Onx draw odds?

nontyp

Well-known member
Joined
Aug 6, 2016
Messages
615
Location
Kansas
Just checked out the Onx draw odds calculator. Not sure what to think. I hope they aren’t correct because the NM odds are way lower than gohunt. It says the calculation is the actual outcome of the 2018 draw. I’d like to know exactly how the calculations are done.
 
  • Like
Reactions: LCH
Wow. You're right. I compared about a dozen on goHunt and Toprut and they were all within a percent of each other. The odds on OnX are all around 40% lower. Shite.
 
It appears OnX is using something closer to what the raw data would suggest. As I understand it, GoHunt runs a draw simulation that gives you the estimated odds. This process results in an "inflated" draw odd of sorts. I've seen some hunts on GoHunt that have a 4x higher draw odd than the raw data from NMGF would suggest, but this is due to their simulation. So by comparison, for the exact same hunt GoHunt shows a 16% odd, OnX shows a 7.1% and if I divide the number of resident tags by the total resident applications I come up with 6.5% on my own.

At the end of the day, with NM's lottery system it is almost impossible to get an accurate draw odd. I find the data is best used to sort out low odd, medium odd, and high odd hunts.
 
I wonder if their numbers use number of applicants vs number of applications? That would lead to lower percentage. Just a thought.
 
Both Gohunt and Toprut buy the applicant databases then run the draw thousands of times. Their odds are the real odds from that data.

OnX is simply dividing success and applicants for their odds.
 
Last edited:
In NM if you divide total tags by total applicants (all 3 choices) you will systematically underestimate your odds, typically by about 1.5 to 2X. It will give you a "worst case" estimate, but it will almost always be significantly lower than the Toprut or Gohunt values, which will be more accurate. The reason you shouldn't include all of the 2nd and 3rd choice applicants is because some of those folks will have drawn their first choice and are no longer in the game. Also, a considerable number of folks still order their choices incorrectly, so their 2nd and 3rd choices are never relevant either.
 
In NM if you divide total tags by total applicants (all 3 choices) you will systematically underestimate your odds, typically by about 1.5 to 2X. It will give you a "worst case" estimate, but it will almost always be significantly lower than the Toprut or Gohunt values, which will be more accurate. The reason you shouldn't include all of the 2nd and 3rd choice applicants is because some of those folks will have drawn their first choice and are no longer in the game. Also, a considerable number of folks still order their choices incorrectly, so their 2nd and 3rd choices are never relevant either.
Doesn’t the excel spreadsheet from the New Mexico sight give a break down of how many applied and drew for 1st 2nd and 3rd choice? That’s how I figured my odds for all three choices. Not the overall but from the choice columns

Edit: I think I read your post to quick.
 
Yeah, OnX, at a glance, appears to be the simple odds of tags/apps. The GoHunt odds are essentially an excellent predictor of LAST YEAR. All odds calculations are based the previous year, but I think that when you start simulating odds based on the choice order of individual applicants in the previous year that you get more fluctuations from year to year even with similar applicant totals. That’s just a guess. Another guess would be that reality is between GoHunt and OnX, but probably closer to GoHunt.

People frequently talk about how much better odds are due to irrationally ordered choices. That’s true if you talking about your odds of drawing a low odds hunt when people put a high odds hunt as their first choice and a low odds hunt as their third, but when considering third choices it goes the opposite direction. The guys drawn early in the draw that should have drawn a glory tag, put a high odds hunt as their first choice and now your odds are lower than you thought due to irrationally ordered choices.
 
Yeah, OnX, at a glance, appears to be the simple odds of tags/apps. The GoHunt odds are essentially an excellent predictor of LAST YEAR. All odds calculations are based the previous year, but I think that when you start simulating odds based on the choice order of individual applicants in the previous year that you get more fluctuations from year to year even with similar applicant totals. That’s just a guess. Another guess would be that reality is between GoHunt and OnX, but probably closer to GoHunt.

People frequently talk about how much better odds are due to irrationally ordered choices. That’s true if you talking about your odds of drawing a low odds hunt when people put a high odds hunt as their first choice and a low odds hunt as their third, but when considering third choices it goes the opposite direction. The guys drawn early in the draw that should have drawn a glory tag, put a high odds hunt as their first choice and now your odds are lower than you thought due to irrationally ordered choices.
Agreed the NM odds calculations are much more complicated than they appear. “incorrectly” ordering your choices doesn’t automatically turn your 2nd and 3rd choices to 0% because supposedly 50% of apps are ordered incorrectly. It all depends on the random draw order. Regardless of who has correct odds, I do think Onx and Gohunt will be entering each other’s market space very soon.
 
while I can see the odds on the mobile version of OnX I don't see it on the website version.....what am I missing?
 
Found it thanks. It’s pretty basic, more along the lines tags divided into applications....not very useful or accurate. I guess it better then nothing and it is free...
 
I was hoping that they would have the odds for the MT deer/big game combo license but they only have the permits. Hopefully they continue to expand it on the future or partner with GoHunt
 
MTNTOUGH - Use promo code RANDY for 30 days free

Latest posts

Forum statistics

Threads
110,805
Messages
1,935,060
Members
34,883
Latest member
clamwc
Back
Top