09 October 2013

Participants in the Q3 2013 PL/SQL Championship


The following players will be invited to participate in the Q3 2013 championship playoff. The number in parentheses after their names are the number of playoffs in which they have already participated.

Congratulations to all listed below on their accomplishment and best of luck in the upcoming competition!

We currently plan to hold the championship on 7 November (again, at a single time world-wide), and are waiting for confirmation from players.

See the FAQ for an explanation of the three ways a player can qualify for the playoff.

Name Rank Qualification Country
swart260 (5)1Top 25Netherlands
Rakesh Dadhich (3)2Top 25India
Oleksiy Varchyn (0)3Top 25Norway
Stelios Vlasopoulos (8)4Top 25Belgium
Ajaykumar Gupta (2)5Top 25Singapore
Janis Baiza (7)6Top 25Latvia
mentzel.iudith (11)7Top 25Israel
Jerry Bull (9)8Top 25United States
Niels Hecker (12)9Top 25Germany
Frank Puechl (2)10Top 25Germany
Siim Kask (11)11Top 25Estonia
Viacheslav Stepanov (10)12Top 25Russia
Justin Cave (9)13Top 25United States
dannyg64 (2)14Top 25United States
Peter Schmidt (4)15Top 25Germany
Andrey Zaytsev (0)16Top 25Russia
Hamid Talebian (0)17Top 25Netherlands
Jeroen Rutte (5)18Top 25Netherlands
Michal Cvan (9)19Top 25Slovakia
Milibor Jovanovic (3)20Top 25Serbia
Zoltan Fulop (5)21Top 25Hungary
Pavel Vorontsov (0)22Top 25Russia
Frank Schmitt (6)23Top 25Germany
_tiki_4_ (4)24Top 25Germany
Ivan Blanarik (6)25Top 25Slovakia
Tony Winn (3)32WildcardAustralia
Giedrius Deveikis (3)38CorrectnessLithuania
Denis Ivin (0)42CorrectnessRussia
Tobias Stark (2)56CorrectnessGermany
Martin_Scholz_Berlin_Berlin (0)57CorrectnessGermany
james su (6)84WildcardCanada
Ingimundur Gudmundsson (1)87CorrectnessNorway
Thierry Poels (5)92CorrectnessBelgium
richb (0)133CorrectnessUnited States
Dan Kiser (4)223CorrectnessUnited States
Telmoc (0)224CorrectnessPortugal
Upendra Motamarri (0)316CorrectnessIndia


14 comments:

  1. Hello Steven,

    I have got 2 questions.
    First one is that I am just wondering, is it possible to rank players by their countries?

    My second question is, can you increase the number of players in championships. Because it is seriously hard for us to join championship.

    Thanks

    ReplyDelete
  2. Rankings by country: I am not sure if this is exactly, what you want but check out:

    Rankings -> Country Rankings

    Increase number of players: well, sure I can do that, but that's the whole point, isn't it? That you have to qualify at a high level of play. Right now up to 45 players can compete. That's already just under 10% of the average daily player count.

    What do you think a better number would be?

    ReplyDelete
  3. Hello Steven,

    I was also thinking for some time about something similar to what the first poster suggested above.

    What about, for example, to allow to qualify in the Playoff ALL the players that fulfill ALL the
    conditions below:

    1. Have played at least 60 quizzes in the quarter.

    2. Have a correctness percentage of at least 90%.

    3. Are ranked in the Top 50% for that quarter.


    This would encourage the players' commitment to playing daily and favorizing correctness upon speed,
    without, however, using an "infinite" playing time".

    I guess that this could add a number of maximum 10 to 20 more players to the Playoff,
    chosen from among the most dedicated ones.

    Thanks a lot & Best Regards,
    Iudith

    ReplyDelete
  4. So is the motivation behind your concerns of "more players in championship" that you want to allow more players who are clearly dedicated (playing a lot) to compete?

    ReplyDelete
  5. Hello Steven,

    The motivation behind my suggestion is two-fold:

    On one side, I can understand the frustration of a player that feels that he will maybe never
    be able to qualify in a Playoff, because of the too limited places.
    If I go 3 years back, I remember when I myself was thinking that I would never be able to do it ...
    so, it could be encouraging for the players to see that not only "half-Gods" can qualify in a Playoff ...

    On the other side, there are many players that maybe even did participate in several Playoffs,
    but however are going to miss another one, in spite of having a very high number of quizzes played,
    a very good correctness percentage, but probably just because of very few quizzes missed in addition
    to a somewhat slower, but still reasonable playing time.

