airhockeyworld.com - The ultimate resource for competitive Air Hockey!

home

forums

rankings

profiles

air hockey chat forums

Forums Home | Log in for Private Messages | Search | View New Posts (Mark All Read) | User List

Forums Home / Tournaments and Challenges / Rating Vs Ranking ( View Older Thread | View Newer Thread)

First | 1 | 2 | 3 | 4 | Last

Poll: Just curious as to which method the players truly do prefer for seeding a tournament?

(5%)

Current Ranking System - Mouseover to see Voters

(9%)

Current Rating System - Mouseover to see Voters

(18%)

Current Ranking System but with modifications - Mouseover to see Voters

(68%)

Current Rating System but with modifications - Mouseover to see Voters

(0%)

Doesn't Matter - Mouseover to see Voters

TheAirHockeyGuy - 13 Aug 2011
Total Posts: 472

Just curious as to which method the players truly do prefer for seeding a tournament?

Chris Lee
Co-Founder, CEO
Air Hockey Players Association (AHPA)

goran - 13 Aug 2011
Total Posts: 428

I was going to post the same poll today - the doesn't matter vote. Great minds think alike.

sjrbat - 13 Aug 2011
Total Posts: 527

As the sport gets bigger, I really do think we will have to switch to the Ratings System.

But for now, I think the Ranking System can be tweaked can put to better use during tournaments.

Question:

Since many players do not play rated sets, how would a Pro-level player rated in the 60's or 70's be treated if there was a tournament/

For example, Albert before the tourny?

goran - 13 Aug 2011
Total Posts: 428

I just came up with a rule where you can challenge your rating. Since seeding is done by tourney finish, you can partition for a rating adjustment if 5 players are rated above you that you finished higher then at your last nationals. This would be good for players who don't play rated sets and only compete in every other nationals.

jasonstevens - 13 Aug 2011
Total Posts: 176

goran said:

So who decides where this person would move to and how is it calculated?

I don't think this is a good move.

goran - 13 Aug 2011
Total Posts: 428

Their rating is re-ran through the last 10 tournament matches they played. Their win loss record decides the outcome. Other Elo system users use similar methods to handle player who jump up big in a tourney. They were obviously under rated when entering the tourney. Players who play several challenge sets wont need to be looped, because they don't enter a tourney underrated.

ajflanagan - 13 Aug 2011
Total Posts: 509

Without a deeper analysis and some examples to back it up, I don't think this is a good idea. You are complicating a system that is otherwise perfectly simple. The system is easy to maintain and easy use to seed tournaments throughout the world on a moment's notice.

What you are entering into the equation is not quite as subjective as classification seed boosts, but it is equally problematic. Remember, there are play style matchup issues in Air Hockey and clustered player issues that Dan mentioned.

At the very heart of the Mitic System is PARTICIPATION. We are trying to encourage recorded activity. If the system had more value to players and was used to seed tournaments, I believe recorded activity would increase.

If Jose Mora v2.0 showed up unannounced to a tournament with to rank, no rating, no seed and no one to vouch for his skill level... oh well. He starts at the bottom like everyone else and works his way up the ratings ladder. Besides, a player like that would theoretically have a lot of tough matches to play in his first tournament and prove himself. If he's as good as he thinks he is, his string of upsets in his first tournament will boost his rating quickly.

goran - 14 Aug 2011
Total Posts: 428

If Jose moral 2.0 started at the bottom and finished 5th at the nationals, he would move up fast in the ratings, but still might not have a top 20 seed for the next tournament. Looping data for players who have great tournaments will put their seed closer to where we have traditionaly seeded players which is by tournament finish.

Darth_Wafu - 14 Aug 2011
Total Posts: 117

I am with Goran on this one. It keeps us closer to our traditions while still utilizing the rating system. Plus, with clearly defined criteria as to when it can be done, there really isn't any subjectivity to the process. The only subjectivity would come from determining what criteria to use for the looping. As long as we can agree on that it should be fine. Once the ratings catch on and more data is available we would very rarely have to do it, or possibly even remove it altogether at that point.

