LMI Tests 2011 - Some Statistics
@ 2011-12-27 7:43 PM (#6263) (#6263) Top

debmohanty




1000500100100100202020
Country : India

debmohanty posted @ 2011-12-27 7:43 PM

We have seen the last LMI test of 2011. In next few days, we will be posting some statistics about LMI tests in 2011. There are certainly many interesting numbers.

As usual, we welcome any idea related to this. You can either post here or send me an PM, and we'll see if we have the required information in the database.

Please note that most of the information will be presented as images; so they won't be translatable, but I hope they don't need to be.
@ 2011-12-27 7:46 PM (#6264 - in reply to #6263) (#6264) Top

debmohanty




1000500100100100202020
Country : India

debmohanty posted @ 2011-12-27 7:46 PM


To start with, here is the summary.


@ 2011-12-27 8:18 PM (#6265 - in reply to #6263) (#6265) Top

Para



Posts: 315
100100100
Country : The Netherlands

Para posted @ 2011-12-27 8:18 PM

Hmmmm, does that mean I was first to report all broken puzzles?
@ 2011-12-27 8:22 PM (#6267 - in reply to #6265) (#6267) Top

debmohanty




1000500100100100202020
Country : India

debmohanty posted @ 2011-12-27 8:22 PM

Para - 2011-12-27 8:18 PM

Hmmmm, does that mean I was first to report all broken puzzles?

Interestingly, yes.
May be I should ask you to take test always as the first player
@ 2011-12-28 10:32 AM (#6278 - in reply to #6263) (#6278) Top

vopani



Posts: 739
50010010020
Country : India

vopani posted @ 2011-12-28 10:32 AM

Looking at those statistics, it seems to be a wrong notion that sudoku is more popular than puzzles!

12 Players have participated in 'all tests'. It will be interesting to know about 'The Perfect Dozen' :-)

Edited by Rohan Rao 2011-12-28 10:35 AM
@ 2011-12-28 10:41 AM (#6279 - in reply to #6278) (#6279) Top

debmohanty




1000500100100100202020
Country : India

debmohanty posted @ 2011-12-28 10:41 AM

Rohan Rao - 2011-12-28 10:32 AM

Looking at those statistics, it seems to be a wrong notion that sudoku is more popular than puzzles!

Are you referring to "total number of unique players"?
In that case, as Akash would point out, we had more puzzle tests including a screen test which had tons of players.
@ 2011-12-28 10:56 AM (#6280 - in reply to #6278) (#6280) Top

debmohanty




1000500100100100202020
Country : India

debmohanty posted @ 2011-12-28 10:56 AM

Rohan Rao - 2011-12-28 10:32 AM

12 Players have participated in 'all tests'. It will be interesting to know about 'The Perfect Dozen' :-)


Here are they (in alphabetical order)
aldentea
deu
hayam
Kota
Nikola
nyuta
Rohan Rao
sknight
skrzypl1
S_Aoki
volxa
yureklis

(Note that 7 of them are from Japan)


Here are the 8 players, who missed just one test out of 31.
HuDu (Nikoli Selection 2011)
mackokajka (Nikoli Selection 2011)
motris (Best of LMI Puzzle Tests)
rob (Tapa Variations Contest - VI)
snakeeyes (Magic Cube)
sugitakukun (Melon's Puzzle Zoo)
tamz29 (Puzzle Hybrids)
zorko (Forsmarts 3² Anniversary)

Edited by debmohanty 2011-12-28 10:57 AM
@ 2011-12-28 11:23 AM (#6281 - in reply to #6280) (#6281) Top

debmohanty




1000500100100100202020
Country : India

debmohanty posted @ 2011-12-28 11:23 AM

@ 2011-12-28 3:28 PM (#6282 - in reply to #6263) (#6282) Top

Administrator



2000100050020
Country : India

Administrator posted @ 2011-12-28 3:28 PM


More information on similar lines - Number of Players Vs Number Tests



@ 2011-12-28 3:29 PM (#6283 - in reply to #6263) (#6283) Top

Administrator



2000100050020
Country : India

Administrator posted @ 2011-12-28 3:29 PM


Breakup for Sudoku and puzzle tests








@ 2011-12-28 3:32 PM (#6284 - in reply to #6263) (#6284) Top

Administrator



2000100050020
Country : India

Administrator posted @ 2011-12-28 3:32 PM

What is common in these 3 graphs : Close to 55% of the players participated in just 1 or 2 tests, and didn't participate again. It will be interesting to understand if we should do something about it.
@ 2011-12-28 3:40 PM (#6285 - in reply to #6284) (#6285) Top

Valezius



Posts: 66
202020
Country : Hungary

Valezius posted @ 2011-12-28 3:40 PM

I see there are at least 392 active member (who participated on at least 4 test), and there are 171 very active member (10+ tests).
In my opinion these numbers are really respectable.

