MUT Mirror 4/2 @ NGCSU

Old college threads.
User avatar
The Ununtiable Twine
Auron
Posts: 1058
Joined: Fri Feb 02, 2007 11:09 pm
Location: Lafayette, LA

Re: MUT Mirror 4/2 @ NGCSU

Post by The Ununtiable Twine »

NGTech wrote:
And now for a brief comment. This comment is my personal opinion.
To the comments on winning. I come originally from a FIRST team before quizbowl. While on that team I learned the values that I use now which contains not valuing winning as the major goal. While on a FIRST team I aided opposing teams to the point of scoring for them and helping with repairs that let them beat my team, but in FIRST helping is more important than winning. I found this to be very fun as all teams try their hardest to be the best, but it never came at the expense of hurting another team. As such the highest honors in FIRST are not to teams who win, but to teams that help and show the best attitude.
So we're supposed to sit on questions. So we're supposed to neg and then give the other team the answer. So we're supposed to give the other teams bonus answers to help them boost their scoring at the risk of losing. It's nice that the highest honors in FIRST are as you've stated them to be, but to be quite honest, they're irrelevant. Teams can show the best attitude and win at the same time. You have no idea how many teammates I've scolded for behaving unnecessarily. FIRST has nothing to do with quizbowl. Nothing.
Aaron Carney wrote:RealHermEdwards is a prime example of not reading the middle sentences.
He was invented because of you! How do you not get that?
Aaron Carney wrote:As a mass response to several others. I'm sure most everyone is aware of the treatise on how nothing but the most basic argument can be proven.
I'm a mathematician. I can vouch for having proven things that would qualify as nontrivial. Same for Jerry, Jon, and maybe some of the others that have posted in this thread as I do not know everyone's educational background. As someone with a degree in a logic-based field speaking to someone who more than likely does not have such a degree (although I cannot "prove" it), I hereby state that you are incorrect, and that will be taken as a fact by every member of this forum who values anything about logic. Your education in the area of logic is mind-boggling and horrid.

If you can name the treatise, then you'll get 10 points and a bonus. If not, 0. But you just might be better off negging because it will help with the tiebreakers.
Jake Sundberg
Louisiana, Alabama
retired
User avatar
Mechanical Beasts
Banned Cheater
Posts: 5673
Joined: Thu Jun 08, 2006 10:50 pm

Re: MUT Mirror 4/2 @ NGCSU

Post by Mechanical Beasts »

Aaron Carney wrote:This will likely be the last time I post in thread as I find it a moot point to argue with people only looking to argue, which is one of the ways some people swim in faulty logic.

Please do not mix what I say with an argument someone else is making. I am only one person.
Please read all of what I say and not skim over the first few sentences and maybe last few.
Please do not assume you know the definition to every word in your vocabulary, you do not use them all correctly.

I say this, and its a problem on many forums, people rush in and respond without putting much thought to what they have read. I apologize to some but there's a certain level of intelligence required to have a conversation with me as I assume a certain level mental activity to be able to analyze and evaluate information on the fly, or at least take time to process it all before responding.
I'm suited up. Try me.
First, I was responding in defense of the use of quality points and then explaining the purpose of quizbowl. I was not arguing for the use of quality points in the MUT or even arguing on behalf of my teams leadership.
So, you were "responding in defense of"--i.e. defending--your new stat. But you were not "arguing for [its] use." What on earth? More importantly, you're "explaining the purpose of quizbowl"... I assume you omitted "as I see it" or something like that, since the tremendous majority of the quizbowl community appears to disagree.
I will openly admit that they screwed up by not informing teams ahead of time that this system would be used to determine ties. I also admit that a mirror should be done with at least a very similar set of rules, at the time of the event I did not know this was a mirror and would have advised otherwise had I known. Do not assume that there are two sides to any argument or even that an argument is restricted to one, two or even three dimensions.
I am glad that you are conceding that you were wrong in the dimensions of the argument that don't have to do with quality points.
Second;
Putting aside the rest of your post, this is also incredibly wrong. The point of quizbowl is to know more than the other team, and therefore win. Is there a social aspect to it? Sure. But if the point is not to win, then why even keep score? I also have no idea what the last sentence means.
Analyze and evaluate...and go enjoy a dictionary or get a psych eval, I'm still hoping no one here needs one.
If you think that there is a point in engaging in the following process:
1) decide on a format based on which there will be determined a tournament winner, and indeed a winner of every game
2) decide on a tiebreaker statistic that attempts to measure how little a team tried to win, and rewards that team for its "sportsmanship"
then you need a psychological evaluation more than any who suggests that we might want to properly determine a winner. You've turned a debate about the merits of a tiebreaker into a referendum on "is competition good?" And, see, if this weren't quizbowl you would smack yourself in the head and say "yes, obviously college-age individuals will not be psychologically harmed by keeping score in a meaningful way when they get together to play basketball." But since this is quizbowl, and one element of quizbowl involves intellectual, rather than physical, pursuits, you have an angle: you can quietly assume that learning is socially isolating or something and suggest that an environment that encourages "trying to learn the most stuff" instead of "trying to learn the most stuff, and then lose narrowly" might actually be somehow awful for you. Try again.
RealHermEdwards is a prime example of not reading the middle sentences. Professional sports has very little parallelism to Quiz Bowl. Their jobs, income and livelihood depend on winning, the more they win, the better their off. We don't get paid to win. We don't get scholarships to win. Yes the purpose of professional basketball is to win. You are hired based on your ability to win the game. The owners increase their money based on their teams winning. One might however make the argument the purpose of professional basketball is to make money but I believe that would be classed as a result.
So, your argument is "if we got paid to win quizbowl, it might be justified to do something horrible like break ties in a meaningful fashion. Since we don't get paid, we should avoid the dread harm to our souls that would result." (With the scary implication that we ought so sacrifice our souls for $$$, which is totally out of keeping with the rest of your feel-good tone.
Now some of you are confused because you hurriedly assumed that purpose and point mean the same thing as goal or aspect. The purpose of the military is not to win battle's, the point of the military is not to win wars. Those are goals, aspects, objectives of the military, not the objective or point. The point and/or purpose of the military is to protect the nation (this is an example on the use of wording).
Okay, so now you're entering into nitpicking semantic lunacy. Having read this, I assume your response to my last objection will be "no, that wasn't my argument. That was my contention. My argument was flibberty gabber ding dong."
Yes, if we are debating or arguing then you should take note in the terminology and what it means. I find it ridiculously amusing that here at the forum for quiz bowl, that terminology which is an important part of being able to answer questions early on for power is not something many people are aware of or even attempt to use.
As someone who has powered many questions, I will tell you that failing to understand your bogus semantic distinctions between purpose, point, goal, and aspect has never actually gotten in my way.
It might have taken some serious analyzing and evaluating to have realized this as I sometimes use people's emotions so they see for themselves when they lass out.
What? Oh, lash out. Right.
Andrew Watkins
User avatar
Marble-faced Bristle Tyrant
Yuna
Posts: 853
Joined: Fri Feb 08, 2008 11:05 pm
Location: Evanston, IL

Re: MUT Mirror 4/2 @ NGCSU

Post by Marble-faced Bristle Tyrant »

Jesus Christ, stop hiding behind semantic arguments to avoid addressing whether the quality rating was terrible.
Farrah Bilimoria
Formerly of Georgia Tech and Central High School (Macon)
User avatar
The Ununtiable Twine
Auron
Posts: 1058
Joined: Fri Feb 02, 2007 11:09 pm
Location: Lafayette, LA

Re: MUT Mirror 4/2 @ NGCSU

Post by The Ununtiable Twine »

Furthermore, I agree with everything Watkins said. The only thing missing from this thread is Mr. Dees, and I'm probably going to agree with everything he's going to say as well as soon as he gets here. That should tell you something.
Jake Sundberg
Louisiana, Alabama
retired
User avatar
Alejandro
Wakka
Posts: 226
Joined: Mon Jul 10, 2006 8:39 pm
Location: Seattle, WA

Re: MUT Mirror 4/2 @ NGCSU

Post by Alejandro »

NGTech wrote:As requested I used the 2010 ACF Nationals to test the stat we used. To test only the stat I did no sorting based on W-L.