Nick Geoffroy

goran - 14 Aug 2011
Total Posts: 428

Nick gets it. I would rather not use it, but I think its needed now with the low amount of recorded matches. If the rating system is used for seeding, more matches will be played and recorded between tourneys, and the need for looping will go away. Just like randomization was once needed then removed. When the rating system is adopted for official use, rules can be voted on.

goran - 15 Aug 2011
Total Posts: 428

Its nice to see no one voted for the current ranking system yet. 3 of the top ten players in the world are in favor of the rating system with modifications. Keep the ideas flowing and lets make a change. Why try fixing the old system when there is a better new system and it's FREE!!! If you haven't voted yet, please do.

fupersly - 15 Aug 2011
Total Posts: 231

I think we all generally agree, Goran - even though I'm voting for the rankings system with modifications, since I think rating is not a clear indicator of tournament performance (which I think is a belief that Phil also subscribes to). I think the biggest surprise that came out of this discussion was the "looping" of ratings, which nobody seems to have heard of before and was never mentioned as part of the ratings model. I know it's probably an afterthought if you really understand the system all the way to its core, but for those of us not entirely in the know, it feels like there might be other "surprise" elements to the system that we aren't aware of that would be nice to know before we move forward and totally adopt it.

chuckdpuck - 16 Aug 2011
Total Posts: 26

I know the current rankings system. But I don't have a good understanding of the Mitic Rating system. Goran, do you have something you could send me or post that would provide a brief overview of how it works. I would appreciate it.Thanks
Chuck

Q - 16 Aug 2011
Total Posts: 497

chuckdpuck said:

Hey Chuck. If you go to the rankings page on AHW there is a link that says "What does AirHockeyWorld Rating mean?" There is the brief explanation of the math/calculations used to adjust players' rating depending upon the type of set or match they are playing. Goran has been working on adding the playing of single rating games and best 2 out of 3s or 3 out of 5s to the system that used to primarily use just 4 of 7 game sets (Mitic sets or tournament sets) and 3 of 5 or 4 of 7 set matches. It's based on the Elo rating system which is a widely used statistical method for assigning a ranking that reflects a players skill level. The system determines a probability of victory for the two players playing based solely on their difference in rating. Then depending upon the value of the match which varies depending upon whether it's a Mitic set or a set at a tournament for example the appropriate percentage, X, of those points would go to the "underdog" if he/she were to win or (100-X) percent would go to the favorite. For example: if someone is an 85% underdog according to the difference in their rating and wins, he/she get 85% of the value of the match from their opponent. If in that set, the 85% favorite won, he/she would get 15% of the value.

travis - 16 Aug 2011
Total Posts: 530

chuckdpuck said:

Read this excellent post from Dan about the differences between the existing USAA seeding method and how the Mitic system works:

http://airhockeyworld.com/forummessages.asp?forum=1&thread=1005&page=1

Travis Luscombe
AirHockeyWorld.com Webmaster
http://twitter.com/air_hockey

goran - 16 Aug 2011
Total Posts: 428

here is the link to the "what does air hockey world rating mean".

http://www.airhockeyworld.com/ahw_rating.asp

all other new adjstments are posted in the "changes to rating sytem" thread.

Thats right no "surprise" elements. Just new ideas.

goran - 16 Aug 2011
Total Posts: 428

fupersly said:

I think rating is not a clear indicator of tournament performance (which I think is a belief that Phil also subscribes to).

I believe rating is a better indicator of tournament performance. The ranking system is a good indicator of the last tournament finish. If it wasn't for challenge matches to fix the inaccurate rankings, we might as well call it a tournament finish instead of a rank. The rating system captures results from national, state, and local tournaments. It even uses one on one tournaments we call challenge matches, and challenge sets.

fupersly - 17 Aug 2011
Total Posts: 231

Hmm... I think I meant to say "predictor" instead of "indicator", but I'll stand by the spirit of my original post and say that just because someone can beat a bunch of people a lot, it doesn't mean they're necessarily going to be able to beat a bunch of people that beat the people they can beat a lot. It's kind of like boxing in that styles make match-ups - so, while you might be able to put some fancy statistics in a formula and come up with what should happen most of the time, I still don't think you can accurately measure someone's heart or desire and how well they might perform (or collapse) under the spotlight.

