Socal's conspiracy corner: PHD statistician proves the draws aren't random!

First topic message reminder :

Clearly, the level of statistical improbability that is taking place at the grandslams bears some mentioning of the financial and pecuniary interest( god I love using those law school words) in protecting the possibility of a fedal final and the ratings boon that would follow is just overwhelming. For the sixth consecutive slam Novak gets Roger and not Andy, this should only be a 50 percent probabiltiy and the odds of Novak getting Roger six straight times is 64 to 1. I have been roundly criticized in the past for my cynical and conspiratorial nature, being a middle eastern man raised in George Bush's America I am powerless to be anything but what my nature dictates. I am calling the fix. Not only does Novak get Roger for the umpteenth time, but he will most likely draw the shortest straw on stupid saturday by having the late match again. Stupid Saturday magnifies the importance of the semi draw, the feature semi (most likely Roger and Novak) will be greatly disadvantaged in the one day turnaround of Stupid Saturday. Now Novak most likely with a bum shoulder, will be required to defeat Roger federer in a night match, then comeback the next day and defeat the winner of the Nadal/Murray semi that took place earlier in the day. In short, Djokovic as the least marketable of the top four is getting jobbed by the tournament committee that is doing their best to protect their superbowl of ratings if Roger and Rafa happen to play in the final.



The brits should be happy as this configuration presents both Andy and Rafa with a wonderful opportunity to steal one from Novak who might as well be rated as the 32 seed as opposed to the number 1 seed with the draw he is getting. Roger also is done no favors with this draw as most likely if he wins out he will get the deathmatch on saturday night. In light of Djoko's shoulder injury and the quick turnaround in the final this grandslam has been decided at the outset by the tournament commitee to be a victory for either Rafa or Andy. So if you got any money right now bet those two to win the tournament, they have all the possible benefits of scheduling and draw imaginable at this point. It is a shame that the second biggest slam has to taint the sport with such a ridiculous fix and its even more ridiculous stupid saturday event that unfairly discriminates against the second semi.

Hi Julius, well I was present on a roulette table when RED came up 13 times in a row and at a blackjack table when the dealer get 5 BJ's in a row. Both times the house took all the cash, as the odds of that happening are rare. So any talk of 6 or 7, to me is run of the mill for permutations of any given flip of a coin.

The fact that it has happened does not make it impossible, does it ? (I am agreeing with you by the way)



And this is the only evidence, if you can call it that my critics have produced. That yes longshots do eventually come in. Well they certainly have come in conveniently for the tournament organizers and TV broadcasters.



One grandslam gets caught redheaded tampering with the draw and two different statisticians looked at it and confirmed the results and the methodology. You guys have produced nothing.



As for the other slams it is quite clear when you look at their conflict of interest that exists for the tournament organizers, combined with the odd draws we have been seeing with Novak repeatedly getting Roger over and over and over and over and over and over again that there is a great deal of circumstantial evidence of fraud. You have conflict of interest, money involved, no transparency and the odd anomaly that PRECISELY THE LONGSHOT THAT SUITS THE FINANCIAL INCENTIVES OF THOSE WHO CONTROL THE PROCESS HAPPENING OVER AND OVER AGAIN. If for example Roger and Rafa were put in the same half over and over again, I wouldn't have as much reason to doubt it, because you don't have the financial incentive and the conflict of interest. But the exact longshot that benefits the ratings comes in.



No I don't have video evidence of each of the 4 draws being fixed. But there is a great deal of fishy circumstances, and the exact fishy circumstance that creates a potential for higher finals ratings. As the expression goes where there is smoke there is fire.

hawkeye wrote:This whole draw conspiracy theme is fueled by the ATP at present having 3 dominant top seeds. Anyone projected to have a semi with the number 4 seed is thought to have an unfair advantage. It used to be "who's got Djokovic". Now its "who's got Federer". There is no advantage being the number 1 seed compared with being the number 2 seed.

The only fair way to solve this problem is to forget about draws and seeding when placing the number 3 seed. Just alternate the match up at every slam.