As such the statistic was able to predict the Preliminary outcome to and overall accuracy of 71%. This is better than the calculated values for other tie breakers, and this is the stat ranking the field not breaking ties.
Using PPB predicts the winner ~90% of the time in the prelims of 2010 ACF Nationals (72/78 correct in Red and 69/78 correct in Black). Granted, this was calculated without removing the game's effect on the PPB, but it seems unlikely that this would decrease its accuracy by 20% or more.
Alejandro
Naperville Central '07
Harvey Mudd '11
University of Washington '17
User avatar
Mechanical Beasts
Banned Cheater
Posts: 5673
Joined: Thu Jun 08, 2006 10:50 pm

Re: MUT Mirror 4/2 @ NGCSU

Post by Mechanical Beasts »

Alejandro wrote:
NGTech wrote:As requested I used the 2010 ACF Nationals to test the stat we used. To test only the stat I did no sorting based on W-L.

As such the statistic was able to predict the Preliminary outcome to and overall accuracy of 71%. This is better than the calculated values for other tie breakers, and this is the stat ranking the field not breaking ties.
Using PPB predicts the winner ~90% of the time in the prelims of 2010 ACF Nationals (72/78 correct in Red and 69/78 correct in Black). Granted, this was calculated without removing the game's effect on the PPB, but it seems unlikely that this would decrease its accuracy by 20% or more.
Well, he never said all other tiebreakers. He only tested shoelace length and girth; what more do you expect?
Andrew Watkins
Aaron Carney
Kimahri
Posts: 4
Joined: Sun Apr 03, 2011 9:48 pm

Re: MUT Mirror 4/2 @ NGCSU

Post by Aaron Carney »

No I don't have a degree in logic, but I will be taking this thread and showing to several professors of mine who are equally amused by these things but do not frequent forums as much as I do. Isolating individual sentences and paragraphs to suite your needs does not prove your side. I see a large number of uninhibited trolls enjoying a good flaming and am rather shocked to see such a thing happen on a forum such as this.

I am not my teams president, I am not part of its leadership, why do several of you insist on tying me down into those roles?

I'm arguing in the context of NGCSU's use of quality points, not the use of quality points in a Mirror, I don't see whats so hard to understand? This is not the first time we've used the method.

No myself and NG_Tech are not sitting next to each other however we have identical MBTI profiles. We have fewer then 7 posts so a moderator has to approve them.

Why do some of you insist on thinking we are ignoring you, I found several individuals very helpful and will be using their reasoning when strongly suggesting how the teams leadership changes things. However most everyone had no intention of providing help and seem to prowl the forums looking for prey to verbally abuse to make themselves feel better. Again, I stated I'm not part of the teams leadership, Why do you accuse me of saying what someone else has said?

Yes, it's my opinion on what the purpose of what Quiz Bowl is. So yes I was wrong in implying it was fact, though I never stated it was and left it open to other possibilities (you would needed to have read everything). However using an incorrect meaning of a word in an argument is much more wrong.

Crazy Andy Watkins, please do this without anger or emotional turmoil, print out a list of logical fallacies and examine your post. You might need to review my two previous posts while your at it. I ask you do so because you did say you were "suited up".
Jesus Christ, stop hiding behind semantic arguments to avoid addressing whether the quality rating was terrible.
Stop bringing up individual examples and personal experiences as reasons why its bad.

User was banned for three days for telling others how to post and ad hominem attacks. - FM
North Georgia College and State University
User avatar
Mechanical Beasts
Banned Cheater
Posts: 5673
Joined: Thu Jun 08, 2006 10:50 pm

Re: MUT Mirror 4/2 @ NGCSU

Post by Mechanical Beasts »

Aaron Carney wrote:Crazy Andy Watkins, please do this without anger or emotional turmoil, print out a list of logical fallacies and examine your post. You might need to review my two previous posts while your at it. I ask you do so because you did say you were "suited up".
I'm sure those professors you mentioned will be proud that your best rebuttal to what I say is "you are angry and emotionally conflicted; please review your post for fallacies" rather than... responding to anything I've said?
Andrew Watkins
User avatar
Xfact115
Kimahri
Posts: 4
Joined: Thu Sep 25, 2008 11:58 pm

Re: MUT Mirror 4/2 @ NGCSU

Post by Xfact115 »

Dudes, you do not get to use your status as a new team as a defense for your mistakes and then continue to defend said mistakes. As this thread continues everything you guys say makes less and less sense. I don't really know what FIRST is and I don't know why it should have anything to do with quizbowl. I certainly don't know where you guys find the gall to suddenly change the fundamental rules of the game out of the blue without notifying anyone or how it's possible that you still haven't realized that the quality stat is not suited to quizbowl. I find it interesting that you guys spent so much time working on this stat and fine tuning it and didn't take two seconds to make sure the teams coming to your tournament were aware of this fundamental change to the standards and practices of quizbowl. This whole discussion should have ended a while ago with you guys admitting that you screwed up and at least a couple people posting “okay, we forgive you, we actually did have fun at this tournament overall, hope you guys can host some more, blah blah blah.” Instead you're digging yourself a bigger hole and making it unlikely that schools will attend any future tournaments at your university.

Too put it bluntly, USC did not pay for a day at a Montessori school, we paid for a tournament. At the risk of sounding like Gordon Gekko pretty much everything good about quizbowl comes from the desire to win. If I didn't want to win tournaments I wouldn't know half the things I know now, because winning requires intensive and close study of the material. And this knowledge comes in handy outside of quizbowl all the time, which is possibly the thing I love most about the game. The desire to win in no way conflicts with networking with intelligent people with similar interests, or learning interesting things, or having fun. It actually does the exact opposite. I thoroughly enjoy winning and when I play I play to win. But USC is attending Nationals where we will lose many times, probably by large margins. And I intend to enjoy every second of that too because I have plenty of fun when I'm losing. I want to watch the best players in the game do their thing, because I know that they got where they are because they wanted to be good at quizbowl and win tournaments. And I know that the teams who know more than we do deserve to win against us by a margin proportional to just how much more they know. The thing about games is that they involve winning and losing. If you can't handle that find another activity.

I also don't know where you guys get the idea that the quizbowl community is made up of Machiavellian cutthroats who are all cruel to each other and need you to come in and teach us how to stop being barbarians. As far as I can tell the quizbowl community is pretty strong. Players do help each other, largely by posting relevant information on the message boards and there is plenty of information and help available for new teams in particular. In fact, the posts in this thread have overall been nothing but helpful, at least at the beginning. Quizbowlers can get pretty vicious but only if the situation calls for it. You guys continue to defend practices that are entirely unacceptable and at one point stated that most of us need mental help. Please do not act like you are saints getting ganged up on by a bunch of jerks. People from all over the country have taken time to offer constructive advice and criticism which is then just thrown back in their face. Pretty much any of them would have been glad to offer advice before the tournament too. And I've never had an issue with sportsmanship at tournaments themselves. I've never known a team to gloat or to be dismissive of less skilled teams (except for perhaps out of anger during forum arguments). And any time that I've asked a player where they got something or how they knew something they've been more than happy to talk about it with me.

