10 March 2014

Results of 2013 Q4 PL/SQL Championship

You will find below the rankings for the 2013 Q4 PL/SQL championship; the number next to the player's name is the number of times that player has participated in a championship.

Congratulations first and foremost to our top-ranked players:

1st Place: Peter Auer of Germany

2nd Place: Mike Pargeter of United Kingdom

3rd Place: Frank Schmitt of Germany


This was Peter's first championship. An impressive rookie performance, Peter!

Next, congratulations to everyone who played in the championship. I hope you found it entertaining, challenging and educational. And for those who were not able to participate in the championship, you can take the quizzes next week through the Practice feature. We will also make the championship as a whole available as a Test, so you can take it just like these players did.

My thanks to all my fine reviewers, but most particularly Elic, whose impact on the quality of our daily quizzes and championships cannot be overstated.

Steven Feuerstein

Note: Below the table of results for this championship, you will find another list showing the championship history of each of these players.

Rank Name Country Total Time % Correct Total Score
1Peter Auer (1)Germany27 mins 25 secs92%2882
2Mike Pargeter (12)United Kingdom22 mins 06 secs87%2788
3Frank Schmitt (8)Germany30 mins 05 secs87%2723
4Karel Prech (3)Czech Republic32 mins 25 secs87%2652
5Michal Cvan (11)Slovakia33 mins 58 secs87%2551
6Zoltan Fulop (7)Hungary29 mins 14 secs82%2485
7Janis Baiza (9)Latvia22 mins 35 secs76%2443
8Jerry Bull (11)United States29 mins 20 secs74%2443
9Ravshan Abbasov (1)Uzbekistan33 mins 11 secs79%2441
10Anna Onishchuk (8)Ireland16 mins 40 secs79%2432
11dmitrysk (1)Russia32 mins 52 secs84%2398
12Manfred Kleander (1)Austria34 mins 52 secs84%2393
13Vincent Malgrat (7)French Republic32 mins 35 secs79%2388
14Lukasz Kubicki (1)Poland22 mins 54 secs71%2362
15james su (8)Canada21 mins 52 secs71%2348
16Chad Lee (10)United States33 mins 00 secs84%2325
17Jeroen Rutte (7)Netherlands34 mins 46 secs82%2320
18Frank Puechl (4)Germany20 mins 50 secs74%2298
19Rytis Budreika (1)Lithuania18 mins 20 secs68%2253
20mentzel.iudith (13)Israel33 mins 13 secs84%2251
21João Barreto (4)Portugal22 mins 17 secs74%2214
22Telmoc (2)Portugal20 mins 30 secs66%2190
23Randy Gettman (12)United States34 mins 52 secs71%2183
24Yuri Pedan (4)Ukraine17 mins 06 secs71%2163
25Viacheslav Stepanov (12)Russia31 mins 00 secs76%2145
26Krzysztof Helbin (3)Poland26 mins 08 secs68%2112
27Anil Jha (3)United States34 mins 56 secs76%2101
28Kevan Gelling (9)Isle of Man30 mins 37 secs68%2078
29Siim Kask (13)Estonia27 mins 48 secs63%2054
30Leszek Grudzień (2)Poland30 mins 20 secs66%1998
31Yuan Tschang (8)United States34 mins 24 secs55%1972
32Niels Hecker (14)Germany35 mins 00 secs71%1960
33Oleksiy Varchyn (2)Norway21 mins 32 secs61%1919
34_tiki_4_ (6)Germany25 mins 06 secs58%1678
35Goran Stefanović (5)Serbia24 mins 33 secs61%1544
36Rakesh Dadhich (5)India26 mins 50 secs55%1443
37Stelios Vlasopoulos (10)Belgium24 mins 13 secs50%1406
38Thierry Poels (7)Belgium28 mins 05 secs58%1343
39Milibor Jovanovic (5)Serbia32 mins 24 secs53%1272
40Dan Kiser (6)United States33 mins 38 secs50%897