Exactly, it is a very simple solution hawkeye that they do not use for this particular reason. In fact why even alternate, why not just give the #3 guy to the number 2 player every time and give #4 to 1, most other tournaments in other sports would not even leave any vagary and questions in the process and just organize it logically like this. Then there would be no, no basis at all for controversy ever. But instead they continue with this process that is very riggable and there are a great deal of oddities appearing in their draws.

socal1976 wrote:

hawkeye wrote:This whole draw conspiracy theme is fueled by the ATP at present having 3 dominant top seeds. Anyone projected to have a semi with the number 4 seed is thought to have an unfair advantage. It used to be "who's got Djokovic". Now its "who's got Federer". There is no advantage being the number 1 seed compared with being the number 2 seed.

The only fair way to solve this problem is to forget about draws and seeding when placing the number 3 seed. Just alternate the match up at every slam.

Exactly, it is a very simple solution hawkeye that they do not use for this particular reason. In fact why even alternate, why not just give the #3 guy to the number 2 player every time and give #4 to 1, most other tournaments in other sports would not even leave any vagary and questions in the process and just organize it logically like this. Then there would be no, no basis at all for controversy ever. But instead they continue with this process that is very riggable and there are a great deal of oddities appearing in their draws.

Is this because the current #3 is a bad match up for the current #1 who is bad match up for current #2 who is a bad match up for current #3? and

I LIKE the word *riggable*



Hows about we put the top four players into a sack, shake it. toss 'em out and see who plays who ?

socal1976 wrote:

hawkeye wrote:This whole draw conspiracy theme is fueled by the ATP at present having 3 dominant top seeds. Anyone projected to have a semi with the number 4 seed is thought to have an unfair advantage. It used to be "who's got Djokovic". Now its "who's got Federer". There is no advantage being the number 1 seed compared with being the number 2 seed.

The only fair way to solve this problem is to forget about draws and seeding when placing the number 3 seed. Just alternate the match up at every slam.

Exactly, it is a very simple solution hawkeye that they do not use for this particular reason. In fact why even alternate, why not just give the #3 guy to the number 2 player every time and give #4 to 1, most other tournaments in other sports would not even leave any vagary and questions in the process and just organize it logically like this. Then there would be no, no basis at all for controversy ever. But instead they continue with this process that is very riggable and there are a great deal of oddities appearing in their draws.

A simple, but flawed solution. How would the alternating be started? At which slam? And the scenario of fixing #1 v #4 is also problematic. If #3 and #4 were each more proficient on hard and clay respectively then this could actually be the worst scenario for the #1. Not to mention that this could lead to players aiming for a certain ranking before a slam as they know their draw from a certain slot. A can of worms is then undoubtedly opened.

Socal, please lay off the capital letters. They are a bit desperate and childish. Also, as stats is my line of business and I know that nothing is given credibility until it is peer-reviewed I stand by that comment. Perhaps if it was your area of expertise you'd understand. I was enjoying the debate but don't appreciate
being scoffed at and bombarded with capitals, you are unrelenting and won't even acknowledge that you might be wide of the mark. I simply mean we should not treat these findings as gospel just yet, you are gullible and cynical as it suits.

Take for example every American team sport, they always for the playoffs (cup elimination for you uninitiated european types) that decide the championship, they simply give the best record during the regular season the #1 seed and then if it is an 8 team playoff they give 1 to 8, 2 to 7, 3 to 6, and 4 to 5. And then they put the 4th seed in the next round to the one seed. The 1 and 2 seeds can't play till the championship just like in tennis in this regards.



Instead another one of the long list of oddities that appears in tennis seedings and draws is this added discretion and riggability factor that grandslam committees use by giving themselves added leeway to play around with the top seed matchups. The system is too complicated, gives too much discretion to these committees, it is resulting in an appearance of impropriety, and at least one slam has already been caught cheating.

OK, so we need evidence to convince you that the draws aren't rigged, but you only need fishy circumstances to convince yourself that they are rigged. Fair enough.
You keep going on about how we've provided no evidence, when you yourself admit that you have no evidence.