In conclusion: quizbowl is not broken, don't fix it. Quizbowlers are very nice to people who aren't trying to ruin quizbowl but do not suffer fools gladly. And as atrocious as the quality stat is the philosophy behind it is twice as bad and nobody else agrees with it or will ever agree with it. Let it go, apologize, and move on.
Robert Harden
University of South Carolina
SamJHarb
Kimahri
Posts: 2
Joined: Mon Apr 04, 2011 3:40 am

Re: MUT Mirror 4/2 @ NGCSU

Post by SamJHarb »

As Chris mentioned earlier in his apology, which I hoped would end this tirade, two more experienced players will be coming in to make sure things stay on track for the foreseeable future. I am not a master of Quiz Bowl or a veteran director, but I am an alum of NGCSU and the previous VP of the their team when it formed. The former President, the other player mentioned, and myself were responsible for orchestrating the first tournaments run by NGCSU which I remember being great successes. My friends have admitted their lack of experience and willingness to improve and learn from the mistakes that were made, including the use of the new scoring system. NG_Tech had a good idea, and it was run in the background originally and actually used in a HS tournament with precise accuracy when compared against the traditional systems. I am not saying it will work every time and for all I know, because I do know my share of math, it could conceivably never work again with the same consistent results. For this reason Chris said it will not be used in any official capacity until it can be shown with definitive proof that it worked to the satisfaction of the Quiz Bowl community. Aaron does not speak for us, which he has admitted. He has made the mistake of making this personal and engaging in the same tired banter that helps no one. If anyone has any further constructive criticism to add, it is always welcome. I apologize for Aaron and he has been asked to cease his posting, but I also ask that the unnecessary back and forth also stop. I would like to offer a most sincere apology to MUT and all its editors for this tournament being run the way it was without informing both them and the competing teams first. I can assure with complete certainty that it will not happen again. Thank you for your time.
Sam Harbison
Gainesville High School '07
NGCSU Physics '10
User avatar
theMoMA
Forums Staff: Administrator
Posts: 5993
Joined: Mon Oct 23, 2006 2:00 am

Re: MUT Mirror 4/2 @ NGCSU

Post by theMoMA »

In a round robin, everyone plays the same schedule. Essentially, everyone's schedule is equally difficult. Your stat rewards teams that make an equally difficult schedule "more difficult" by performing worse against the common opponents. How can you not see the complete insanity in this?
Andrew Hart
Minnesota alum
SamJHarb
Kimahri
Posts: 2
Joined: Mon Apr 04, 2011 3:40 am

Re: MUT Mirror 4/2 @ NGCSU

Post by SamJHarb »

Xfact115, Chris has already apologized, twice, and now I have added my piece. As for theMoMA, it will not be used unless there is certainty it will work. There is an obvious disconnect between the mentality of FIRST and that of Quiz Bowl, so while it was interesting to compare, we acknowledge it should never have been used in actual tournament play of this level without more data. Quiz Bowl is about learning, excelling and if you can, winning, absolutely. Who doesn't love winning? This system will only be used in unofficial capacities, likely never to again see full use in a tournament.
Sam Harbison
Gainesville High School '07
NGCSU Physics '10
nadph
Rikku
Posts: 256
Joined: Mon Jun 08, 2009 4:21 pm

Re: MUT Mirror 4/2 @ NGCSU

Post by nadph »

EDIT: I had this post sitting for a long time and didn't see some of the later comments. These are just my own thoughts as to why the quality metric wouldn't be as useful here; I am not trying to be vitriolic towards NGCSU or anybody else. I am sorry if this post comes across as such.

If I recall correctly, FIRST's ranking formula awards something like [winning score]+2*[losing score] for the winning team, and [winning score] for the losing team. I don't think it actually involves the ratio of scores. I didn't see the actual formula for the metric used here posted anywhere in the thread, so my apologies if something different was used.

Anyway, in my opinion as a member of a FIRST team (for whatever that's worth), the FIRST metric has little to no applicability to quizbowl. The point of the metric in FIRST is to (a) reward teams that by chance play more competitive schedules (and have less of a chance of doing spectacularly), since pairing of teams is essentially random and each team plays only a small subset of other teams in the competition; and (b) to reward teams that create high-scoring matches for both sides and are willing to score for the other team to make that happen. Neither of these have clear analogues in quizbowl.

"Strength of schedule" correction is necessary in FIRST because initial team pairings are done essentially randomly before the competition begins, and (since the average FRC regional tournament attracts about 80-90 teams) there is no realistic way to allow any team to play with or against any more than half of the other teams competing (in the best case). Even within FIRST, I think many regard the formula as something that's pretty arbitrary and merely "better than the obvious alternative" (i.e. just valuing the winning alliance's score, which give equal weight to blowouts and close victories), and feel often constrained by it. Furthermore, this "quality metric" isn't used for the final rankings, but only serves to seed the teams picking playoff alliances; most FIRST teams recognize that this metric can and does produce bullshit results, and usually ignore them in favor of their own data when making selections.* As Andrew said, trying to correct for "strength of schedule" in a quizbowl tournament is in most cases pointless: except for pretty ridiculous formats, and occasional really large tournaments like the HSNCT (which sometimes get Swiss-paired), essentially every tournament will consist of a round robin on either the entire field or brackets formed from that field, followed by tiebreakers, rebracketing if necessary, and a final/3rd place game if necessary. There is no need to correct for strength of schedule when the schedule is the same for everybody, plus/minus the two teams being compared. The only time this metric could have conceivable use would be when comparing teams across brackets, but because in general TDs don't mess up brackets so that a disproportionate number of highly competitive teams are in one of them, stuff like PPB usually makes for a better indicator.

The idea of "coopertition" (FIRST's word for sportsmanlike conduct in competition) valued by the metric also cannot find a sensible analogue in quizbowl. The FIRST metric essentially wants strong teams to help weaker teams score, by giving them scoring objects, assisting their robots, or even scoring in their areas if the rules permit it. On the other hand, you cannot cannot buzz in during a quizbowl match and say "I'd like these points to go to the other team," or provide the correct answers to bonus parts the other team doesn't know, without invalidating the competitive aspect of the current paradigm of the competition. If you feel that quizbowl doesn't promote cooperation within rounds and want to use this metric to change it, I can understand your view, although I feel that a change to promote this would not be a net positive. But the metric is not advantageous if it tries to reward something which cannot happen within a match. There are plenty of other ways people can help: without exception, the best high school and college teams out there are enthusiastic at welcoming new teams and players, teaching them about the game, and showing them how to improve as players - this forum alone is a testament to that. Giving strong teams incentives to score low against weak or new teams can be harmful, both because it prevents the better team from scoring highly and showing others what they know, and because it prevents the weaker team from seeing how strong, experienced teams play.

Sorry for the long post. One more thing, the FRC national tournament this year is in St. Louis and not Atlanta, as far as I know.

*In FIRST, all matches are 3-on-3, with each group of 3 being called an "alliance." A bunch of qualification matches are played, with random selection of teams to form alliances for each game, and two quality metrics are calculated (one for the winning alliance, one for the losing) and given to the appropriate teams. A team's "seed" is determined by summing its calculated quality in all matches it has played; the top 8 seeds, in order, pick two "alliance partners" for single-elimination playoffs, where each pairing is best-of-three.
Nikhil Desai
Bellarmine College Prep '12
Stanford '18 (leave of absence 2016-17)
NGTech
Lulu
Posts: 13
Joined: Sun Apr 03, 2011 12:30 am

Re: MUT Mirror 4/2 @ NGCSU

Post by NGTech »

Alejandro wrote:Using PPB predicts the winner ~90% of the time in the prelims of 2010 ACF Nationals (72/78 correct in Red and 69/78 correct in Black). Granted, this was calculated without removing the game's effect on the PPB, but it seems unlikely that this would decrease its accuracy by 20% or more.
I made my comparison based on data from a thread linked to earlier. I also did not use the stat to predict outcomes of matches, only the end ranking. I may try a match by match check in the next day or so. I know from just watching as I entered the data it predicted every match I checked. Granted I think it has a flaw with teams that lose more but have high overall score games.

