Wednesday, August 19, 2009

Hockey: An Element of Rizk when trusting CIS stats

Meet Marc-André Element.

He's a Concordia assistant captain who was awarded a Guy Lafleur Award of Merit. His coach calls him "one of the top defensive forwards in CIS hockey" and said "I've witnessed firsthand how his philosophy and ethics influence everyone around him." Stat-wise, last year, Element had five goals and eight assists.

And he only played one game.

Well, at least that's what the CIS would have you believe. Concordia's website has him in 31 games, season and playoffs, but the CIS statistics only have one regular-season game. (EDIT: The OUA's official totals are correct, so it seems something was lost in transition.)

It's not clear what happened until you gather all the stats together and notice that another Stinger's name appears in 55 games--an incredible accomplishment with a 28-game schedule. That someone else is Marc-André Rizk, who not only shares a first name with Element, but also the first digit of their sweater numbers (Rizk is 17, Element is 15).

And then you look a little more and find games like this, where Rizk apparently cloned himself. One version of him had 12 penalty minutes; the other: a goal and an assist. What makes this all the more confusing is that Element is named twice in the penalty summary, with a total of 12 PIM.

So I have some questions: If the scorekeeper was awake enough to record a penalty, twice, for #15, shouldn't there also be an entry for him below? How can a player's penalty be recorded for him but not show up in his stats? Why aren't shots recorded for any players in the OUA? (Wait, that's another topic.) How many other players lost their stats to someone with a similar name?

The CIS says Marc-André Element's official total is one goal in one game. Who knows which game that was...
posted by Rob Pettapiece at 1:40 PM
Labels: , ,

4 comments:

  1. Ridiculous. Good to highlight the fact that statistics at the CIS level can at times be terribly inaccurate or almost totally in complete. More of a focus needs to be put on accuracy - it makes the league look second class when things like this happen, even though the product is first rate. Stats are important for fans to follow players and teams and need to be 100% accurate.
    ReplyDelete
  2. This is due to the CIS using a primitive DOS program called TASHK. for the stats to "line up" between each teams submissions the player needs to be named the exact same thing, right down to spacing,hyphens and capitals. Obviouslly with names like that it makes it difficult. Why the player would only show up as having points/pims but not actually "played" is because for the player to have a "GP" added to his stats the home team must put in all active players at the start of the game. So even though you may have scored the game winner, if your name wasn't checked off earlier you wouldn't of "played". Long story short CIS needs to either train all stats people across the country the right way (which would be difficult because they are mainly students) or get a stats program that is a little more flexible. The third option is to be more like the AUS and have someone do quality control/ fax the stats. The AUS has John Keefe to "crack the whip" and compile the stats properly. Check out the difference in the stats on the CIS website and the AUS website..they were all done with the same software.
    ReplyDelete
  3. I would concur that there are issues with the software, and I can't understand why any database related program would apparently use the entire name as the key field (which of course is subject to typing errors)and not their team and number.

    One other issue with TASHK has to do with overtime. The software (or maybe it is the operator) doesn't recognize 10 minute overtimes as played in the AUS, and often posts the wrong time for the winning goal (often off by 5 minutes).

    I also agree that it is unfortunate that in order to get data integrity someone at the AUS has to double-check and manually correct the game stats. Some of this self-inflicted: teams are under the gun to get the stats uploaded to the CIS and their conference as soon as they can after the end of the game, while team coaches may be still hashing out goal and assist corrections on the official, manual scorekeeper's sheet. These are often the changes faxed or e-mailed after the CIS has already posted game results online, and I don't understand why the CIS doesn't seem to make corrections (or is this the responsibility of the home team?). Attendance is often left off the initial game sheet submission (which I'm not condoning ...), which is why there is such a disparity between the CIS and AUS attendance statistics.

    Fianlly, one issue that drives me batty is that the OUA apparently doesn't track shots consistently for every game (Ok, that's not that big a deal) and doesn't tally plus/minus stats (OK, that to me is a bid deal). I feel plus/minus is an important comparative measure of an individual or team's play, particularly even strength, yet we keep seeing in the scoring leader lists all of these zeros for teh OUA players.
    ReplyDelete
  4. the OUA apparently doesn't track shots consistently for every game

    Actually, when you get over to the CIS side of things, there are no shots recorded for OUA teams whatsoever, throughout all of last year.

    The whole-name-as-key thing is not a problem in and of itself, but any system that requires the scorer to type everyone's name in, rather than populating the scoresheet automatically from the team's roster, is asking for problems.

    Apparently there's a new stats system coming in the fall for the CIS...we'll see how much of an improvement it is.
    ReplyDelete

Subscribe to: Post Comments (Atom)