Curious case of draws at the ITF Grand Slam tournaments (2008-2011) [Archive] - MensTennisForums.com

Curious case of draws at the ITF Grand Slam tournaments (2008-2011)

RIboy
10-20-2011, 10:29 AM
http://www.playthegame.org/fileadmin/image/PTG2011/Presentation/thursday/Katarina_Pijetlovic_-_6_Oct_at_PLAY_THE_GAME_2011.pdf

we all know that but still...:rolleyes:

MuzzahLovah
10-20-2011, 01:39 PM
http://www.playthegame.org/fileadmin/image/PTG2011/Presentation/thursday/Katarina_Pijetlovic_-_6_Oct_at_PLAY_THE_GAME_2011.pdf

we all know that but still...:rolleyes:

Honestly, I'd be more shocked if I saw evidence to the contrary. Fed and Djokovic have had a semi together almost every major since he entered the top 3, way more than at random. This is obviously because Djokovic had a better chance against Nadal on grass/hardcourts, and the organizers wanted Fedal finals. Same thing with Murray not being in Fed's half. If Murray had more semis with Fed, he probably would have extended his winning record. (His only meetings with Fed were in finals, where Fed obviously had a vast edge in experience). There is a huge incentive for Fedal finals, an with results so heavily skewed, any one who flat out denies draw fixing seems pretty naive.

Now that Nadal and Fed are both on the decline, it probably won't matter so much. Hopefully they give up on the draw fixing.

alex_pgh
10-20-2011, 02:01 PM
Hmmm, let me start this reply by saying that I'm one of biggest, most dedicated Nole's fans (since 2007). I wake up, if needed, at crazy early hours just to watch Nole's matches....

Having said that, the math from this presentation is horribly, horribly wrong. Every draw is an independent event. Hence, the probability for every new draw that Nole will play Federer is *always* exactly 50%. The draw has no memory and its probabilities do not change according to past results. So even if Nole drew Federer ten consecutive times, the probability that the next draw will produce Federer vs Nole matchup is still 50%.

People get confused because of the fallacy of the "law of averages". I'll paraphrase Wikipidia:
"As invoked in everyday life, this "law" usually reflects bad statistics or wishful thinking rather than any mathematical principle. While there is a real theorem that a random variable will reflect its underlying probability over a VERY LARGE sample, the law of averages typically assumes that unnatural short-term "balance" must occur."

Source: http://en.wikipedia.org/wiki/Law_of_averages

hotdog
10-20-2011, 02:03 PM
I didn't even open that page because I know it must be a load of tosh.

Li Ching Yuen
10-20-2011, 02:10 PM
Hmmm, let me start this reply by saying that I'm one of biggest, most dedicated Nole's fans (since 2007). I wake up, if needed, at crazy early hours just to watch Nole's matches....

Having said that, the math from this presentation is horribly, horribly wrong. Every draw is an independent event. Hence, the probability for every new draw that Nole will play Federer is *always* exactly 50%. The draw has no memory and its probabilities do not change according to past results. So even if Nole drew Federer ten consecutive times, the probability that the next draw will produce Federer vs Nole matchup is still 50%.

People get confused because of the fallacy of the "law of averages". I'll paraphrase Wikipidia:
"As invoked in everyday life, this "law" usually reflects bad statistics or wishful thinking rather than any mathematical principle. While there is a real theorem that a random variable will reflect its underlying probability over a VERY LARGE sample, the law of averages typically assumes that unnatural short-term "balance" must occur."

Source: http://en.wikipedia.org/wiki/Law_of_averages

The law of averages is more telling of everything in this world than statistics on a whole could ever wet dream about.

Stop bullshiting and underestimating people's intelligence here.

If I take a coin and flip it in every house of a neighborhood I would guarantee you that the proportion of the outcome between the two faces would be closer to equal than not. Now don't come in here and tell me that the fact that I flip the fucking coin in a different house it has an effect on the overall outcome. It doesn't. In between 15 flips there's a higher probability that there's gonna be a 40%/60% or a relatively close number to that rather than an skewed one.

Chris Kuerten
10-20-2011, 02:13 PM
Pure genius, to be honest. Fix it in a way so obvious that nobody actually believes it's fixed.

Li Ching Yuen
10-20-2011, 02:18 PM
Since august of 2007 (since Djokovic played his first grand slam as world No3), Novak was on Federer's side in grand slams tournaments EVERY TIME... but two times... And guess on which two grand slams Novak was on Nadal's side? French open 2008 and 2010.


So, since august 2007, there's been 13 hard court and grass court grand slam tournaments (where Roger is favorite). And EVERY TIME Novak is in Federer's draw side. Chances of that happening are 1 against 8192.

That's 0.000122%!

So in other words, I'm gonna go play the lottery today, stupidly easy chance for me to win with similar numbers like these.
Say hello to your new overlord, MTF. By tomorrow this site is mine.

DUUUURRR DURRRR LOGICZZZZZ!!!

abraxas21
10-20-2011, 02:19 PM
I didn't even open that page because I know it must be a load of tosh.

like you, when a meal smells bad and looks rather dodgy, i close my eyes and shut my nose and pretend it's a fucken banquet

MuzzahLovah
10-20-2011, 02:22 PM
Hmmm, let me start this reply by saying that I'm one of biggest, most dedicated Nole's fans (since 2007). I wake up, if needed, at crazy early hours just to watch Nole's matches....

Having said that, the math from this presentation is horribly, horribly wrong. Every draw is an independent event. Hence, the probability for every new draw that Nole will play Federer is *always* exactly 50%. The draw has no memory and its probabilities do not change according to past results. So even if Nole drew Federer ten consecutive times, the probability that the next draw will produce Federer vs Nole matchup is still 50%.

People get confused because of the fallacy of the "law of averages". I'll paraphrase Wikipidia:
"As invoked in everyday life, this "law" usually reflects bad statistics or wishful thinking rather than any mathematical principle. While there is a real theorem that a random variable will reflect its underlying probability over a VERY LARGE sample, the law of averages typically assumes that unnatural short-term "balance" must occur."

Source: http://en.wikipedia.org/wiki/Law_of_averages

But the question here is are the draws fixed? Are they independent and giving anything close to a 50-50 result, and they are not. The chance of getting a coin to flip 15 times on heads out 15 times is incredibly low, precisely because each flip should be independent of one another.

alex_pgh
10-20-2011, 02:23 PM
I didn't say if it was fixed or not ;) I just said that it's mathematically possible....

To the "genius" who suggested flipping coin in different houses: Coin flips are already independent events. That example that you gave shows that you have no idea what an independent event is.

Again, I'm not saying anything about draws being fixed or not. I'm just saying that "0.000122%" chance from the article is far from making sense.

Ilovetheblues_86
10-20-2011, 02:29 PM
Li.

13 of 15 would be more correct

Lopez
10-20-2011, 02:33 PM
But the question here is are the draws fixed? Are they independent and giving anything close to a 50-50 result, and they are not. The chance of getting a coin to flip 15 times on heads out 15 times is incredibly low, precisely because each flip should be independent of one another.

But precisely because each event is independent the past events won't affect each other. That is to say, the chance of getting 2 tails in a row is 0,25 before you throw the first but after you've thrown it and gotten the tails you can't say that the chance of the second being tails is anything other than 50%... Unless the coin isn't fair of course ;)

Sunset of Age
10-20-2011, 02:38 PM
Stop bullshiting and underestimating people's intelligence here.

You are 'bullshiting' yourself. alex_pgh's and Lopez's explanation is purely sound.

MuzzahLovah
10-20-2011, 02:42 PM
But precisely because each event is independent the past events won't affect each other. That is to say, the chance of getting 2 tails in a row is 0,25 before you throw the first but after you've thrown it and gotten the tails you can't say that the chance of the second being tails is anything other than 50%... Unless the coin isn't fair of course ;)

The probability of getting results consecutively is dependent. 13 times you get heads from flipping a coin in a row, the probability is (1/2)^13- a very very small percentage.

Lopez
10-20-2011, 02:48 PM
The probability of getting results consecutively is dependent. 13 times you get heads from flipping a coin in a row, the probability is (1/2)^13- a very very small percentage.

The probability of getting heads on the fourteenth is still 50% after you've thrown the 13 though.

Before the fact we can say where we expect the results to settle (about 50% each) and CLT does predict that that will happen: After about an infinite number of repetitions.

Ever been to a roulette table? There are times when the past result are only a certain color.

Or you can test this yourself with Excel for example: Run a simulation of a coin toss about 10 000 times and calculate the number of heads and tails, then you can try to look for consecutive results from the simulation, I bet you'll get runs of many heads or tails if you look at the data.

echf
10-20-2011, 02:50 PM
All the "math" in those slides is horribly wrong. One would expect a lawyer to be more careful before making claims outside their field of expertise but...

