The minimum log length statistics is very wrong. Why?

Question

The log listed as the shortest one with only four words is actually on of my longest logs with hundreds of words and many images. Why?

The log: http://coord.info/GLMZ8KTD (on GCM00X)

The stats: Longest: GC2FY5W 478, Shortest: GCM00X 4

The computer is confused, isn't it?

Comments

In my statistics the minimum log length also is wrong. It says the shortest log has 14 words. But according to Targets nice tool http://project-gc.com/Challenges//19085 (thank you, Target!) there are also logs with 10, 11, 12 and 13 words (shame on me...). So the shortest log has 10 words (which is true), I logged it in 2010 and didn't touch it afterwards (in the meantime the cache is archived and locked, so it cannot be changed). Moreover I refreshed my founds with the self-support a few days ago...

---

and this answers the question how?

---

I just wanted to confirm what Kub4 described...

---

I would guess that the algorithm to count words differs in both cases (though it is weird). There was some question about this some time ago describing how the words are counted in the stats.

---

In the thread http://project-gc.com/qa/?qa=10700/do-chinese-charcaters-count-towards-the-the-author-badge Target writes that his tool counts the number of words the way that pgc counts it...

Answer 1

Did you edit the log - ie originally start with a short message (Full log to follow.) and then rewrite it fully? If so it may be the new log hasn't been updated and the old length is still cached?

Comments

I did, but I do it with most my logs - a short log in a field from a phone and then longer log at home using a computer. Both older and newer logs are counted correctly, my average is around 100 words, but this one is just somehow stuck.

---

It's not stuck, just new - the log is only a couple of days old and the task to refresh old logs only runs occasionally (weekly possibly) so if you wait longer it'll probably refresh. You were probably unlucky this time that the job that detected the log ran between your original publish and update, generally if you update them quickly project-gc won't even see the original log and so grab the updated text on the first attempt.

If you want to speed it up you could try using the 'refresh found geocaches' in the self support section of the support system.

---

A better way might be to dont log the cache online in the field but make a fieldnote instead. And use the filednotes on the webpage to post a full log directly

The there is more likely that the CO reads the log because there is nog notification for log changes if you dont use pgc.
I only do online logs for FTF and then I delete the log and post a new log at home.
And fieldnotes for or other finds