Thank you for this additional information.

I also plan to start a discussion of my equation in the theory section in the near future.
Nicolas A. Collins
North Georgia College Academic Team Technical Director
User avatar
jonpin
Auron
Posts: 2266
Joined: Wed Feb 04, 2004 6:45 pm
Location: BCA NJ / WUSTL MO / Hackensack NJ

Re: MUT Mirror 4/2 @ NGCSU

Post by jonpin »

nadph wrote:If I recall correctly, FIRST's ranking formula awards something like [winning score]+2*[losing score] for the winning team, and [winning score] for the losing team. I don't think it actually involves the ratio of scores. I didn't see the actual formula for the metric used here posted anywhere in the thread, so my apologies if something different was used.
Thanks; I'd tried to find the metric on the FIRST website, but had no success. I believe the metric used here was posted Saturday at 11pm ET:
North_GA_ATeam wrote:The basis of this statistic is to determine the difficulty of a match/round based on the relative scores of the two teams. That is, to compare how close those two scores are to each other in order to determine a relative level of ability and thus, a relative level of difficulty.

The statistic is based on a simple percent difference calculation. The points scored by the opposing team (B) is subtracted from the team’s (A’s) score. This is divided by the sum of both scores (the total amount scored by A and B). A constant of 1 is added in order to adjust the normalization.

What this means in practicality is that if Team A scores 20 and Team B scores 10 then we would, to determine Team A’s difficulty rating, subtract 10 from 20 (getting 10) and dividing by the sum of both scores (30) getting (1/3). We then add the constant in order to normalize the scores (getting 1.333). To determine Team B’s, we do the same thing for B. Taking 10-20=-10. Then dividing it by 30 getting -1/3. Adding the constant, however, returns us into the wonderful realm of positive numbers (which is why this constant exists), giving us a difficulty rating of .667.

If you notice, the sum of both difficulty ratings for ANY match will equal 2. (Without the normalization, it would equal zero.)

What this means in terms of application is that, the winner’s number will always be 2>x>1 and the loser’s number will always be 1>x>0. The closer the difficulty rating is to 1, the more evenly matched the two teams are. The lower the number, the more difficult that particular match was for that team. The higher the number, the easier that particular match was for that team.

Because of the way that this statistic accurately and fairly illustrates difficulty, we “reward” teams who have proven to have more difficult matches.
In other words, Q-rating for Team A = (A-B)/(A+B) + 1 = 2A/(A+B) = 2*(the fraction of points scored by A) = 2*R/(R+1), where R is the ratio of A's score to B's score.

I'll again note that it's still not 100% clear to me whether a higher Q-rating or a lower Q-rating was privileged by this tournament. However, I took the ten minutes to enter all 28 games into a spreadsheet and calculate the ratings. USC's score is 9.892 and Georgia's is 8.956. Chatta's score is 8.756, GT's is 8.313. So it looks like a lower "difficulty rating" was preferred. Wacky.
Additional insane incentive inspired by this insane metric: Once victory is out of reach, mathematically or practically, the best strategy is to stop trying and neg (if possible). Conversely, the winning team should stop trying to extend their lead, which would increase and therefore damage their Q-rating, so they should be trying to neg.

Did you not see how this was fucking crazy and completely antithetical to any idea of sportsmanship or learning or anything useful? Did you not spend ten minutes looking at the possible consequences of your tiebreaking system?
Jon Pinyan
Coach, Bergen County Academies (NJ); former player for BCA (2000-03) and WUSTL (2003-07)
HSQB forum mod, PACE member
Stat director for: NSC '13-'15, '17; ACF '14, '17, '19; NHBB '13-'15; NASAT '11

"A [...] wizard who controls the weather" - Jerry Vinokurov
User avatar
Mechanical Beasts
Banned Cheater
Posts: 5673
Joined: Thu Jun 08, 2006 10:50 pm

Re: MUT Mirror 4/2 @ NGCSU

Post by Mechanical Beasts »

jonpin wrote:Did you not see how this was fucking crazy and completely antithetical to any idea of sportsmanship or learning or anything useful? Did you not spend ten minutes looking at the possible consequences of your tiebreaking system?
Their idea wasn't sportsmanship. Their concept was sportsmanship. Their idea was winning battle's.
Andrew Watkins
NGTech
Lulu
Posts: 13
Joined: Sun Apr 03, 2011 12:30 am

Re: MUT Mirror 4/2 @ NGCSU

Post by NGTech »

jonpin wrote: I'll again note that it's still not 100% clear to me whether a higher Q-rating or a lower Q-rating was privileged by this tournament. However, I took the ten minutes to enter all 28 games into a spreadsheet and calculate the ratings. USC's score is 9.892 and Georgia's is 8.956. Chatta's score is 8.756, GT's is 8.313. So it looks like a lower "difficulty rating" was preferred. Wacky.
I know my terminology was very confusing, I tend to phrase things different from the norm.

The idea behind picking the lower Q-rating was that a lower number corresponds to a team that had harder fought/tougher matches.

The design of the stat was based around rewarding the team that had to work the hardest for their wins.

I would also like to point out while the team stat favored closer wins the individual stat favored the player with the best ratio of points scored to all tossup points in game. This ratio is usually improved through high scoring matches where one player scores most of the points.
Nicolas A. Collins
North Georgia College Academic Team Technical Director
User avatar
AlphaQuizBowler
Tidus
Posts: 695
Joined: Mon Dec 03, 2007 6:31 pm
Location: Alpharetta, GA

Re: MUT Mirror 4/2 @ NGCSU

Post by AlphaQuizBowler »

Something I'd like to add about the "quality" or "difficulty rating," as a player at this tournament: nobody actually adjusted their play to get a better "quality," because nobody knew what "quality" was until the end of the tournament. So if the point was to promote "sportsmanship" (defined, I guess, as beating teams by the lowest possible margin) then it didn't work at all.

Anyway, I've attached the spreadsheet that NGCSU used to keep stats if anybody's interested in the actual calculations.
Attachments
MUTscore(1).xlsx
(72.89 KiB) Downloaded 331 times
William
Alpharetta High School '11
Harvard '15
Kanga-Rat Murder Society
Wakka
Posts: 209
Joined: Fri Mar 07, 2008 10:52 pm
Location: NW Suburbia, IL

Re: MUT Mirror 4/2 @ NGCSU

Post by Kanga-Rat Murder Society »

One obvious thing that has yet to be stated is the fact that encouraging good teams to tank at the end of games takes away one of the major sources of joy for non-elite teams, which is getting a tossup against an elite team. I am by no means an elite, or even a good, player. Off of the top of my head, I can think of four games where I have gotten beat by at least 500 points. The reason that these games are still in my head, though, is not because I am damaged by the memory of getting crushed. In fact, it is quite the opposite. Those games stick in my head because those are the games where I (or my teammates) got really impressive buzzes to beat really good teams, if only for one question. I remember being a beginning player and being impressed by how good the elite teams were. I would go into games against Rockford Auburn or Loyola or Carbondale expecting to get crushed, but the one reassurance that I had was the fact that I could get a tossup. Even though these teams knew seemingly everything, it reminded me that I had a niche, no matter how small, that I could get off of them.