First, they decide to ignore the French Open data just because it looks balanced. Great way to start, selecting the data that supports the conclusion they want to reach ;) Then there is the nonsensical statement in page 5: they say Fed and Djokovic were placed in the same half in 12 consecutive slams, which is indeed rather suspicious, but then they stress that it also includes the surfaces on which Federer was harder to beat, as if this were any further indication of rigging (of course they are included as all the slams are included, what's that supposed to even mean?). The probability on page 6 is wrong by a factor of two, it should be 2^{-11} and not 2^{-12} (which is of course not very important but shows how much thinking has gone into this). Finally, the claim that the ranking switching would decrease the probabilities is totally groundless.


Stop bullshiting and underestimating people's intelligence here.

I have no idea what you think your point is, but alex_pgh was right here. If you toss a balanced coin 100 hundred times and it comes heads every time, the probability of getting another head the next time is still 1/2. Not only does the law of large numbers not contradict this, but it is proven using this. (What does hold is that if you did actually come up with 100 consecutive heads right from the start, you would have astronomically high statistical evidence that the coin is biased or the tosses are not independent. But that's a different matter).

hotdog
10-20-2011, 02:54 PM
The lawyer is from Estonia...not exactly the hotbed of legal expertise...

abraxas21
10-20-2011, 03:04 PM
The lawyer is from Estonia...not exactly the hotbed of legal expertise...

why do brits always have to put down eastern euro nations?

hotdog
10-20-2011, 03:10 PM
My statement is fact.

MuzzahLovah
10-20-2011, 03:37 PM
The probability of getting heads on the fourteenth is still 50% after you've thrown the 13 though.

Before the fact we can say where we expect the results to settle (about 50% each) and CLT does predict that that will happen: After about an infinite number of repetitions.

Ever been to a roulette table? There are times when the past result are only a certain color.

Or you can test this yourself with Excel for example: Run a simulation of a coin toss about 10 000 times and calculate the number of heads and tails, then you can try to look for consecutive results from the simulation, I bet you'll get runs of many heads or tails if you look at the data.

The draws including Djokovic and Federer will never occur a thousand times. Under your standard, on can never cast doubt on any draw, even one so obviously skewed as this, because of of insufficient sample size. There have only been a 100 or so draws in the Open era.

We must start by recognizing the draw as a whole at the US Open highly unlikely to be random- as seen here: http://espn.go.com/espn/otl/story/_/id/6850893/espn-analysis-finds-top-seeds-tennis-us-open-had-easier-draw-statistically-likely

Now given the draw fixing goes on at the ITF level, and the astronomically low chance of getting these consecutive results in a fair draw, let's look at this. (Examining the probability consecutive results requires dependence. The chance of getting 3 heads in a row is 1/8, not 1/2. If the draw was fair, and we asked what's the probability of Djokovic and Federer being in the same half given their 13/13 previous meetings on grass and hard, it would still be 1/2. That isn't what we're talking about here. We're looking at the 13/13 as asking if the draw is likely to be fair).

Getting 15 heads in a row when flipping millions of times is not unlikely, and as you said it must happen in an infinite amount of flips- but getting a run of 15 in out of a hundred draws, that is significantly less likely. It's not impossibly that this run of draws is a random outlier, just very, very unlikely. Given the financial incentive for Fedal finals, and the propensity for draw fixing in other cases, draw fixing here again is the simplest explanation.

dodo
10-20-2011, 03:45 PM
My statement is fact.
your statement demonstrates the same level of thinking as that of the presentation author's.
way to do your country proud, champ.

echf
10-20-2011, 05:15 PM
The draws including Djokovic and Federer will never occur a thousand times.
Getting 15 heads in a row when flipping millions of times is not unlikely, and as you said it must happen in an infinite amount of flips- but getting a run of 15 in out of a hundred draws, that is significantly less likely. It's not impossibly that this run of draws is a random outlier, just very, very unlikely. Given the financial incentive for Fedal finals, and the propensity for draw fixing in other cases, draw fixing here again is the simplest explanation.

For reference, the probability of getting at least 13 consecutive heads OR at least 13 consecutive tails out of 100 tosses is 1.08% (I have computed it precisely). Which is unlikely, but not ``astronomically'' so. The most likely outcome is to get a run of length 7. In general the expected size of the longest run after n tosses can be shown to be roughly log_2(n).

LawrenceOfTennis
10-20-2011, 05:16 PM
Called 'rigged'.

MuzzahLovah
10-20-2011, 05:22 PM
For reference, the probability of getting at least 13 consecutive heads OR at least 13 consecutive tails out of 100 tosses is 1.08% (I have computed it precisely). Which is unlikely, but not ``astronomically'' so. The most likely outcome is to get a run of length 7. In general the expected size of the longest run after n tosses can be shown to be roughly log_2(n).

So, all your saying is 99% of the time you wouldn't get 13 heads in a row out of all your tosses. I guess not astronomically unlikely, but
"very, very unlikely."

echf
10-20-2011, 05:35 PM
So, all your saying is 99% of the time you wouldn't get 13 heads in a row out of all your tosses. I guess not astronomically unlikely, but
"very, very unlikely."
It is quite unlikely, I don't think anyone disagrees on that. But a chance of 1% is still high enough not to rule out the possibility of it happening by mere chance in my opinion.

On the other hand, even if it is not random that doesn't mean the draw is fixed. It might be that the process used as a source of randomness (picking pieces of paper from a urn or whatever they do, does anyone know?) is just not good enough.

Henry Chinaski
10-20-2011, 05:46 PM
if the draws are fixed (and you'd have to be seriously naive to rule it out completely) it's likely to be IMG doing the fixing, not nike as the presenter implies.

there are obviously a lot of massive holes in the presentation

The Magician
10-20-2011, 06:05 PM
Even if we didn't have overwhelming statistical evidence that the draws are fixed, which we do, the fact that there is so much money in tennis, so little oversight, and so much incentive to fix draws means it would take an extraordinary effort by all of the corporations to come together and promise not to bribe the ATP. Everyone knows that matches are fixed, everyone's a doper, and certain players have a larger influence than others, but if you talk about the fixed draws you get outrage :eek: It's obvious the draws are fixed, the interesting thing is why they're fixed to favor Nadull and disfavor Murray so much when you would think that British tennis, which has a lot of money, would be more powerful than Spanish corporations. I can't wait for Uncle Tony's tell all book about how be bought and controlled the all of tennis.

Pirata.
10-20-2011, 07:37 PM
Can't believe that nobody has commented on the fact that only the top two are pictured with their slam trophies :haha: Novak gets Rome (I think?) while Murray's is big and looks like a slam trophy, but it's only from Queens.

Lopez
10-20-2011, 08:01 PM
The draws including Djokovic and Federer will never occur a thousand times. Under your standard, on can never cast doubt on any draw, even one so obviously skewed as this, because of of insufficient sample size. There have only been a 100 or so draws in the Open era.

We must start by recognizing the draw as a whole at the US Open highly unlikely to be random- as seen here: http://espn.go.com/espn/otl/story/_/id/6850893/espn-analysis-finds-top-seeds-tennis-us-open-had-easier-draw-statistically-likely

Now given the draw fixing goes on at the ITF level, and the astronomically low chance of getting these consecutive results in a fair draw, let's look at this. (Examining the probability consecutive results requires dependence. The chance of getting 3 heads in a row is 1/8, not 1/2. If the draw was fair, and we asked what's the probability of Djokovic and Federer being in the same half given their 13/13 previous meetings on grass and hard, it would still be 1/2. That isn't what we're talking about here. We're looking at the 13/13 as asking if the draw is likely to be fair).

Getting 15 heads in a row when flipping millions of times is not unlikely, and as you said it must happen in an infinite amount of flips- but getting a run of 15 in out of a hundred draws, that is significantly less likely. It's not impossibly that this run of draws is a random outlier, just very, very unlikely. Given the financial incentive for Fedal finals, and the propensity for draw fixing in other cases, draw fixing here again is the simplest explanation.

The sample sizes are way too small to assume that CLT holds anyway :shrug:. And besides if we're looking at the draws made in the open era, then we can merely look whether there's a difference between 1 and 3 meeting in the semis opposed to 1 and 4 meeting (assuming that draws were made similarly back then as they are now). Federer and Djokovic were 2 and 3 before the USO, 1 and 3 at the USO.

Given also the small sample size, deviation can have dramatic influence on the results. We could calculate the confidence intervals if you wish ;).