Surely, the best way to ensure a Rafa/Fed final is to give Murray and Djoko a tough quarter to get them knocked out/worn out before the semis. So why give Djoko a qualifier in the first round, when it could have been fixed to give him someone in the top 40?

How exactly would the draw that Rafa did be rigged so as not to be noticed by the media present or the TV cameras or Rafa himself? Or are they all in on the conspiracy? How is this process 'very riggable'? (Saying 'it just is' doesn't count)

Has anyone checked the stats for the Masters series draws? Why would they fix the GS draws, but not the Masters (when the Masters need the TV viewers more than the GS)?

How about checking all the other players/combinations? Surely that would be fair on the tournament organisers? If Mardy Fish was always in the same quarter as Fernando Gonzales, would that be a fix? Or just a longshot that miraculously came in?

Incidentally, who would be involved in the Wimbledon draw that would also benefit financially from a Wimbledon Rafa/Fed final? The TV rights are sold before the tournament, the tickets are sold out before the tournament and the BBC don't get any advertising revenue. Do ESPN show the final in the States - do they rig the Wimby draw?

Positevly, I understand this your line of work, I don't attempt to get into the technicalities. All that i know is that not one but two different statisticians, who it is their line of work and area of expertise looked at these results for the USO and determined that with an exceedingly high probablity (.3 percent) these draws were not random. This is not my conclusion this is the conclusion of two different statisticians.



And this is no court of law or scientific journal, the standard proof is obviously much lower. Whether or not you agree that it proves anything one thing that it does do is raise serious suspicions that should be addressed. Suscipion and investigation at this point should be called for, not a conviction that is a different level of proof. But a prima facie showing has been made of serious issues of concern that I would think tennis fans would want looked at.

Incidentally, who would be involved in the Wimbledon draw that would also benefit financially from a Wimbledon Rafa/Fed final? The TV rights are sold before the tournament, the tickets are sold out before the tournament and the BBC don't get any advertising revenue. Do ESPN show the final in the States - do they rig the Wimby draw?



julius



The Tv rights are sold and the price paid in the new contract is based almost entirely on the average ratings lets say of the last few events and particularly the final that draws by far the most interest and viewers.





Has anyone checked the stats for the Masters series draws? Why would they fix the GS draws, but not the Masters (when the Masters need the TV viewers more than the GS)?



julius



Yes oddily enough the the master fair better in regards to Djoko/Fed matchup rigging than the grandslams. But EVEN AT THE MASTERS LEVEL THERE SEEMS TO BE A DJOKO FED BIAS JUST LESS PRONOUNCED. 15 times out of the last 24 masters.

OK, so we need evidence to convince you that the draws aren't rigged, but you only need fishy circumstances to convince yourself that they are rigged. Fair enough.
You keep going on about how we've provided no evidence, when you yourself admit that you have no evidence.


julius



I hate to break it to you there are a lot of people sitting in jail because of fishy circumstances surrounding conflict of interest laws. At the very least with the limited ability I have to investigate these issues there is a lot of smoke. At this stage more serious questions need to be asked and prima facie showing for investigation and reform of the process exists. Open it up, lets look to limit the discretion of these committees, and I would for once like the sports media to start looking at some of these issues. The attitude you seem to take is that nothing should be done, honor system, and all that and lets just pretend that there are no odd circumstances that bear explanation or investigation.

Wimbledon final 2009 - Federer/Roddick had more viewers than any of the Rafa/Fed finals (NBC figures).
You'd think the draw-riggers would research things like that.

http://tvbythenumbers.zap2it.com/2009/07/09/wimbledon-mens-final-is-most-viewed-in-10-years/22595/

Rafa/Berdych 2010 did quite poorly though.

ESPN had already paid for the next 12 years, so whoever made the final this year was irrelevant to them by then. Unless it was NBC who rigged the draw? These deals are rarely (never?) made on a year-by-year basis.

