| goran
- 12 Aug 2011
Total Posts: 428
|
| forefitting only effects rating if the match was started. Vic didnt lose any points for not playing the spinoff, but he didn't gain any either.
|
|
|
| Mike C
- 12 Aug 2011
Total Posts: 459
|
| Dan, I agree with your summation regarding UltraPro for those guys, but all three of you(Goran, Dan and Q) are in the same UltraPro boat in your progression right now. Albert Ortiz and Evan Siegworth can be added to that list as well.
|
|
|
| fractalzoom
- 12 Aug 2011
Total Posts: 201
|
Let me directly answer this as best I can. You are talking about the "match value" that is the multiplier for how many points swing in one direction or the other.
This value is the same for everyone. The bigger this value, the more a point swing we have due to perceived importance - which is why a set in a tourney is worth 3x the points in a regular Mitic set. This value is somewhat subjective, but it's applied evenly to everyone. I mean, why drop people 3.5 spots for missing a tourney - that value is subjective, but it doesn't seem unfair. Same thing with this multiplier. This 3.5 drop is actually applied only to people who miss a tourney. This doesn't reflect true skill. We saw Tim ranked lower than some people and probably unjustifiably so, and you could look at the Mitic ratings to justify that opinion.
Other than how we treat new players and inactive players, this value is the only subjective value in the system. If everyone played hundreds of sets each year, it wouldn't matter what this number is - everyone would get rated similarly.
Let me ask you a question. Let's use Travis as an example. Looking at the last 5 national tourneys, Travis finished 4th, 6th, 12th, 14th, and 17th. Since his last tourney finish was 14th and he missed the last Vegas tourney, he'd be seeded 3.5 spots lower (or between 17th and 18th). It was determined that this was too low and he got a 12 seed. This is based on his past performance - and certainly seems justified, but I'm not sure if there's a procedure to determine who gets moved and who doesn't. This seems incredibly arbitrary. Q finished Master this tourney (hard seeded 12th) and justified his seed and then some... but last year he finished 45th. If the next tourney Q finishes in the 20s - would we move him up to Master level for the next tourney? Has Q established himself as a master? How many finishes outside the top 16 does Travis have to have before he's not automatically seeded 12th? How often do we move someone's seeding outside the top 32? Proper seeding is just as important to lower level players than as it is to the top 16.
To me - this part of the current seeding system is arbitrary at best. At least with the Mitic rating factor you called out as subjective - it's applied to *everyone* exactly the same.
If we had a tournament of 200 players from all over the globe - how would this seeding system work if we're going to subjectively label someone as pro or ultra-pro or master? Is there someone going through the novice and pro-a groups to see if there are any inconsistencies in seeding?
carolina phil said:
The final formula for a ratings system creates a number than can be used objectively, but that number arises from a collection of subjective decisions made by a small group of people in a context of discussion and compromises.
It is a socially constructed reality. It is clearly subjective.
|
|
|
|
| carolina phil
- 12 Aug 2011
Total Posts: 1084
|
Excellent discussion, guys. There are some good points being made from many different perspectives. Personally, I wish there was a way to use a rating system while at the same time allowing a challenge match victory over someone higher than you to allow you to take that person's "seed" position.
Is there some way to continue that tradition while allowing a ratings sytem to be used for seeding at the same time?
This may be too hybrid of a concept, I know, and would erode the "elegance" of the mathmatics, but creative minds like yours may discover a way.
Let's say person X defeats person A in a challenge match and takes his ranking position and also is awarded a Mitic number (rating/seeding) of just one point higher than the person he just defeated in the match. It would be as if his new number is a new starting Mitic rating position for him.
Yes, this distorts the "pure" mathmatices, but since the mathmatics are somewhat arbitrary decisions to begin with, what's one more? Plus it would satisfy those players who want match play for ranking to determine seeds. And it would put in place a form of the Mitic rating/seeding system that would still use sets and tourney finishes to designate numbers for a vast number of players who do not play matches for ranking/seedings.