goran - 17 Aug 2011
Total Posts: 428

Joe, how does style match up work differently in the current ranking system? The ratings just "predicted Donovan to beat Chris 4-3. It is a better predictor then any human. Ratings can see player improvement faster then any of us.

travis - 17 Aug 2011
Total Posts: 530

fupersly said:

It's kind of like boxing in that styles make match-ups - so, while you might be able to put some fancy statistics in a formula and come up with what should happen most of the time, I still don't think you can accurately measure someone's heart or desire and how well they might perform (or collapse) under the spotlight.

So, should tournaments be random-draw, then, if we have no way of predicting who is more likely to win?

Travis Luscombe
AirHockeyWorld.com Webmaster
http://twitter.com/air_hockey

chuckdpuck - 17 Aug 2011
Total Posts: 26

Ok, Dan Meyer, thanks for your detailed comparison between the Mitic Rating System and the Current Ranking System or whatever it is officially called. So, one question I have for Dan or Mr. Mitic, it seems that if you happpen to live in a area such as Dallas or somewhere else that has no top 10 players in close proximity that this would be a detriment to that players Mitic Rating unless they were able to travel and play several Mitic Matches against those with significantly higher Mitic Ratings. I may be completely off with this question so I appreciate your patience in explaining. Secondly, I look at my Mitic rating which is 131-Chuck Forrest -1283 then I look at the 130 above me. There are a great number of these players that I have played in the last year and beat in either a National Tournament or in a challenge match. So, this is where I am confused by the math involved in the Mitic system. However,if I look at the USAA rankings which although I believe I am actually a better player than the rankings indicate, I have clarity in how it works,and it seems to be a much more accurate depiction of my personal skill level. My current USAA Ranking (which is based on my finish in Houston a bit ago) is 45th. Now, although because of various factors I feel my ranking isn't as accurate as it should be, I have the ability to make my argument very easily by challenging someone who I believe I have a decent or better chance of defeating in a challenge match. For instance, last year in Houston, I finished #25 at the top of ProB which is why I was seeded at the bottom of the Pro level in this year's tourney in Houston. I can now challenge someone who I feel I have a good chance of defeating and take their rank, if I loss then I can try someone else. So, now back to my Mitic rating of 131 and 1283pts, I have 2 questions, if the Mitic Ratings were to be used for an upcoming tournament in the next week and my ratings were as they are now, how would put me in relation to seeding? Secondly, if I wanted to climb to a Mitic Rating that I felt was closer to my actual skill level based on my recent performances in tournaments and challenge matches, let's say I challenged someone like Mr Donavan Brown,currently ranked 21st in the USAA Rankings and 43rd with 1564pts on the Mitic rating,to an official challenge match and won the match, how would this affect my Mitic rating? I know the answer already with regards to the USAA Rankings, if I were to win the match, I would then jump from 45 to 21. However, I have no idea how this would effect my Mitic Rating. I'm sure there's an easy answer being that it's based on math but could one of you please help me out with the answer?
At this point, I look at both rankings the USAA and the Mitic, and the Mitic seems very skewed compared to my skill level by the fact that I have not been keeping track of scores for the matches I have been playing because I just recently became aware of the impact this had on the Mitic Ratings.

chuckdpuck - 17 Aug 2011
Total Posts: 26

By the way, I reserve the right to change my vote in the polling via Write In Vote lol, if my boys, Mr. Meyer or Mr. Mitic, respond to my questions and observations so that I can better understand how my personal rating may be an anomaly or would be much more accurate to my actual skill level once I begin keeping track of my scores. Thanks in advance.

Chuck

travis - 17 Aug 2011
Total Posts: 530

chuckdpuck said:

At this point, I look at both rankings the USAA and the Mitic, and the Mitic seems very skewed compared to my skill level by the fact that I have not been keeping track of scores for the matches I have been playing because I just recently became aware of the impact this had on the Mitic Ratings.

That's because the Mitic ratings currently have no impact on seed right now. If Mitic ratings WERE used for seeding a tournament, then people would be more diligent in reporting them.