So the Masters series shows no statistically relevant bias. In other words - no evidence of rigging.

Are there really people in jail without any evidence against them? Whatever happened to the burden of proof?

JuliusHMarx wrote:Wimbledon final 2009 - Federer/Roddick had more viewers than any of the Rafa/Fed finals (NBC figures).
You'd think the draw-riggers would research things like that.

http://tvbythenumbers.zap2it.com/2009/07/09/wimbledon-mens-final-is-most-viewed-in-10-years/22595/

Rafa/Berdych 2010 did quite poorly though.

ESPN had already paid for the next 12 years, so whoever made the final this year was irrelevant to them by then. Unless it was NBC who rigged the draw? These deals are rarely (never?) made on a year-by-year basis.

So the Masters series shows no statistically relevant bias. In other words - no evidence of rigging.

Are there really people in jail without any evidence against them? Whatever happened to the burden of proof?

The final is not being shown by Espn, the final is always shown as the entire last weekend is by CBS. Actually, what the masters showed was a bias combined with the massive bias shown at the slams. It didn't come out 50/50 or close to it. It is in no way as fishy as in the slams but they still appeared in 62.5 percent of Master's draws as well.



On your last point, like i said I have limited tools of investigation at this point there is enough of a showing and enough oddities that this needs to be looked at and frankly reformed. What I oppose is your attitude that nothing is odd here, and that there is no basis to tighten up the discretion of these committees or look at these questions. I find the media in particulars lack of attention to these questions very lazy and disturbing. This is a huge, huge issue. Remember what the tour did to davy for the appearance of impropriety, they investigated him for over a year? Who is investigating or auditing the work of these committees on an issue that strikes at the integrity of the game?

My assumption is that the media's lack of interest is because it isn't a huge issue. Statistical analyses rarely are. Should further analyses be done - possibly. Certainly a peer-review to ensure the stats are correct, as P4 says is the norm.

Any evidence, from statistical analyses or otherwise that Wimby, AO, or FO are rigged - none whatsoever.

You say the draw of 1 & 2 seeds vs 3 & 4 seeds is 'very riggable' but offer no explanation of how.
Seeds are drawn out of a hat/trophy, by a tennis player, in front of media and other officials, and on TV. What exactly would you improve on?

If you were to combine the Masters and GS draws the Fed/Djoko combination would actually show less indication of a rigged draw than the GS alone - P4 can correct me if I'm wrong

Well I disagree in my opinion there is a real appearance of impropriety or conflict of interest with these committees that have been given a lot of discretion. Remember how long Davydenko was investigated for an issue that bore on the credibility of the game like this does over the appearance of impropriety before he was cleared, they practically ruined his performance that year during the investigation. Answer me one question who are these people's watchdogs, where is the check in the process. And when there is no check on the process why give them so much discretion with the seeding, give them some hard and fast rules for seed matchups.

Bribing the USO defending champion or putting the same two numbers (either #3 or #4 but not both) in the #3/#4 draw is only method by which Socal's conspiracy can be addressed.

Socal, have you considered the remote possibility of Niland beating Djokovic? Can you help with a 'unriggable' statistical analysis?

P4thS... Did you see the ESPN links that I mentioned? They provide the algorithm used for the analysis.

socal1976 wrote:Positevly, I understand this your line of work, I don't attempt to get into the technicalities. All that i know is that not one but two different statisticians, who it is their line of work and area of expertise looked at these results for the USO and determined that with an exceedingly high probablity (.3 percent) these draws were not random. This is not my conclusion this is the conclusion of two different statisticians.

And this is no court of law or scientific journal, the standard proof is obviously much lower. Whether or not you agree that it proves anything one thing that it does do is raise serious suspicions that should be addressed. Suscipion and investigation at this point should be called for, not a conviction that is a different level of proof. But a prima facie showing has been made of serious issues of concern that I would think tennis fans would want looked at.

Hi socal,

Fair enough. I am going to read the work done in full and try to replicate it myself. Of course, if their conclusion is robust then this is worrying. I am a little cautious as I have acted as a reviewer in the past and seem some flawed work.