So, in effect, a match victory would be weighted so heavily that it would automatically trump all the other Mitic factors in this hybrid model. A match defeat of the challenger would still cost the challenger Mitic points, and would still affect the defender as now.
But, still, the seeding process for a tournament would benefit due to the advantages of the Mitic rating system, while at the same time match play would maintain its value and special appeal as a way of quickly jumping up in the rankings/seedings.
Just brainstorming with you genius gurus to see if we might just come up with a mutant quantum leap concept:)
Thanks for the feedback,
Phil
|
|
|
| brain
- 13 Aug 2011
Total Posts: 578
|
Here are the Singles Seeding Procedures as passed in the 2005 Vegas USAA meeting. They have been used in every tournament since then as is and implemented the same way each time.
1. Create a list. Begin the list by including all players having a current USAA National/World Singles Ranking. Insert the players in order of increasing value of their USAA National/World Singles rankings in the list. Hence forth, all players will be inserted into the appropriate location according to assigned "seeding value" in ascending order from lowest to highest.
2. If any players still remain without a seeding value, then all players participating in the tournament will be classified according to the official USAA Classification Process.
3. Any player who is not currently ranked but participated in a previous USAA National/World Singles Tournament will be assigned a seeding value. Each player's value will be the sum of his/her most current USAA National/World Singles Ranking plus a value of 3 per each USAA Sanctioned National Singles Tournament missed since their last USAA Sanctioned National Tournament performance plus 0.5.
4. Any USAA-assessed ranking penalties will now be applied to any players currently possessing a seeding value as required by USAA mandate.
5. Any players who had a previous USAA National/World Singles Ranking whose seed would result in their being among a set of players of a lower classification will be adjusted according to the following algorithm:
- Each player, in order of value, will be inserted at the bottom of the largest group of their corresponding classification.
- A classification group is defined as a set of same-classified players, which has no discontinuities of 3 or more players.
- A group is considered contiguous if 2 or fewer players of another classification are embedded within a set of same-classified players.
- If a classification group's continuity is broken by 3 contiguous different classified players, then the original set of classified players will be considered 2 different groups, and the player would be inserted at the bottom of the larger of these groups.
6. For players without a current seeding value, apply the same classification insertion process described in #5. Players of the same classification should be randomly seeded at the bottom of the largest group of their classification.
7. In the event that any set of players has the same value, the player with the most current USAA National/World Singles Tournament performance will be given precedence. After that, a coin toss will be used to resolve conflicts.
|
|
|
| Q
- 13 Aug 2011
Total Posts: 497
|
Thanks Brian! One follow up question to this is what officially defines the classifications/USAA classification procedure? I haven't checked the rules/constitution here on AHW yet.
|
|
|
| ajflanagan
- 13 Aug 2011
Total Posts: 509
|
Mike C said: | Dan, I agree with your summation regarding UltraPro for those guys, but all three of you(Goran, Dan and Q) are in the same UltraPro boat in your progression right now. Albert Ortiz and Evan Siegworth can be added to that list as well. |
Mike, your posts illustrate the exact problem with the current system. It's COMPLETELY and UTTERLY subjective. So now Evan is an ultra-pro? Why? Because you believe he is? For the same reason that Travis is an ultra-pro? Why? Because Brian said so? And Vic is now a Pro who should have a higher seed? Why? because you believe it to be true? Dan, Q and Goran are ultra-pros too? Why? Because you think they will beat most Houston "pros" who are mostly over-rated (in the non-mathematical sense)?
It's not my job to tell you what level Vic or anyone else is playing at. Frankly, it's not your job or Brian's job either. A promoter should not have control to "fudge" the numbers to give anyone an advantage or disadvantage. Make no mistake about it. It is "fudging" the numbers to arbitrarily decide if someone is considered an ultra-pro or not.
A player's results on the table tell the story just fine.
