GSAK Cache Quality Filter

[+Jamie Z]

Wasn't sure if this should go in the general forums or the software forum... although I think it's more a topic of general interest, it does relate to software, and I think the topic is less likely to be dragged through the mud tucked away amongst the geeks.

 

In this post, Lil Devil stated a feeling that has been stated before. Essentially, that so-called good caches tend to have longer logs than so-called lame caches, which tend to have logs of only a few words, or often not even words but abbreviations.

 

His comment sparked an idea, and I went to the GSAK forums to ask Clyde if he would consider adding a feature to GSAK which would sort caches by their average log length. I suggested that only Found It and Didn't Find It logs be counted, since Notes and other logs generally aren't associated with hunting for the cache.

 

Well, not surprisingly, Clyde responded quickly and advised me that this was already possible. At least the tools were there to do it, although I don't have the knowledge. Clyde suggested that a macro could be written to perform this task.

 

Then, he made another post a few hours later with just such a macro! Holy cow!

 

So... here it is. Now you can sort caches by the average length of its logs. Try it on your local caches to see if there is a correlation.

 

Clyde warns:

1. This macro will update the user sort column in your current database. If you are already using this column then I suggest you copy this database to a new temp database, then run the macro

 

2. It is no speed demon. All the logs need to be analysed and some databases can have thousands of logs. I suggest you first set a filter on the "likely suspects" to narrow down the number of caches/logs searched. When testing I ran this macro on a small databasse of 800 caches with around 9,000 logs and it took 15 seconds.

If you want, go into View-->Add/Delete Columns and check User Sort. Now you can even see the results and see how many caches in your area have log which average more than 500 characters. Do any average more than 1000? What about those less-than-100-character caches?

 

You can find the macro here: http://gsak.net/board/index.php?act=Attach...st&id=20343

 

Just save that web page as a text file and open it from GSAK with Macro-->Run

 

Thanks Clyde.

Edited November 30, 2006 by Jamie Z

[+ThePropers]

Sounds great! I'll be checking that out when I get back home. Thanks for the heads up.

[+Jamie Z]

Man,

 

I thought this new macro would turn the geocaching world upside down.

 

Perhaps it better belongs in "general" for discussion where it's not hidden amongst the geeks.

 

Jamie

Edited December 1, 2006 by Jamie Z

[+DocDiTTo]

I just tested it on a database I had loaded in GSAK, and it appears to be worth looking at. The first cache on the list happened to be one of mine, and the one I consider to be my best hide, so that was pretty cool. The bottom of the list contained event caches, probably cause they have a lot of little logs like "bug drop". But a number of roadside micros and other C&Ds hit the bottom too, so that's good. This method isn't a guarantee of quality by any means, but it will certainly help while trying to sort the wheat from the chaff. Caches closer to the top of the list would definitely be worth looking at. Thanks for sharing this!

[+thedeadpirate]

Preliminary results are fuzzy. I know for a fact that some of the caches it picked out are great caches. However, a few of the caches with high results were so/so caches with one in particular having a very long log that was not exactly flattering.

 

I'm gonna filter so I can run this on just caches I have been to before so I can get a better idea of how well I agree with the results in my database. Overall it does look promising.

[+TrailGators]

I just ran the macro using my finds PQ. I can honestly say that the better caches all showed up at the top of the list. Amazingly the bottom of the list was loaded with 1/1 caches. The very bottom were all the events but if GSAK only counted the "attended" logs then those would rise above the 1/1s. Thanks Clyde!

 

Edit: I would also only count "found" logs. DNF logs don't have much to do with quality of the cache....

Edited December 1, 2006 by TrailGators

[+budd-rdc]

Man,

 

I thought this new macro would turn the geocaching world upside down.

 

Perhaps it better belongs in "general" for discussion where it's not hidden amongst the geeks.

 

Jamie

I prefer it discussed here. Don't want to change the psychology of the sampling data.

 

Thanks for taking the initiative and sharing this info with us. Also, thanks Clyde.

[+TrailGators]

Is it true that PQs only give you the last 5 logs?

[+Jamie Z]

Edit: I would also only count "found" logs. DNF logs don't have much to do with quality of the cache....

In part, I agree. Typically, my own DNFs have the story about my experience looking for the cache. In many ways, my own DNFs come out exactly as would a Found It log, except for the last paragraph. Example, even if I don't find the cache, I often talk about the great hike, or mention the cool scenery. That part of the caching experience doesn't change.

 

Clyde did omit notes, which rarely have anything to do with the cache hunt (except for those people who log DNFs as notes). "Attended" and "Picture Taken" logs aren't counted as "Found It," that's why you'll see webcams and events on the bottom, since they don't have any "Found It" logs, their effective log length is zero.

 

Is it true that PQs only give you the last 5 logs?

Mostly, yes. You get the most recent five logs, plus any others that are yours. That is, if you were the most recent person to log the cache, you would receive your log plus the next four. If you were the sixth-to-last person to log the cache, you would get the last five, plus yours. If you also had a note or DNF prior to that, you would get the five previous logs plus all of your logs. Make sense?

 

If you download an individual GPX from a cache page, you get the last 20 logs.