I don't think it necessarily follows that the semi-finals are rigged even if this early round business is shown to be corrupt. As I said on another thread, looking at previous draws for the US Open one could make a case that, up to last year, the alternative composition of Nadal v Djokovic and Federer v Murray would be more likely to yield a Nadal v Federer final. Why? Well, until that point Nadal had never lost in a slam to Djokovic, but had to Murray (including their only US Open encounter) and Federer had never lost in a slam to Murray, but had to Djokovic (albeit at the Aus Open, and their US Open matches were tight).

laverfan wrote:P4thS... Did you see the ESPN links that I mentioned? They provide the algorithm used for the analysis.

Hi LF. I did, thanks, and am going to have a tentative look when I return to work tomorrow.

Fair enough. I am going to read the work done in full and try to replicate it myself. Of course, if their conclusion is robust then this is worrying. I am a little cautious as I have acted as a reviewer in the past and seem some flawed work.

I don't think it necessarily follows that the semi-finals are rigged even if this early round business is shown to be corrupt. 4th street



Here I disagree with you, now it is no longer a question of mathematical technicalities and a matter of human nature and here I have plenty of experience. When you prove that the same group of people manufactured the first two rounds, and you see a long shot coming in that suits their financial interests in the semi portion of the draw; you no longer give them any benefit of the doubt whatsoever.

Positively 4th Street wrote:

laverfan wrote:P4thS... Did you see the ESPN links that I mentioned? They provide the algorithm used for the analysis.

Hi LF. I did, thanks, and am going to have a tentative look when I return to work tomorrow.

Thanks. You may have to use Wiki and ATP websites to extract the data used in the analysis going back to 2001 (40 slams). Depending on how much time you have on hand, do you want to consider going back to the beginning of the Open Era?

The challenge I see is the advent of Federer in 2003 may also be a significant factor till 2011, which may have been ignored and can be normalised by going further back to 1969.

JuliusHMarx wrote:Any evidence, from statistical analyses or otherwise that Wimby, AO, or FO are rigged - none whatsoever.

You say the draw of 1 & 2 seeds vs 3 & 4 seeds is 'very riggable' but offer no explanation of how.
Seeds are drawn out of a hat/trophy, by a tennis player, in front of media and other officials, and on TV. What exactly would you improve on?

I'm quoting myself, megalomaniac that I am.
Hoping to get a response on these points.

JuliusHMarx wrote:If you were to combine the Masters and GS draws the Fed/Djoko combination would actually show less indication of a rigged draw than the GS alone - P4 can correct me if I'm wrong

I haven't looked at the details, but I think so. Certainly, Federer and Nadal have been in the same half at times this season. Picking out just the US Open, or indeed the slams, is akin to looking at a selection of outcomes from a long sequence. The small selection may show what looks like non-randomness even when the underlying process is random. The variance from a small sample of what are essentially Bernoulli trials is also quite large.

http://en.wikipedia.org/wiki/Bernoulli_trial

JuliusHMarx wrote:

JuliusHMarx wrote:Any evidence, from statistical analyses or otherwise that Wimby, AO, or FO are rigged - none whatsoever.

You say the draw of 1 & 2 seeds vs 3 & 4 seeds is 'very riggable' but offer no explanation of how.
Seeds are drawn out of a hat/trophy, by a tennis player, in front of media and other officials, and on TV. What exactly would you improve on?

I'm quoting myself, megalomaniac that I am.
Hoping to get a response on these points.

Actually, wrong the french open showed significant signs of non-randomness in the woman's portion of their draw. Maybe they just did a better and more subtle job of touching the draw up on the men's side. The woman's draws for the FO also was determined to be problematic



Here is what i would suggest. Either like hawkeye says they simply alternate each slam the seed matchups vis a vis the top 4 seeds in the semis, or they just limit the discretion of these unchecked committees by saying look; 1 is always going to play 4, 2 is always going to play 3 that is the rule period.



Now answer my question who exactly is the watchdog of these tournament organizing committees? Who polices them?

