Hey, draw odds dorks…

[MTGunner]

For NR that truly want the experience to hunt bighorn sheep I recommend unit 500 and 501. Both OTC tags. Quota, 2 each for these units. But, you can purchase an OTC tag. Hire a guide, roll the dice. Someone has to score. I went along with a friend that purchased one of these tags. Went in a week early. Hike up, hiked down. Spent hours on a spotting scope. He did not take a sheep. Luck of the hunter. BUT, we got the experience. Let’s face it. There is little to slim chances to hunt sheep, goats and moose in the lower 48. Possible? Yes! Probable? One can only go and put in the effort.
I put in for a unit that is 45 minutes to my west. Getting a tag in this unit is extremely difficult. I know the country. Have seen many trophy animals. I have 20 bonus points that are squared as a resident. I have yet to draw a tag for sheep. I drew a mountain goat tag in 7 years of applying. Great hunt. Good goat. I believe it should be a once in a lifetime draw for each species. So, I will continue to apply until I reach 80 years of age. If I draw, fantastic. If not, so be it!
For those that ABSOLUTELY must hunt MSG I say look a Canada, Alaska, Russia and other countries along with the U.S. There are other places to satisfy your desires. This is NOT a dress rehearsal. You go this way but once.
I have it on good information that “There are NO ATM machines in HELL”! MTG

 

[PRO MO]

and the draw odds. which adds up to a total of 30.44%. The percentages should add up to 100%, right? because there's guaranteed to be one non resident tag?

 

[SAJ-99]

and the draw odds. which adds up to a total of 30.44%. The percentages should add up to 100%, right? because there's guaranteed to be one non resident tag?
View attachment 320498

That is about what I come up with when I do it myself. I think the total doesn't add to 100% because there are multiple applicants at the same point level. What the applicant cares about is their chance of drawing with X pts not the chance of someone with Xpts drawing. It can get skewed from year to year because of the number of applicants at a specific point level. I hope I understand it correctly.

 

[rmyoung1]

and the draw odds. which adds up to a total of 30.44%. The percentages should add up to 100%, right? because there's guaranteed to be one non resident tag?
View attachment 320498

Is this still for goat unit 329? If so, this proves that my suspicion is valid. Let’s take my point level as an example. With 19 points in 2023, GoHunt says “2.5%” And toprut (see post 19) says “17%”

That’s a bit of a difference. Unfortunately, I believe GoHunt

 

[SAJ-99]

Is this still for goat unit 329? If so, this proves that my suspicion is valid. Let’s take my point level as an example. With 19 points in 2023, GoHunt says “2.5%” And toprut (see post 19) says “17%”

That’s a bit of a difference. Unfortunately, I believe GoHunt

Gohunt is correct. I don't know what the hell is going on in the Toprut algorithm.

 

[Stone_Ice_1]

You guys are nuts paying money for this, makes me think I should start a business creating these for people. Takes about 5 minutes to put together. I just have a simple excel spreadsheet that calculates all my draw odds for every species in Montana. Its just a simple calculation based off of the people that apply last year. I could even create a template where you just copy the data from the district you are interested in and it will give you your percent. I do exactly that every year, copy the draw results into the spreadsheet every year and automatically know the chance I will draw. Granted my job is a data analyst so I do this for work.

For example, this year I have a 12.3% of drawing one of out the big 3. My odds of drawing a sheep tag are a whopping .78%.

 

[SAJ-99]

You guys are nuts paying money for this, makes me think I should start a business creating these for people. Takes about 5 minutes to put together. I just have a simple excel spreadsheet that calculates all my draw odds for every species in Montana. Its just a simple calculation based off of the people that apply last year. I could even create a template where you just copy the data from the district you are interested in and it will give you your percent. I do exactly that every year, copy the draw results into the spreadsheet every year and automatically know the chance I will draw. Granted my job is a data analyst so I do this for work.

For example, this year I have a 12.3% of drawing one of out the big 3. My odds of drawing a sheep tag are a whopping .78%.

The good news is that Toprut is free with OnX. So I guess I get what I pay for. LOL. My only real complaint is that a lot of the stuff from agencies comes in Pdf and the copy/paste into excel can be a PITA.

 

[rmyoung1]

You guys are nuts paying money for this, makes me think I should start a business creating these for people. Takes about 5 minutes to put together. I just have a simple excel spreadsheet that calculates all my draw odds for every species in Montana. Its just a simple calculation based off of the people that apply last year. I could even create a template where you just copy the data from the district you are interested in and it will give you your percent. I do exactly that every year, copy the draw results into the spreadsheet every year and automatically know the chance I will draw. Granted my job is a data analyst so I do this for work.

For example, this year I have a 12.3% of drawing one of out the big 3. My odds of drawing a sheep tag are a whopping .78%.

I’m sure you’re good at your job, but hardly anything about calculating NR draw odds across the American West is as simple as you just suggested. Even the no-point states have nuances that muddy the waters (see NM 1st/2nd/3rd choice). If it were easy as dumping last year’s data into Excel, these app services wouldn’t be creating computer models to simulate the draws hundreds of thousands of times.