Please address the problem. There needs to be a "no questions asked" system for seeding tournaments. A system that can by used anywhere at any time by any one. The rating system is the best possible option on the table right now.
|
|
|
| fupersly
- 13 Aug 2011
Total Posts: 231
|
Q said: Thanks Brian! One follow up question to this is what officially defines the classifications/USAA classification procedure? I haven't checked the rules/constitution here on AHW yet.
|
To AC's point, Q, there really isn't one - at least not a well-defined one, anyway. Historically, anyone who has finished Master in the recent past gets categorized as an Ultra Pro, as well as anyone finishing 11-16 in the previous tournament. Now, I haven't been privy to a lot of seeding sessions so I'm not an expert (heh!), but I can say that I've heard this discussion before and there have been enough answers to basically get this right in the past - it's just not anything that the "masses" have exposure to and can easily understand, and I agree with AC that is exactly what's most wrong about this part of the seeding process.
|
|
|
|
| brain
- 14 Aug 2011
Total Posts: 578
|
Though the USAA has been around for many years, it is very immature in the way of its processes. Player Classifications is one of them.
Travis and I created the seedings based on the approved seeding procedures and submitted to the committee for approval. All but Phil looked them over, made some corrections and then approved the final list. I had no requirement to follow this procedure, but I did because I believe in process and know that fewer mistakes arise out of a process that includes the eyes of many experts.
As for a USAA Classification procedure: I have asked for more than a decade for a USAA Classification procedure. One has never been developed. The USAA talks and talks and talks but nobody ever actually does anything productive (except Travis).
As there is no system, I rely on my own analytical observations for local players. I am there nearly every week at the weeklies. I observe the skills and performance of all houston players, analyze their finishes and of course look at their weekly classification. For instance, Travis, as a level 5, has numerous top 4 finishes and usually does well. Albert Ortiz is also ranked as a 5 (out of 6) and has had many top finishes. Whether or not anybody agrees, I consider myself an expert on skill evaluation because I have been involved in raking players almost every week for 16 straight years. I had said that Albert was also an untra-pro, and the tournament committee said no. Then Albert finished 8th.
Now for players outside of Houston, I rely on a "primaty contact". Goran is my contact in Illinois. Like me, Goran is deeply involved in Chicago air-hockey. He is there every week and knows who is who and how people play. If Goran tells me that so and so is a Pro, and I take his word for it. In the realm of PMI Project Management, this concept is known as "Expert Judgment". It is a concept relied upon heavily in the business world because few decisions can be made without some element of subjectivity. You minimize the probability of error by relying on the judgment of somebody who has expertise in the realm in which you investigate. That is the approach that I utilize.
Outside of Chicago and Houston, most other players bases are 7 players or less and the degree of classification is fairly evident. Mark Nizzi is a reliable contact for Colorado. Davis Huynh or Jae Cain are reliable classification experts from California. Etc.
Until the USAA can develop a clearer methodology for classifying or produce a rating system that will be accepted and passed, this is how I will operate my events.
For the record, I am happy to share my methods with anybody that asks politely. It would be awesome to see more major tournaments spring up and new faces taking on promotion. More events would also lend to a much more reliable rating system.
Brian
|
|
|
|
| goran
- 14 Aug 2011
Total Posts: 428
|
| Great post Brian.
|
|
|
|
| Mike C
- 14 Aug 2011
Total Posts: 459
|
No Andrew, your answers are the problem! Nah nee naw boo boo. I didn't classify Syed, August as Ultra Pros - that was Goran. My OPINION that Albert and Evan are in the same category with the Chicago pros is because that is how they finished this past tournament & last year and they played great and finished high. It was a comment.
The solution I am proposing is use the rating system as a tool but it should not be the only factor used to seed a tournament. Because believe it or not, the rating system is a great measurement to use, but like in every sport people will find ways to cheat the system and take advantage of the system.
|
|
|
|
| fractalzoom
- 14 Aug 2011
Total Posts: 201
|
If you were to seed a tournament based on ratings, I don't think there really is a need to classify players at all.
Why argue about who is what class, or how many ratings points you need to be in a different class?
All of these are subjective observations - even though I agree that your expert opinions are going to be spot on.