If someone did police them, how could you be sure they themselves were not also corrupt, if they found no evidence of corruption? A true conspiracy theorist trusts no-one except those that agree with them.

I didn't make clear in my most recent post (although I had done so previously) that I was referring to the FO men's draw.

Thought I'de check the first round draws in slams of both Nadal and Federer for past 3 years. These are the sort of matches that have been questioned by the research in the linked article.

Players are seeded to 32 at all slams so a seeded player can only meet a player outside the top 32 in first round.

In 2011 Nadal played Isner (39) in first round of FO. He played Gasquet (46) in first round of 2009 US Open. Also he was drawn to play Hewitt in first round of Wimbledon 2009 but withdrew with injury.

In 2011 Federer played Lopez (41) in first round of FO. Andreev (37) in first round of AO in 2010 and Sepi (35) in first round of AO 2009.

Without any stastical analysis it looks like they've had their fair share of tough matches. The study looked at a ten year period. Wonder what 10 year period?

socal1976 wrote:Now answer my question who exactly is the watchdog of these tournament organizing committees? Who polices them?

Tennis Australia runs the AO under the supervision of ITF.

FFT runs the FO under the supervision of ITF.

AELTC run the W under the supervision of ITF.

USTA runs US Open under the supervision ITF.

Masters and the rest of the tour is ATP/WTA.

Well then why isn't the ITF policing this stuff or looking at reform, tennis' seeding system in my opinion from start to finish gives them too much leeway and discretion to play around with the matchups their rules are too complex and leave to much room for the luck of the draw. This is something that needs investigation similar to what they did to Davydenko, a real thorough look at why the numbers don't add up and the facts surrounding how the process went about.

socal1976 wrote:Well then why isn't the ITF policing this stuff or looking at reform, tennis' seeding system in my opinion from start to finish gives them too much leeway and discretion to play around with the matchups their rules are too complex and leave to much room for the luck of the draw. This is something that needs investigation similar to what they did to Davydenko, a real thorough look at why the numbers don't add up and the facts surrounding how the process went about.

1. How do you know ITF does not?
2. It is done in public domain within public purview. 'Too complex' is an arbitrary statement, very much like 'random draws'. Can you define it?
3. They do add up. You have a different opinion, or perhaps a different mathematical approach. This does not mean one is right or the other is wrong.

For example, Ryan Harrison (WC) plays Cilic in R1. Michael Yani plays Tomic (QF @ W). If USTA was involved in 'rigging', why would such matches happen?

Laverfan, well it seems that if the ITF does police them they kind of missed policing the USO for oh the last 10 years.



There are many ways to do a single elimination tournament much simpler with much less discretion and luck of the draw than what the grandslams committees generally do. The complexity gives them too much discretion and as I have pointed out discretion with little to no oversight is a recipe for abuse.



Laverfan, they don't add up, one narrow look was made at randomness and 3 of the 8 draws at the grandslams look to have failed. USO men and woman's draw, and the FO woman's draw also failed the randomness analysis for the first two rounds.



As to matter's of particular draws they have to be very subtle of course they can't completely and obviously play favorites from start to finish maybe just a little tinkering here and there around the edges to maximize some potential matchups for them. And have you seen the way Cilic has played recently that may not be that bad of matchup for harrison?

socal1976 wrote:And have you seen the way Cilic has played recently that may not be that bad of matchup for harrison?

Cilic this year is W/L 30-17, Harrison is 13-15.

laverfan wrote:

socal1976 wrote:And have you seen the way Cilic has played recently that may not be that bad of matchup for harrison?

Cilic this year is W/L 30-17, Harrison is 13-15.

Yeah, I guess you are right about that one, but harrison has been playing better in the summer on current form I think Harrison can and probably will win this one, they can't really give him anything but a difficult matchup when he is ranked in the 70s. I think this would be an obvious controversy if they gave harrison a wildcard or qualifier.

Harrison just lost to Cilic.