Congratulation.
@ 2011-12-29 1:29 AM (#6289 - in reply to #6263) (#6289) Top

motris



Posts: 199
10020202020
Country : United States

motris posted @ 2011-12-29 1:29 AM

Well, I think the biggest factor regarding participation rates is online solving ability. My guess is a lot of the single occurrence puzzle players are from the Screen Test (only paperless Puzzle test), and a lot of the lower occurrence Sudoku players are specifically from the tests that had online solving option. Not everyone has a printer.
@ 2011-12-29 6:43 AM (#6290 - in reply to #6289) (#6290) Top

debmohanty




1000500100100100202020
Country : India

debmohanty posted @ 2011-12-29 6:43 AM

There are certainly lot of players who participated only in the Screen Test, but it is not that high.

I'm breaking up the 304 number of players who played just one test. Here is the break up by the "Only Test"
(This is to be read as, for example, 11 players participated in Twist, and didn't participate in any other puzzle test)

Similar break up for Sudoku tests as well.




So, my 55% number is certainly not correct. We can ignore Screen Test and April Sudoku test. We can also ignore the two "Best of" tests since they were the last tests of the year.
@ 2011-12-29 7:09 AM (#6291 - in reply to #6263) (#6291) Top

debmohanty




1000500100100100202020
Country : India

debmohanty posted @ 2011-12-29 7:09 AM

#of Paper Solvers / #of Online Solvers in Sudoku tests


@ 2011-12-29 1:04 PM (#6292 - in reply to #6263) (#6292) Top

prasanna16391



Posts: 1791
100050010010020202020
Country : India

prasanna16391 posted @ 2011-12-29 1:04 PM

Well I think we can consider some of the participants in the 2 Best of tests as newcomers who might stay on. :) I personally feel its due to the tough difficulty mainly, as I think that most of the 1-test participants are Indians(correct me if I'm wrong.) Considering the difficulty of the National Finals, LMI tests are way more difficult. I don't really know an apt solution to that though. Probably request the authors to do some easy puzzles for an added first-timer version.
@ 2011-12-29 1:50 PM (#6293 - in reply to #6292) (#6293) Top

Administrator



2000100050020
Country : India

Administrator posted @ 2011-12-29 1:50 PM

prasanna16391 - 2011-12-29 1:04 PM

Well I think we can consider some of the participants in the 2 Best of tests as newcomers who might stay on. :) I personally feel its due to the tough difficulty mainly, as I think that most of the 1-test participants are Indians(correct me if I'm wrong.) Considering the difficulty of the National Finals, LMI tests are way more difficult. I don't really know an apt solution to that though. Probably request the authors to do some easy puzzles for an added first-timer version.


The numbers suggest so. Here is the break up of 1-test participants (sudoku or puzzle) by countries.

@ 2011-12-29 2:36 PM (#6295 - in reply to #6293) (#6295) Top

prasanna16391



Posts: 1791
100050010010020202020
Country : India

prasanna16391 posted @ 2011-12-29 2:36 PM

Serbia's 7 might be mostly due to the last two Championship tests. They may continue participating now. Lets hope so. For India at least, based on some friends' reactions, I think its because they come, give the test, get a zero score or something just a bit more, see the rest of the scores and get discouraged. Maybe we should have a stat for everyone's first puzzle/sudoku tests up to show them that a lot of the top players started off bad too :P
@ 2011-12-29 5:53 PM (#6297 - in reply to #6295) (#6297) Top

rakesh_rai




Posts: 774
500100100202020
Country : India

rakesh_rai posted @ 2011-12-29 5:53 PM

I would be interested in a stat like this: "Amongst those who have submitted more than 100 answers, who are the ones with highest % of correct answer submissions" for puzzle tests, and for sudoku tests. Probably, we can exclude tests like Screen test here.
@ 2011-12-29 7:59 PM (#6298 - in reply to #6297) (#6298) Top

Administrator



2000100050020
Country : India

Administrator posted @ 2011-12-29 7:59 PM

rakesh_rai - 2011-12-29 5:53 PM

I would be interested in a stat like this: "Amongst those who have submitted more than 100 answers, who are the ones with highest % of correct answer submissions" for puzzle tests, and for sudoku tests. Probably, we can exclude tests like Screen test here.



