There is alot of discussion about how the preference point system works. THere are literally 100's of ways to design a points system. The premise is ( and you alligator hunters asked for it, I might add, although not all of you asked for it) that the more you apply and the less that you get drawn over time then the more likely you are to get drawn.
Jason Price, our IT Biologist, who produced our points scoring formula, explained to me me that the system is kinda like golf, the lower your score the better your chances. This being our first year, found a tremendous number of applicants with a very low score (actually a negative number ie, 0.75) however some people had a score as high as 198 because they had applied maybe 5 times and had been drawn twice in the same zone. By the way, you have a score for the Pearl and the Pascagoula if you have applied for both and neither score will effect the other. In other words if you got drawn in the past on the Pascagoula, but never on the Pearl, then your score would be higher for the Pasc and lower for the Pearl.
This year there were more than enough people with a negative digit score to fill all available slots. So, if you had a positive number score, you would not get drawn this year, but your score will drop each time you apply and do not get drawn in the future. Each time you get drawn your score will get significantly higher which will greatly decrease your chance the following year.
JASON DESCRIBES HIS SYSTEM VERY WELL HERE:
1. The weighting system is based entirely on your average win/loss percentage across each hunt on a specific area. It does not include attendance if you were drawn the previous year and did or did not show up for your hunt. It is weighted per WMA (or Pearl / Pascagoula) for gators so your win / loss percentage for Pascagoula does not affect your score for Pearl. The "algorithm" weights your applications based on number of years applied, total winning applications and total applications. It also gives a little extra weight to people who put in consistently every year. In other words first and second year applicants don't automatically get drawn just because they were not drawn before. For example if hunter A put in 5 years and has been drawn 1 previous year his score will be better than hunter B who has only put in 1 or 2 years but has never been drawn.
2. The drawing/weighting system is not what was being tested for alligators. It is the same weighting system that I developed for the WMA drawings last year so it has been tested extensively. However, this is the first year for the alligator draw to use the weighting system because it was not in place until after the gator draw last year. What Flynt and I were testing was the previous year?s history data to ensure it was correctly in the system and that correct id numbers were used to score all the previous year?s applications. I certainly could have created those histories before yesterday morning, but I was too busy playing Atari with TheStump to mess with them. [laugh]
3. I am not a computer programmer, I am a wildlife biologist that can hack at a computer with enough hammer to eventually get the thing to do what it is supposed to. I have been hacking at computers and developing database applications long enough now to understand a little about business rules though so the statement -
?but when you go to an input screen and it can't handle leading zeroes in a drivers license number or dashes in a SSN, I just go.. what tha ????? This is embarrassing that this is the best web interface my state can muster?
makes me chuckle. Sure we could make the SSN field a text field and allow every crazy number combination in the world with dashes, leading 0s, commas, colons, spaces, etc. That would certainly make initial development easier and data entry easier, but what do you think would happen if your key relational field didn?t have business rules to ensure data integrity? Would the text 123-45-6789 match in a SQL relate to 123456789? How much overhead do you think it requires to store that text based number versus the integer number? Those business rules are designed to stop ?garbage in? problems. I could write a java app that would clean up all the different ways people enter SSN/DL, but in my common sense ******* mind it is easier to just ask the applicant to enter a number. (On a side note the gator applications are done through POS and not in-house so I have no control over how they behave).
4. I?m sure someone trained in the computer sciences could make the drawing program more efficient, but they would also demand much more money that I do. Would you be willing to pay an additional $50,000 to speed up the drawing process from 15 minutes to 2 minutes? The biggest time consuming part is going through the applications and making sure there are no duplicates and that the data is clean. The most time consuming part of the actual computer drawing (in terms of processing time) is a) checking applicants for more than 3 WMAs, b) checking to make sure someone doesn't get drawn on the same day on different WMAs and c) creating the temp tables to store each applicants score. For the alligator drawing the program wouldn't have to perform a and b, but in my mind it isn't worth modifying the application to save a few minutes of processing time on just those 2 hunts.