Yes, Laverfan I made the wrong pick, but has little to do with the argument regarding these draws. With harrison's ranking he couldn't hope to do much better than cilic.

1 in 64 doesn't prove anything at all.

Anyone with a robust knowledge of statistics knows that outlier events are an intrinsic part of the population.

"To further measure opponent difficulty, ESPN assigned scores to a player's draw by using the opponent's rank compared to all possible opponent ranks that the player could have faced in that round. So if a top two seed faced the 33rd-ranked player in the first round, he/she would get a difficulty score of 0.995 for that round; if he/she faced the 128th-ranked player in the first round, the score for that round would be 0.005. An average opponent (ranked around 80th or 81st), would correspond to a difficulty score near 0.500, which should be the average difficulty score over several years of draws.

Once ESPN had the first-round draw difficulty scores for each of the top two seeds, the scores were averaged by Grand Slams across the 10-year span (11 years for Wimbledon) to see if the top two seeds got relatively easy or difficult draws at each individual Grand Slam. The findings in the first round of the U.S. Open stood out."

http://espn.go.com/espn/otl/story/_/id/6854000/how-espn-analyzed-us-open-tennis-tournament-draw

Yesterday, Falla beat Troicki (seed #15). By the above technique, Troicki is considered a more difficult opponent (0.995) compared to Falla (ATP rank - current 119, highest 58). This is what I call arbitrary weight assignment in a statistical sense. The whole methodology seems flawed by collapsing ATP ranks into a linear scale of 128 (total slam players) without any statistical model for performance.

Here is another 'statistical' study which proves an interesting result.

http://tennis.com/articles/templates/news.aspx?articleid=10798&zoneid=25

http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0017249

Well Laverfan, this a totally different type of analysis, I don't see how this prestige ranking analysis has anything to do with what we are talking about. Of course getting troicki is a tougher draw than getting Falla, Falla happened to win but there is a reason he is rated much lower in the rankings.



Bogbrush not only are we talking about the 1 in 64 longshot of Novak drawing Roger. There is also the analysis done by two sets of statisticians that finds that the USO open draw for the top two seeds in the first two rounds is not random. That statistical analysis found only .3 percent chance that the USO draw was randomly generated and not rigged. So is your position that both 1 to 64 longshot isn't a big longshot and 3 out 1000 isn't a big longshot either. Funny how these longshots keep coming in that favor the financial interests of those who control the process.

bogbrush wrote:1 in 64 doesn't prove anything at all.

Anyone with a robust knowledge of statistics knows that outlier events are an intrinsic part of the population.

And anyone who knows human nature should be very concerned when the longshots start coming in that favor the financial interests of those controlling the process, especially since this process has little transparency and little oversight.

socal1976 wrote:

bogbrush wrote:1 in 64 doesn't prove anything at all.

Anyone with a robust knowledge of statistics knows that outlier events are an intrinsic part of the population.

And anyone who knows human nature should be very concerned when the longshots start coming in that favor the financial interests of those controlling the process, especially since this process has little transparency and little oversight.

if 1 in 64 chances were coming out in the casino's, the pit boss would probably take you outside and give you a shiny black eye!

No JM not just a 1 in 64 longshot there is also the 3 in 1000 longshot of the US open draw being random for the last ten years as well. I bet if you were in a casino and you hit a back to back 1 in 64 longshots and then 3 in 1 thousand longshot, they would certainly have a little chat with you and look at you suspiciously to investigate what is going on. That is the situation we are in now and everybodies attitude is ho-hum nothing to see here.

It's juvenile to keep on about this; as I keep telling you anyway your analysis is flawed.

You can choose to point out the apparant remoteness of chance that Federer & Djokovic meet up, but what about the expected distribution of the #1 and #3 meeting? After all, this is the first event that Djokovic and Federer have been #1 and #3 for ages, so perhaps it was well beyond due that they met.

The thing about statistics is that people who don't understand them frequently misinterpret them. Take that last point, for instance; which should we focus on? Djokovic/Federer or #1/#3?