There are only <15 players who have submitted 100+ sudokus. So I reduced the limit to 50 in case of Sudoku.












@ 2011-12-29 8:01 PM (#6299 - in reply to #6263) (#6299) Top

Administrator



2000100050020
Country : India

Administrator posted @ 2011-12-29 8:01 PM

Not sure if Screen Test should have been excluded. ByronosaurusRex has the most correct submission percentage and interestingly, he was the only player to have all puzzles correct in ST2.
@ 2011-12-29 9:38 PM (#6300 - in reply to #6299) (#6300) Top

vopani



Posts: 739
50010010020
Country : India

vopani posted @ 2011-12-29 9:38 PM

Gotroch!

Also, surprisingly, Serkan appears only in least % for both together.


Btw, just wanted to know what is my %s?

EDIT:Graphs are re-created.

Edited by Rohan Rao 2011-12-30 10:50 AM
@ 2011-12-29 10:02 PM (#6301 - in reply to #6300) (#6301) Top

Administrator



2000100050020
Country : India

Administrator posted @ 2011-12-29 10:02 PM

Rohan Rao - 2011-12-29 9:38 PM
Btw, just wanted to know what is my %s?

@ 2011-12-29 10:10 PM (#6302 - in reply to #6300) (#6302) Top

Administrator



2000100050020
Country : India

Administrator posted @ 2011-12-29 10:10 PM

Rohan Rao - 2011-12-29 9:38 PM

Gotroch!

Also, surprisingly, Serkan appears only in least % for both together.
Something wrong with the numbers. I'll paste the correct numbers later.
@ 2011-12-29 10:16 PM (#6303 - in reply to #6302) (#6303) Top

vopani



Posts: 739
50010010020
Country : India

vopani posted @ 2011-12-29 10:16 PM

Administrator - 2011-12-29 10:10 PM

Rohan Rao - 2011-12-29 9:38 PM

Gotroch!

Also, surprisingly, Serkan appears only in least % for both together.
Something wrong with the numbers. I'll paste the correct numbers later.

It need not be wrong if Serkan has <75 submissions in any one of puzzle/sudoku.
@ 2011-12-29 10:16 PM (#6304 - in reply to #6263) (#6304) Top

prasanna16391



Posts: 1791
100050010010020202020
Country : India

prasanna16391 posted @ 2011-12-29 10:16 PM

Is it too much to ask about my %s too then?

Edited by prasanna16391 2011-12-29 10:16 PM
@ 2011-12-30 5:06 AM (#6305 - in reply to #6298) (#6305) Top

Administrator



2000100050020
Country : India

Administrator posted @ 2011-12-30 5:06 AM

Graphs are uploaded again.


There are only <15 players who have submitted 100+ sudokus. So I reduced the limit to 50 in case of Sudoku.











@ 2011-12-30 5:14 AM (#6306 - in reply to #6304) (#6306) Top

Administrator



2000100050020
Country : India

Administrator posted @ 2011-12-30 5:14 AM

prasanna16391 - 2011-12-29 10:16 PM

Is it too much to ask about my %s too then?



You could be leading the right side table next year if you continue like this
@ 2011-12-30 5:18 AM (#6307 - in reply to #6267) (#6307) Top

Para



Posts: 315
100100100
Country : The Netherlands

Para posted @ 2011-12-30 5:18 AM

debmohanty - 2011-12-27 8:22 PM

Para - 2011-12-27 8:18 PM

Hmmmm, does that mean I was first to report all broken puzzles?

Interestingly, yes.
May be I should ask you to take test always as the first player


I doubt I'd notice them all. I mean I missed one of them while solving and only found out when I realised my answer key was wrong (all my deductions had been correct, it just lead to a problem at the end in a rule I hadn't been using to finish off the puzzle). I also still have a few puzzles I haven't managed to solve (yet).
@ 2011-12-30 12:53 PM (#6310 - in reply to #6306) (#6310) Top

prasanna16391



Posts: 1791
100050010010020202020
Country : India

prasanna16391 posted @ 2011-12-30 12:53 PM

Administrator - 2011-12-30 5:14 AM

prasanna16391 - 2011-12-29 10:16 PM

Is it too much to ask about my %s too then?