 

[Smtn10pt]

and the draw odds. which adds up to a total of 30.44%. The percentages should add up to 100%, right? because there's guaranteed to be one non resident tag?
View attachment 320498

i thought it was up to 10% COULD go to NR, but none HAVE to go to NR

 

[Stone_Ice_1]

The good news is that Toprut is free with OnX. So I guess I get what I pay for. LOL. My only real complaint is that a lot of the stuff from agencies comes in Pdf and the copy/paste into excel can be a PITA.

Yes, this is true, PDF's are not user friendly at all.

I’m sure you’re good at your job, but hardly anything about calculating NR draw odds across the American West is as simple as you just suggested. Even the no-point states have nuances that muddy the waters (see NM 1st/2nd/3rd choice). If it were easy as dumping last year’s data into Excel, these app services wouldn’t be creating computer models to simulate the draws hundreds of thousands of times.

I only do it for Montana, as that is all I apply for so I don't have to worry about the various rules in other states. I am not familiar with the methods these companies use but I would be shocked if they use computer models and use anything much more complicated than my method. There would be no benefit to using computer models here with the randomness, it's just a probability based on the number of people. The only good a computer model would be for here is to model the number of people applying. I build models like you mention for my job, my background is in statistics, or more specifically I have a masters in Ecological and Environmental Statistics. A "model" can be anything. For example I can say my method of removing the successful participants from the pool from last year, adding one point to all the unsuccessful applicants and going from there is a "model".

 

[BuckRut]

Yes, this is true, PDF's are not user friendly at all.

I only do it for Montana, as that is all I apply for so I don't have to worry about the various rules in other states. I am not familiar with the methods these companies use but I would be shocked if they use computer models and use anything much more complicated than my method. There would be no benefit to using computer models here with the randomness, it's just a probability based on the number of people. The only good a computer model would be for here is to model the number of people applying. I build models like you mention for my job, my background is in statistics, or more specifically I have a masters in Ecological and Environmental Statistics. A "model" can be anything. For example I can say my method of removing the successful participants from the pool from last year, adding one point to all the unsuccessful applicants and going from there is a "model".

How are you accounting for the point at which all the NR get thrown out of a specific draw and suddenly your odds as a resident are higher for the remaining draws than they were for the previous draws?

 

[jrabq]

The toprut figures changed significantly from 2022 to 2023, more than I would suspect from just another year of draw system maturation.

Yes, something is screwed up with the 2023 data. Rumor is the guy who created Toprut sold everything to OnX last year and it's likely something is not being implemented correctly since he left.

 

[Stone_Ice_1]

How are you accounting for the point at which all the NR get thrown out of a specific draw and suddenly your odds as a resident are higher for the remaining draws than they were for the previous draws?

Well that's easy, I don't count NR's. They are a subhuman population and are disregarded.

Seriously though, I kid about the NR's. There is only one draw for each animal and HD. So, there isn't something where one draw affects another. I think what you are referring to is how do I handle the 10% cap on NR's successful applicants. The answer is I don't do anything special for it. The change that makes in the probabilities is very small. Many times that cap isn't even reached and at times it is, it is after most successful applicants have already been selected. Remember this is all an estimate, there is no way to know exactly how many people are going to apply in a district. Does a difference in the probability of 4.3% to 4.4% really change where you apply, or in the case of sheep or something hard to draw like that a difference from .77% to .81%? Anyway, if you are consistent in your method across each district you are still comparing apples to apples (although some districts obviously have more NR apps).

 

[rmyoung1]

. I think what you are referring to is how do I handle the 10% cap on NR's successful applicants. The answer is I don't do anything special for it. The change that makes in the probabilities is very small.

I believe this assumption is false, which is why this gets interesting. Last year in MT R3 for goat in the districts where an NR was eligible, about 3400 NRs applied and about 9000 Rs applied, which makes NRs about 27% of the applicant pool. So if about 27% of the pool is NRs but they are limited to 10% of the region’s tags, that 10% cap absolutely will affect the odds of a NR pulling a tag. After all, isn’t that what it is supposed to do in the first place?

 

[SAJ-99]

I believe this assumption is false, which is why this gets interesting. Last year in MT R3 for goat in the districts where an NR was eligible, about 3400 NRs applied and about 9000 Rs applied, which makes NRs about 27% of the applicant pool. So if about 27% of the pool is NRs but they are limited to 10% of the region’s tags, that 10% cap will affect the odds of a NR pulling a tag.

The data comes in R and NR split, so calcing it isn't hard. The only issue for big3 is that some years a unit may have a NR tag and the next year it won't. They list those units that are eligible for a tag in the MSG book. The actual tags are not allocated to the draw pool until later. That is where a service might be useful. To the best of my knowledge, no one tries to make predictive odds. They are only doing simple stuff on data from last year. Apparently even Toprut can't get the simple stuff right.