Lies, Damn Lies, and Statistics anyone?

bogbrush wrote:It's juvenile to keep on about this; as I keep telling you anyway your analysis is flawed.

You can choose to point out the apparant remoteness of chance that Federer & Djokovic meet up, but what about the expected distribution of the #1 and #3 meeting? After all, this is the first event that Djokovic and Federer have been #1 and #3 for ages, so perhaps it was well beyond due that they met.

The thing about statistics is that people who don't understand them frequently misinterpret them. Take that last point, for instance; which should we focus on? Djokovic/Federer or #1/#3?

Lies, Damn Lies, and Statistics anyone?

No its not juvenille as serious questions are raised by the analysis of these STATISTICIANS who are professionals in the field. I would think that tennis fans would want to know if the draws are random and truely fair or if they are tinkered with. At this point, due to this study and other statistical oddities that seem to benefit the financial interests of those controlling the process then I think it isn't to much to ask to have these issues be properly investigated. Maybe looking at options of reform and of limiting the discretion of these committees that do the draw.

right I've stayed out of this debate until now, but as a mathematician (well doing a PhD in maths) in probability theory I just can't stay out of this one any more as it's getting silly, so a few points.

1) 1 in 64 probabilities DO occur. They actually occur every time you toss a coin 6 times in a row. Yes that's right, every 6-long sequence of two events with an even chance has a 1 in 64 chance of occuring. In other words, had Djokovic's last six draws (in order) read Fed-Fed-Nadal-Fed-Nadal-Fed (or whatever) it ould also have been a 1 in 64 chance.

2) 64? Have you any idea just how small 64 is? it's tiny. 1 in 64 is nothing.

3) Statistics. I admittedly only have a working knowledge of statistics, but one of the main things is that you can't actually "prove" anything conclusively (i.e. with 100% certainty). At best you get a result stipulating "there is an at least X% chance that hypothesis A is correct/false".

4) The sample size here is 6. 6!!!! that's a ridiculously small sample size to infer anything from I'm afraid, you'd be laughed at by your peers if you claimed that coffee contained caffeine based on a sample size of 6.

MfC, I did Stats up to "A" Level may years ago and completely sailed an "A" grade, poartly because I always found it a naturally easy subject to grasp. Everything you say about those odds is correct, this is a total non-story.

bogbrush wrote:MfC, I did Stats up to "A" Level may years ago and completely sailed an "A" grade, poartly because I always found it a naturally easy subject to grasp. Everything you say about those odds is correct, this is a total non-story.

I'm sure Fed is happy that he won't have to meet Nadal in a semi, because as we all know, it's 7-2 in Nadal's favour in their slam battles. The ATP and ESPN are not stupid, and they realise by keeping Federer in Novak's half it would make for a grand finale should Nadal get the expected walkover route to the final!

MFC, all the things you say I am clearly aware of. Yes long shots do come in, I am more worried about the fact that the EXACT longshot that benefits a fedal final possibility is the one that comes in. As to the work of the TWO statisticians that looked at the USO draw and the woman's draw of the FO for the last few years they found an exceedingly high probability that these were not random draws. For the USO I think they stated that in 997 random simulations were more difficult than the draws that actually happened. Out of 1000 random sims, 997 were more difficult than the actual draws the top seeds got. Those two strange longshots and the financial incentives involved give cause for a closer analysis and an investigation of these selection committees.

M4C, thank you for reiterating a post I made earlier on, that the sample size is useless as a vehicle to prove a statistical average. (which is also pointless in any determination of proof)

If you were to use the entire history of the US open Draw results and make a comparison, even then all you could show are the averages of certain permutations, which are not predictions or any kind of proof that a certain minute section of those statistics are able to prove a fact, because as you say, they are just random events compiled into a statistic which can only provide results for the analysis of averages, AKA, to make a prediction.

The only real proof of a draw being rigged is not a statistic, but visual proof of tampering.

A random event is only perceived as random after the event has happened. Until you flip a coin 6 times in a row, it doesn't become a statistic until the 6th event is over and the stats can then be compiled.