By encouraging the great teams to tank, you are essentially telling beginners that they cannot beat them to any tossups. To me, there is not a more unsportsmanlike or disheartening message out there. You are taking away the possibility for that joyous moment of taking a tossup off of them, and instead telling teams that they aren't good enough to score at all within the normal rules of the game. Thus, even if we ignore the feelings of the elite teams, it seems pretty obvious to me that your system is a terrible idea.
Nicholas Bergeon
Buffalo Grove High School '09
UW-Madison '12
WUSTL Law'15
User avatar
Habitat_Against_Humanity
Rikku
Posts: 476
Joined: Sun Jan 21, 2007 8:51 pm
Location: Syracuse, NY

Re: MUT Mirror 4/2 @ NGCSU

Post by Habitat_Against_Humanity »

Aaron Carney wrote: Analyze and evaluate...and go enjoy a dictionary or get a psych eval, I'm still hoping no one here needs one.

Yes, I intend to insult people who have intelligence but refuse to use it or become so wrapped up in themselves or something that they insist on being narrow minded. It might have taken some serious analyzing and evaluating to have realized this as I sometimes use people's emotions so they see for themselves when they lass out.
I don't care that you want to insult people - it's your prerogative to dig yourself into a hole. I do however take issue with the fact that you decided to be derogatory towards the mentally ill.
Rachel
UChicago 09
User avatar
gwaustin4
Lulu
Posts: 69
Joined: Wed Jun 17, 2009 2:08 am

Re: MUT Mirror 4/2 @ NGCSU

Post by gwaustin4 »

Gonzagapuma1 wrote:
Aaron Carney wrote:For those of you who think the point of Quiz Bowl is to win seriously need to get a psychological evaluation. I'm serious, the point of Quiz Bowl is to enjoy yourself while forming a social network with other intelligent people. If you are being paid based on how much you win, the purpose is to win. If your job or lifestyle depends on it, the purpose is to win. If you have condition that demands you collect trophies, the purpose is to win. If your social life depends on your ability to brag about how many tournaments you've won, the purpose is to win. Otherwise it's much more likely what I said, or simply one of the two. Given there are some exceptions, like picking up dates...
Putting aside the rest of your post, this is also incredibly wrong. The point of quizbowl is to know more than the other team, and therefore win. Is there a social aspect to it? Sure. But if the point is not to win, then why even keep score? I also have no idea what the last sentence means.

Also, this: http://www.youtube.com/watch?v=fEkWH8DB7b0.
Like.
Wes Austin
Chattahoochee '09
UGA '09-
NGTech
Lulu
Posts: 13
Joined: Sun Apr 03, 2011 12:30 am

Re: MUT Mirror 4/2 @ NGCSU

Post by NGTech »

Alejandro wrote:
NGTech wrote:As requested I used the 2010 ACF Nationals to test the stat we used. To test only the stat I did no sorting based on W-L.

As such the statistic was able to predict the Preliminary outcome to and overall accuracy of 71%. This is better than the calculated values for other tie breakers, and this is the stat ranking the field not breaking ties.
Using PPB predicts the winner ~90% of the time in the prelims of 2010 ACF Nationals (72/78 correct in Red and 69/78 correct in Black). Granted, this was calculated without removing the game's effect on the PPB, but it seems unlikely that this would decrease its accuracy by 20% or more.
I ran the round by round prediction for the 2010 ACF Nationals. My stat was able to predict 173/182 matches. This gives me an accuracy of ~95%.

On a side note, my accuracy for ranking is incorrect, I counted teams flipped, ie. 2,3 to 3,2, as two errors where it is really just one. With that in mind my stat was able to rank with ~85% accuracy without accounting for W-L.

Thank you for the better data, it is aiding to improve what I feel is a system with great potential.
Nicolas A. Collins
North Georgia College Academic Team Technical Director
User avatar
theMoMA
Forums Staff: Administrator
Posts: 5993
Joined: Mon Oct 23, 2006 2:00 am

Re: MUT Mirror 4/2 @ NGCSU

Post by theMoMA »

Did you use it so that the team with the lower "quality rating" is expected to win the game? Because that's apparently how you used it to break ties.
Andrew Hart
Minnesota alum
NGTech
Lulu
Posts: 13
Joined: Sun Apr 03, 2011 12:30 am

Re: MUT Mirror 4/2 @ NGCSU

Post by NGTech »

theMoMA wrote:Did you use it so that the team with the lower "quality rating" is expected to win the game? Because that's apparently how you used it to break ties.
Unfortunately I had to flip it. Without the W-L being accounted for all the stat would tell me is who had the hardest schedule.

In the tweak I used to compare to other stats it is similar in that it rewards teams for having easy matches and running up the score on weaker opponents.

I will point out that in my opinion rewarding teams who have easy matches and run up scores on weaker opponents is nothing more than praising the bullies for being bullies instead of rewarding the teams that fought hard for the wins they received. That said I believe W-L should always be the first comparison, unlike what is currently done in college football.
Nicolas A. Collins
North Georgia College Academic Team Technical Director
Ethnic history of the Vilnius region
Auron
Posts: 1000
Joined: Tue Feb 17, 2004 12:50 am
Location: Columbia, SC

Re: MUT Mirror 4/2 @ NGCSU

Post by Ethnic history of the Vilnius region »

NGTech wrote: I will point out that in my opinion rewarding teams who have easy matches and run up scores on weaker opponents is nothing more than praising the bullies for being bullies instead of rewarding the teams that fought hard for the wins they received. That said I believe W-L should always be the first comparison, unlike what is currently done in college football.
You're wrong, and as someone who spends hours a week working with a team that works its butt off and won several matches in a convincing manner as a result of working its butt off at your tournament that they paid money to play in, it's pretty damn offensive that you keep labeling them or other good teams bullies. You have to do a lot to make me mad, but by Jove you've done it.
Eric D.
University of South Carolina Alum
User avatar
Bartleby
Rikku
Posts: 315
Joined: Wed Dec 02, 2009 5:45 pm
Location: Philadelphia, PA

Re: MUT Mirror 4/2 @ NGCSU

Post by Bartleby »

NGTech wrote:
theMoMA wrote:Did you use it so that the team with the lower "quality rating" is expected to win the game? Because that's apparently how you used it to break ties.
Unfortunately I had to flip it. Without the W-L being accounted for all the stat would tell me is who had the hardest schedule.

In the tweak I used to compare to other stats it is similar in that it rewards teams for having easy matches and running up the score on weaker opponents.