    Once a player reaches a high correction percentage, then each additional quiz answered 100% correctly
    will hardly add 0.1% or 0.2% to his average correctness, so, at the end of a quarter,
    having a correctness percentage 5% higher than another player does have a significance.

    Also, I see many players that, by answering light-fast and making one mistake, do still receive a higher
    score than a player who answered 100% correctly in a longer, though still reasonable time.

    So, I think that a player who did achieve a high enough correctness percentage (in absolute terms also, like for example
    at least 90%, and not just by comparison to other players), along a high number of quizzes played and still
    within a reasonable playing time, maybe does deserve to qualify in a Playoff.

    Of course, any of these qualification criteria may be changed over time, depending on the total number
    of players, total number of players with high correctness percentages, and so on.

    It would be interesting to hear other players' opinion on this issue as well,
    several opinions are always better than fewer ones.

    And, like with anything else, anyone can be right or mistaken.

    Thanks a lot & Best Regards,
    Iudith

    ReplyDelete
    Replies
    1. Hi Iudith,

      Thanks for sharing your thoughts and your kind concern for others who are just on the borderline for qualifying to the quarterly championship.

      I am personally not in favor of relaxing the rule only for high correctness percentage.

      Most of us play the quiz during office hours and it is practically not possible to focus more than 3-5 minutes (To me, even this becomes challenging sometimes). Yes, of course, some of us may have sufficient time to open the quiz, lookup the documents, manually type the code to verify before submitting the answers (This may take additional 10-15 minutes). But, If we relax the rules, more players will follow this strategy and identification of best players (that have knowledge imbibed in their minds alongwith reasonable ability to analyze and correctly apply it) will become more tough.

      Also, in competitive environment, one's ability is measured on the basis of balance between time, knowledge and ability.

      I am playing on PL/SQL challenge since its inception. For most quarters my rank has been between 50 to 100. I have qualified only once for the quarterly championship, but that has not resulted in any grudges. I can understand that there are better PL/SQL professionals in this world who qualify for the quarterly championship.

      Just for my personal situation when I need to compromise correctness over time:

      1) If I could not get the time to during office hours and I know that I will not be able to take quiz after going back home. The only time that I have got is on my way back home in Taxi and the only device is my little Android phone.

      2) Please do not take it offensively but this has happened to me sometimes. Consider one busy day when I have very little time but I open the quiz and pressed "Play Now". The quiz page is opened and the author name is "Mentzel Iudith". I quickly realize that this is going to take me at least 5 minutes to understand the quirks behind the quiz and if time is not permitting I just need to guess the answers. You can validate this by analyzing the average time spend by players on quizzes submitted.

      In summary, yes percentage correctness is important, but so is the time. And, to me, the current method of choosing the qualifiers correctly maintains this balance.

      If we want to relax the rules, we can increase the number in each category (e.g. Top 40 + 15 based on Percentage Correctness + 15 based on Wildcard)

      Thanks,
      Naresh

      Delete
  6. I am going to put up a poll on the Challenge site, asking for feedback on this issue. What do you think of these choices (offer suggestions for changes, removal, additions):

    Leave the rules as they are. A maximum of 45 players is fine.
    Allow the top 50 ranked players to participate.
    Allow the top 100 ranked players to participate.
    All players with a sufficiently high % correct (say >= 90%) and minimum number of quizzes can play.

    ReplyDelete
  7. Hello Steven,

    perhaps you should increase the number of randomly picked playes whose percentage correct is at least as high as the (in pseudo SQL) "Least( 90, 'lowest percentage correct in the top 10') in the quarter. For 2013-Q3 the lowest percentage correct of the top 10 players was 85.4%. So perhaps it would be sufficient to increase the number to 25 or 50 players.

    Kind regards,
    Niels Hecker

    ReplyDelete
  8. Hello Steven,

    I just took the poll, and voted for choice 5, because that seemed to most resemble Naresh's suggestion, which seems the best option to me.
    I also follow the same strategy, I quick scan the quiz as a whole and start guessing when I know I probably will not grasp the essence of a quiz good enough to be fully correct in a reasonable amount of time.
    The good thing about that is that the following day, upon reviewing the results, I will have learned something new in PL/SQL. I play the quiz for the competitive element, but also to learn.

