PQ Results not matching up

[+nericksx]

Around on June 20th I ran my Mt Hood PQ and loaded it into my offline mapping app for use on my Android tablet. I've been working off this same revision of the PQ for 3 trips to the mountain now, including this past weekend. Today on my lunch hour I decided to log my caches from this weekend and since I didn't have my GPS with me, I pulled up my Mt. Hood PQ on the website to refresh my memory. I was really confused when I found that most of the caches we did or talked about doing where not on the PQ list. Dumbfounded, I grabbed my tablet and pulled the GC numbers from the 6/20 PQ.

 

So the question is.... why are my downloaded PQ results from 6/20 different from the results that show on the PQ preview? I haven't changed the PQ parameters at all.

 

This makes me a little nervous because after I pull a PQ and load it into my tablet or GPS, when I'm in the woods I'm pretty confident that if there isn't a cache listed on my GPS or tablet, there is no cache there. Now I'm wondering how many potential caches I've been missing because my PQ isn't what I thought it was.

[+StarBrand]

Simply put - there MUST be some difference in the PQ (now vs then). I do not think the site has changed anything that dramatic recently. Is there any correlation at all in the 'missing' caches (some size or attribute or rating??)

[+nericksx]

Simply put - there MUST be some difference in the PQ (now vs then). I do not think the site has changed anything that dramatic recently. Is there any correlation at all in the 'missing' caches (some size or attribute or rating??)

 

Nope, no difference. The only attributes I specified was a 61 mile radius, traditional & multis only, and small/reg/large size only. No terrain or difficulty rating restriction.

 

I just tinkered with the PQ to see if I could tweak it to make the "missing" caches show up. The only change that would force the missing ones to show was to increase the result output from 500 to 1000. Then the missing caches showed up. Now I understand that if you try to pull a 500 result PQ from an area with 1000 caches, the system is going to pick and choose for you. I'm just surprised that the system-pick would change, if you haven't re-run the PQ.

[+JPatton]

It sounds like some new caches have been placed in the area, which are showing up in the preview. The preview is showing the caches that would be included in the query if it ran today, not what was in it the last time the query ran.

[+nericksx]

It sounds like some new caches have been placed in the area, which are showing up in the preview. The preview is showing the caches that would be included in the query if it ran today, not what was in it the last time the query ran.

 

Nope, because the missing caches are in my downloaded PQ file from 6/20, but not showing in the preview now. Plus they were all placed 2011 or earlier - and no, they're not all archived. Good thought, tho.

[+The A-Team]

You still haven't answered StarBrand's question, though. If you pick out a handful of the missing caches, can you see anything in common? Are they close to the center of the PQ, or far away? Are they all the same size/D&T/owner/etc.?

Also, what JPatton meant was that if a bunch of new caches were placed closer to the center of the PQ than the "missing" caches, the PQ may now hit the 500 limit before getting to the missing ones, whereas it wouldn't have before those new caches were published.

[jholly]

Simply put - there MUST be some difference in the PQ (now vs then). I do not think the site has changed anything that dramatic recently. Is there any correlation at all in the 'missing' caches (some size or attribute or rating??)

 

Nope, no difference. The only attributes I specified was a 61 mile radius, traditional & multis only, and small/reg/large size only. No terrain or difficulty rating restriction.

 

I just tinkered with the PQ to see if I could tweak it to make the "missing" caches show up. The only change that would force the missing ones to show was to increase the result output from 500 to 1000. Then the missing caches showed up. Now I understand that if you try to pull a 500 result PQ from an area with 1000 caches, the system is going to pick and choose for you. I'm just surprised that the system-pick would change, if you haven't re-run the PQ.

Yes if you restrict the PQ to 500 caches and there are 600 candidates the system will trim the result to 500. It is, after all, what you asked for. The preview page does not have that restriction. It is always best to check the preview page to see if you exceeded the 500 limit so you have the option of decreasing the range or increasing the limit. It did exactly what you asked so I don't see a problem.

[+Markwell]

Are any of the caches archived?

[+nericksx]

No none of the caches are archived.

 

To answer StarBrand's question: the missing caches are in the same area (about a 2 mile stretch along a river) but it is a variety of terrains & difficulty, all small or regular, a variety of COs. The caches are well within the radius of the search, as caches farther away from the center show up.