I will point out that in my opinion rewarding teams who have easy matches and run up scores on weaker opponents is nothing more than praising the bullies for being bullies instead of rewarding the teams that fought hard for the wins they received. That said I believe W-L should always be the first comparison, unlike what is currently done in college football.
The problem with this is that it only accounts for an end-result (answering a question), and does not account for what actually happens in-game. If I am winning 400-100, I am liable to let the entire question be read before I buzz (and this has been my experience when playing Quiz Bowl), but if I know the answer, I'm not going to sit on the question solely to make the other team feel better about themselves. As has already been elucidated, the primary goal in the game of Quiz Bowl is to answer the most amount of questions correctly. If you can do this better than your opponents, you deserve to win; if you can do this to a substantially greater degree than your opponents, then they need to improve. This is coming from someone who averages ~15 ppg in "Academic" Quiz Bowl, and who could sorely stand to do some improving.
Brian McNamara
Western University '13
University of Waterloo '14
Temple University '20
Tanay
Rikku
Posts: 427
Joined: Wed Jun 03, 2009 5:05 pm

Re: MUT Mirror 4/2 @ NGCSU

Post by Tanay »

NGTech wrote:I will point out that in my opinion rewarding teams who have easy matches and run up scores on weaker opponents is nothing more than praising the bullies for being bullies instead of rewarding the teams that fought hard for the wins they received.
Assuming a round-robin format or something comparably fair, teams with "easy matches" are invariably the best teams, since they face essentially the same strength of schedule as their opponents and manage to win by more. In fact, this format becomes even worse if the losing team starts to neg the questions that they've essentially been "granted" by the winning team. You fundamentally assume that if a better team holds off on buzzes, there will be a statistically significant effect on the margin of victory which you believe improves gameplay. Sadly, bad teams are bad. They often get questions wrong. In that fashion, a weaker team could systematically choke a good team out of contention at a tournament by increasing the margin of victory (which by your paradigm is awful, unsportsmanlike, and constitutes bullying), whether or not they intend to do so.
Tanay
ex-Berkeley, ex-Bellarmine
User avatar
Maxwell Sniffingwell
Auron
Posts: 2164
Joined: Sun Feb 12, 2006 3:22 pm
Location: Des Moines, IA

Re: MUT Mirror 4/2 @ NGCSU

Post by Maxwell Sniffingwell »

NGTech wrote:I ran the round by round prediction for the 2010 ACF Nationals. My stat was able to predict 173/182 matches. This gives me an accuracy of ~95%.
NGTech wrote:
theMoMA wrote:Did you use it so that the team with the lower "quality rating" is expected to win the game? Because that's apparently how you used it to break ties.
Unfortunately I had to flip it. Without the W-L being accounted for all the stat would tell me is who had the hardest schedule.
So in your test of the stat (as used at the tournament) on the ACF Nationals stats, your stat then predicted 9/182 games correctly?
Greg Peterson

Northwestern University '18
Lawrence University '11
Maine South HS '07

"a decent player" - Mike Cheyne
User avatar
Marble-faced Bristle Tyrant
Yuna
Posts: 853
Joined: Fri Feb 08, 2008 11:05 pm
Location: Evanston, IL

Re: MUT Mirror 4/2 @ NGCSU

Post by Marble-faced Bristle Tyrant »

Aaron Carney wrote:
Jesus Christ, stop hiding behind semantic arguments to avoid addressing whether the quality rating was terrible.
Stop bringing up individual examples and personal experiences as reasons why its bad.
Seriously, you're using perceived bad arguments made by people who are genuinely trying to convince you of something as an excuse for making nonsensical, evasive posts? Everyone in this thread has explained why it's terrible: we want fair competitions that show which teams are the best. We aren't trying to make every Southeastern tournament an automatic first place for South Carolina, but if they're the best team in attendance, then they deserve first place, period.

Furthermore, I imagine the producers of a set can't really get good data on whether they hit their target difficulty if teams aren't playing to the best of their ability. On a novice set like MUT, you'd expect top teams like USC to have PPG/PPB within a certain range, then middle-tier teams within another range, etc. Plus, a team/individual really cannot adequately track their own progress if their scoring is all over the place within a single tournament.

Imagine the sort of gamesmanship required to take advantage of the quality rating. You'd have to keep track of the score yourself (some people already do this, though not everyone) and make certain you maintain a 5- or 10-point lead on your opponent (or to play it safer, a ~40-point lead). You would then have to sit on every tossup waiting for your opponent to answer. If it's a really weak team you're playing, then multiple questions will last until the giveaway, or even go dead. This will lengthen every round considerably (like the aforementioned rounds against the chairs). You'd only answer questions after your opponent managed to convert some points. If this happens near the end of the round, the stronger team could get into a situation where they wouldn't be able to catch up with their opponent because there aren't any questions left. You'd end up with a lot of matches with the opposite W/L result of what should have happened, never mind how different the actual scoring would have been. I say "should have happened" not, like I said before, because all of USC's matches should be automatic wins, but because it's simply a fact that some teams are better than other teams, and we (and all other competitive activities) should have a system that yields which ones are which.
Ethnic history of the Vilnius region wrote:You have to do a lot to make me mad, but by Jove you've done it.
As far as I can recall, this is the first time I've ever felt belligerent on hsqb.
Farrah Bilimoria
Formerly of Georgia Tech and Central High School (Macon)
User avatar
The Ununtiable Twine
Auron
Posts: 1058
Joined: Fri Feb 02, 2007 11:09 pm
Location: Lafayette, LA

Re: MUT Mirror 4/2 @ NGCSU

Post by The Ununtiable Twine »

cornfused wrote:
NGTech wrote:I ran the round by round prediction for the 2010 ACF Nationals. My stat was able to predict 173/182 matches. This gives me an accuracy of ~95%.
NGTech wrote:
theMoMA wrote:Did you use it so that the team with the lower "quality rating" is expected to win the game? Because that's apparently how you used it to break ties.
Unfortunately I had to flip it. Without the W-L being accounted for all the stat would tell me is who had the hardest schedule.
So in your test of the stat (as used at the tournament) on the ACF Nationals stats, your stat then predicted 9/182 games correctly?
So if we just invert the result of the stat, then we might have a great stat here. Interesting...the statistic measures precisely everything that could be wrong with quizbowl, so just flip the result and bang, we have the indicator that we've been looking for all these years. Just a thought, it just might work.
Jake Sundberg
Louisiana, Alabama
retired
User avatar
Maxwell Sniffingwell
Auron
Posts: 2164
Joined: Sun Feb 12, 2006 3:22 pm
Location: Des Moines, IA

Re: MUT Mirror 4/2 @ NGCSU

Post by Maxwell Sniffingwell »

DarkMatter wrote:
cornfused wrote:
NGTech wrote:I ran the round by round prediction for the 2010 ACF Nationals. My stat was able to predict 173/182 matches. This gives me an accuracy of ~95%.
NGTech wrote:
theMoMA wrote:Did you use it so that the team with the lower "quality rating" is expected to win the game? Because that's apparently how you used it to break ties.
Unfortunately I had to flip it. Without the W-L being accounted for all the stat would tell me is who had the hardest schedule.
So in your test of the stat (as used at the tournament) on the ACF Nationals stats, your stat then predicted 9/182 games correctly?
So if we just invert the result of the stat, then we might have a great stat here. Interesting...the statistic measures precisely everything that could be wrong with quizbowl, so just flip the result and bang, we have the indicator that we've been looking for all these years. Just a thought, it just might work.
Well, I mean, it looks like this stat is a fancier version of margin of victory.
Greg Peterson

Northwestern University '18
Lawrence University '11
Maine South HS '07

"a decent player" - Mike Cheyne
NGTech
Lulu
Posts: 13
Joined: Sun Apr 03, 2011 12:30 am

Re: MUT Mirror 4/2 @ NGCSU

Post by NGTech »

cornfused wrote: So in your test of the stat (as used at the tournament) on the ACF Nationals stats, your stat then predicted 9/182 games correctly?
The stat was never design for the purpose of determining match winners, it was designed to find the team with the hardest schedule in a set.
It was only meant as a method for breaking ties based on Schedule Strength after W-L had been applied.
It just happened to work as an effective win picker when reversed.
DarkMatter wrote: So if we just invert the result of the stat, then we might have a great stat here. Interesting...the statistic measures precisely everything that could be wrong with quizbowl, so just flip the result and bang, we have the indicator that we've been looking for all these years. Just a thought, it just might work.
While it may be more effective in that form that was not my intended use of the code. As such I will not allow for use of my algorithms in this corrupted manner in actual play.