Besides, as someone pointed out, it's stupid to rule out some of the draws as "outliers" if they don't fit your perspective, that totally defeats the purpose of statistical analysis.

The pdf. looks very unprofessional btw :lol:

ivanban
10-20-2011, 09:29 PM
You are 'bullshiting' yourself. alex_pgh's and Lopez's explanation is purely sound.

....and then I saw your signature.....

Li Ching Yuen
10-20-2011, 10:30 PM
Don Tiriac had this to say about the clay masters in Madrid around the Wimbledon final this year: We're doing it for 3 years now, and two of them have been complete successes with Nadal - Federer finals with both of them winning one.

Now in Tiriac's language (which is that of an elite "bussiness-man") that means something very important for the money side of things (read, actual interest). There's nothing sweeter than making people that actually don't really follow the sport jump on it like it's the hottest rodeo in town.
And mind you, this is in Spain where Nadal is numero uno right?...so translating this to a certain...let's say...US Open with the kind of "bussiness-stuff" going around that tournament and the various parties implicated I would like to propose the thought of a certain duo reaching the finals and making the most of the beloved sport on it's biggest stage...ahhh la la sounds like poetry, sounds like zing-zing doesn't it?
And since that never happened *there* and with the years going by and the seeds placement ever changing...ahhh I see the sour look of a grey-fluffy haired gentlemen looking in dismay at all the Monday finishes and thinking of "what could have been"... a fairy-tale vision I tell ya.

Of course, this is just an exercise of imagination from my part, everyone being entitled to it.

Or of course, we could dwelve into the mathematical side of it and just shout-out at each other about how seeds have changed, how Irene hit, how lightning strikes twice in the same place and how Alice really exists not just only in Wonderland.

Oh nevermind me, I'm just thinking out loud.
Carry on.

MIMIC
10-20-2011, 11:45 PM
1 can draw 3/4/5 evenly and the PLAYERS associated with those seeds could still draw each other 1,000 times in a row.

Some people just do not get it. It's not about the freakin seeds.

Rodre Fegassi
10-21-2011, 09:35 AM
Who cares if it's rigged or not, if the tournament organizers want to put Federer/Djokovic and Nadal/Murray as the semifinal of every tournament - why shouldn't they? It's their tournament, not yours.

And if it's meant to be random, then IT IS random. Simple as. Why bother doing the math, when they've already done the math better than you ever could.

fabolous
10-21-2011, 10:27 AM
Having said that, the math from this presentation is horribly, horribly wrong. Every draw is an independent event. Hence, the probability for every new draw that Nole will play Federer is *always* exactly 50%. The draw has no memory and its probabilities do not change according to past results. So even if Nole drew Federer ten consecutive times, the probability that the next draw will produce Federer vs Nole matchup is still 50%.
another one who got the principle of independent chances wrong (i see that sunset of age still is among them, too). of course, the chance for the new draw is always 50%. but that is not how you calculate the chance for an event to happen like 12 of 12 times or 13 out of 15 times. as li ching yuen has rightly stated, if you flip a coin 15 times, it is much more likely to get a balanced outcome than an unbalanced, although the chance for every single flip is 50% for head and 50% for tail.

the way you are talking seems to implicate that drawing djokovic and fed together 12 out of 12 times has the same probability than drawing them 6 out of 12 times or 1 out of 12 times. this is not true.

barbadosan
10-21-2011, 12:40 PM
1 can draw 3/4/5 evenly and the PLAYERS associated with those seeds could still draw each other 1,000 times in a row.

Some people just do not get it. It's not about the freakin seeds.

Exactly. Somebody elsewhere did a listing of how the numbered seeds 1-4 were drawn for last 4 years - by the NUMBERS, which is how they're drawn. Turns out 1 drew 4, and 1 drew 3 almost even number of times, and similarly for 2 drawing 3 and 2 drawing 4. What is happening is that none of the people associated with the seeding numbers is remaining constant.

BTW, don't some draws not pick numbers by hand, but actually use those lotto ball machines? So you're telling me that they're weighting ball 4 so ball 3 would come up when they're picking for one half??

Sunset of Age
10-21-2011, 12:49 PM
....and then I saw your signature.....

Care to explain what my signature has to do with this subject? :rolleyes:

Exactly. Somebody elsewhere did a listing of how the numbered seeds 1-4 were drawn for last 4 years - by the NUMBERS, which is how they're drawn. Turns out 1 drew 4, and 1 drew 3 almost even number of times, and similarly for 2 drawing 3 and 2 drawing 4. What is happening is that none of the people associated with the seeding numbers is remaining constant.

BTW, don't some draws not pick numbers by hand, but actually use those lotto ball machines? So you're telling me that they're weighting ball 4 so ball 3 would come up when they're picking for one half??

Always nice to see someone around with some common sense.

tnosugar
10-21-2011, 02:21 PM
The law of averages is more telling of everything in this world than statistics on a whole could ever wet dream about.

Stop bullshiting and underestimating people's intelligence here.

If I take a coin and flip it in every house of a neighborhood I would guarantee you that the proportion of the outcome between the two faces would be closer to equal than not. Now don't come in here and tell me that the fact that I flip the fucking coin in a different house it has an effect on the overall outcome. It doesn't. In between 15 flips there's a higher probability that there's gonna be a 40%/60% or a relatively close number to that rather than an skewed one.

:yeah:

tnosugar
10-21-2011, 02:42 PM
Hmmm, let me start this reply by saying that I'm one of biggest, most dedicated Nole's fans (since 2007). I wake up, if needed, at crazy early hours just to watch Nole's matches....

Having said that, the math from this presentation is horribly, horribly wrong. Every draw is an independent event. Hence, the probability for every new draw that Nole will play Federer is *always* exactly 50%. The draw has no memory and its probabilities do not change according to past results. So even if Nole drew Federer ten consecutive times, the probability that the next draw will produce Federer vs Nole matchup is still 50%.

People get confused because of the fallacy of the "law of averages". I'll paraphrase Wikipidia:
"As invoked in everyday life, this "law" usually reflects bad statistics or wishful thinking rather than any mathematical principle. While there is a real theorem that a random variable will reflect its underlying probability over a VERY LARGE sample, the law of averages typically assumes that unnatural short-term "balance" must occur."

Source: http://en.wikipedia.org/wiki/Law_of_averages

the point is that each separate event (Djokovic and Fed in the same half) has a a 50% chance

however, when you treat 13 draws in a row as linked events (ie as one event) then there is no way you can apply your logic. the probability decreases by 50% with each draw so:


(1) 50% -> (2) 25% -> (3) 12.5% -> (4) 6.25% -> (5) 3.125& -> (6) 1.5625% -> (7) 0.78125 -> (8) 0.390625 -> (9) 0.1953125 -> (10) 0.09765625 -> (11) 0.048828125 -> (12) 0.0244140625 -> (13) 0.0122070312

so it's quite obvious they were fixed.

Sophocles
10-21-2011, 03:03 PM
IF the results are random, then there is indeed a 50% chance of Djoker's & Fed's being on opposite sides of the draw in the next slam. But we're not discussing the chances of their being on opposite sides in the next slam. We're discussing the probability that they ended up on the same side in 13 out of the last 15 slams with random draws.

barbadosan
10-21-2011, 03:12 PM
the point is that each separate event (Djokovic and Fed in the same half) has a a 50% chance

however, when you treat 13 draws in a row as linked events (ie as one event) then there is no way you can apply your logic. the probability decreases by 50% with each draw so:


(1) 50% -> (2) 25% -> (3) 12.5% -> (4) 6.25% -> (5) 3.125& -> (6) 1.5625% -> (7) 0.78125 -> (8) 0.390625 -> (9) 0.1953125 -> (10) 0.09765625 -> (11) 0.048828125 -> (12) 0.0244140625 -> (13) 0.0122070312

so it's quite obvious they were fixed.

Except that they are not drawing Djokovic, Nadal, Murray or Fed, or any other named player for that matter --- they're drawing 3 or 4, to play 1, or 3 or 4 to play 2 - they aren't drawing NAMES, they draw NUMBERS. So why don't you take a look yourself and see how often no 3 is drawn with no 1, or no 4 with no 1.

The top 4 players, or other groups for that matter, did not remain static in their numbered seeding for the last 4 years

MuzzahLovah
10-21-2011, 03:36 PM
Except that they are not drawing Djokovic, Nadal, Murray or Fed, or any other named player for that matter --- they're drawing 3 or 4, to play 1, or 3 or 4 to play 2 - they aren't drawing NAMES, they draw NUMBERS. So why don't you take a look yourself and see how often no 3 is drawn with no 1, or no 4 with no 1.

