New GoHunt Predictive Odds

[S.Clancy]

I'm interested how they are deciding it's a 95% CI. They can't have enough data pts to justify 95%, so I'm not understanding how the model is robust enough.

 

[Tilzbow]

Interesting, I had never heard of Gaia when I got in initially and started out with OnX and have developed the familiarity there.

So you think OnX has caught up to/passed Gaia in terms of strictly mapping& GPS? I'm not in any hurry to switch but always curious what's out there.

I don’t know if OnX is better than GAIA now since I haven’t used it for at least 10 years (it wasn’t better 10 years ago and actually didn’t work that great) but I’ve read a lot of good reviews on OnX the past year or two and it seems they have some cool newer features for hunters so I thought I’d give it a try. I also liked Top Rut back in the day. Plus add in the fact that GAIA got bought by Outside.com and that required a new account and some setting adjustments to keep my data from being shared and I was thinking about ditching GAIA and moving to OnX. But with all the issues with the draw odds on that platform I’ll stick with what I’ve got for now.

 

[sivart]

Theres no way it can be accurate with the 2025 CO changes and major changes in 3 years. There is no way even AI can predict that much variability.

 

[Tilzbow]

Theres no way it can be accurate with the 2025 CO changes and major changes in 3 years. There is no way even AI can predict that much variability.

Marketing hyperbole….. It would’ve been more convincing if they started with AZ since those quotas were set and the draw hasn’t changed in several years. Utah is good too other than quotas won’t be set until May 1. I’m not sure if NV will be included but our quotas also aren’t set until May and could vary widely from past years, especially for sheep.

 

[wingmaster]

GoHunt is the first big company to offer predictive odds like this. I believe that because of this, they will actually influence the draw odds somewhat significantly by publishing "predictive odds". They may have been 95% accurate had they not published their predictions.

Unless AI has already factored this in. This sounds like a joke but I'm not so sure these days

 

[WiscoHoundsman]

I’m guessing AI is being used and they can plug in data from many years back if they chose. They could also write it to review tag allocations from the past to account for any trends there.

I’m guessing someone that likes to play with AI could do this on their own.

 

[thegoosano]

Unless AI has already factored this in. This sounds like a joke but I'm not so sure these days

Totally, the way I’ve heard them talk on podcasts etc I know they are aware they have an effect on applications so the impact predictive odds would have on draws wouldn’t be new to them.

Reminds me of someone having the power to sway the stock market through massive followings and telling everyone to “buy” a meme stocks and artificially increasing value. Predictive odds potentially gives a lot of power to a company who knows they have a massive following.

I trust the folks at GoHunt and am not worried, but it’s an interesting thought to imagine having that kind of influence.


Sent from my iPhone using Tapatalk

 

[Beendare]

Looked around at the new predictive odds for Colorado. Really surprised me on the prediction for my go to deer unit, they’re touting 95% confidence levels which is a bold claim! Very interested to see the odds report for next year and look back at accuracy.

What does it mean when I plug in my points for a species and it displays a bunch of laughing emojis?

 

[thegoosano]

What does it mean when I plug in my points for a species and it displays a bunch of laughing emojis?

Sent from my iPhone using Tapatalk

 

[Extrapale]

You want predictive data?

Check out the Death Clock.
Then you’ll know how many hunts you have left.

The Death Clock

Death Clock. Have you ever asked yourself 'when will I die?', use our advanced life expectancy calculator in combination with ai to accurately predict your death date and receive your own death countdown clock.

www.death-clock.org

104 and change for me. Almost exactly halfway there

 

[Jethro]

Gohunt doesn’t know CO quotas, doesn’t know who is going to apply for what, and can’t know what NR all draw archery will do to odds. But they have a marketable product. Tag competition has gotten so intense people will look at, and pay for, any number that may provide some intel. Even numbers that can’t be backed up with math.

 

[180ls1]

Looking at a CO hunt.

2024 - 3 points - 80% drew out.

25' Predicting - 0% at 4pts - 100% at 5pts.

I am seeing other massive jumps that I can't wrap my head around. Is anyone seeing the same or have thoughts?

 

[svivian]

Looking at a CO hunt.

2024 - 3 points - 80% drew out.

25' Predicting - 0% at 4pts - 100% at 5pts.

I am seeing other massive jumps that I can't wrap my head around. Is anyone seeing the same or have thoughts?