Any to any comments on the fact that the algorithm is out in the open and available to anyone I will point out that as the original creator with the original form of both I hold a copyright of my algorithm.
Nicolas A. Collins
North Georgia College Academic Team Technical Director
User avatar
Auks Ran Ova
Forums Staff: Chief Administrator
Posts: 4295
Joined: Sun Apr 30, 2006 10:28 pm
Location: Minneapolis
Contact:

Re: MUT Mirror 4/2 @ NGCSU

Post by Auks Ran Ova »

NGTech wrote:While it may be more effective in that form that was not my intended use of the code. As such I will not allow for use of my algorithms in this corrupted manner in actual play.

Any to any comments on the fact that the algorithm is out in the open and available to anyone I will point out that as the original creator with the original form of both I hold a copyright of my algorithm.
Something tells me this won't be an issue.
Rob Carson
University of Minnesota '11, MCTC '??, BHSU forever
Member, ACF
Member emeritus, PACE
Writer and Editor, NAQT
User avatar
Cody
2008-09 Male Athlete of the Year
Posts: 2891
Joined: Sun Nov 15, 2009 12:57 am

Re: MUT Mirror 4/2 @ NGCSU

Post by Cody »

NGTech wrote:While it may be more effective in that form that was not my intended use of the code. As such I will not allow for use of my algorithms in this corrupted manner in actual play.

Any to any comments on the fact that the algorithm is out in the open and available to anyone I will point out that as the original creator with the original form of both I hold a copyright of my algorithm.
Not that anyone would bother to use it, but you can't copyright an algorithm. Perhaps you want to patent it, but you can't patent algorithms--first, you have to implement it in software, then patent it; then you'd be able to stop us from using it in software. However, that's probably above your pay grade, and that still wouldn't stop us using it at a tournament if we wished to (we don't).

Edit: It has been brought to my attention that Bilski did not, in fact, forbid patenting all business methods, so you could try that as well.
Last edited by Cody on Mon Apr 04, 2011 8:27 pm, edited 1 time in total.
Cody Voight, VCU ’14.
User avatar
The Toad to Wigan Pier
Tidus
Posts: 528
Joined: Mon Oct 10, 2005 6:58 pm
Location: Seattle

Re: MUT Mirror 4/2 @ NGCSU

Post by The Toad to Wigan Pier »

NGTech wrote:
cornfused wrote: So in your test of the stat (as used at the tournament) on the ACF Nationals stats, your stat then predicted 9/182 games correctly?
The stat was never design for the purpose of determining match winners, it was designed to find the team with the hardest schedule in a set.
It was only meant as a method for breaking ties based on Schedule Strength after W-L had been applied.
It just happened to work as an effective win picker when reversed.
DarkMatter wrote: So if we just invert the result of the stat, then we might have a great stat here. Interesting...the statistic measures precisely everything that could be wrong with quizbowl, so just flip the result and bang, we have the indicator that we've been looking for all these years. Just a thought, it just might work.
While it may be more effective in that form that was not my intended use of the code. As such I will not allow for use of my algorithms in this corrupted manner in actual play.

Any to any comments on the fact that the algorithm is out in the open and available to anyone I will point out that as the original creator with the original form of both I hold a copyright of my algorithm.
"Copyright protection is not available for ideas, program
logic, algorithms, systems, methods, concepts, or layouts"(http://www.copyright.gov/circs/circ61.pdf , emphasis is mine)
William Butler
UVA '11
Georgia Tech 13
NGTech
Lulu
Posts: 13
Joined: Sun Apr 03, 2011 12:30 am

Re: MUT Mirror 4/2 @ NGCSU

Post by NGTech »

SirT wrote: Not that anyone would bother to use it, but you can't copyright an algorithm. Perhaps you want to patent it, but you can't patent algorithms--first, you have to implement it in software, then patent it; then you'd be able to stop us from using it in software. However, that's probably above your pay grade, and that still wouldn't stop us using it at a tournament if we wished to (we don't).
Software is copyrighted under the Berne Convention as a literary work as soon as it is placed in a written form. The U.S. is a member of the Berne Convention.
Nicolas A. Collins
North Georgia College Academic Team Technical Director
User avatar
mhayes
Rikku
Posts: 308
Joined: Thu Feb 16, 2006 3:15 pm

Re: MUT Mirror 4/2 @ NGCSU

Post by mhayes »

NGTech wrote:Software is copyrighted under the Berne Convention as a literary work as soon as it is placed in a written form. The U.S. is a member of the Berne Convention.
Yes, the *software* is copyrighted, but algorithms themselves are not patentable.
Matt Hayes

Unaffiliated
User avatar
Cody
2008-09 Male Athlete of the Year
Posts: 2891
Joined: Sun Nov 15, 2009 12:57 am

Re: MUT Mirror 4/2 @ NGCSU

Post by Cody »

NGTech wrote:Software is copyrighted under the Berne Convention as a literary work as soon as it is placed in a written form. The U.S. is a member of the Berne Convention.
Copyright only protects the form of expression of an idea, not the idea itself (for that, you need a patent). Thus, any algorithm used within a computer program ("software") is not eligible for copyright. Additionally, I see no software in this thread.
Cody Voight, VCU ’14.
User avatar
Marble-faced Bristle Tyrant
Yuna
Posts: 853
Joined: Fri Feb 08, 2008 11:05 pm
Location: Evanston, IL

Re: MUT Mirror 4/2 @ NGCSU

Post by Marble-faced Bristle Tyrant »

NGTech wrote:The stat was never design for the purpose of determining match winners, it was designed to find the team with the hardest schedule in a set.
Doesn't NAQT already do something like this? The S value? But they use that to compare teams at different SCT sites. It's not really necessary in a single tournament since, as people have been saying, you schedule prelims so that everyone has schedules of roughly the same difficulty. And it's definitely not necessary in a single-bracket prelim where everybody plays everybody. You should have just rebracketted for playoffs based on prelim W/L standings to determine final standings, since you had some packets remaining.
Last edited by Marble-faced Bristle Tyrant on Mon Apr 04, 2011 8:31 pm, edited 1 time in total.
Farrah Bilimoria
Formerly of Georgia Tech and Central High School (Macon)
NGTech
Lulu
Posts: 13
Joined: Sun Apr 03, 2011 12:30 am

Re: MUT Mirror 4/2 @ NGCSU

Post by NGTech »

mhayes wrote:
NGTech wrote:Software is copyrighted under the Berne Convention as a literary work as soon as it is placed in a written form. The U.S. is a member of the Berne Convention.
Yes, the *software* is copyrighted, but algorithms themselves are not patentable.
While it is true that the algorithm itself isn't protected the equation is part of the source code and cannot be copied or distorted. If you wish to run the calculation by hand without ever copying down the equation then good luck.
Nicolas A. Collins
North Georgia College Academic Team Technical Director
User avatar
Bartleby
Rikku
Posts: 315
Joined: Wed Dec 02, 2009 5:45 pm
Location: Philadelphia, PA

Re: MUT Mirror 4/2 @ NGCSU

Post by Bartleby »

List of wrestling-based comic books wrote:
NGTech wrote:The stat was never design for the purpose of determining match winners, it was designed to find the team with the hardest schedule in a set.
Doesn't NAQT already do something like this? The S value? But they use that to compare teams at different SCT sites. It's not really necessary in a single tournament since, as people have been saying, you schedule prelims so that everyone has schedules of roughly the same difficulty. And it's definitely not necessary in a single-bracket prelim where everybody plays everybody. You should have just rebracketted for playoffs to determine final standings, since you had some packets remaining.
The "S-Value" is one of many calculations which goes into determining a team's average PP20TUH (this is called D-Value, I think) if they were playing what NAQT deems an "average" schedule (if I understand it correctly). While it's a similar idea, it is not the only metric used to determine how good a team is. Teams that score a lot of points generally have pretty good D-Values, no matter what.
Brian McNamara
Western University '13
University of Waterloo '14
Temple University '20
User avatar
Mechanical Beasts
Banned Cheater
Posts: 5673
Joined: Thu Jun 08, 2006 10:50 pm

Re: MUT Mirror 4/2 @ NGCSU

Post by Mechanical Beasts »

NGTech wrote:
mhayes wrote:
NGTech wrote:Software is copyrighted under the Berne Convention as a literary work as soon as it is placed in a written form. The U.S. is a member of the Berne Convention.
Yes, the *software* is copyrighted, but algorithms themselves are not patentable.
While it is true that the algorithm itself isn't protected the equation is part of the source code and cannot be copied or distorted. If you wish to run the calculation by hand without ever copying down the equation then good luck.
It is not true that any equation contained in a piece of copyrighted source code cannot be used by other persons; otherwise, it would be literally impossible for both DirectX and OpenGL to exist (for example).

