The Mythical 100%

You have 13 200 series elk points.  You find a hunt that looks interesting, 286Q East Hebo.  Here are the draw odds:

286Q East Hebo
Preference Points:0 – 910111213+
2019 Resident Actual Odds3%50%100%100%100%

Let’s say this hunt matches your parameters, season dates, weapon, bag limit and the location is all good. With 13 points and the information listed above, how certain are you that you’ll draw this tag?

The Mythical 100%

You choose a hunt where your odds of drawing are 100% at your point level.  You apply.  You plan and forego other hunting opportunities to prepare for this particular hunt because you are certain you will draw.  Your fall schedule is based on drawing this tag, it has 100% draw odds, you are all in, and then you do not draw.

This is what happened to me 20 years ago. 

The truth is, people see 99% and 100% so differently.  People see 100% as a guarantee and they want a guarantee when it comes to drawing a tag.  However, when it comes to the Oregon controlled draw, there will always be a level of unpredictability and uncertainty and there are no guarantees.  

Sometimes 100% draw odds is not a sure bet.

Last fall I applied for an Oregon antelope tag. My listed odds were 100% at my point level.  While I knew the math was right, I also knew that I likely would not draw . . . and I did not draw.  An aspect to the hunt I had seen before indicated that a small change in application behavior might result in not drawing the tag.  You can call it “human intuition” but it’s more of an educated observation.

I cannot inject “intuition” or observation into the predictions.  What I can do is share some of the factors and traits that make listed odds, particularly 100%, vulnerable to change.  I’d like to share some of the things you need to consider when you see 100%.  If you are aware of other traits that should be considered, please send them to me and I will add them to this list.

The number of tags

Looking again at East Hebo:

86Q East Hebo   Tags: 3
Preference Points:0 – 910111213+
2019 Resident Actual Odds3%50%100%100%100%

I added a new piece of information – the number of tags.  The number of tags allocated for a hunt is by far the single biggest factor to consider when it comes to 100% draw confidence.  East Hebo has 3 tags, which means two tags in 75% pool and 1 in the 25% pool.   

When you see 100% always ask:  How many people would it take to impact my odds? 

I’ll phrase this another way, how many people jumping into this hunt at 14 points will it take to ruin your chances of drawing? 

With 13 points and East Hebo, it would take just 2 applicants with 14 or more points.   

The second question to ask: How many people are capable of impacting my odds?

How many people in Oregon have 14 or more elk points?  There are hundreds of people with 14 or more points. 

Given that there are a lot of people capable of ruining your odds and that it will only take two of them to achieve this, the prediction for this hunt should be viewed with low confidence.  The fact is, even with 17 points you cannot be confident of drawing. Two people could jump in at 18 or more points.

Here is a fair question to ask – if the listed odds are too high (100%), why don’t I fix them? 

The odds are based on what has happened in the past It is the best predictor of what will happen next.    If the listed odds are 100% with 11 points, as in this example, it means that people have not been jumping into this hunt with 12 or more points – not from 299 pool or from another hunt.  However, this could change at any moment, and it just takes two people.

The math says 100% at 13 points, human intuition says maybe, maybe not.

Your distance from the 25% pool

I am going to change a few attributes on 286Q.  Let’s say there are 300 tags, you have 11 points and here are the odds:

86Q East Hebo 
Preference Points:0 – 910111213+
2019 Resident Actual Odds3% 5%100%100%100%

When the draw occurs, its starts with the 75% pool draw.  It starts at the highest point level and looks at first choice applicants, if any, at that level.  It gives out tags and once everyone at the highest level has received a tag, it moves to the next lower point level.  The draw continues to work its way down the point levels until it runs out of 75% pool tags.  I call the point level where the 75% tags run out “the break point”.  Every hunt has a break point (unless its 100% to draw with zero points), every point level above the break point is 100% odds, every point level below the break point is the 25% pool odds. 