 

Of course, as a couple others have mentioned, this isn't a foolproof technique. It's just a generalization--and especially if each cache has only five logs, the sample size is quite small. But all other things being equal, I think it's a good way to go to a new area and quickly identify probable caches that have a greater "fun" factor, or something else which makes people talkative.

 

I'm not inherently anti-micro, but I did note that for my local area, not counting those caches which have a zero log length, the bottom nineteen caches in my database as ranked by log length are micros. After that, there are a few regulars sprinkled in, but it's clear that the vast majority of caches with the shortest average logs are micros. No wonder they get such a bad reputation.

 

Jamie

[+Maingray]

Nice thinking, thanks for the heads up. I'll give it a whirl.

Edited December 1, 2006 by Maingray

[+TrailGators]

Edit: I would also only count "found" logs. DNF logs don't have much to do with quality of the cache....

In part, I agree. Typically, my own DNFs have the story about my experience looking for the cache. In many ways, my own DNFs come out exactly as would a Found It log, except for the last paragraph. Example, even if I don't find the cache, I often talk about the great hike, or mention the cool scenery. That part of the caching experience doesn't change.

 

Clyde did omit notes, which rarely have anything to do with the cache hunt (except for those people who log DNFs as notes). "Attended" and "Picture Taken" logs aren't counted as "Found It," that's why you'll see webcams and events on the bottom, since they don't have any "Found It" logs, their effective log length is zero.

 

Is it true that PQs only give you the last 5 logs?

Mostly, yes. You get the most recent five logs, plus any others that are yours. That is, if you were the most recent person to log the cache, you would receive your log plus the next four. If you were the sixth-to-last person to log the cache, you would get the last five, plus yours. If you also had a note or DNF prior to that, you would get the five previous logs plus all of your logs. Make sense?

 

If you download an individual GPX from a cache page, you get the last 20 logs.

 

Of course, as a couple others have mentioned, this isn't a foolproof technique. It's just a generalization--and especially if each cache has only five logs, the sample size is quite small. But all other things being equal, I think it's a good way to go to a new area and quickly identify probable caches that have a greater "fun" factor, or something else which makes people talkative.

 

I'm not inherently anti-micro, but I did note that for my local area, not counting those caches which have a zero log length, the bottom nineteen caches in my database as ranked by log length are micros. After that, there are a few regulars sprinkled in, but it's clear that the vast majority of caches with the shortest average logs are micros. No wonder they get such a bad reputation.

 

Jamie

 

Thanks! How about the our total finds query. Does it has the last five logs and our log or just our log?

[+Jamie Z]

Thanks! How about the our total finds query. Does it has the last five logs and our log or just our log?

The all-find query contains only those caches which you've found, and for those caches you get only your logs.

 

For example, say you logged 4 DNFs, 2 notes, an SBA, and a find on a particular cache. You'd get all those logs. On another cache, you've logged 6 DNFs, 3 notes, and a Needs Maintenance. That cache wouldn't show up in your all-finds query.

 

I've appealed to TPTB to include at least DNFed caches in the all-finds query to no avail.

 

As such, the All-find query would be pretty useless for this particular macro. All you will do is get the average number of characters in your own log. Unless you're into that kind of thing.

 

Jamie

Edited December 1, 2006 by Jamie Z

[+TrailGators]

Thanks! How about the our total finds query. Does it has the last five logs and our log or just our log?

The all-find query contains only those caches which you've found, and for those caches you get only your logs.

 

For example, say you logged 4 DNFs, 2 notes, an SBA, and a find on a particular cache. You'd get all those logs. On another cache, you've logged 6 DNFs, 3 notes, and a Needs Maintenance. That cache wouldn't show up in your all-finds query.

 

I've appealed to TPTB to include at least DNFed caches in the all-finds query to no avail.

 

As such, the All-find query would be pretty useless for this particular macro. All you will do is get the average number of characters in your own log. Unless you're into that kind of thing.

 

Jamie

Actually, I ran it to see how accurate the macro was for my own logs and it was pretty good! Some of my favorite caches showed up high in the list and the very common so-so ones that I can never remember when it comes thime to log showed up at the bottom of the list. Of course, I tend to write longer logs for caches I like and very short ones for ones I don't so it should have worked!

 

I still think the DNF logs are not needed. I know I have written some very long logs on DNFs because there were so many issues with the cache. So why should these caches get ranked higher?

[+The Jester]

<snip>

 

I still think the DNF logs are not needed. I know I have written some very long logs on DNFs because there were so many issues with the cache. So why should these caches get ranked higher?

It's pretty easy to modify the macro to remove the DNF logs, leaving just the Found logs - just change this line:

If $d_lType = "Found it" or $d_lType = "Didn't find it"

to

If $d_lType = "Found it" 

Edited December 1, 2006 by The Jester

[+TrailGators]

<snip>

 

I still think the DNF logs are not needed. I know I have written some very long logs on DNFs because there were so many issues with the cache. So why should these caches get ranked higher?

It's pretty easy to modify the macro to remove the DNF logs, leaving just the Found logs - just change this line:

If $d_lType = "Found it" or $d_lType = "Didn't find it"

to

If $d_lType = "Found it" 

 

Thanks!