|
Post by Richie3Jack on Nov 14, 2014 10:43:06 GMT -5
And I showed you really clearly why guys who got really close avg can make totally different earnings and that got nothing to do with adjustment avg. No, you did not. Having a difference of 0.27 strokes on the PGA Tour is not close when it comes to a measurement. It's like saying that 'how can somebody be closer if they are 100 inches away than if they are 1 mile away?' You are completely disregarding how the measurement is taken. And you have completely ignored what I showed...yes...it is very possible for there to be a $700k difference in earnings when there is 'only' a 0.27 difference in scoring average. I showed that the projected earnings on the PGA Tour show it to be roughly $550k off of 25 events. And that was based off of Adjusted Scoring Average because that gives a more accurate depiction of the correlation between stroke average and earnings per event. If I were to use non-adjusted Scoring Average, the potential to be gained from lowering the scoring average by 0.27 strokes would be *more*. The correlation in non-adjusted scoring average dips to -0.65 (from -0.78) because non-adjusted scoring average does not depict what the player is shooting based on the level of difficulty of the courses they are playing. You won't answer that and keep claiming that you are, so I feel no need to allow you to continue in this thread. You do realize there is something called standard deviation and you can use that with the mean to figure it out? www.mathsisfun.com/data/standard-deviation-formulas.htmlThe issue is that is not how the best players on Tour make their money. And it's not how the players that have the best seasons of their career making money. The best players on Tour have better averages and score consistently lower. That's why there is a strong correlation between Purse Size per Event and non-adjusted Scoring AVERAGE and an even stronger correlate between earnings per event and Adjusted Scoring AVERAGE. This means that there is a strong indirect 1:1 relationship. Which means that even if somebody reduces their Adjusted Scoring Average by .01 strokes per round, they are going to likely increase their earnings per event. That is a strong mathematical and undeniable correlation and I didn't even factor in the size of the purse in the events each player played in. Which would cancel out those players that earned more earnings per event despite having a worse Adjusted Scoring Average simply because they had a favorable schedule. State your educational background in statistics then. When you tell me that the average is worthless, don't understand standard deviation and come up with cockamamie reasoning that can't explain me showing that 0.27 strokes can make a difference of $700k in a season, your credibility in statistics looks dreadful. When the current exchange rates are equated from Euros/Pounds/Australian Dollars/Yen, the average purse size of each event on the European Tour came to $3,679,607.14 US Dollars 1 standard deviation in purse size per event was a whopping $2,690,348 US DollarsThat means the from 1 standard deviation, the vast majority of the purse sizes on the European Tour were vastly wide ranging, from roughly $989K to $6.4 million US DollarsOn the PGA Tour, the average purse size last season was $6,714,286. 1 standard deviation was only $1,583,880. (purses mostly ranged from $5.1 million to $8.3 million) The standard deviation for the PGA Tour was only 24% of the mean purse size. Conversely, 1 standard deviation on the European Tour was 73% of the mean purse size It is critical to adjust the scoring average to get a more accurate depiction. If the data shows that it iss possible to show that 0.27 strokes can equate to $550k on the PGA Tour which has a much smaller deviation in purse size per event, then on the European Tour where the purse sizes can be very high or very low, it is absolutely possible to see a 700k increase in earnings based off of 0.27 difference in scoring average Guys like Lee Westwood have a considerable advantage since they can play in purses that can easily be 6 times higher than the amount that many of the unproven European Tour players are involved with. And he can play worse on average, even with an adjusted score, and still make out better in earnings because he is playing for much more money. He can also play worse on average and do it extremely consistently and earn more money than young, unproven players that are very inconsistent and get hot because he is playing for much larger money. 3JACK
|
|
|
Post by teeace on Nov 14, 2014 12:02:35 GMT -5
State your educational background in statistics then. 3JACK Its in economical math and statistics is big part of that. So you have claimed I don't have education for that and even when you no zero about my education and background. And for sure you are right. If player plays better, he earns more, but still avg can change differently and that's my point. It's same in economy that you have to look graphs to understand what changes the average. Is it upper quartile, median, or lower.. or something else. Like said, I made my own statical analyze work tools at middle of 90's when I found what all those were missing and to really understand better how scoring and results are gained. I found it very important to understand players good or great days instead of average. Finding those factors when he/she could go low, even if it happens only few times of year. Like I showed, 1.107.000 in 4 tournaments, 1.293 in five with total of 24 tournaments and high numbers with MC. Finally those 4/5 tournaments were all that really matter. Check those quartiles from few players and I'm sure you find the same relation and it's stronger than average. You got tools for that, so it doesn't take a lot to check it.