You could be leading the right side table next year if you continue like this


hehe luckily I've almost never got something I solved as wrong in either the Nationals or the WCs. I guess its more to do with online submissions, without a printer it can get pretty confusing
@ 2011-12-30 2:02 PM (#6312 - in reply to #6263) (#6312) Top

rakesh_rai




Posts: 774
500100100202020
Country : India

rakesh_rai posted @ 2011-12-30 2:02 PM

Thanks Deb... those numbers are quite interesting.
Could we have an India list also with a lesser number constraint.
@ 2011-12-30 3:25 PM (#6314 - in reply to #6290) (#6314) Top

BohemianCoast



Posts: 16

Country : United Kingdom

BohemianCoast posted @ 2011-12-30 3:25 PM

I wouldn't worry about people participating once or twice and going away. First of all, there are lots of fun things you can do on this site that aren't participating in the timed tests -- like grabbing puzzle sets to go and do at leisure. My kids like the easier puzzles from the sets, but they'd hate the stress and perceived lack of success of the timed tests. You might want to consider whether you have regular visitors who aren't participating in the timed tests.

But also, the tests are really *very difficult*. I'm a pretty experienced puzzler and, with the exception of the Screen Test (which played to my strengths), I'm consistently coming about 3/4 of the way down the field. Lots of people would find that discouraging and go find an easier site. But that's actually the unique thing about this site so you wouldn't want to get rid of it.

You might want to think about occasionally having a set of easier puzzles for novices/kids. Or not; there's plenty of other stuff out there.
@ 2011-12-30 7:46 PM (#6317 - in reply to #6263) (#6317) Top

tamz29



Posts: 225
10010020
Country : Thailand

tamz29 posted @ 2011-12-30 7:46 PM

2nd most incorrect submission percentage... yay?
I remembered rushing my assignment and miscalculating the time I have left to play Puzzle Hybrids and ended up missing it.
Was a shame because Bram's puzzles are really good. Kudos to the Perfect Dozen - you're truly a fan of puzzles!
@ 2011-12-30 7:46 PM (#6318 - in reply to #6312) (#6318) Top

Administrator



2000100050020
Country : India

Administrator posted @ 2011-12-30 7:46 PM

rakesh_rai - 2011-12-30 2:02 PM

Thanks Deb... those numbers are quite interesting.
Could we have an India list also with a lesser number constraint.

I didn't add any filters. Just took top 10 submissions from India, and sorted by the %age correctness.


@ 2011-12-30 10:09 PM (#6320 - in reply to #6263) (#6320) Top

prasanna16391



Posts: 1791
100050010010020202020
Country : India

prasanna16391 posted @ 2011-12-30 10:09 PM

Tejal makes 2 extremes :P I'm equally bad at both. At least there's uniformity.
@ 2011-12-30 10:35 PM (#6321 - in reply to #6263) (#6321) Top

neerajmehrotra



Posts: 328
10010010020
Country : India

neerajmehrotra posted @ 2011-12-30 10:35 PM

Really interesting statistics.....
I am happy to find my name here amongst top Indians......

Edited by neerajmehrotra 2011-12-30 10:40 PM
@ 2011-12-31 5:46 AM (#6323 - in reply to #6317) (#6323) Top

Administrator



2000100050020
Country : India

Administrator posted @ 2011-12-31 5:46 AM

tamz29 - 2011-12-30 7:46 PM

2nd most incorrect submission percentage... yay?
I can also confirm that most number of incorrect submissions are from you (40). I'm not going to put a table for that.

Moving on to the next category, here is some puzzle related stats. The first one is easy to find out from the tests' score page.


@ 2011-12-31 5:48 AM (#6324 - in reply to #6263) (#6324) Top

Administrator



2000100050020
Country : India

Administrator posted @ 2011-12-31 5:48 AM

Most 10 stars ratings received by puzzles (Minimum 5 10-Star ratings)



@ 2011-12-31 5:56 AM (#6325 - in reply to #6263) (#6325) Top

Administrator



2000100050020
Country : India

Administrator posted @ 2011-12-31 5:56 AM

Most 10 stars ratings received by sudokus (Minimum 5 10-Star ratings)



@ 2012-01-01 3:31 PM (#6331 - in reply to #6263) (#6331) Top

Administrator



2000100050020
Country : India

Administrator posted @ 2012-01-01 3:31 PM

Some more stats based on ratings


Here is the list of players who rate most frequently





List of top 3 tests just based on average of puzzle ratings


@ 2012-01-02 1:21 AM (#6332 - in reply to #6293) (#6332) Top

Valezius



Posts: 66
202020
Country : Hungary

Valezius posted @ 2012-01-02 1:21 AM

Administrator - 2011-12-29 1:50 PM

The numbers suggest so. Here is the break up of 1-test participants (sudoku or puzzle) by countries.