Travis Luscombe
AirHockeyWorld.com Webmaster
http://twitter.com/air_hockey

goran - 17 Aug 2011
Total Posts: 428

Chuck, if you can beat Donovan a challenge match your new rating would be 1427-1433 depending on the set count. That would put you around #65 out of everyone in the world. There is a rating calculator on the "what is an ahw world rating" section above where the ratings are posted. Being away from top players, you would have to travel to improve your ranking, but not to improve your rating. Just beat your local player base and record your matches. Although traveling will help your rating and skill. Just look how much Donovan improved.

fractalzoom - 17 Aug 2011
Total Posts: 201

Hey Chuck. I give Travis the credit for this.. but go here:

http://www.airhockeyworld.com/rating_calc.asp

You can do some calculations for how your rating would change here.

Also - every time you play, record your scores and shoot an email to Goran or Q and they can put in your scores. You, Chris, and the Keller bros are already rated and in the system, so if you beat up on these guys your rating will go up.

I'm currently rated 25th, but ranked 13th - which is a big difference, but there are a lot of people ahead of me in the ratings that haven't been active (but are really good).

I think you're also a little low because you had a bad tourney this year (unlike last year where you cleaned house in your spinoff). Looking at your page, your only win was against Christopher Clements, but lost to Chris, Fernando, and Justin. In major tourneys your rating can change quite a bit (the multiplier is higher). The best way to boost your ratings is to play a challenge match.

For instance, if you were to play Fernando Guillen a challenge match and beat him 4-0... your rating would go from 131st (1283 points) to 68th (1418 points)... which is a huge leap (but would be impressive to 4-0 Fernando). Big leaps in rating can happen because a lot of players in this range have ratings that are close to each other. In one day, you could shoot up 53 spots.

The more you play and record your matches, the closer you'll get to your true rating. It looks like you're in a downswing, but I'm confident that if you kept playing people, you'd skyrocket up the ratings in very little time.

fractalzoom - 17 Aug 2011
Total Posts: 201

travis - 17 Aug 2011
Total Posts: 530

fractalzoom said:

Hey Chuck. I give Travis the credit for this.. but go here:

http://www.airhockeyworld.com/rating_calc.asp

You can do some calculations for how your rating would change here.

Thanks.

Keep in mind that I may need to update the calculator for the things that the ratings committee changes.

Travis Luscombe
AirHockeyWorld.com Webmaster
http://twitter.com/air_hockey

chuckdpuck - 17 Aug 2011
Total Posts: 26

fractalzoom said:

Also, Chuck - having just a few people to play every week isn't a detriment. I've played Billy what seems to be hundreds of times and I've yet to beat him. Because our rating is so lopsided, Billy usually doesn't get any points at all.

Mostly - Q, Goran, and I have been exchanging points... with a few beginners thrown in as well (and I don't get many/any points from them). I was lucky in that there were some regular IL air hockey players that I was able to beat and start to move up the scale early on, but Q was in sub 1200 territory for a long time before he started to take sets off Goran and I. It didn't take long for him to jump above me in the ratings.

If you were to take just 1 or 2 pilgrimages to Houston (I know it's a heck of a drive) a year (and win)... or *any* Dallas player, really. There will be more points in your pool of regular players to fight over.

- First of all thanks Dan,Travis, and Goran for your answers. I really appreciate it. Now, I agree Donovan is a great example of how a player can quickly increase his skill level as are Dan and Q in my opinion. However, Donavan has a pretty large number of top 10/Master level players to learn from and compete against which obviously by his increase in skill level, has helped a great deal. Now in the case of Q and yourself, Dan, comparing my situation here in Dallas with your situation in Chicago is significantly different. You have the GREAT BILLY STUBBS to learn from and compete against along with Goran who is a what an Ultra Pro? (not certain) anyway this is where I see issues with the Mitic rating system being solely used for seeding at major tournaments. Because in my situation here in BIG D, currently we have no Master level players within 200 miles which may not seem like a big deal but in some situations due to economic reasons or various circumstances some people including myself can not afford or justify traveling solely for the purpose of increasing my Mitic Rating. So, I see this as a disadvantage to us if the Mitic ratings are used. Please understand, the will is there, personally, for me to travel, I simply can't justify spending the money when I have other more pertinent obligations that take priority. With that being said, my situation may in fact be somewhat of an anomaly. Also, I am hoping that with the traction we are generating with two national level sponsors being willing to put up a Cash Prize of $100-200 per bi-weekly tourneys that will be starting the week of Sept 12 (not quite firm on exact date), and the audience I am getting with my meetings with Dynamo (one coming on Friday) I am confident that we can begin to build the talent base here locally in the DFW area. I will also make certain that we immediately start keeping account of our games for M ratings. Thanks again guys for the answers.

