More sharing options...
yobobo Posted February 18, 2000 Author Share Posted February 18, 2000 Reply to Eridani, What if victory conditions where set prior to the game. What is needed for a victory, a draw, or loss. This could solve some problems? Hmm just thnking outloud. Thanks, yobobo Link to comment Share on other sites More sharing options...
Eridani Posted February 18, 2000 Share Posted February 18, 2000 Yes there are mission objectives... but these are only a part of the total decision of who victorious... As in real life, if you do manage to win a battle, but in the process of doing so castrate yourself so badly that you are in-effectual for the rest of the war, then you didn't really win. This also brings in the crucual fact that there are different levels of victory... There is a BIG difference between a commander who just manages to edge out his opponent and a command who stomps on the enemy...There is too much of a factor of luck to make simply win/loss the determining factor, a great general could have hit a string of bad luck which would through off the rating (and luck does play a factor in CM as it does in real warfare... all you can do as a commander is play the odds to your favor) The only common demoninator (that I can see, I may be off kilter Here... Fionn is the true CM expert and the one to ask) is the final score, which is in effect "percent victory"... So in order for the ladder system to be accurate, it will have to use this final score, not simply who won and who lost. -EridanMan Link to comment Share on other sites More sharing options...
Fionn Posted February 18, 2000 Share Posted February 18, 2000 I think it can be taken as read that if I've given this issue some thought (as I have) and if I've looked at various rating systems for over 3 months (as I have) and have come to the conclusion that something more multi-factorial than the chess rating system is necessary then probably I have some method to my madness I'll post some info about this rating thing once things are a bit firmer. Link to comment Share on other sites More sharing options...
Stryker Posted February 18, 2000 Share Posted February 18, 2000 I've thought about this subject for quite some time as well. My conclusion is that there is no 'right' answer to player ranking. I took the approach that *who* you beat is more important than by how much you beat them. The main reason for this was the win by surrender outcome. I couldn't figure out how best to point score it. Sometimes you surrender because you are really getting destroyed, other times it is near the end of the game and you are only losing by a small margin and you don't have enough left to change the outcome. Are all surrenders judged equally? The other potential issue is system performance. In my case after every completed game the program makes an *adjustment* to the score of everyone that ever played one of the players that just completed the game. This dynamic ranking takes into account how well your previous opponents preform after you have played them. I would have liked to have done more with the dynamic ranking, but the need to keep if fast limited my options. Regards, Stryker www.pantherworks.com [This message has been edited by Stryker (edited 02-17-2000).] Link to comment Share on other sites More sharing options...
yobobo Posted February 18, 2000 Author Share Posted February 18, 2000 Reply to EridanMan, Yes that is what i mean, what constitutes a win, a draw, and a loss. It could be based on points, %, anything. Hmm still thinking outloud Link to comment Share on other sites More sharing options...
Recommended Posts