That’s approximately a 1 point per year creep which is not massive If you’ve been following many units in CO

Additionally, as higher point units become further and further out to draw you’ll see large jumps in the lower point units as people “burn” points

I think with the changes coming in 2028 it will be a crap shoot to guess what units will spike in points.

At the same time, as I mentioned before, I’m surprised it took this long. Anyone with a spreadsheet keeping track of the stats can figure a pretty close average and estimate the points needed.

Where I see go hunt going, is to the point where it’s similar to fantasy football stats. What I mean by that is being able to filter what units are trending up in points or down. Using this information to decide whether a unit is on the decline and to be potentially avoided or a unit on the up that you should try to apply for before it’s too late.

What I’m also surprised by is the lack of AI use at this point in the game. I mean really, if they uploaded as much info as possible into their system, they could have an “AI hunt planner”. All you would have to do is have a thorough questionnaire and from there have a customized hunt plan that tells you what states, units, and year to hunt for any species or weapon type based on your preferences. Not only that but you could fine tune it to application costs willing to spend, success rates, number of tags drawn etc….

If done correctly, based on your preferences, it would tell you what states to apply for rather than someone trying to figure out which ones to apply for.

There would be thousands of people willing to spend extra for a system that tells them what to do and when to do it.

 

[realunlucky]

I had a long chat at the gohunt booth about the new predictive odds and how they calculate it. They played it pretty close to the vest on the exact procedure but did convince me it wasn't a total WAG(wild ass guess) using trends and simulated draws to test those predictions.

Many changes impacting where and how people apply that have limited to no historical data to refer to. I'll be following to see how close thier predictive odds end up being.

I agree this is the beginning and only going to advance more and more.



Sent from my SM-S926U using Tapatalk

 

[wapitibob]

There is a lot of info most people don't ever see. As an example, last year in WY there were ~27,200 full price NR Elk apps, and ~139,900 people who never applied for an elk hunt, only buying the point. Those 139K people are invisible, and don't show up on the odds reports. Having all that info, knowing the point levels of those 139K, and the draw databases for the last 5+ years, Gohunt could get a pretty good idea of future draw trends.

NE​	20026​
NSE​	6157​
NYE​	1050​
TOTAL FP Elk​	27233​
​	​
​	​
NEPF​	156973​
NYEPF​	10173​
TOTAL points​	167146​
​	​
POINT ONLY, never applied for a hunt​	139913​

 

[Zzyzx]

GoHunt predictive odds - even if perfectly modeled - can only take in to account known factors.

So many unknowns at play here that can’t be accounted for. The announced 2028 CO change to 50/50 is going to change applicant behavior for 2025. That is just one of many unknowns that will screw up their models.

The CO change to 50/50 is such a huge game changer that I’m surprised they still moved forward with rolling out CO predictions. Could be a lot of egg on face.

 

[Jstumbaugh]

Looking at a CO hunt.

2024 - 3 points - 80% drew out.

25' Predicting - 0% at 4pts - 100% at 5pts.

I am seeing other massive jumps that I can't wrap my head around. Is anyone seeing the same or have thoughts?

I’m looking at a similar tag. Historical data shows point creep being less than 33% change in the last few years. They now predict a jump of more than a point. Not saying it can’t happen or won’t happen. I’m hoping it causes people to not apply.

 

[180ls1]

I’m looking at a similar tag. Historical data shows point creep being less than 33% change in the last few years. They now predict a jump of more than a point. Not saying it can’t happen or won’t happen. I’m hoping it causes people to not apply.

Yeah, if they were smart they would error on the side of hunters being pleasantly surprised rather than let down. I just don't see the jump as big as they project.

 

[thegoosano]

Yeah, if they were smart they would error on the side of hunters being pleasantly surprised rather than let down. I just don't see the jump as big as they project.

Ya my unit has been over 50% with 1 pt for 4-5yrs and next year it predicts single digit percentage with 1pt. Definitely a reasonable reduction if there is state wide pt creep from non res tag reductions, but some units that are similar don’t have the same drop in draw odds. So maybe they are factoring in something else like harvest statistics, recent trophy potential, etc to guess at increased demand for good units? Or it’s just off lol, can’t want to see.

And I’m still throwing my name in the hat, not letting AI scare me away lol


Sent from my iPhone using Tapatalk

 

[thegoosano]

If AI listens to every hunting podcast, scours forums, and every social media influencers YouTube and social media accounts for hints at units they hunted…NOW we’re talking some advanced pt creep predictions!


Sent from my iPhone using Tapatalk