The top 4 players, or other groups for that matter, did not remain static in their numbered seeding for the last 4 years

Which makes the fixing more evident. Any two players being on the same side of the draw 13/15 is very low probability, but when they players are top players and you have a huge financial incentive to put them there, you have to been hopelessly naive to not at east have some doubt to the fairness of the draw.

Sophocles
10-21-2011, 04:32 PM
The top 4 players, or other groups for that matter, did not remain static in their numbered seeding for the last 4 years

But Djoker & Fed have been potential semi-final opponents in all of them (I think).

MIMIC
10-21-2011, 09:54 PM
Except that they are not drawing Djokovic, Nadal, Murray or Fed, or any other named player for that matter --- they're drawing 3 or 4, to play 1, or 3 or 4 to play 2 - they aren't drawing NAMES, they draw NUMBERS. So why don't you take a look yourself and see how often no 3 is drawn with no 1, or no 4 with no 1.

The names are ATTACHED to the numbers well before the draw ceremony so they might as well be drawing names. It's not like we don't know who they're talking about when they say "Number 1 will play Number 4 in the semis".


The top 4 players, or other groups for that matter, did not remain static in their numbered seeding for the last 4 years

And in past eras, other players' seedings didn't remain fixed, either. Yet, they weren't drawn in the same half nearly as many times as Fed-Djokovic & Nadal-Murray.

echf
10-21-2011, 11:27 PM
The sample sizes are way too small to assume that CLT holds anyway [...] Given also the small sample size, deviation can have dramatic influence on the results. We could calculate the confidence intervals if you wish ;).

The point is that the sample IS large enough to draw some conclusions. You speak about the possibility of calculating the confidence intervals, but I already did: as I wrote, with a confidence of 99%, if we take a period of 25 years (= 100 Grand Slam draws) the longest run of consecutive seedings where (say) the 3rd seeded player lands in the same half is 13 or less. So yes, we are speaking about an event that is unlikely (1%), but not outrageously so. (I guess the open era has been around for longer than 25 years but the results won't change much).
On the other hand, I haven't checked if that 13/13 figure includes the French Open or not. If it doesn't this computation is meaningless.

echf
10-21-2011, 11:31 PM
1 can draw 3/4/5 evenly and the PLAYERS associated with those seeds could still draw each other 1,000 times in a row.

Some people just do not get it. It's not about the freakin seeds.

You are quite mistaken here. Whether the draw is made on the numbers or the players does not matter at all. In both cases, the event "Fed and Djoko are in the same half for t consecutive times" has exactly the same probability (assuming they do not become #1 and #2).
You're right that the draw can be "balanced" with regard to the seeds and not with regard to the players. But a truly random draw will be balanced with regard to BOTH.

echf
10-21-2011, 11:35 PM
the point is that each separate event (Djokovic and Fed in the same half) has a a 50% chance

however, when you treat 13 draws in a row as linked events (ie as one event) then there is no way you can apply your logic. the probability decreases by 50% with each draw so:


(1) 50% -> (2) 25% -> (3) 12.5% -> (4) 6.25% -> (5) 3.125& -> (6) 1.5625% -> (7) 0.78125 -> (8) 0.390625 -> (9) 0.1953125 -> (10) 0.09765625 -> (11) 0.048828125 -> (12) 0.0244140625 -> (13) 0.0122070312

so it's quite obvious they were fixed.
This calculation is only correct if you want to compute the probability that Djokovic and Fed are in the same half in the next 13 slams starting right NOW (or at any fixed point in time). But this doesn't mean much unless you had a separate reason to start counting specifically from that point in time. What is relevant is the probability that we witness something like this over the period of several decades, which is much larger (see other posts).

echf
10-21-2011, 11:39 PM
The law of averages is more telling of everything in this world than statistics on a whole could ever wet dream about.

Stop bullshiting and underestimating people's intelligence here.

If I take a coin and flip it in every house of a neighborhood I would guarantee you that the proportion of the outcome between the two faces would be closer to equal than not. Now don't come in here and tell me that the fact that I flip the fucking coin in a different house it has an effect on the overall outcome. It doesn't. In between 15 flips there's a higher probability that there's gonna be a 40%/60% or a relatively close number to that rather than an skewed one.

:yeah:

How can you :yeah: this post? It doesn't make any sense at all. It's not even clear what it's trying to argue or in response to what... Who the heck said that the outcome of one flip will affect the other? What I and everybody else is saying is that they are independent, which is precisely the opposite.

MIMIC
10-22-2011, 12:06 AM
You are quite mistaken here. Whether the draw is made on the numbers or the players does not matter at all. In both cases, the event "Fed and Djoko are in the same half for t consecutive times" has exactly the same probability (assuming they do not become #1 and #2).
You're right that the draw can be "balanced" with regard to the seeds and not with regard to the players. But a truly random draw will be balanced with regard to BOTH.

But that's what this thread is about. With regard to the players (Fed/Djokovic & Nadal/Murray), it isn't balanced.

Say Hey Kid
10-22-2011, 05:31 AM
Actually ESPN did an investigation into the draws of the US Open. Some of the top statisticans and probability experts in the country did the math.
http://espn.go.com/espn/otl/story/_/id/6850893/espn-analysis-finds-top-seeds-tennis-us-open-had-easier-draw-statistically-likely


"Outside the Lines" analyzed the average difficulty -- determined by the players' ATP or WTA rankings before the draws -- of those who played the top two seeds in all Grand Slams over 10 years. That was compared to 1,000 random simulations of 10 years of Grand Slam draws -- or the equivalent of producing 10,000 random draws taken 10 years at a time.


Only three of OTL's 1,000 simulations produced first-round opponents as easy as those the top two men's seeds have actually faced on average over 10 years in the U.S. Open. In none of the 1,000 simulations did OTL get the extreme results found in 10 years of actual opening matchups for the top two women's Open seeds.

Statiscan Dr. Andrew Swift then did his own simulations. Only four times in 1 million simulations did Dr. Andrew Swift come up with an average ranking equal to or easier than what was actually observed in the men's and women's draws over the last 10 years.


"There's always the chance that, yes, freak occurrences happen. But you're telling me a freak occurrence has happened with the men and the women?" he asked. "Double freak occurrences?" Any way you want to look at these, there is significant evidence here that these did not come from a random draw
-Dr. Andrew Swift. Former chairman of the American Statistical Association's Section on Statistics in Sports.


No evidence of draw fixing in the USO, what so ever.:haha:

echf
10-22-2011, 12:02 PM
Actually ESPN did an investigation into the draws of the US Open. Some of the top statisticans and probability experts in the country did the math.
http://espn.go.com/espn/otl/story/_/id/6850893/espn-analysis-finds-top-seeds-tennis-us-open-had-easier-draw-statistically-likely


"Outside the Lines" analyzed the average difficulty -- determined by the players' ATP or WTA rankings before the draws -- of those who played the top two seeds in all Grand Slams over 10 years. That was compared to 1,000 random simulations of 10 years of Grand Slam draws -- or the equivalent of producing 10,000 random draws taken 10 years at a time.


Only three of OTL's 1,000 simulations produced first-round opponents as easy as those the top two men's seeds have actually faced on average over 10 years in the U.S. Open. In none of the 1,000 simulations did OTL get the extreme results found in 10 years of actual opening matchups for the top two women's Open seeds.

Statiscan Dr. Andrew Swift then did his own simulations. Only four times in 1 million simulations did Dr. Andrew Swift come up with an average ranking equal to or easier than what was actually observed in the men's and women's draws over the last 10 years.


"There's always the chance that, yes, freak occurrences happen. But you're telling me a freak occurrence has happened with the men and the women?" he asked. "Double freak occurrences?" Any way you want to look at these, there is significant evidence here that these did not come from a random draw
-Dr. Andrew Swift. Former chairman of the American Statistical Association's Section on Statistics in Sports.


No evidence of draw fixing in the USO, what so ever.:haha:
The problem with the link you provide is that it is way too vague. Nowhere is it said what they mean precisely by "draws as easy as the actual draws". You cannot take those figures at face value without knowing what they mean. (By the way, it doesn't say either that top statisticians or probability theorists were involved in any way. At any rate the conclusion seems based on simulations rather than actual calculations).

echf
10-22-2011, 12:13 PM
By the way, hereīs an interesting article

http://www.significancemagazine.org/details/webexclusive/1327211/Advantage-top-seeds---the-US-Open-random-draw.html

It makes exactly one of the points Iīve been trying to make:

It is completely reasonable to suspect that the US Open draw is neither rigged nor random.