EDIT: I just realized that I'm debating whether it's possible to copyright something that no one would ever mind if you copyrighted because no one wants to use this bad thing.
Andrew Watkins
User avatar
grapesmoker
Sin
Posts: 6345
Joined: Sat Oct 25, 2003 5:23 pm
Location: NYC
Contact:

Re: MUT Mirror 4/2 @ NGCSU

Post by grapesmoker »

Are we seriously having this discussion? Nicolas, your algorithm is not copyrightable. If someone were to use it, you would have no legal recourse against them. Not that anyone would want to, but still.
Jerry Vinokurov
ex-LJHS, ex-Berkeley, ex-Brown, sorta-ex-CMU
presently: John Jay College Economics
code ape, loud voice, general nuissance
User avatar
AKKOLADE
Sin
Posts: 15782
Joined: Thu Apr 24, 2003 8:08 am

Re: MUT Mirror 4/2 @ NGCSU

Post by AKKOLADE »

Posting a general reminder aimed at all parties to calm the hell down.
Fred Morlan
University of Kentucky CoP, 2017
International Quiz Bowl Tournaments, CEO, co-owner
former PACE member, president, etc.
former hsqbrank manager, former NAQT writer & subject editor, former hsqb Administrator/Chief Administrator
User avatar
theMoMA
Forums Staff: Administrator
Posts: 5993
Joined: Mon Oct 23, 2006 2:00 am

Re: MUT Mirror 4/2 @ NGCSU

Post by theMoMA »

The way you broke a tie appears to pick reward the team less likely to win a head-to-head match about 95% of the time. Also, all teams in a round robin play what amounts to the same schedule, so by trying to find the "most difficult schedule" you're "solving" a problem that doesn't even exist in your own format. But yeah, the biggest problem here is clearly the copyright issue.
Andrew Hart
Minnesota alum
NGTech
Lulu
Posts: 13
Joined: Sun Apr 03, 2011 12:30 am

Re: MUT Mirror 4/2 @ NGCSU

Post by NGTech »

theMoMA wrote:Also, all teams in a round robin play what amounts to the same schedule, so by trying to find the "most difficult schedule" you're "solving" a problem that doesn't even exist in your own format.
While it may seem to be trying to solve a problem that doesn't exist the problem does in fact exist. The round robin format is assumed to be an even format based on the fact that every team faces every other team. The problem is that teams are made of people not computers, as such the strength of a team changes from round to round. The only ways to counteract this is to let each team face each other team multiple times, or use a mathematical method to adjust the field.
Nicolas A. Collins
North Georgia College Academic Team Technical Director
Tanay
Rikku
Posts: 427
Joined: Wed Jun 03, 2009 5:05 pm

Re: MUT Mirror 4/2 @ NGCSU

Post by Tanay »

NGTech wrote:The problem is that teams are made of people not computers, as such the strength of a team changes from round to round.
Can you explain how this is true? The questions obviously change, but the concept of "team strength" has more to do with overall performance at a given tournament than fluctuations between packets of questions.
Tanay
ex-Berkeley, ex-Bellarmine
User avatar
Mechanical Beasts
Banned Cheater
Posts: 5673
Joined: Thu Jun 08, 2006 10:50 pm

Re: MUT Mirror 4/2 @ NGCSU

Post by Mechanical Beasts »

NGTech wrote:
theMoMA wrote:Also, all teams in a round robin play what amounts to the same schedule, so by trying to find the "most difficult schedule" you're "solving" a problem that doesn't even exist in your own format.
While it may seem to be trying to solve a problem that doesn't exist the problem does in fact exist. The round robin format is assumed to be an even format based on the fact that every team faces every other team. The problem is that teams are made of people not computers, as such the strength of a team changes from round to round. The only ways to counteract this is to let each team face each other team multiple times, or use a mathematical method to adjust the field.
But your mathematical method introduces additional problems, including the fact that it rewards teams for less impressive wins. A team unable to score many points, playing weak opponents, advances over a team able to score many points and many more than strong opponents. And if teams are aware of the system, which they ought to be, they can game it to maximize their tiebreakers while maintaining an equal number of wins.

I absolutely don't see why "sometimes people play different in different rounds" requires a response of "why don't we institute a system where one can optimize ones tiebreakers by doing anything other than trying their best." I furthermore don't see how encouraging people to do anything other than try their best encourages them to engage in any kind of sportsmanship. More like "toy with bad teams that can't touch you, tricking them into answering questions because it can only help your tiebreakers."
Andrew Watkins
User avatar
grapesmoker
Sin
Posts: 6345
Joined: Sat Oct 25, 2003 5:23 pm
Location: NYC
Contact:

Re: MUT Mirror 4/2 @ NGCSU

Post by grapesmoker »

NGTech wrote:While it may seem to be trying to solve a problem that doesn't exist the problem does in fact exist. The round robin format is assumed to be an even format based on the fact that every team faces every other team. The problem is that teams are made of people not computers, as such the strength of a team changes from round to round. The only ways to counteract this is to let each team face each other team multiple times, or use a mathematical method to adjust the field.
No, that's not what happens. What happens is that the questions change from round to round. And in fact we already have a statistic that averages team strength over multiple rounds. It's called points per bonus, and it's not prone to pathologies such as penalizing you for beating weaker teams. In fact, the better the team, the better its PPB will reflect its strength because the sample size will be larger.
Jerry Vinokurov
ex-LJHS, ex-Berkeley, ex-Brown, sorta-ex-CMU
presently: John Jay College Economics
code ape, loud voice, general nuissance
gatsby
Kimahri
Posts: 5
Joined: Fri Dec 25, 2009 1:54 pm

Re: MUT Mirror 4/2 @ NGCSU

Post by gatsby »

NGTech wrote:
While it may seem to be trying to solve a problem that doesn't exist the problem does in fact exist. The round robin format is assumed to be an even format based on the fact that every team faces every other team. The problem is that teams are made of people not computers, as such the strength of a team changes from round to round. The only ways to counteract this is to let each team face each other team multiple times, or use a mathematical method to adjust the field.
ACF rules, which this tournament was supposed to follow, dictate that the teams with the best records play another game or games to determine the winner. Surely, instead of using your so-called mathematical method that has no significant empirical evidence, the solution would be to "let each team face each other team multiple times?" I am dissatisfied with the tournament's handling of the tie-breaking and how I paid to only play a handful of legitimate rounds. I feel that a finals round or even a playoffs would have solved all the problems more accurately than using an arbitrary mathematical system.
Phong Nguyen
Southside High School '09
University of South Carolina '13
Locked