Championship Performance History

After each name, the quarter in which he or she played, and the ranking in that championship.
Name History
Peter AuerQ4 2013:1st
Mike PargeterQ4 2010:22nd, Q1 2011:15th, Q2 2011:8th, Q4 2011:5th, Q1 2012:5th, Q2 2012:17th, Q3 2012:5th, Q4 2012:20th, Q2 2013:12th, Q4 2013:2nd
Frank SchmittQ4 2011:22nd, Q2 2012:2nd, Q3 2012:21st, Q4 2012:3rd, Q1 2013:13th, Q2 2013:1st, Q3 2013:9th, Q4 2013:3rd
Karel PrechQ3 2012:36th, Q4 2012:23rd, Q4 2013:4th
Michal CvanQ3 2010:22nd, Q4 2010:25th, Q3 2011:21st, Q1 2012:11th, Q3 2012:14th, Q4 2012:14th, Q1 2013:14th, Q2 2013:11th, Q3 2013:1st, Q4 2013:5th
Zoltan FulopQ1 2012:15th, Q2 2012:26th, Q3 2012:22nd, Q4 2012:17th, Q1 2013:22nd, Q3 2013:22nd, Q4 2013:6th
Janis BaizaQ2 2010:3rd, Q4 2010:7th, Q3 2011:9th, Q4 2011:1st, Q3 2012:23rd, Q1 2013:4th, Q2 2013:4th, Q3 2013:3rd, Q4 2013:7th
Jerry BullQ2 2011:32nd, Q3 2011:8th, Q1 2012:12th, Q2 2012:11th, Q3 2012:13th, Q4 2012:15th, Q1 2013:9th, Q2 2013:15th, Q4 2013:8th
Ravshan AbbasovQ4 2013:9th
Anna OnishchukQ1 2011:5th, Q2 2011:20th, Q3 2011:19th, Q4 2011:17th, Q1 2012:21st, Q2 2012:8th, Q4 2013:10th
dmitryskQ4 2013:11th
Manfred KleanderQ4 2013:12th
Vincent MalgratQ4 2011:9th, Q1 2012:13th, Q2 2012:13th, Q4 2012:6th, Q1 2013:1st, Q2 2013:14th, Q4 2013:13th
Lukasz KubickiQ4 2013:14th
james suQ3 2010:56th, Q2 2011:7th, Q3 2011:11th, Q4 2011:14th, Q2 2012:15th, Q1 2013:7th, Q3 2013:15th, Q4 2013:15th
Chad LeeQ2 2011:26th, Q3 2011:17th, Q4 2011:12th, Q1 2012:1st, Q2 2012:23rd, Q3 2012:28th, Q4 2012:22nd, Q1 2013:20th, Q2 2013:8th, Q4 2013:16th
Jeroen RutteQ3 2010:19th, Q3 2012:10th, Q4 2012:11th, Q1 2013:8th, Q2 2013:18th, Q3 2013:4th, Q4 2013:17th
Frank PuechlQ3 2012:26th, Q1 2013:15th, Q3 2013:12th, Q4 2013:18th
Rytis BudreikaQ4 2013:19th
mentzel.iudithQ4 2010:4th, Q1 2011:17th, Q2 2011:23rd, Q3 2011:5th, Q4 2011:4th, Q1 2012:7th, Q2 2012:16th, Q3 2012:33rd, Q4 2012:4th, Q1 2013:2nd, Q2 2013:7th, Q3 2013:2nd, Q4 2013:20th
João BarretoQ3 2010:18th, Q4 2010:21st, Q2 2011:3rd, Q4 2013:21st
TelmocQ4 2013:22nd
Randy GettmanQ3 2010:8th, Q1 2011:25th, Q2 2011:10th, Q3 2011:3rd, Q4 2011:11th, Q1 2012:18th, Q2 2012:19th, Q3 2012:30th, Q4 2012:12th, Q1 2013:3rd, Q2 2013:13th, Q4 2013:23rd
Yuri PedanQ3 2010:45th, Q4 2010:9th, Q1 2011:2nd, Q4 2013:24th
Viacheslav StepanovQ1 2011:8th, Q2 2011:4th, Q3 2011:13th, Q4 2011:19th, Q1 2012:17th, Q2 2012:9th, Q3 2012:6th, Q4 2012:21st, Q1 2013:18th, Q2 2013:22nd, Q3 2013:11th, Q4 2013:25th
Krzysztof HelbinQ1 2012:10th, Q4 2012:18th, Q4 2013:26th
Anil JhaQ2 2012:28th, Q4 2012:19th, Q4 2013:27th
Kevan GellingQ2 2011:19th, Q3 2011:1st, Q4 2011:6th, Q1 2012:6th, Q2 2012:22nd, Q4 2012:16th, Q1 2013:12th, Q4 2013:28th
Siim KaskQ1 2011:28th, Q2 2011:6th, Q3 2011:10th, Q4 2011:3rd, Q1 2012:8th, Q2 2012:4th, Q3 2012:31st, Q4 2012:10th, Q1 2013:11th, Q2 2013:16th, Q3 2013:17th, Q4 2013:29th
Leszek GrudzieńQ2 2013:24th, Q4 2013:30th
Yuan TschangQ2 2012:24th, Q3 2012:24th, Q4 2012:27th, Q2 2013:30th, Q4 2013:31st
Niels HeckerQ2 2010:2nd, Q3 2010:1st, Q4 2010:15th, Q1 2011:6th, Q3 2011:7th, Q4 2011:10th, Q1 2012:2nd, Q2 2012:3rd, Q3 2012:2nd, Q4 2012:7th, Q1 2013:6th, Q2 2013:3rd, Q3 2013:13th, Q4 2013:32nd
Oleksiy VarchynQ3 2013:28th, Q4 2013:33rd
_tiki_4_Q4 2011:26th, Q1 2012:16th, Q2 2012:20th, Q4 2012:8th, Q3 2013:8th, Q4 2013:34th
Goran StefanovićQ1 2012:31st, Q2 2012:27th, Q3 2012:34th, Q4 2013:35th
Rakesh DadhichQ2 2012:10th, Q1 2013:23rd, Q2 2013:28th, Q3 2013:27th, Q4 2013:36th
Stelios VlasopoulosQ4 2010:37th, Q4 2011:20th, Q1 2012:27th, Q2 2012:30th, Q3 2012:1st, Q4 2012:1st, Q1 2013:27th, Q2 2013:23rd, Q3 2013:19th, Q4 2013:37th
Thierry PoelsQ3 2011:22nd, Q1 2012:24th, Q1 2013:26th, Q3 2013:25th, Q4 2013:38th
Milibor JovanovicQ4 2012:29th, Q1 2013:30th, Q3 2013:18th, Q4 2013:39th
Dan KiserQ4 2013:40th