For example, this tennis situation reminds me of a 1971 paper by S.E. Fienberg, "Randomization and Social Affairs: The 1970 Draft Lottery".Fienberg examined the results of the 1970 Selective Service draft lottery and demonstrated that the lottery was inconsistent with the statistical definition of randomness. This lottery was performed by actually placing each number 1 through 366 in a capsule into a bowl and drawing, rather than using a computer to generate random numbers. Based on Fienberg's paper, this lack of randomness likely occurred because of inadequate mixing of the capsules, rather than some malicious intentions.

This seems to answer my question of how the draws are picked too: a computer program is used. It may just be that the pseudorandom generator used in the program is not good enough for this application.

Benny_Maths
10-22-2011, 02:12 PM
But that's what this thread is about. With regard to the players (Fed/Djokovic & Nadal/Murray), it isn't balanced.

Spot on. If draw rigging occurs then it violates the 'coin flip' independence assumption. You can only have one or the other. If you were to assume independence then you're implicitly asserting that no rigging occurs.

echf
10-22-2011, 02:51 PM
But that's what this thread is about. With regard to the players (Fed/Djokovic & Nadal/Murray), it isn't balanced.
Yes, I think I know what the thread is about. Do you have a point? We are trying determine just how unlikely the imbalance is. Saying "it isn't balanced" without estimating the probabilities is useless.

Spot on. If draw rigging occurs then it violates the 'coin flip' independence assumption. You can only have one or the other. If you were to assume independence then you're implicitly asserting that no rigging occurs.
But that's exactly the way to proceed. You assume independence, then check whether this assumption is consistent with the data or not. (On the other hand what you said is not quite correct, the draws could be independent yet still be rigged, the same way a coin can be biased and land heads most of the time even if the tosses are independent).

Lopez
10-22-2011, 03:25 PM
The point is that the sample IS large enough to draw some conclusions. You speak about the possibility of calculating the confidence intervals, but I already did: as I wrote, with a confidence of 99%, if we take a period of 25 years (= 100 Grand Slam draws) the longest run of consecutive seedings where (say) the 3rd seeded player lands in the same half is 13 or less. So yes, we are speaking about an event that is unlikely (1%), but not outrageously so. (I guess the open era has been around for longer than 25 years but the results won't change much).
On the other hand, I haven't checked if that 13/13 figure includes the French Open or not. If it doesn't this computation is meaningless.

Didn't know you made the calculations :). But yeah there never was a run of 13. From 2008, Djokovic and Federer were in the same halves in:
AO 2008

W 2008
USO 2008
AO 2009
FO 2009
W 2009
USO 2009
AO 2010

W 2010
USO 2010
AO 2011
FO 2011
W 2011
USO 2011

So that's one run of 7 and one of 6. During this time their seedings didn't remain constant though.

MIMIC
10-22-2011, 09:08 PM
Yes, I think I know what the thread is about. Do you have a point? We are trying determine just how unlikely the imbalance is. Saying "it isn't balanced" without estimating the probabilities is useless.

Well I'm not one for calculating probabilities but I know just by going from my work in this thread (http://www.menstennisforums.com/showthread.php?t=189197&highlight=) that this type of thing has never happened before (or even come close to happening).

I just know that the odds of this are extremely low, and that combined with the huge incentive to do it (that is, rig the draw) makes it very suspicious to me.

Sapeod
10-22-2011, 11:26 PM
The atp obviously wanted Nadal to win more slams. They wanted him to face Federer in the final, so he could win. Federer started to become mentally weak against Nadal but was able to beat Djokovic more often than not. Therefore, they placed Murray in Nadal's half (hoping Murray would lose beforehand or just hoping Nadal would win) and Federer in Djokovic's half so he could beat Djokovic. Then, when Federer and Nadal faced each other in the final, the atp could sit back and watch their idol win another slam. Seems Djokovic has blown all of that up. The one thing Djokovic has done right: stopped Nadal from winning slams. Not that it's a good thing he's now winning slams. Djokovic is just as bad him. Nadal or Djokovic winning slams? Either way, tennis loses.

sicko
10-22-2011, 11:31 PM
Nadal or Djokovic winning slams? Either way, tennis loses.

So get used to tennis losing a lot in the next years.

Roadmap
10-22-2011, 11:44 PM
So get used to tennis losing a lot in the next years.

:spit: Nadull might scrape himself another French Open. Faker? :haha: That clown will do well to win another slam. Luckovic won't win Roland Garros or retain his Wimbledon title. If eggman does not win down under then his only chance is at Flushing Meadows. Can he handle the pressure? :D

Roadmap
10-22-2011, 11:47 PM
Faker is a flash in the pan.

sicko
10-22-2011, 11:52 PM
They won 7 out of the last 8 slams. You better wake up.

Saberq
10-22-2011, 11:52 PM
Faker is a flash in the pan.

hope my young Padawan.......Novak will first clean up AO like it's nothing and then RG while your boys will hold the trays or watching the final at home ....

n8
10-23-2011, 12:43 AM
They won 7 out of the last 8 slams. You better wake up.

That's correct, but it sounds more impressive saying that they've won the last 7 Slams.

The Magician
10-23-2011, 02:17 AM
:spit: Nadull might scrape himself another French Open. Faker? :haha: That clown will do well to win another slam. Luckovic won't win Roland Garros or retain his Wimbledon title. If eggman does not win down under then his only chance is at Flushing Meadows. Can he handle the pressure? :D

Interesting point. If Fakervic doesn't win at the AO, which seems very likely since he is very bad at defending titles, there's an excellent chance he won't win a slam all year. In that case we could be in for another 2009-2010 when he fell apart except this time his peak years will be passing him by. I really feel like Fakervic is done, it simply seems impossible for him to win another slam for at least 2 years :shrug:

MIMIC
10-23-2011, 06:09 AM
:spit: Nadull might scrape himself another French Open. Faker? :haha: That clown will do well to win another slam. Luckovic won't win Roland Garros or retain his Wimbledon title. If eggman does not win down under then his only chance is at Flushing Meadows. Can he handle the pressure? :D

Bitterness: 9/10
Flame bait: 6/10
Defiance of logic: 8/10
Subtly: 0/10

6/10 troll attempt

sicko
10-23-2011, 07:54 AM
That's correct, but it sounds more impressive saying that they've won the last 7 Slams.

True.

Btw, Rafole are closing in big time on Fedray's GS count. Now 2 short.

EddieNero
10-23-2011, 08:08 AM
The fact of rigging draws is more than obvious, personally I just got over this.

Lopez
10-25-2011, 12:02 PM
:lol:

This "study" made the sports websites in Finland. Pretty embarrassing... They could at least use a more credible source for that piece of news or do their own math instead of using this piece of junk :p

abraxas21
10-25-2011, 01:20 PM
Fakervic might win another GS. it's such a mug era that any semi decent player willing to put enough dedication to compensate for their lack of talent can make a GS these days.

but in any case, he might win at most 2. he'll be remebered as the guy who won a few GS in a transitional era of what already was a pretty mug era overall to begin with. but i think other things he did will make him more famous. if he manages to get the career retirement GS, well then that's something to be proud of. then again, with his "gluten free treatment", who knows? he might last for a long time. i just wonder if there are any nasty side effects derived from that "treatment"...

Rodre Fegassi
10-25-2011, 01:29 PM
They won 7 out of the last 8 slams. You better wake up.

They won 100% of the last 7 slams, and I'd bet my house that they'll win 20 out of the next 20 slams ie taking us up to end of 2016.

networthy
10-25-2011, 07:58 PM
why do brits always have to put down eastern euro nations?

Umm, isn't it pretty obvious...?

GSMnadal
10-25-2011, 08:29 PM
I just love how the French Open is scratched, simply because it doesn't fit the whole conspiracy theory :rolleyes:

PureStorm
10-29-2011, 05:48 PM
Some unbelievable statistic that make you question some of the Federer's 16 Grand Slam titles...and partly explain why Murray has not captured a Grand Slam title so far.
http://www.facebook.com/pages/Tennis-biggest-public-secret/284730451547505?sk=info

In short, it says that "in 12 out of 12 Grand Slam tournaments played on hard and grass courts between 2008-2011, Federer and Djokovic were always drawn to the same half of the draw, while Nadal and Murray were drawn to the other half. In addition, in five of those 12 tournaments Murray was not among the first four seeded players, so his draw was conducted separately five times. Thus, the statistics are as follows: to get the same result 12 out of 12 times, probability is 1 in 4096. For the case of Murray's separate draws which produced the same result 5 out of 5 times, it is 1 in 32. THEREFORE, THE PROBABILITY TO OBTAIN DRAW RESULTS AS OBTAINED AT THE 12 GRAND SLAM TOURNAMENTS IS 131072 TO 1 (4096 x 32 = 131072). Combined with the study conducted by ESPN on the draws of unseeded players at US Open it is ca. 32 BILLIONS TO 1."