ajflanagan - 17 Aug 2011
Total Posts: 509

Chuck, it's going to be easy to find discrepancies or anomolies in any system if you look for them. The point here is to develop a system that is the MOST FAIR and LEAST SUBJECTIVE that will allow for a smooth transition into the future.

You guys are organizing another tournament in Dallas, right? How are you planning to seed that event? Do you or Chris Green have a solid understanding of the current seeding procedures? Not very many people do.

I contend that the discrepancies in the ranking system are far more extreme. The system is inaccurate. A tournament finish is a tournament finish. That will never change. A champion is a champion. But the subsequent rankings are garbage. I mean... look at the current rankings. Look how many amazing and active players are not even on that list. Look at where some players who forfeited matches are now ranked. Is that what you consider a good representation of our sport? No way!

The rating system is designed to encourage activity. Right now, if you want to improve your seed, you really just have to play one match. Like you said, you could jump from 40-something to 20-something in one day. How could that possibly be remotely accurate? One day you are 40-something and the next you are 20-something? Really? Something is off. With a rating system, if you want drastically improve your seed, you have to show up and you have to play. Any way you try to slice it, that is a GOOD thing.

ajflanagan - 17 Aug 2011
Total Posts: 509

First of all thanks Dan,Travis, and Goran for your answers. I really appreciate it. Now, I agree Donovan is a great example of how a player can quickly increase his skill level as are Dan and Q in my opinion. However, Donavan has a pretty large number of top 10/Master level players to learn from and compete against which obviously by his increase in skill level, has helped a great deal. Now in the case of Q and yourself, Dan, comparing my situation here in Dallas with your situation in Chicago is significantly different. You have the GREAT BILLY STUBBS to learn from and compete against along with Goran who is a what an Ultra Pro? (not certain) anyway this is where I see issues with the Mitic rating system being solely used for seeding at major tournaments. Because in my situation here in BIG D, currently we have no Master level players within 200 miles which may not seem like a big deal but in some situations due to economic reasons or various circumstances some people including myself can not afford or justify traveling solely for the purpose of increasing my Mitic Rating. So, I see this as a disadvantage to us if the Mitic ratings are used. Please understand, the will is there, personally, for me to travel, I simply can't justify spending the money when I have other more pertinent obligations that take priority. With that being said, my situation may in fact be somewhat of an anomaly. Also, I am hoping that with the traction we are generating with two national level sponsors being willing to put up a Cash Prize of $100-200 per bi-weekly tourneys that will be starting the week of Sept 12 (not quite firm on exact date), and the audience I am getting with my meetings with Dynamo (one coming on Friday) I am confident that we can begin to build the talent base here locally in the DFW area. I will also make certain that we immediately start keeping account of our games for M ratings. Thanks again guys for the answers.

Correct me if I'm off base here... but it sounds to me like you are arguing against a system that promotes growth because you don't yet have a huge player base in Dallas or top level players. Right? You are kind of putting the cart before the horse. Build it and they will come. Help develop a system that is easy to use and promotes growth and activity... then USE that system. Keep doing what you're doing and use these promotional tools and players will come. It's a win-win for everyone Chuck. Don't shoot down the system because it doesn't fit your circumstances right now at this moment. Embrace the system for what it will do for your player base in the long run.

First | 1 | 2 | 3 | 4 | Last

Forums Home / Tournaments and Challenges / Rating Vs Ranking

HOME | SPONSORS | PRIVACY POLICY | TERMS OF USE