Discussions

different popularity

Alterna
different popularity
Risposta
09/10/19 22.50
 Hello everybody,
 I haven't enetered this site for long time because of problems with my computer, and I've recently solved the trouble (buying a new PCemoticon) and I could enter. 
 I used to list the griddlers by popularity in order to solve the most popular first, and so I have done these last days, but I found something that surprised me in more than one griddler. That is, different popularity between the list and the solving page.
I atach an image illustrating
 My question is, why does this happen?   Doesn't the popularity change in the list after a long time? 
Thank you for answer me. 
0 (0 Voti)

RE: different popularity
Risposta
10/10/19 19.38 come risposta a maría ..
seems to be fine. Try refreshing your browser.
0 (0 Voti)

RE: different popularity
Risposta
11/10/19 0.10 come risposta a Elad.
I have refreshed my browser and it continues the same, but not only with this one, with more ones. The problem is I don't know what's the real popularity
0 (0 Voti)

RE: different popularity
Risposta
16/10/19 8.55 come risposta a maría ..
Can you give me more puzzle id's with this problem?
0 (0 Voti)

RE: different popularity
Risposta
16/10/19 21.25 come risposta a Elad.
I'll do. In the moment I find more, I'll do it. And I¡ll search for more in the weekend
0 (0 Voti)

RE: different popularity
Risposta
02/11/19 10.47 come risposta a maría ..
The same for me. In the overview the following griddler showed 88%. When clicking on the griddler itself in "only" showed 71%
https://www.griddlers.net/nonogram/-/g/62480
0 (0 Voti)

RE: different popularity
Risposta
04/11/19 17.32 come risposta a User52044 User52044.
DB reindexed - fixed
0 (0 Voti)

RE: different popularity
Risposta
23/12/19 7.04 come risposta a Griddlers Team.
Fixed?  Doesn't appear so.  Just sort all puzzles by popularity, descending...  look at some of the puzzles rated 100%.  Many are rated much lower when you look at detail.
0 (0 Voti)

RE: different popularity
Risposta
23/12/19 17.57 come risposta a Carol.
please send id's to team@griddlers.net
0 (0 Voti)

RE: different popularity
Risposta
23/12/19 18.36 come risposta a Griddlers Team.
Do the popularity descending sort and click on some links!  It's very evident!

First puzzle (130760) says 100%, click on link = 75%
Second puzzle (136471) says 100%, click on link = 73%
   ---- skip more parts of above puzzle ----
Third puzzle (115517) says 100%, click on link = 77%
on and on and on.........................................
195124 - click on link = error
203856 - 100% = 76%
every one I click on that should be 100% is NOT 100% when you look at the actual puzzle.  I will stop list.  Check it out yourself?  There are obviously too many for me to list for you.  It is not 1 or 2 individual puzzles that exhibit this problem.

I notice that when I click on "go back to list", the list displayed is not the same as previous!  Puzzles missing.  Refresh page redisplays missing puzzles.
0 (0 Voti)

RE: different popularity
Risposta
28/12/19 20.57 come risposta a Carol.
I've checked this -
ID's – 136471, 115517, 203856 are correct!
ID 195124 - There is no puzzle under this id - please send a screenshot to team@griddlers.net of the page of the search results list displaying a link to this puzzle.
ID 130760 - is still under investigation.
0 (0 Voti)

RE: different popularity
Risposta
29/12/19 5.29 come risposta a Griddlers Team.
no time for screen shot (not my job) but still curious.  On my computer the puzzle list shows 136471 (and ALL other parts) as 100%.  When displaying that multi screen, the popularity is 73%.  Which one is correct?
0 (0 Voti)

RE: different popularity
Risposta
29/12/19 7.17 come risposta a Carol.
So finally I found two bugs which caused wrong popularity display.
It is a bit tricky for multi griddlers becasue votes goes to entire bundle not only for that one part you have just solved.
Rasto
0 (0 Voti)

RE: different popularity
Risposta
30/12/19 0.57 come risposta a Ra100.
hmmm, don't think so.

Take 1 that is NOT multi (204100 is at top of list if sorted descending by puzzle number).

On list, 204100 shows 68.75% looking at puzzle shows 48%.  ALL (that I have looked at) popularity is different on list and individual puzzle page, has been for a LONG time!
0 (0 Voti)

RE: different popularity
Risposta
04/01/20 15.42 come risposta a Carol.
I've deserved, there many of them, and I refreshed many times, and it's still the same. I think lists of puzzles don't change to show some of them first!
0 (0 Voti)

RE: different popularity
Risposta
10/01/20 19.33 come risposta a maría ..
NEW puzzle 204327, popularity on list 100%, detail shows 6 solvers, 80% popularity.  This is not a multi.  With only 6 solvers, perhaps someone could look into individual ratings/popularity and find the problem.  It really should not be difficult!
0 (0 Voti)

RE: different popularity
Risposta
10/01/20 19.37 come risposta a Carol.
Hi
there is a time delay. Now both values are 80%.
R
0 (0 Voti)

RE: different popularity
Risposta
10/01/20 23.12 come risposta a Ra100.
Not so.  I doubt very much that my computer/browser/location/etc?!? is the cause of the values remaining 80% in one place and 100% in the other?

Time delay makes NO SENSE!
0 (0 Voti)

RE: different popularity
Risposta
11/01/20 6.42 come risposta a Carol.
Actually I know how ti works, I made it. Detail page is shown directly from master database. Puzzle is reads data from Elasticsearch index and it is updated with some delay.
You know with every solved and voted puzzle that one puzzle record should be recalculated ( popularity, solved users, average time ) user statistic are recalculaed and it takes some time and resources. And it also blocks reading from database a bit. So the data are processed in queue and some calculations are made in batch every few minutes.

There was really a bug in popularity calculation which I fixed few days ago. Now I am almost sure it is correct with some delay.

Rasto
0 (0 Voti)

Moderatori del forum griddlers_team, elad, Ra100, chefmomster2, domi77, dreamtheater, elimaor, ElinaMaria, Jeltje, sslug, cosmictrombonis, raist.
Please read the Board-wide Policies before you start using this forum.