|
|
|
Post by Richie3Jack on Nov 14, 2014 12:42:32 GMT -5
And you still have yet to prove anything I have written and showed the math on as being wrong. And for somebody that supposedly knows 'Economic Math', your lack of knowledge with regards to understanding the importance of the mean, using regression analysis, ANOVA, standard deviation makes me question your credibility on statistics.
I've shown how 0.27 strokes can equate to $700k difference in earnings. I've shown how the European Tour's variance in event purse sizes can mean that 0.27 strokes can equate to even more than $700k difference in earnings. And you have never proven how Adjusted Scoring Average is a 'smoke curtain' while you seem to think that using non-Adjusted scoring average is okey-dokey. I've shown that the best players play better on average than the other players and the players that have their best seasons normally play better on average.
If the typical Tour player is making 80% of their money in 20% of their events, then they will make more money if their average earnings per event is higher. It is basic, simple math.
3JACK
|
|
|
Post by teeace on Nov 14, 2014 12:55:13 GMT -5
I've shown how 0.27 strokes can equate to $700k difference in earnings. If the typical Tour player is making 80% of their money in 20% of their events, then they will make more money if their average earnings per event is higher. It is basic, simple math. 3JACK Of course it can, but it's not sure. Sometimes they can make more money with worsening 0,27 even with adjusted scoring average. That's the whole point that in average also worst rounds are included even if you make zero money then... and there can be lot of them. So scoring avg doesn't prove that as we can see when mirroring stroke avg and earnings table. They just don't fit together.
|
|
|
Post by teeace on Nov 14, 2014 12:59:17 GMT -5
And you still have yet to prove anything I have written and showed the math on as being wrong. And for somebody that supposedly knows 'Economic Math', your lack of knowledge with regards to understanding the importance of the mean, using regression analysis, ANOVA, standard deviation makes me question your credibility on statistics. 3JACK Once again you talk about something you don't really know. We haven't have any discussion about those. Those are really basic things that doesn't change the truth that there can be correlation between those two or not. Check those quartiles Richie and you see what I mean. It's already in your 80/20 model pretty much. Those bad rounds makes scoring avg much higher than their effect to the earnings. Drop at least lowest quartile away and you get closer.