In the 286Q example above, the break point odds (5%) is very close to the 25% pool odds.  This is telling you something important.  It means that while there are 75% pool tags predicted for the 10 point level, there will not be very many.  This also means that if just a few people decide to apply with 11 or more points, the 11 point level will become the break point.  If you have 11 points and that level becomes the break point, then you may not draw. 

286Q East Hebo Tags: 300
Preference Points:0 – 1011121314+
2019 Resident Actual Odds3%87%100%100%100%

This is what happened with my hunt last year. I was listed as 100%, but the actual draw showed it became 87% and I fell into the 13% that didn’t draw. 

Draw Odd Consistency

When you see a hunt like this, run in the other direction…

2014 Resident Actual Odds3%38%100%100%100%100%
2015 Resident Actual Odds2%2%2%2%87%100%
2016 Resident Actual Odds3%89%100%100%100%100%
2017 Resident Actual Odds4%4%71%100%100%100%
2018 Resident Actual Odds14%100%100%100%100%100%
2019 Resident Actual Odds1%1%51%100%100%100%

… unless you know why, in which case, use that inside knowledge to your advantage. 

Oregon has hunts where the draw odds bounce around.  This makes a mathematical prediction for the coming years an impossible task.  It’s like predicting the next winning lottery tickets based on what numbers have happened in the past.

My algorithm is going to produce a prediction, but for hunts that bounce around, without a numerical pattern, that prediction is likely going to be wrong. 

A change by the ODFW

Every year the ODFW changes hunts.  Sometimes the ODFW changes the hunt id but the hunt parameters remain the same.  Sometimes the ODFW keeps the hunt id but changes the parameters of the hunt. 

When a hunt has a change you need to decide if it is really the same hunt and will draw the same applicants or whether it now is a new hunt.

Ask this: Has the hunt changed so much that whatever happened in the past with this hunt has no bearing on what will happen next? 

Here is a (hypothetical) example:

Hunt 286V is the Starkey Exp. Forest cow tag for the past 5 years.  This year the ODFW makes a small change, just to the bag limit, it changes from “cow elk” to “any elk”.  A change like this makes this a completely different hunt and past applicant numbers are meaningless.

The Odds Mess with the Odds

I recently received a question on this real hunt:

Preference Points:012+
2013 Resident Actual Odds35%100%100%
2014 Resident Actual Odds25%91%100%
2015 Resident Actual Odds25%86%100%
2016 Resident Actual Odds25%78%100%
2017 Resident Actual Odds24%74%100%
2018 Resident Actual Odds27%81%100%
2019 Resident Actual Odds39%99%100%
2020 Resident Prediction77%100%100%

The question was simple, do you really believe the odds will be 77% with zero points?  My answer was simple as well, no the odds will be lower, something like the 20-30% range.  If you look at the numbers, you can see the hunt is getting easier to draw with 1 point.  The math is right, if that trend continues the prediction will be spot on.  However, the trend will not continue because folks will see these odds, particularly applicants with 1 point, and focus on the 100%. It will pull them in.

Here is another actual hunt (there are lots of hunts like this):

Preference Points:01+
2014 Resident Actual Odds100%100%
2015 Resident Actual Odds62%100%
2016 Resident Actual Odds39%100%
2017 Resident Actual Odds71%100%
2018 Resident Actual Odds61%100%
2019 Resident Actual Odds27%100%
2020 Resident Prediction100%100%

This hunt is listed as 100% with zero points.  I am certain that this hunt will not be 100% with zero points in the actual draw.  If you look at the numbers you can see how the math arrives at 100%.  Yet, these odds will catch attention of people with no points, applicants that are looking for something to apply for. They will see this hunt, and it will pull enough of them in that the odds will change. 

The point level is important, keep in mind that every applicant has at least zero points and can impact the odds of a hunt at the zero point level.

When you see that mythical 100% draw prediction, please look a little closer.