"Facts and statistics strongly indicate that draws at the Grand Slam tournaments 2008-2011 have been fixed at the very top of men's tennis"

TBkeeper
10-29-2011, 06:02 PM
Yes they are :(

Dougie
10-29-2011, 06:08 PM
This is getting old...Several threads about this already.

PureStorm
10-29-2011, 06:17 PM
Sorry I'm new, I did not know that it has been discussed already... but this is the first time I see the research done...there is a FB page about it. It is time someone did something concrete so that's why this is new, talking on forums is old, you're right, and it is not going to change anything...So seems like something concrete is finally taking place. Le Monde, Der Spiegel, Daily Mail...have already reported it. Interesting to see the developments. All the organisers and ITF and ATP are still not responding to any of the journalists' questions, and seems like they want to brush everything off as a coincidence... they think people are dumb and blind.

xdrewitdajx
10-29-2011, 06:33 PM
-insert reply about flawed statistical analysis
-insert rebuttal, defense of statistical analysis
-insert example of similar statistical "anomalies" that go unnoticed
-insert conspiracy theory about why draws are rigged
-insert joke about conspiracy theorists
-insert response about that kind of attitude being the very reason things like this are allowed to happen, and nothing is done
-repeat as necessary

MuzzahLovah
10-29-2011, 06:40 PM
I think the real question is, why hasn't ESPN or some sports journalism outfit investigated this or reported on it? I've commentators mention it in passing like- wow, Fed and Djokovic have played in the same half almost every time in the slams, even though the draws are supposed to be 50-50 and their rankings have fluctuated. But I'd rather have a reputable source investigate this and confirm or debunk it- the arguments here go nowhere.

PureStorm
10-29-2011, 06:42 PM
This study looks just fine to me... http://www.facebook.com/pages/Tennis-biggest-public-secret/284730451547505?sk=wall
In fact, I do not see how can anyone argue against it...

The presentation does not fit into the study...maths in presentation were fine but incomplete... This study is the same lawyer's study and it completes the presentation. Seems like the facts were just made too simple in the presentation, perhaps to not confuse the audience :-)

xdrewitdajx
10-29-2011, 06:42 PM
I think the real question is, why hasn't ESPN or some sports journalism outfit investigated this or reported on it? I've commentators mention it in passing like- wow, Fed and Djokovic have played in the same half almost every time in the slams, even though the draws are supposed to be 50-50 and their rankings have fluctuated. But I'd rather have a reputable source investigate this and confirm or debunk it- the arguments here go nowhere.

think about this a bit, though.

PureStorm
10-29-2011, 06:47 PM
This is all from the same study and I cannot see a thing wrong with it!

INCREDIBLE RESULTS OF THE DRAWS:

AUSTRALIAN OPEN 2011
1st seed NADAL
2nd seed FEDERER
3rd seed DJOKOVIC
5th seed MURRAY
Result of the draw:
DJOKOVIC IN FEDERER’s HALF; MURRAY IN NADAL’s HALF

WIMBLEDON 2011
1st seed NADAL
2nd seed DJOKOVIC
3rd seed FEDERER
4th seed MURRAY
Result of the draw:
DJOKOVIC IN FEDERER’s HALF; MURRAY IN NADAL’s HALF

US OPEN 2011
1st seed DJOKOVIC
2nd seed NADAL
3rd seed FEDERER
4th seed MURRAY
Result of the draw:
DJOKOVIC IN FEDERER’s HALF; MURRAY IN NADAL’s HALF

AUSTRALIAN OPEN 2010
1st seed FEDERER
2nd seed NADAL
3rd seed DJOKOVIC
5th seed MURRAY
Result of the draw:
DJOKOVIC IN FEDERER’s HALF; MURRAY IN NADAL’s HALF

WIMBLEDON 2010
1st seed FEDERER
2nd seed NADAL
3rd seed DJOKOVIC
4th seed MURRAY
Result of the draw:
DJOKOVIC IN FEDERER’s HALF; MURRAY IN NADAL’s HALF

US OPEN 2010
1st seed NADAL
2nd seed FEDERER
3rd seed DJOKOVIC
4th seed MURRAY
Result of the draw:
DJOKOVIC IN FEDERER’s HALF; MURRAY IN NADAL’s HALF

AUSTRALIAN OPEN 2009
1st seed NADAL
2nd seed FEDERER
3rd seed DJOKOVIC
4th seed MURRAY
Result of the draw:
DJOKOVIC IN FEDERER’s HALF; MURRAY IN NADAL’s HALF

WIMBLEDON 2009
1st seed NADAL (withdrew)
2nd seed FEDERER
3rd seed MURRAY
4th seed DJOKOVIC
Result of the draw:
DJOKOVIC IN FEDERER’s HALF; MURRAY IN NADAL’s HALF (Nadal withdrew due to injury but this is irrelevant in this context)

US OPEN 2009
1st seed FEDERER
2nd seed MURRAY
3rd seed NADAL
4th seed DJOKOVIC
Result of the draw:
DJOKOVIC IN FEDERER’s HALF; MURRAY IN NADAL’s HALF

AUSTRALIAN OPEN 2008
1st seed FEDERER
2nd seed NADAL
3rd seed DJOKOVIC
9th seed MURRAY
Result of the draw:
DJOKOVIC IN FEDERER’s HALF; MURRAY IN NADAL’s HALF

WIMBLEDON 2008
1st seed FEDERER
2nd seed NADAL
3rd seed DJOKOVIC
12th seed MURRAY
Result of the draw:
DJOKOVIC IN FEDERER’s HALF; MURRAY IN NADAL’s HALF

US OPEN 2008
1st seed NADAL
2nd seed FEDERER
3rd seed DJOKOVIC
6th seed MURRAY
Result of the draw:
DJOKOVIC IN FEDERER’s HALF; MURRAY IN NADAL’s HALF

A CASE OF ROLAND GARROS (the only clay court Slam)

Results on ALL clay tournaments are as follows: Murray never played Federer on any clay court tournament. Nadal leads 4:0 against Murray on clay. Nadal was likely to win whomever he got on clay, Djokovic or Murray. Djokovic played 11 matches on clay with Nadal of which he lost 9 first matches. Federer vs. Djokovic on clay - 3:1. Federer was likely to win on clay against Djokovic.

ROLAND GARROS 2011
1st seed NADAL
2nd seed DJOKOVIC
3rd seed FEDERER
4th seed MURRAY
Result of the draw:
DJOKOVIC IN FEDERER’s HALF; MURRAY IN NADAL’s HALF

ROLAND GARROS 2010
1st seed FEDERER
2nd seed NADAL
3rd seed DJOKOVIC
4th seed MURRAY
Result of the draw:
MURRAY IN FEDERER’s HALF; DJOKOVIC IN NADAL’s HALF (the only draw that does not fit the formal pattern perfectly. But substance is different. Regardless of the draw Murray was likely never to make it to the semis on the this Slam, and he barely passed the first round match against Gasquet)

ROLAND GARROS 2009
1st seed NADAL
2nd seed FEDERER
3rd seed MURRAY
4th seed DJOKOVIC
Result of the draw:
DJOKOVIC IN FEDERER’s HALF; MURRAY IN NADAL’s HALF

ROLAND GARROS 2008
1st seed FEDERER
2nd seed NADAL
3rd seed DJOKOVIC
10th seed MURRAY
Result of the draw:
BOTH DJOKOVIC AND MURRAY IN NADAL’s HALF (let’s not forget here who is “king of clay”)
***Roland Garros statistics were not taken into consideration for the purposes of research but it does not mean that it is excluded from the possible fixing, and in fact it would make the statistics worse if we look at the substance of the draws.

Alex999
10-29-2011, 06:53 PM
Sorry I'm new, I did not know that it has been discussed already... but this is the first time I see the research done...there is a FB page about it. It is time someone did something concrete so that's why this is new, talking on forums is old, you're right, and it is not going to change anything...So seems like something concrete is finally taking place. Le Monde, Der Spiegel, Daily Mail...have already reported it. Interesting to see the developments. All the organisers and ITF and ATP are still not responding to any of the journalists' questions, and seems like they want to brush everything off as a coincidence... they think people are dumb and blind.
It's OK. Here is the original thread/discussion about this 'issue'.
http://www.menstennisforums.com/showthread.php?t=191678&highlight=rigged+draw

PureStorm
10-29-2011, 06:56 PM
I think that Roland Garros was taken out of the conference presentation for the sake of convenience and simplicity. In fact, seems like it is not out at all if you read the full study.

xdrewitdajx
10-29-2011, 06:56 PM
This is all from the same study and I cannot see a thing wrong with it!