|
|
|
Post by Richie3Jack on Nov 14, 2014 15:06:39 GMT -5
You told me that it couldn't. Now, you're saying of course it canWhich is it? That's what the correlation coefficient is there to show. The strength of the relationship. If it was between 0 to 0.5 then we would have a much weaker relationship between scoring average (regardless if it is adjusted or not) and earnings per event. Instead, the correlation coefficient is a mathematical fact that shows a very strong 1:1 relationship between adjusted scoring average and earnings per event. Essentially, you are at best using anecdotal evidence Again, the correlation coefficient proves that at best this is anecdotal. And usually when you see it there is a large difference between the average purse size per the event. In the case of the European Tour, the average purse size per event can be staggeringly different from player to player This shouldn't be difficult to figure out as the average obvious accounts for the great rounds as well. You can simply use the median instead, which comes close to the mean in most cases and helps negate the effect of the extremes of great rounds versus awful rounds . Either way, the deviation of Tour players in scoring rounds is not that great and the MATH shows the strength of the relationship between Adjusted Scoring Average and Earnings per Event. You are arguing against a mathematical fact. And if you still feel the need to argue against mathematical facts where I have shown the math...I'm afraid I'm going to have to cut you loose from this forum. It is becoming too apparent that you just want to argue with me for the sake of arguing and not admitting that you're wrong. 3JACK
|
|
|
Post by teeace on Nov 15, 2014 2:57:50 GMT -5
You told me that it couldn't. Now, you're saying of course it canWhich is it? So you as a stats guy don't understand that it can is not a prove of anything. Its coincidence. Also it can be true in some player profiles, and total lie with some others. That's what the correlation coefficient is there to show. The strength of the relationship. If it was between 0 to 0.5 then we would have a much weaker relationship between scoring average (regardless if it is adjusted or not) and earnings per event. Instead, the correlation coefficient is a mathematical fact that shows a very strong 1:1 relationship between adjusted scoring average and earnings per event. So why that scoring avg and money list is not even close to 1:1? to even if divided to played tournaments?Essentially, you are at best using anecdotal evidence Again, the correlation coefficient proves that at best this is anecdotal. And usually when you see it there is a large difference between the average purse size per the event. In the case of the European Tour, the average purse size per event can be staggeringly different from player to player This shouldn't be difficult to figure out as the average obvious accounts for the great rounds as well. You can simply use the median instead, which comes close to the mean in most cases and helps negate the effect of the extremes of great rounds versus awful rounds . Either way, the deviation of Tour players in scoring rounds is not that great and the MATH shows the strength of the relationship between Adjusted Scoring Average and Earnings per Event. You are arguing against a mathematical fact. And if you still feel the need to argue against mathematical facts where I have shown the math...I'm afraid I'm going to have to cut you loose from this forum. It is becoming too apparent that you just want to argue with me for the sake of arguing and not admitting that you're wrong.3JACK [/quote] It's you Richie who argue and I try to tell you just what I found already 20 years ago.But your ears are def and your eyes are blind because you are angry that someone you didn't know got education for that area also and tells you what is wrong in some stats basics... not your work actually, but also statistical thinking when having linear and non linear factors included. I have to say I wonder really your comments like this " This shouldn't be difficult to figure out as the average obvious accounts for the great rounds as well What a heck, are you really saying that? For sure they are counted there, but their effect to the money is extremely different. That's why it should be observed at better half, top quartile and top 10% of rounds to get better view. Or even by graph with score levels frequency. I give you one more example: 100 rounds and avg of 50 best is 68. Lowest 30 are 74 and those missing 20 are 73. His AVG is 70,6. Then he gets "better" by changing those 74's to 72,5 and those missing ones to 72. Great improvement and new average is 70,1. Then there comes a stat guy and wonders why this guy earned only 1000 more that year even his AVG get so much better. When someone explains it and tells that 1000 came because he made one more cut and was T58 instead of MC in one tournament. So the stat guy gets angry because someone is using logic against cold numbers of stats The whole thing is that two or three shots in right place and right tournament can make so huge difference, that one can totally forget all other tournaments.