Can we see a full nationality list about the 1041 unique players?
@ 2012-01-02 2:32 AM (#6333 - in reply to #6298) (#6333) Top

utkaarsh



Posts: 89
20202020
Country : India

utkaarsh posted @ 2012-01-02 2:32 AM

Yipee I topped a list
But wait isnt it the wrong one

So for 2012, less of mistakes is my aim !!

Edited by utkaarsh 2012-01-02 2:34 AM
@ 2012-01-02 6:24 AM (#6334 - in reply to #6332) (#6334) Top

Administrator



2000100050020
Country : India

Administrator posted @ 2012-01-02 6:24 AM

Valezius - 2012-01-02 1:21 AM

Can we see a full nationality list about the 1041 unique players?

Here we go

CountryTotalNonZeroNewIn2011
India312195181
USA14611683
Poland644533
Japan494326
Germany484118
Italy423021
Turkey372219
France362812
United Kingdom353118
Slovakia323020
Czech Republic21204
China20156
Serbia19166
Thailand14912
Hungary12125
South Korea12811
Netherlands11107
Canada943
Bulgaria992
Finland985
Russia874
Estonia774
Australia776
*UnKnown*534
Romania520
Austria530
Switzerland552
Greece423
Argentina433
Republic of Kosova414
Belgium442
Croatia440
Belarus432
Malaysia331
Bosnia and Herzegovina331
Taiwan330
Spain321
Peru210
Israel201
Uzbekistan221
Macedonia221
Luxembourg210
Denmark220
Indonesia111
Zhongguo111
Mexico111
Portugal111
Lithuania100
Slovenia110
Cyprus111
Sweden110
Malta101
Brazil111
Morocco110
Philippines110



Please note : We don't have a drop down for filling countries. So some of the names below could be duplicates. Please post here if you see any.
@ 2012-01-02 6:46 AM (#6335 - in reply to #6263) (#6335) Top

Administrator



2000100050020
Country : India

Administrator posted @ 2012-01-02 6:46 AM

A related graph - most from any country for any test.

@ 2012-01-02 7:35 PM (#6344 - in reply to #6263) (#6344) Top

Administrator



2000100050020
Country : India

Administrator posted @ 2012-01-02 7:35 PM

Some stats about individual performances



















@ 2012-01-05 10:15 AM (#6359 - in reply to #6263) (#6359) Top

Administrator



2000100050020
Country : India

Administrator posted @ 2012-01-05 10:15 AM

Last Set of stats











@ 2012-06-30 6:15 PM (#7638 - in reply to #6333) (#7638) Top

debmohanty




1000500100100100202020
Country : India

debmohanty posted @ 2012-06-30 6:15 PM

Administrator - 2011-12-30 5:06 AM



utkaarsh - 2012-01-02 2:32 AM

Yipee I topped a list
But wait isnt it the wrong one

So for 2012, less of mistakes is my aim !!


What happpened to your goal Utkaarsh?
@ 2012-06-30 6:33 PM (#7639 - in reply to #7638) (#7639) Top

prasanna16391



Posts: 1791
100050010010020202020
Country : India

prasanna16391 posted @ 2012-06-30 6:33 PM

debmohanty - 2012-06-30 6:15 PM

What happpened to your goal Utkaarsh?


That goal was in itself a wrong submission.

Yay, I'm not on the wrong list! One step forward to the right list.

Just realized thats the same list as the one at the top of the page O.o oops. Now I'm confused.

Edited by prasanna16391 2012-06-30 6:35 PM
@ 2012-07-01 12:41 AM (#7646 - in reply to #6263) (#7646) Top

utkaarsh



Posts: 89
20202020
Country : India

utkaarsh posted @ 2012-07-01 12:41 AM

Thanks PS for rubbing it in !!
i sincerely hope my mistakes take a time out during ISC !!
@ 2012-07-01 12:50 AM (#7647 - in reply to #7646) (#7647) Top

prasanna16391



Posts: 1791
100050010010020202020
Country : India

prasanna16391 posted @ 2012-07-01 12:50 AM

utkaarsh - 2012-07-01 12:41 AM

Thanks PS for rubbing it in !!
i sincerely hope my mistakes take a time out during ISC !!


I'm in the same boat mate :P I correct most of my errors, I solve almost everything twice :D