INCREDIBLE RESULTS OF THE DRAWS:

AUSTRALIAN OPEN 2011
1st seed NADAL
2nd seed FEDERER
3rd seed DJOKOVIC
5th seed MURRAY
Result of the draw:
DJOKOVIC IN FEDERERís HALF; MURRAY IN NADALís HALF

WIMBLEDON 2011
1st seed NADAL
2nd seed DJOKOVIC
3rd seed FEDERER
4th seed MURRAY
Result of the draw:
DJOKOVIC IN FEDERERís HALF; MURRAY IN NADALís HALF

US OPEN 2011
1st seed DJOKOVIC
2nd seed NADAL
3rd seed FEDERER
4th seed MURRAY
Result of the draw:
DJOKOVIC IN FEDERERís HALF; MURRAY IN NADALís HALF

AUSTRALIAN OPEN 2010
1st seed FEDERER
2nd seed NADAL
3rd seed DJOKOVIC
5th seed MURRAY
Result of the draw:
DJOKOVIC IN FEDERERís HALF; MURRAY IN NADALís HALF

WIMBLEDON 2010
1st seed FEDERER
2nd seed NADAL
3rd seed DJOKOVIC
4th seed MURRAY
Result of the draw:
DJOKOVIC IN FEDERERís HALF; MURRAY IN NADALís HALF

US OPEN 2010
1st seed NADAL
2nd seed FEDERER
3rd seed DJOKOVIC
4th seed MURRAY
Result of the draw:
DJOKOVIC IN FEDERERís HALF; MURRAY IN NADALís HALF

AUSTRALIAN OPEN 2009
1st seed NADAL
2nd seed FEDERER
3rd seed DJOKOVIC
4th seed MURRAY
Result of the draw:
DJOKOVIC IN FEDERERís HALF; MURRAY IN NADALís HALF

WIMBLEDON 2009
1st seed NADAL (withdrew)
2nd seed FEDERER
3rd seed MURRAY
4th seed DJOKOVIC
Result of the draw:
DJOKOVIC IN FEDERERís HALF; MURRAY IN NADALís HALF (Nadal withdrew due to injury but this is irrelevant in this context)

US OPEN 2009
1st seed FEDERER
2nd seed MURRAY
3rd seed NADAL
4th seed DJOKOVIC
Result of the draw:
DJOKOVIC IN FEDERERís HALF; MURRAY IN NADALís HALF

AUSTRALIAN OPEN 2008
1st seed FEDERER
2nd seed NADAL
3rd seed DJOKOVIC
9th seed MURRAY
Result of the draw:
DJOKOVIC IN FEDERERís HALF; MURRAY IN NADALís HALF

WIMBLEDON 2008
1st seed FEDERER
2nd seed NADAL
3rd seed DJOKOVIC
12th seed MURRAY
Result of the draw:
DJOKOVIC IN FEDERERís HALF; MURRAY IN NADALís HALF

US OPEN 2008
1st seed NADAL
2nd seed FEDERER
3rd seed DJOKOVIC
6th seed MURRAY
Result of the draw:
DJOKOVIC IN FEDERERís HALF; MURRAY IN NADALís HALF

A CASE OF ROLAND GARROS (the only clay court Slam)

Results on ALL clay tournaments are as follows: Murray never played Federer on any clay court tournament. Nadal leads 4:0 against Murray on clay. Nadal was likely to win whomever he got on clay, Djokovic or Murray. Djokovic played 11 matches on clay with Nadal of which he lost 9 first matches. Federer vs. Djokovic on clay - 3:1. Federer was likely to win on clay against Djokovic.

ROLAND GARROS 2011
1st seed NADAL
2nd seed DJOKOVIC
3rd seed FEDERER
4th seed MURRAY
Result of the draw:
DJOKOVIC IN FEDERERís HALF; MURRAY IN NADALís HALF

ROLAND GARROS 2010
1st seed FEDERER
2nd seed NADAL
3rd seed DJOKOVIC
4th seed MURRAY
Result of the draw:
MURRAY IN FEDERERís HALF; DJOKOVIC IN NADALís HALF (the only draw that does not fit the formal pattern perfectly. But substance is different. Regardless of the draw Murray was likely never to make it to the semis on the this Slam, and he barely passed the first round match against Gasquet)

ROLAND GARROS 2009
1st seed NADAL
2nd seed FEDERER
3rd seed MURRAY
4th seed DJOKOVIC
Result of the draw:
DJOKOVIC IN FEDERERís HALF; MURRAY IN NADALís HALF

ROLAND GARROS 2008
1st seed FEDERER
2nd seed NADAL
3rd seed DJOKOVIC
10th seed MURRAY
Result of the draw:
BOTH DJOKOVIC AND MURRAY IN NADALís HALF (letís not forget here who is ďking of clayĒ)
***Roland Garros statistics were not taken into consideration for the purposes of research but it does not mean that it is excluded from the possible fixing, and in fact it would make the statistics worse if we look at the substance of the draws.

doesnt mean there isnt. first of all, draws aren't determined by name, but ranking/seed. Try analyzing in terms of #1/2 and #3/4. Also, RG has been excluded "for the purposes of research". Lmao. Why? Oh, right, because it doesn't fit in with the theory.

MuzzahLovah
10-29-2011, 06:57 PM
think about this a bit, though.

I have- have you? All that proves is that it isn't an innocent computerized draw error favoring 1vs3 semi instead of 1vs4, but makes purposeful rigging more likely. The argument is that the draws are rigged, not the seeding or the rankings- the same players play the same round regardless of the rankings/seedings, that makes the rigging more obvious.

MuzzahLovah
10-29-2011, 07:00 PM
doesnt mean there isnt. first of all, draws aren't determined by name, but ranking/seed. Try analyzing in terms of #1/2 and #3/4. Also, RG has been excluded "for the purposes of research". Lmao. Why? Oh, right, because it doesn't fit in with the theory.

The seeds/rankings are known before the draws are made... and there is no incentive in rigging based on seeds, but based on players.

PureStorm
10-29-2011, 07:01 PM
doesnt mean there isnt. first of all, draws aren't determined by name, but ranking/seed. Try analyzing in terms of #1/2 and #3/4. Also, RG has been excluded "for the purposes of research". Lmao. Why? Oh, right, because it doesn't fit in with the theory.

You apparently do not understand the tennis draws or statistics... There is an answer to this also on the same facebook page for the study. It goes like this:

"ANSWER TO READER OF "DER SPIEGEL"

Dear Ms. Pijetlovic,

Today i recieved a reader's letter, who criticised your study. The reader has looked at the drawings of the u.s. open, the australian open and wimbledon in the past four years and found out, that at these 12 tournaments, the number 1 of the drawing list was placed to the number 3 six times, as often as to the number 4. The reader says, this is a normal statistical outcome.

As i would like to answer the reader, what is your explanation. Why is manipulation still possible?
_____

Dear A.d.e..,

Your reader's argument would make sense only if there weren't any other factors attached to those numbers. But in tennis draws we have players names attached to each of the numbers, and that brings another, equally valid dimension into probability calculations with its own independent statistics. The names were changing the number to which they are attached.

Under the system of draws, every player seeded from 3 to 32 (and all unseeded players) has 1 in 2 ( or 50%) chance to be placed in top, and 1 in 2 (50%) chance to be placed in the bottom of the draw.

For example, if we have this seeding

1 Federer
2 Nadal
3 Djokovic
4 Murray

or

1 Nadal
2 Federer
3 Murray
4 Djokovic

or any other seeding that happened 2008-2011 (see the post below on the wall, or the table in the study document)

you have 1 in 2 (50%) chance that you will draw Djokovic/Federer and Murray/Nadal combination, and you have 1 in 2 (50%) to draw Federer/Murray and Djokovic/Nadal combination. At the same time, you have a 1 in 2 chance that numbers 1 and 3 will be matched and numbers 2 and 4 will be matched, and conversely you have 1 in 2 chance that it will be 1 and 4, and 2 and 3 combination.

If you obtain normal statistics under number-dimension, that does not affect the statistics under name-dimension of the draw. They are two different things.