I realize the drawing system is not perfect and never will make everyone happy, but I have tweaked and made changes to it every year to make it more fair, speed up the process and save money. Good luck to those drawn and hopefully I will see you out on the water again this year. I got some good pictures last year.
Jason Price, our IT Biologist, who produced our points scoring formula, explained to me me that the system is kinda like golf, the lower your score the better your chances. This being our first year, found a tremendous number of applicants with a very low score (actually a negative number ie, 0.75) however some people had a score as high as 198 because they had applied maybe 5 times and had been drawn twice in the same zone. By the way, you have a score for the Pearl and the Pascagoula if you have applied for both and neither score will effect the other. In other words if you got drawn in the past on the Pascagoula, but never on the Pearl, then your score would be higher for the Pasc and lower for the Pearl.
This year there were more than enough people with a negative digit score to fill all available slots. So, if you had a positive number score, you would not get drawn this year, but your score will drop each time you apply and do not get drawn in the future. Each time you get drawn your score will get significantly higher which will greatly decrease your chance the following year.
JASON DESCRIBES HIS SYSTEM VERY WELL HERE:
1. The weighting system is based entirely on your average win/loss percentage across each hunt on a specific area. It does not include attendance if you were drawn the previous year and did or did not show up for your hunt. It is weighted per WMA (or Pearl / Pascagoula) for gators so your win / loss percentage for Pascagoula does not affect your score for Pearl. The "algorithm" weights your applications based on number of years applied, total winning applications and total applications. It also gives a little extra weight to people who put in consistently every year. In other words first and second year applicants don't automatically get drawn just because they were not drawn before. For example if hunter A put in 5 years and has been drawn 1 previous year his score will be better than hunter B who has only put in 1 or 2 years but has never been drawn.
2. The drawing/weighting system is not what was being tested for alligators. It is the same weighting system that I developed for the WMA drawings last year so it has been tested extensively. However, this is the first year for the alligator draw to use the weighting system because it was not in place until after the gator draw last year. What Flynt and I were testing was the previous year?s history data to ensure it was correctly in the system and that correct id numbers were used to score all the previous year?s applications. I certainly could have created those histories before yesterday morning, but I was too busy playing Atari with TheStump to mess with them. [laugh]
3. I am not a computer programmer, I am a wildlife biologist that can hack at a computer with enough hammer to eventually get the thing to do what it is supposed to. I have been hacking at computers and developing database applications long enough now to understand a little about business rules though so the statement -
?but when you go to an input screen and it can't handle leading zeroes in a drivers license number or dashes in a SSN, I just go.. what tha ????? This is embarrassing that this is the best web interface my state can muster?
makes me chuckle. Sure we could make the SSN field a text field and allow every crazy number combination in the world with dashes, leading 0s, commas, colons, spaces, etc. That would certainly make initial development easier and data entry easier, but what do you think would happen if your key relational field didn?t have business rules to ensure data integrity? Would the text 123-45-6789 match in a SQL relate to 123456789? How much overhead do you think it requires to store that text based number versus the integer number? Those business rules are designed to stop ?garbage in? problems. I could write a java app that would clean up all the different ways people enter SSN/DL, but in my common sense ******* mind it is easier to just ask the applicant to enter a number. (On a side note the gator applications are done through POS and not in-house so I have no control over how they behave).
4. I?m sure someone trained in the computer sciences could make the drawing program more efficient, but they would also demand much more money that I do. Would you be willing to pay an additional $50,000 to speed up the drawing process from 15 minutes to 2 minutes? The biggest time consuming part is going through the applications and making sure there are no duplicates and that the data is clean. The most time consuming part of the actual computer drawing (in terms of processing time) is a) checking applicants for more than 3 WMAs, b) checking to make sure someone doesn't get drawn on the same day on different WMAs and c) creating the temp tables to store each applicants score. For the alligator drawing the program wouldn't have to perform a and b, but in my mind it isn't worth modifying the application to save a few minutes of processing time on just those 2 hunts.
I realize the drawing system is not perfect and never will make everyone happy, but I have tweaked and made changes to it every year to make it more fair, speed up the process and save money. Good luck to those drawn and hopefully I will see you out on the water again this year. I got some good pictures last year.