17 comments:

  1. Hi, Steven!

    It is quite strange for me:
    12 Manfred Kleander (1) Austria 34 mins 52 secs 84% 2393
    20 mentzel.iudith (13) Israel 33 mins 13 secs 84% 2251

    The correctness of 2 players is the same, but the one who took more time to answer is placed 8 positions higher than the second one who took less time?

    ReplyDelete
  2. Ravshan, you have a sharp eye. The score you see above and on the rankings page is actually the weighted score, which applies a formula involving time, to the "base" score, which is not shown.

    If, however, you drill down to this championship via View All Past Competitions (bottom of Take a Quiz on home page), you will see that Manfred has a base score of 618, while Iudith's is 583. This is not a big gap; the #1 winner has a base score of 686.

    So your question really comes down to: how can the base score be different with the same %correct?

    And the answer is that the total score is a sum of scores of individual quizzes, and the penalty paid for a wrong choice varies depending on how many choices there are in that quiz. So likely Manfred was wrong more on quizzes with lots of choices, while Iudith was wrong on quizzes with fewer choices.

    Ah, the joy of algorithms!

    ReplyDelete
  3. Oh, now it is more obvious for me. Thanks for the clarification, Steven!

    ReplyDelete
  4. Hello Steven, Ravshan, All,

    If we look at the ranking list, something might look very strange:

    Players with correctness of 70 to 80% are "spread" from Rank 7 to Rank 32,
    which means 25 ranking positions out of the 40 participants.

    I think that this also somehow suggests that the playing time still "weighs" too much
    in comparison with the correctness.

    These two factors are also not completely independent, in the sense that the playing time being limited,
    you should play faster for making sure that you will be at least able to see ALL the quizzes
    before the time expires ... and doing things faster normally means being more error prone,
    we don't even need a Playoff to acknowledge this, real life proves it almost every day.

    I completely agree that those who are able to be both fast and correct at the same time should definitely
    rank higher ... but, especially in the Playoff, a relatively low correctness most probably means
    lack of time (mainly for becoming aware to all the small details and differences between the often
    very long choices) rather than lack of knowledge.


    I am pretty sure that, if instead of having to "decipher" so long pieces of code, the very same knowledge
    would have been used for asking the very same question, but formulated "in words only",
    probably many players would have achieved a much higher correctness percentage.

    For me it is still very pleasant to remember the first 3 Playoffs, back in 2010,
    with those much shorter and much more focused quizzes.

    It's true that the world becomes more and more complex with the time, but, unfortunately,
    the human brain still has some limitations regarding what is it able "to consciously assimilate"
    in a very short time.

    Thanks a lot & Best Regards,
    Iudith

    ReplyDelete
  5. From Viacheslav Stepanov:

    Hello All,

    I absolutely agree with Iudith. Playoff time weighs too much.

    I would propose one or all of the following:

    1) Value correctness more than time, use time to resolve ties only.
    2) Make quizzes shorter, more focused and reduce time. For example 10 quizzes, 20 minutes.

    Anyway time should not matter too much. There is nothing great in say 5 minutes and 50% correctness. You could click randomly and get similiar results. But now this result could outperform 80% and 35 minutes.