 

Again, I understand the system pulling 500 caches randomly if there are 600 in the area. Although, I didn't think of new caches being placed in the area in the last month changing the returned data set (per A-Team and JPatton). So far, that is the explanation that makes the most sense. I would love it if a Lackey developer would weigh in and let us know how the PQ algorithm works.

 

I'm not complaining at ALL, I just want to understand the system so I can use it more effectively (and not feel like I'm going crazy!).

 

Thanks for the input everyone!

[+Markwell]

If there is an origin, the farthest ones disappear. If it's for a political boundary (like a state), the newest ones drop off.

 

Routes can be anyone's guess

 

Attributes really mess things up.

[+The A-Team]

Again, I understand the system pulling 500 caches randomly if there are 600 in the area.

Are you using a distance from a specified point in your PQ? If so, there's nothing random about it. Like Markwell said, it will return the closest 500 caches to that point. If some new caches are then published within that range, the limit will be hit sooner that it did previously, and therefore some of the farther caches will no longer be included. It sounds like this is what's happening in your case.

[+nericksx]

Also, what JPatton meant was that if a bunch of new caches were placed closer to the center of the PQ than the "missing" caches, the PQ may now hit the 500 limit before getting to the missing ones, whereas it wouldn't have before those new caches were published.

 

If there is an origin, the farthest ones disappear. If it's for a political boundary (like a state), the newest ones drop off.

 

Are you using a distance from a specified point in your PQ? If so, there's nothing random about it. Like Markwell said, it will return the closest 500 caches to that point. If some new caches are then published within that range, the limit will be hit sooner that it did previously, and therefore some of the farther caches will no longer be included. It sounds like this is what's happening in your case.

 

Patton/Marwell/A-Team...you nailed it. The PQ was originally constructed by ensuring all of Mt. Hood Nat'l Forest was on the screen, then clicking the link to make a PQ out of it. It turned out it was a radius of 61 miles. Then I set it to only return 500 trads/multis, and small/reg/large. As I have now learned, it returned 500 caches starting from the center of the search area (a point somewhere south of Mt. Hood) and radiating out until it hit 500. No way did it reach a 61 mile radius, but it included all the caches in all the areas I was planning to frequent, so I didn't think that hard about it. I ran my PQ on 6/20 and loaded it into my tablet and GPS. Apparently between 6/20 and today, quite a few new caches must have been placed closer to the center of my search circle, and I just happened to have been caching this weekend on the very edge (unknowingly) of that search. I just checked the PQ, and the "missing" caches are indeed outside of what I can now see is a smaller circle, but less than a half mile from where I was in person this weekend.

 

It was kind of a perfect storm really, to have this happen. There just happened to have been enough new caches placed in 3 weeks to shrink my 500 cache circle, and I just happened to have found a couple that were inside the circle 3 weeks ago, but outside now.

 

I'm really thankful to all of you to helped me think this out. I hate not understand how something I rely on (even just for fun) works. I just extended the return count to 1000 to widen the circle, and all is well.

[+Don_J]

Simply put - there MUST be some difference in the PQ (now vs then). I do not think the site has changed anything that dramatic recently. Is there any correlation at all in the 'missing' caches (some size or attribute or rating??)

 

Nope, no difference. The only attributes I specified was a 61 mile radius, traditional & multis only, and small/reg/large size only. No terrain or difficulty rating restriction.

 

I just tinkered with the PQ to see if I could tweak it to make the "missing" caches show up. The only change that would force the missing ones to show was to increase the result output from 500 to 1000. Then the missing caches showed up. Now I understand that if you try to pull a 500 result PQ from an area with 1000 caches, the system is going to pick and choose for you. I'm just surprised that the system-pick would change, if you haven't re-run the PQ.

 

It does not pick and choose. It will give you the 500 closest geocaches to the center point. If the caches you are concerned about are towards the outside of the radius circle and a bunch of new ones were placed closer to the center, the new ones could push the other caches above the 500 mark.

[+The A-Team]

I'm really thankful to all of you to helped me think this out. I hate not understand how something I rely on (even just for fun) works. I just extended the return count to 1000 to widen the circle, and all is well.

I'm glad we got to the bottom of it. There's always a reason why something happens, it's just a matter of figuring out what that reason is. Like you, I hate not knowing why something isn't working as expected.

[+Markwell]

Also - check out my second "tip" on my PQ Tips and Tricks