I just don't see a need to classify people at all to seed a tourney. It seems the only reason you would do this is to fudge the seed list for people having a poor prior tourney (I haven't heard of anyone dropping down because of overreaching in a prior tourney).
The great thing about the ratings system is that we can can remove all subjectivity.
|
|
|
|
| fractalzoom
- 14 Aug 2011
Total Posts: 201
|
I should add that classifying players for the weeklies is something that is awesome and that I appreciate how the weeklies are handicapped.
I just don't think we should rely on classifying people like this should be the way we run an International tournament.
|
|
|
| Darth_Wafu
- 14 Aug 2011
Total Posts: 117
|
The reason that players need to be classified for international level tournaments is because their entry fee is determined by skill classification. Any given promoter could choose not to do it that way, but it is pretty standard procedure for our major tournaments and I see no good reason to go away from it. Once again, the promoter can use whatever method they like to denote skill levels for entry fee purposes, but a rating system could be a very useful tool. Plus, it would add transparency to the tournament to help prevent people from feeling like they are being unfairly charged too much. I am not aware of this happening now, but it is an issue that could come up at some point.
Nick Geoffroy
|
|
|
| airhockeyjedi
- 16 Aug 2011
Total Posts: 102
|
Just a bit on Ultra Pro.
This term was started by me many years ago during the HAP Chapter USAA when I was running tournaments both weeklies and majors. This classification was incorporated to designate players who were master prior to a US/Nat'l/WC/Int'l tournament in which they didn't compete. So then an Ultra Pro is a pseudo-master in terms of classification. The classification master is an honorary title for those who are in the top ten.
When Owen G was helping with the seeding of the tournaments, he understood and used this bit of classification knowledge and seeded players according.
Players like Keith Fletcher are now Ultra Pro Classification (pseudo-master). Players like Syed are now Pro Classification since they finished out of the top ten and they competed in the tournament.
Ultra Pro was created because more than ten players deserved to be recognized to be playing master level. It is not something you keep forever. Players like myself who consistently finished master previously but now don't are Pro's at best. A player's classification is earned and fluctuates according to his performance.
Vince Schappell
2007-2009 USAA Treasurer
USAA Member since 1980
|
|
|
|
| Darth_Wafu
- 16 Aug 2011
Total Posts: 117
|
airhockeyjedi said: A player's classification is earned and fluctuates according to his performance.
|
I think that should read "should fluctuate" because I am not so sure it does the way we seed now
Nick Geoffroy
|
|
|
|
| Q
- 16 Aug 2011
Total Posts: 497
|
Darth_Wafu said:
I think that should read "should fluctuate" because I am not so sure it does the way we seed now |
Agreed. And as has been suggested previously, what if the ratings were used to define the classifications? I'm not going to speculate what ranges would be what levels (Ams, Pros, Ultrapros, etc.) but if the classifications were defined by the Mitic rating and the Mitic rating fluctuates with performance then your classification would follow.
|
|
|
|
| Mike C
- 16 Aug 2011
Total Posts: 459
|
| I agree also. The classification is what people comprehend in Houston. What are you? A level 3, a level 4 etc. If I tell someone their Mitic rating is a 1320 today, the response currently will be not comprehend what that means. If they know that is classified as a level 2, they may be upset. To get to a rating system people need to know and speak in the rating language. A person's rating will fluctuate upwards and downwards over time. I may have a rating of 1520 today, but it may be 1400 next week if I don't play well. A big difference to the I'm a 4 and I'll always be a 4 no matter how poorly I am playing!
|
|
|
| travis
- 16 Aug 2011
Total Posts: 530
|
Mike C said: | I may have a rating of 1520 today, but it may be 1400 next week if I don't play well. A big difference to the I'm a 4 and I'll always be a 4 no matter how poorly I am playing! |
Not necessarily true. If you are at a certain level and you play poorly over a long period of time, or continually go two-and-out in weeklies no matter how hard you try, there's a possibility of getting demoted. It's usually at the player's request, and the people running the weeklies review the player's recent history. Demotions aren't common, but they happen.. usually after a too-hasty promotion or someone comes back after a hiatus. Jason Sherman was promoted to a level 5 after he finished Master in Vegas a couple years back, but I think he's playing back at as a level 4 now, right? On the other side of that coin, new players are usually promoted fairly quickly, at least up the lower levels from 0 to 1 to 2.