    Best regards,
    Viacheslav Stepanov

    ReplyDelete
  6. And do you think this different approach should also be used for SQL quizzes and any other championships?

    ReplyDelete
  7. My opinion - for all playoffs correctness should take precedence over time

    Viacheslav

    ReplyDelete
  8. I do strongly agree with Viacheslav, time should be used to solve the base score ties.

    Although, Q4 2013 was my first play-off (I've registered in 2011), I have disappointed to see among questions the one on the feature I have never used before. It was Quiz #4 with a very long setup code, and (omg!) 14 choices! Moreover, about one third of the choices were long like a standalone programs and they were cumulative, meaning that you had to decide not only the correctness of the code, but should decide if it has to be applied. Anyway I'm happy I had a chance to participate in the play-off.

    I assume I might be wrong and the rest of the world uses actively RLS/VPD, but my suggestion, in general, is to avoid (or at least try to avoid) very rare to use features for the play-offs.

    ReplyDelete
  9. Hello All,

    We once discussed in the past various variants of SELECT statements for ordering the Playoff results
    for ranking.
    Mainly, the idea was to allow the playing time to be relevant only when the correctness
    does not differ by more than "N%".

    What Viacheslav does suggest, namely the 10 quizzes in 20 minutes, is exactly the Playoff format
    used in 2010, in the first 3 Playoffs.

    I agree that any decision regarding Playoff rules will have its pros and cons.

    It is hard to suggest an ideal format, that will cover in the best possible way ALL the aspects
    involved: number of quizzes, playing time, length of quizzes, topics to be touched or not,
    ranking algorithms, and so on.

    After many playoffs played, however, I can definitely say that for me, at least,
    the main problem was always that of the too long choices / too much code for the given time.

    In my opinion, as I have always expressed it, the variant of 10 shorter quizzes has the advantage
    of covering more topics and affording a better chance to knowledge to prevail upon the
    "fast eye debugging of long code" capability.
    For allowing knowledge to be relevant, you should fully comprehend the problem,
    and, for me at least, this is very difficult to do when you have to assimilate (or deduce)
    a full scenario by just looking at code and required output.

    We all know that real life sometimes does require such abilities from a developer,
    mainly for correcting bugs,but, even in those cases, the right way to go is to reverse-engineer
    the problem, before attempting to correct a buggy solution.
    Otherwise, chances are that new bugs will just replace the old ones.


    Regarding the other playoffs, it is hard to say yet, because we all have less experience.

    However, based on my personal feeling only, considering quizzes complexity versus playing time,
    the SQL playoff looked to me very well balanced, while the Logic playoff was far more difficult,
    and, in spite of the longest time of all, there was a certain moment when you could not do
    anything better but start to guess.

    Thanks a lot & Best Regards,
    Iudith

    ReplyDelete
  10. I also think correctness is more important then time

    ReplyDelete
  11. I think the timing should have less weight than it does now.
    Personally I would prefer shorter questions as by the time I reach the last line of the quiz, I don't remember the first line anymore and I lose concentration and very often misclick the answer as I have to scroll backward and forward. But it's my personal opinion. Always glad to participate in Play offs though. Makes my quarter. :)

    Best Regards

    Anna Onishchuk

    ReplyDelete
  12. With a set of Advanced quizzes, as we get for all championships, I think the different base score for the same % correct is as much of an issue as having too much weight given to time. My suggestion for ranking championships would be to have the primary ranking based on number of choices correct (what is already displayed as % correct), with the time taken purely being a tiebreaker.

    To me, championships of 10 presumably advanced quizzes in 20 minutes is forcing the time issue to affect the % correct, rather than encouraging players to improve their % correct. They are vague recollections, but I thought that was part of the reason why the format changed.

    On the other championships, I thought that the SQL championship was good, but the Logic one needed to use significantly harder quiz formats than the weekly ones because there are the generators for standard quiz format.

    Thanks,

    Tony

    ReplyDelete
  13. I think we are loosing the focus about the real thing here. The tests with predefined answers have ONLY one advantage: they are very easy to be evaluated. The problem is that any test taker has no chance to express its creativity and way of thinking. For this reason I think that putting a higher or a lower weight to the time is quite unimportant. They are already too artificial to look like "real life" does.

    ReplyDelete
  14. Hello all,

    I've just re-read one of the older threads on the same topic to refresh my memory
    ( see http://plsql-challenge.blogspot.de/2012/11/change-scoring-algorithm-for-playoffs.html ).

    Since I've never played one of the older playofss (10 questions in 20 minutes? Sounds ... interesting) , I can't comment
    on whether the old system was better or worse than the one we're currently using.