    Kind regards,

    Jeroen

    ReplyDelete
  9. Anonymous said: "Because it is seriously hard for us to join championship."
    Well, yes, it *should* be reasonably hard to qualify, else it would not be a championship :-)

    But one of the reasons why it is *seriously* hard to qualify could be that the challenge has attracted a good number of top notch PL/SQL developers who play regularly and qualify often. Thus the chance is slimmer for players just below this "elite" of regular championship players - these would at present probably mostly hope for a wildcard.

    I used the Player Rankings page and selected Q3 2013 with a filter with the expression "percentile >= 91" (or in other words "in the top 10%") - that is 95 players who got to be in those 10% in different ways.

    One player "only" played 51 quizzes out of 66 but got a great score on those quizzes he played - should he qualify?
    One player "only" got 74.3 percent correct but diligently played all 66 quizzes to accumulate score - should he qualify?

    So in one way I think it should be "hard" to qualify - but the presence of an "elite" group of regular players that often take a good part of the fixed number of seats in the championship can discourage other players as they compete for the remaining few seats. With that in mind it could make sense to broaden the scope of qualification so more get a chance for participating in the championship (and might possibly make the very top ranking players need to "sharpen their wits" when there is more competitors in championship.

    But actually I don't quite like the extra rules that if you did not get high rank, then you can qualify by correctness or wildcard. If the challenge site can support 95 simultaneous players in championship (might consider shutting down regular playing while championship is running?) I would support the simple rule of "percentile >= 91", because the normal ranking algorithm based on weighted score already *is* providing players means of good score either by being fast, or by being correct, or by being diligent and play all quizzes in a quarter. KISS also can be applied here ;-)

    Then we might discuss if the present weighting algorithm weighs speed too high above accuracy or vice versa - that would be a different discussion. But we would only have *one* place for a rule that determines how much accuracy is weighed compared to speed in calculating rank. Rank (percentile) would then be the simple way to determine if you're in the playoff or not.

    If you are not very top-notch in percent correct, you can improve your score by being diligent and play every quiz - if you can gather sufficiently high score to be in the top 10%, your reward will be a chance to pit your wits against the elite in the championship.

    ReplyDelete
  10. Hello Kim, All,

    If you used "percentile >= 91" and got 95 players, I think that this result is a little bit misleading,
    because, though there were 1058 players that played in Q3-2013, the daily average of players
    was around 500, which means that there were many players that only played very few quizzes.

    As a result, your 95 players is much higher than a "real daily" Top10%.

    This does not mean, however, that I oppose the idea of a Playoff with 95 players :):)

    This way, at least, it will not be such a "failure" to rank around the 30-th place ...
    which, as we know, can happen even to a former Playoff winner.


    You say:
    "the normal ranking algorithm based on weighted score already *is* providing players means of good score
    either by being fast, or by being correct, or by being diligent and play all quizzes".

    My suggestion was to use "and" instead of those "or"-s, that is,
    to try to prevent the case in which a player has ALL of the
    "high enough correctness" + "many enough quizzes played" + "reasonably fast",
    but, however, by the current rules he/she did not qualify.


    And, yes ... you are right ... the speed versus accuracy discussion is still one that will probably continue,
    especially in the context of the Playoff.

    Thanks a lot & Best Regards,
    Iudith

    ReplyDelete
  11. It has always been unclear to me why the Wildcard entries are ranked based on weighted score with the top 10 gaining playoff spots, yet the Correctness entries are random. With the Correctness entries, why not do likewise and rank them based on the attribute (Correctness) and take the top 10 for playoff entries. That way it would be consistent. For 2013 Q3, only 1 of the top 10 ranked on Correctness (out of pool of 41 players) actually made the playoff, which seems counter to the purpose of rewarding Correctness. This also removes the non-deterministic nature of this entry method.

    I also like the idea of simply taking the top 45-50 players based on Score, as accuracy and timing are already addressed in the scoring algorithm.

    As a side note, it would be interesting to have easy access to the underlying scoring data in order to determine how strong the correlation is between Period Rank, Correctness, Top30 and the final Playoff Rank. That might provide some insight on how playoff spots should be allocated.


    Chad

    ReplyDelete
  12. Hello Steven, All,

    Just a short time after the Playoff, seeing that only 28 players out of the 37 qualified
    did ultimately play, I still think that there is place for increasing the number of Playoff participants,
    even just a little bit.

    In light of the other discussion about the PL/SQL Challenge, I guess that the Playoff participation
    rules will also be reviewed.

    Thanks a lot & Best Regards,
    Iudith

    ReplyDelete
  13. Good point, I guess. :-)

    37 players confirmed their participation, and a few more started but did not finish (when rankings are published, tho, they WILL be shown).

    ReplyDelete