|
|
|
Post by teeace on Nov 15, 2014 2:58:50 GMT -5
You told me that it couldn't. Now, you're saying of course it canWhich is it? So you as a stats guy don't understand that it can is not a prove of anything. Its coincidence. Also it can be true in some player profiles, and total lie with some others. That's what the correlation coefficient is there to show. The strength of the relationship. If it was between 0 to 0.5 then we would have a much weaker relationship between scoring average (regardless if it is adjusted or not) and earnings per event. Instead, the correlation coefficient is a mathematical fact that shows a very strong 1:1 relationship between adjusted scoring average and earnings per event. So why that scoring avg and money list is not even close to 1:1? to even if divided to played tournaments?Essentially, you are at best using anecdotal evidence Again, the correlation coefficient proves that at best this is anecdotal. And usually when you see it there is a large difference between the average purse size per the event. In the case of the European Tour, the average purse size per event can be staggeringly different from player to player This shouldn't be difficult to figure out as the average obvious accounts for the great rounds as well. You can simply use the median instead, which comes close to the mean in most cases and helps negate the effect of the extremes of great rounds versus awful rounds . Either way, the deviation of Tour players in scoring rounds is not that great and the MATH shows the strength of the relationship between Adjusted Scoring Average and Earnings per Event. You are arguing against a mathematical fact. And if you still feel the need to argue against mathematical facts where I have shown the math...I'm afraid I'm going to have to cut you loose from this forum. It is becoming too apparent that you just want to argue with me for the sake of arguing and not admitting that you're wrong.3JACK [/quote] It's you Richie who argue and I try to tell you just what I found already 20 years ago.But your ears are def and your eyes are blind because you are angry that someone you didn't know got education for that area also and tells you what is wrong in some stats basics... not your work actually, but also statistical thinking when having linear and non linear factors included. I have to say I wonder really your comments like this " This shouldn't be difficult to figure out as the average obvious accounts for the great rounds as well What a heck, are you really saying that? For sure they are counted there, but their effect to the money is extremely different. That's why it should be observed at better half, top quartile and top 10% of rounds to get better view. Or even by graph with score levels frequency. I give you one more example: 100 rounds and avg of 50 best is 68. Lowest 30 are 74 and those missing 20 are 73. His AVG is 70,6. Then he gets "better" by changing those 74's to 72,5 and those missing ones to 72. Great improvement and new average is 70,1. Then there comes a stat guy and wonders why this guy earned only 1000 more that year even his AVG get so much better. When someone explains it and tells that 1000 came because he made one more cut and was T58 instead of MC in one tournament. So the stat guy gets angry because someone is using logic against cold numbers of stats The whole thing is that two or three shots in right place and right tournament can make so huge difference, that one can totally forget all other tournaments.
|
|
|
Post by Richie3Jack on Nov 16, 2014 10:11:21 GMT -5
Tapio, I'm done with this because you don't understand statistics and you keep ignoring the facts and how they work. I would recommend reading this so you can understand the correlation coefficient and how it works to determine the mathematical relationship between two variables (in this case: Earnings per Event and Adjusted Scoring Average). en.wikipedia.org/wiki/Pearson_product-moment_correlation_coefficientWhen the correlation coefficient between these 2 variables is at -0.8 (a very strong indirect correlation), it proves that your scenarios are completely anecdotal and not very likely to happen. Let's say we see a correlation between temperature and the # of people that come to a community pool of +0.8. Of course, there could always be those instances where people come out to the pool on a 50-degree (Fahrenheit) day than the days of 75 degrees. But, if we have a large enough sample size and come up with the correlation of +0.8, that tells us that *not* every single day that the temperature is warmer that more people would come to the community pool, but by and large the vast majority of the time the warmer the temperature, even if it is 1 degree warmer, the more likely people are coming to the community pool. I showed the math and showed the correlation coefficient of -0.8 (lower Adjusted Scoring Average higher the earnings per event). And that *mathematically* shows that in all likelihood your examples *could* happen, but the overwhelming majority of the time they do not happen. Better players earn more money and better players usually have better Adjusted Scoring Averages. And the deviation in their scores does *not* change enough from player-to-player as you think they do. If they did as much as you think they do, the correlation coefficient would be much lower and closer to 0. Linear regression (correlation coefficient) is basically Statistics 101. The fact that you I've explained it to you, showed you the math and given you the linear regression formula and you still don't get it can only be interpreted as you have a dearth of statistical knowledge despite your claims and that everything you claim when it comes to statistics should be considered suspect. 3JACK
|
|