    However, regarding the proposal to use time only as a tie-breaker between players with same % correctness,
    I don't think that's the right approach; to quote Chris Saxon from the thread mentioned before:

    "I think that the scores should still be weighted. If one player manages to get 90-95% in half the time of a player who
    got 100% I would rate the first as "better". I can't decide how low the quicker player's correctness % would need to be
    before I thought the second better, but it would probably be in the 80s."

    It might make sense to decrease the weight of time compared to % correctness, though.

    Nevertheless, there are indeed things that IMHO could be improved:
    - don't use excessively long questions in the playoffs. At least one of the questions in the last playoff falls into this
    category
    - if using shorter questions, perhaps increase the number of questions (as apparently was the case
    in the old playoffs). Of course, this might not be viable, because it causes additional work for Steven and the
    other people who write the championship quizzes.
    - it always bothered me that players having the same % correctness might or might not be ranked according to time.
    The scoring of the questions / answers should me modified, so that equal % correctness = equal base score.
    I find results like this:

    9 Ravshan Abbasov (1) Uzbekistan 33 mins 11 secs 79% 2441
    10 Anna Onishchuk (8) Ireland 16 mins 40 secs 79% 2432

    where two players have the same % correctness and the player who took twice as long (!) gets a higher rank
    very, very confusing.

    Just my 0.02€
    Frank

    ReplyDelete
  15. Hello All,

    Just to add 1 more penny to those of Frank above:

    I think that the "equal base score" can be achieved by ranking ALL the choices of ALL the quizzes
    together, that is, as if the entire competition contained "one single big quiz", with all the choices
    belonging to it.

    As Tony said above, all the quizzes being Advanced level, this looks suitable.

    An objection might be raised that some choices may look much simpler in comparison with others.

    I think that following Frank's advise of not using excessively long quizzes,
    will also render all the choices as "comparable", and then the "scoring together" may work.

    By the way, in the 10 quizzes variant Playoff, I remember that part of the quizzes used to be Intermediate level. Anyway, the 2 minute per quiz somehow automatically excluded the excessively long quizzes anyway.

    Thanks a lot & Best Regards,
    Iudith


    ReplyDelete
  16. Somehow this reminds me of the discussions we have had in my company about bonus calculations for sales people. 8 people involved in the decision having at least 10 different opinions and arguments about "what is fair". We managed to code a dynamic system that allowed manager to setup rules that could handle 6 different "fairness" parameters. But noone could figure out why their bonus was such-and-such and support was swamped trying to explain algorithms. So it was soon stopped and replaced by a very much simpler algorithm - might be a few "unfair" bonuses here and there, but people can understand it.

    For the quizzes in the weighted-score calculation it has been attempted to include multiple parameters to be "fair." Naturally that makes it more difficult to understand - just as our bonus system ;-)

    So we score each quiz according to how many correctly marked choices out of the number of choices in the quiz. Correctness obviously is most important - fairness parameter 1 :-)

    Now should a correctly marked choice in a quiz with 4 longish choices count as much as a correctly marked choice with 14 short choices? What is most fair? Fairness parameter 2 here is that each quiz totally counts equally, so in a 5 quiz championship a correct answer in the first case is worth 1/4 * 1/5 of total base score, while in the second case a correct answer is worth 1/14 * 1/5. This is the reason why two players can have same % correct choices but different base scores - if one of the players had two wrong choices in the 4-choice quiz and the other had two wrong choices in the 14-choice quiz, for example.

    And then fairness parameter 3 is weighing the base scores according to time used - which is meant to give higher scores to players who can quickly answer correctly as they probably know their stuff and haven't been spending time googling it. (This parameter can be argued is more important for the daily quiz and less for the championship, as the championship is time limited anyway.)

    All this seemingly complex calculation has evolved over time as response to players crying "unfair" and then some rule was invented to try to address each issue and make it "fair". But the price of the fairness is complexity and making it hard to understand why one player gets higher rank than another. You can't have both - you must choose a suitable middle ground thats reasonably fair without being too complex.

    My personal opinion is I haven't heard any new arguments that warrants much change to the "middle ground" that is the current set of rules. Only thing I would consider is that for championships specifically I can understand that time factor could be decreased as the championship already has a timelimit to preclude too much googling.

    ReplyDelete
  17. Hello Kim,

    thanks for the explanation. Might I then suggest that - additional to the overall % correctness - the rankings display the "weighted % correctness" (for lack of a better term) / % correctness for each individual quiz? I guess this would make the ranking process much more transparent.

    Kind regards,
    Frank

    ReplyDelete