It's just a matter of figuring out what range of ratings might fit within the ideas of our classifications. That range of ratings may adjust slightly over time, but that could easily be reviewed and updated. At the very least, it would be a lot easier to determine which rating values are roughly associated to classifications than it would to go through every single player in the AH database before each and every tournament and figure out what classification they are.
Travis Luscombe
AirHockeyWorld.com Webmaster
http://twitter.com/air_hockey
|
|
|
|
| ajflanagan
- 16 Aug 2011
Total Posts: 509
|
With anything in life, change can be difficult and takes some adapting. But to completely turn your head to the idea of change because you are comfortable with the way things have been is not so smart. The system is clearly broken and it is clearly not going to function for a growing and evolving player base.
Despite the Mitic Ratings not being "official", I have been conscious of my position all along. I am struggling to get my rating back up over 1600 where it was this time last year. Where I feel I belong. I encourage everyone to do the same. Study your rating. Look at the players above and below you. Play some sets and REPORT them. See how it effects your position. I know that tons of sets are being played in Houston... and none of them... not ONE has been reported that I'm aware of. Even by players who are in suppport of the Rating system!
|
|
|
|
| carolina phil
- 18 Aug 2011
Total Posts: 1084
|
From this discussion, the concerns of the community over the importance of seeding exemplifies the fact that it cannot be left in the hands of promoters or regional biases, whether in California, Texas, Illinois, or Colorado. The factors that constitute the process (ranking, classifications, 3.5 calculations)need to be applied in the context of USAA certification.
Otherwise, the community of players will never have the confidence to know that the criteria have been applied according to the guidelines of the community as expressed through the Association.
Andrew and others are correct to point out the classification process has evolved/de-volved now from a more clearly defined process (as Owen, Vince, Mark, and I had) to a more subjective tool that could be used to overturn both the Mitic ratings and the current Ranking methods. After all, if promoters declare a player "better" than their rating points or ranking number and elevate that player to a higher cluster of seeded players, then the system that the community sanctions could very well be violated.
Therefore, logic dictates that future events should continue to have a seeding list that is posted and approved by the USAA in advance of the event. This seeding list should include a statement of classifications for each participant, and an explanation of how their classification has affected any change in their numerical ranking, rating, seeding.
Once the USAA tournament seeding committee reviews this proposed seeding list, and determines that the method employed represents what the community has authorized, then that list can be certified for use in the event.
This does two things: 1. forces, as Brian urges, the USAA to write down its authorized "classification" criteria. 2. assures the worldwide player base that major events are seeded according to player expectation no matter where or who holds the event.
Phil
|
|
|
|
| travis
- 18 Aug 2011
Total Posts: 530
|
carolina phil said: | Therefore, logic dictates that future events should continue to have a seeding list that is posted and approved by the USAA in advance of the event. This seeding list should include a statement of classifications for each participant, and an explanation of how their classification has affected any change in their numerical ranking, rating, seeding. |
This has been impossible in the past, in that the seeding process requires knowing all of the players that are playing in the event in order to determine the "largest groups" of each player classification to do the classification "move ups". So, you can't fully do the seeding until registration is closed.