Whatever the combination of the seeds, Djokovic was always drawn with Federer and Murray with Nadal, and the chances for that each time were 1 in 2 or 50%. This happened 12 out of 12 times. So it is (1x2) x (1x2) x (1x2) x (1x2) x (1x2) x (1x2) x (1x2) x (1x2) x (1x2) x (1x2) x (1x2) x (1x2) x (1x2) = 4096. 1 in 4096 times is not the final result because you will have noticed that Murray was not among the first four seeds and was draw separately 5 of those 12 times (in all three Grand Slam tournaments of 2008, and in Australian Open 2010 and 2011). Therefore, in each of those five separate draws he obtained a new 1 in 2 possibility to be drawn in Nadal's half. And each of the 5 times he was drawn in Nadal's half. It adds another (1x2) x (1x2)x (1x2) x (1x2) x (1x2) to the calculations. So it is 4096 x (1x2) x (1x2) x (1x2) x (1x2) x (1x2) = 131072. The result is therefore 1 in 131072.

So the name-value and the number-value of the draw are statistically separate one from another. In some hypothetical draw, you could go a step further and add another dimension to the calculations. Let's say you can have four different colours switching randomly between four players' numbers in each draw, in addition to players switching the numbers. In that case, matching two of the colours when performing draws will become a value of its own, statistically independent of the number-value and the name-value of the draw.

Let's assume that in 12 Grand Slam draws it was the numbers that were matched in 12 out of 12 draws: then players would not be matched because their seeded position was always changing. In that case, you could use the converse but equivalent argument as your reader to say: "numbers 1 and 3 were always matched in 12 out of 12 draws but that does not mean that there is any irregularity because we had a different pair of players 6 out of those times". The only important point in tennis draws is the name-value, not the number-value, and if the numbers were matched 12 out of 12 times, nobody would say anything because there is no substantive value or any point in matching the same numbers in tennis draws, as much as there would be no substantive value to matching the colours - it is the name-value of the draw that is important and it is the name statistics that will matter. That is why the fact that number-value of the draw produced statistically normal results does not matter at all. If anything, it just proves my point: when you are not fixing certain value of the draw it produces statistically normal results, which is what should have happened with the name-value as well.

If your reader still does not understand, I would suggest he/she does the following: there is the so-called Monte Carlo simulation that you can find on the internet. Insert a correct and complete information into the programme, let the programme perform draw simulations and see how many attempts it will take to produce only one result as obtained at the Grand Slams under investigation: I would bet all my money on approximately 131072 attempts :-)

Best regards
Katarina"


Sorry, but this is how the things are, get your statistic course straight - the draws are apparently fixed!

xdrewitdajx
10-29-2011, 07:39 PM
I have- have you? All that proves is that it isn't an innocent computerized draw error favoring 1vs3 semi instead of 1vs4, but makes purposeful rigging more likely. The argument is that the draws are rigged, not the seeding or the rankings- the same players play the same round regardless of the rankings/seedings, that makes the rigging more obvious.

or does it

xdrewitdajx
10-29-2011, 07:40 PM
You apparently do not understand the tennis draws or statistics... There is an answer to this also on the same facebook page for the study. It goes like this:

"ANSWER TO READER OF "DER SPIEGEL"

Dear Ms. Pijetlovic,

Today i recieved a reader's letter, who criticised your study. The reader has looked at the drawings of the u.s. open, the australian open and wimbledon in the past four years and found out, that at these 12 tournaments, the number 1 of the drawing list was placed to the number 3 six times, as often as to the number 4. The reader says, this is a normal statistical outcome.

As i would like to answer the reader, what is your explanation. Why is manipulation still possible?
_____

Dear A.d.e..,

Your reader's argument would make sense only if there weren't any other factors attached to those numbers. But in tennis draws we have players names attached to each of the numbers, and that brings another, equally valid dimension into probability calculations with its own independent statistics. The names were changing the number to which they are attached.

Under the system of draws, every player seeded from 3 to 32 (and all unseeded players) has 1 in 2 ( or 50%) chance to be placed in top, and 1 in 2 (50%) chance to be placed in the bottom of the draw.

For example, if we have this seeding

1 Federer
2 Nadal
3 Djokovic
4 Murray

or

1 Nadal
2 Federer
3 Murray
4 Djokovic

or any other seeding that happened 2008-2011 (see the post below on the wall, or the table in the study document)

you have 1 in 2 (50%) chance that you will draw Djokovic/Federer and Murray/Nadal combination, and you have 1 in 2 (50%) to draw Federer/Murray and Djokovic/Nadal combination. At the same time, you have a 1 in 2 chance that numbers 1 and 3 will be matched and numbers 2 and 4 will be matched, and conversely you have 1 in 2 chance that it will be 1 and 4, and 2 and 3 combination.

If you obtain normal statistics under number-dimension, that does not affect the statistics under name-dimension of the draw. They are two different things.

Whatever the combination of the seeds, Djokovic was always drawn with Federer and Murray with Nadal, and the chances for that each time were 1 in 2 or 50%. This happened 12 out of 12 times. So it is (1x2) x (1x2) x (1x2) x (1x2) x (1x2) x (1x2) x (1x2) x (1x2) x (1x2) x (1x2) x (1x2) x (1x2) x (1x2) = 4096. 1 in 4096 times is not the final result because you will have noticed that Murray was not among the first four seeds and was draw separately 5 of those 12 times (in all three Grand Slam tournaments of 2008, and in Australian Open 2010 and 2011). Therefore, in each of those five separate draws he obtained a new 1 in 2 possibility to be drawn in Nadal's half. And each of the 5 times he was drawn in Nadal's half. It adds another (1x2) x (1x2)x (1x2) x (1x2) x (1x2) to the calculations. So it is 4096 x (1x2) x (1x2) x (1x2) x (1x2) x (1x2) = 131072. The result is therefore 1 in 131072.

So the name-value and the number-value of the draw are statistically separate one from another. In some hypothetical draw, you could go a step further and add another dimension to the calculations. Let's say you can have four different colours switching randomly between four players' numbers in each draw, in addition to players switching the numbers. In that case, matching two of the colours when performing draws will become a value of its own, statistically independent of the number-value and the name-value of the draw.

Let's assume that in 12 Grand Slam draws it was the numbers that were matched in 12 out of 12 draws: then players would not be matched because their seeded position was always changing. In that case, you could use the converse but equivalent argument as your reader to say: "numbers 1 and 3 were always matched in 12 out of 12 draws but that does not mean that there is any irregularity because we had a different pair of players 6 out of those times". The only important point in tennis draws is the name-value, not the number-value, and if the numbers were matched 12 out of 12 times, nobody would say anything because there is no substantive value or any point in matching the same numbers in tennis draws, as much as there would be no substantive value to matching the colours - it is the name-value of the draw that is important and it is the name statistics that will matter. That is why the fact that number-value of the draw produced statistically normal results does not matter at all. If anything, it just proves my point: when you are not fixing certain value of the draw it produces statistically normal results, which is what should have happened with the name-value as well.

If your reader still does not understand, I would suggest he/she does the following: there is the so-called Monte Carlo simulation that you can find on the internet. Insert a correct and complete information into the programme, let the programme perform draw simulations and see how many attempts it will take to produce only one result as obtained at the Grand Slams under investigation: I would bet all my money on approximately 131072 attempts :-)

Best regards
Katarina"


Sorry, but this is how the things are, get your statistic course straight - the draws are apparently fixed!

you sound awfully confident for someone who's just quoting other people

xdrewitdajx
10-29-2011, 07:45 PM
I have- have you? All that proves is that it isn't an innocent computerized draw error favoring 1vs3 semi instead of 1vs4, but makes purposeful rigging more likely. The argument is that the draws are rigged, not the seeding or the rankings- the same players play the same round regardless of the rankings/seedings, that makes the rigging more obvious.

just to actually give a real response as well, you said:

"Fed and Djokovic have played in the same half almost every time in the slams, even though the draws are supposed to be 50-50 and their rankings have fluctuated."

when I said think about it, I meant, the draws are supposed to be 50-50, based on seeds, so you can't just dismiss the seed thing.
I understand what you're saying about the rigging being for names, not seeds.
But what you complained about here, the draws supposed to be 50-50 but not appearing so, you have to incorporate the "seed" as a factor, to make such a claim, not the name.

that probably wasnt as clear as i could make it, i cant now, its my fault. fuck it

PureStorm
10-29-2011, 08:03 PM
you sound awfully confident for someone who's just quoting other people

Any other comments, like, concerning the issue addressed, as opposed to personal remarks?

xdrewitdajx
10-30-2011, 01:29 AM
Any other comments, like, concerning the issue addressed, as opposed to personal remarks?

no, i think i'm good. feel free to quote it in later posts in case you get low on ammo.

abraxas21
10-30-2011, 01:43 AM
the ITF draws are as legit as this:

http://davecunning.files.wordpress.com/2010/10/free_candy_van.jpg