 

[BuckRut]

Well that's easy, I don't count NR's. They are a subhuman population and are disregarded.

Seriously though, I kid about the NR's. There is only one draw for each animal and HD. So, there isn't something where one draw affects another. I think what you are referring to is how do I handle the 10% cap on NR's successful applicants. The answer is I don't do anything special for it. The change that makes in the probabilities is very small. Many times that cap isn't even reached and at times it is, it is after most successful applicants have already been selected. Remember this is all an estimate, there is no way to know exactly how many people are going to apply in a district. Does a difference in the probability of 4.3% to 4.4% really change where you apply, or in the case of sheep or something hard to draw like that a difference from .77% to .81%? Anyway, if you are consistent in your method across each district you are still comparing apples to apples (although some districts obviously have more NR apps).

The data comes in R and NR split, so calcing it isn't hard. The only issue for big3 is that some years a unit may have a NR tag and the next year it won't. They list those units that are eligible for a tag in the MSG book. The actual tags are not allocated to the draw pool until later. That is where a service might be useful. To the best of my knowledge, no one tries to make predictive odds. They are only doing simple stuff on data from last year. Apparently even Toprut can't get the simple stuff right.

Both of these are off way worse than the draw odds on Toprut. For a lot of draws once a NR draws the number of applicants left for the remaining (what could be 90% of the remaining tags) can be cut more than if half.

Take 622-02 for example. Total tickets in the hat to start the draw was 530,913 to start the draw. Say the first name out of the hat is a NR. Suddenly the R are now in a draw with 322,728 tickets for the remaining 9 tags.

 

[rmyoung1]

Thanks for the screenshot @PRO MO Sorta confirms my suspicions

 

[SAJ-99]

Both of these are off way worse than the draw odds on Toprut. For a lot of draws once a NR draws the number of applicants left for the remaining (what could be 90% of the remaining tags) can be cut more than if half.

Take 622-02 for example. Total tickets in the hat to start the draw was 530,913 to start the draw. Say the first name out of the hat is a NR. Suddenly the R are now in a draw with 322,728 tickets for the remaining 9 tags.

It is my understanding that the R and NR draws are separate. I may be misinformed. There is no reason to put the applicants together because they have different constraints. The NRs are only assigned 1 tag in your example and don't affect the R draw at all. Like I said, I may misunderstand how it works, so don't quote me. MT does some of the dumbest things so it wouldn't surprise me.

 

[Stone_Ice_1]

Both of these are off way worse than the draw odds on Toprut. For a lot of draws once a NR draws the number of applicants left for the remaining (what could be 90% of the remaining tags) can be cut more than if half.

Take 622-02 for example. Total tickets in the hat to start the draw was 530,913 to start the draw. Say the first name out of the hat is a NR. Suddenly the R are now in a draw with 322,728 tickets for the remaining 9 tags.

This is making it more complicated than it needs to be. Your odds of drawing are whatever they are at the beginning of the draw. You don't care what the odds change to after 1 ,2 or 100 people are selected. Think of it this way, you don't get to choose after they draw the 1st person if you want to keep going or not. Your overall odds of drawing don't change. Research the Monte Carlo problem (google it, lots of stuff out there) if you are insterested in researching this more. Or a more fun way might be to watch the movie "21", it has a segment on the Monte Carlo problem (probably find it on youtube if you wanted). Hint, always change the door in the Monte Carlo problem.

It is my understanding that the R and NR draws are separate. I may be misinformed. There is no reason to put the applicants together because they have different constraints. The NRs are only assigned 1 tag in your example and don't affect the R draw at all. Like I said, I may misunderstand how it works, so don't quote me. MT does some of the dumbest things so it wouldn't surprise me.

NR and R can't be done separate. They take UP to 10% of the tags, they don't allocate exactly 10% of the tags. So, they can't say in a district with 10 tags that 1 is for NR and set that aside and do a separate draw for that one tag. They draw everybody together and then once the non-residents achieve their 10% of the tags, they are removed. This is part of what BuckRut is showing above. However you can't account for when the non-residents are removed (if it occurs, many districts it doesn't) so your overall odds of drawing is what it was at the beginning of the draw.

Mostly this is making it way more difficult than it needs to be. As applicants we don't care about precise percentages, mostly we care which district gives us the best chance of drawing and this is much easier to figure out and doesn't need to be as precise.

 

[BuckRut]

This is making it more complicated than it needs to be. Your odds of drawing are whatever they are at the beginning of the draw.

Not arguing this part any farther but the above statements are false. You are over simplifying the complexity of the system and downplaying the task entities like GoHunt and TopRut have to deal with to make it simple enough to present to the public.

In the example I had.

If the first number drawn is a NR then you as a resident had approx. 1.5751% chance to draw assuming we don't try to analyze how many points each person had as they were selected.

If no NR are drawn you had approx. 0.9965% chance with the same assumption.

Numbers are so small it seems meaningless but you as a resident have a 58% greater chance of getting a tag if the first number picked is a NR