What you can do is provide an "unofficial" seed list prior to the close of registration, showing how the seeds are expected to pan out. But until you know every name on the chart, you can't finish the seeding process.
| Andrew and others are correct to point out the classification process has evolved/de-volved now from a more clearly defined process (as Owen, Vince, Mark, and I had) to a more subjective tool that could be used to overturn both the Mitic ratings and the current Ranking methods. |
The USAA seeding method has not been changed in over 10 years, so I'm not sure what you mean by "de-volved". The current USAA seeding system has always had a subjective aspect to it. Go re-read Brian's post of his written procedures, and tell me that it doesn't.
http://www.airhockeyworld.com/forummessages.asp?forum=11&thread=945&page=4
| After all, if promoters declare a player "better" than their rating points or ranking number and elevate that player to a higher cluster of seeded players, then the system that the community sanctions could very well be violated. |
The IAHC seeding was run under the same seeding guidelines as the previous two nationals at SRO. The seed list was presented to all available tournament committee members appointed by the USAA prior to drawing the main tournament bracket.
I agree that the classifications need to be defined, and it probably wouldn't hurt to have some kind of review of players' classifications prior to the tournament if we continue to use the same seeding process. Until then, it's still a judgement call. That's the way it's been since I started playing. And as far as "worldwide player expectation" goes, there are always going to be players that disagree on the skill levels of particular players.
Travis Luscombe
AirHockeyWorld.com Webmaster
http://twitter.com/air_hockey
|
|
|
|
| carolina phil
- 18 Aug 2011
Total Posts: 1084
|
.....So, you can't fully do the seeding until registration is closed......
Yes, that is what we have done. But we can improve the situation by having a preliminary list that includes PROPOSALS for player classifications. That list can be generated prior to the event by local directors for the USAA seeding committee to provissionally certify. Then after Registration closes, the USAA seeding committee can made a final certification based on the imput from the local directors/promoters/etc.
......The USAA seeding method has not been changed in over 10 years, so I'm not sure what you mean ...........
I have been to all national/world tournaments and have identified a tendency to "correct" numerically earned rankings by the use of subjective classifications by a smaller number of decision makers than we had in the past. We usually used regional reps to serve on the tournament committee which also supervised, or was involved, in the seeding process.
http://www.airhockeyworld.com/forummessages.asp?forum=11&thread=945&page=4
[
.......The IAHC seeding was run under the same seeding guidelines as the previous two nationals at SRO. ........
Although that was helpful, it was done after registration had been closed for two days; and was done hours after the board meeting. A future goal would be to submit the classifications at the Board Meeting so that there is time for consideration. Regional reps could give some really valuable imput on players they know better. The Seeding Committee could remain after the meeting to meet with local reps and promoters to finalize the list.
"...........And as far as "worldwide player expectation" goes, there are always going to be players that disagree on the skill levels of particular players. |
............."
They may disagree on specifics, but they will all agree that the USAA Seeding Committee had a good look at it before implementation.
I will do my best to be sure that the USAA will step its game up to help formulate criteria for classification and to forge a pro-active Seeding Committee. I would like the officers to serve on that committee along with regional "ad hoc" member volunteers.
Phil
|
|
|
|
| goran
- 18 Aug 2011
Total Posts: 428
|
| There are a lot of adjustments needed if we continue to use the current ranking system. If we choose to use the rating system, we don't need to discuss all the ways we can make the seeding process more fair with rankings. But for shits and giggles here they are... seed everyone with a current ranking, classify all players based on their past tournament finish or other non subjective criteria. Place all non ranked players below the largest group of the same classification in order of their last tournament finish with ties going to most recent tournament. This will be the master seed list and no other adjustments need to be made depending on who shows up. Get rid of the 3.5 penalty. If we can't come up with a fair way to classify players, or we don't want to classify and reclassify 700 players, just use the ratings and be done with it.
|
|
|
|
| carolina phil
- 18 Aug 2011
Total Posts: 1084
|
Thanks for the step by step method to improve our current traditional seeding process. And, your flexibility in adjusting the Mitic ratings sytem may now include the most important elements of the current traditional system. I am impressed with the positive implications.
I look forward to imput from others on both sides, in private or here, on whether this modified rating system incorporates the best of both worlds so that what emerges satisfies the concerns of all and maintains the advantages of each seeding system.
Appreciatively,
Phil
|
First |
1 | 2 |
Last
Forums Home / Tournaments and Challenges / 2011 IAHC Seed List
|
