Page 1 of 1

NAQT Statistics Database

Posted: Thu Aug 21, 2008 4:20 pm
by rhentzel
I'm pleased to announce that a beta version of NAQT's high school statistics database is now available. This is our attempt to record the results of every match that is ever played on our questions (though, as of yet, it is woefully incomplete).

The home page for the site is here:

http://www.naqt.com/stats/

You can see a list of 2007-2008 tournaments here:

http://www.naqt.com/stats/results.jsp?year_code=2007

You can see Elo ratings for the 2007-2008 year here:

http://www.naqt.com/stats/ratings.jsp

You can see a small number of "all-time" records for teams and individuals here:

http://www.naqt.com/stats/records.jsp

We've started to make an effort to clean this data up, but there are probably lots and lots of errors remaining; we'd welcome any information from players and coaches that help us correct those issues.

As a starting point, if you search for your school:

http://www.naqt.com/stats/school-search.jsp

and then click on "Players" in the lower menu list, you can see a list of current and former players for that school's teams. If any of them are actually the same person, please let us know so we can unify their statistics.

If you have results from past tournaments that aren't loaded, please send us the SQBS files or whatever results you do have.

Finally, if there are additional views of this data, records, statistics, or search options that you would like to see, please let us know. We hope to expand this section greatly as time goes on.

Re: NAQT Statistics Database

Posted: Thu Aug 21, 2008 5:08 pm
by Gonzagapuma1
This is really cool, except the Elo Rankings kinda suck, especially when you get out of the top 15.

Re: NAQT Statistics Database

Posted: Thu Aug 21, 2008 5:21 pm
by rhentzel
Gonzagapuma1 wrote:This is really cool, except the Elo Rankings kinda suck, especially when you get out of the top 15.
I might not go so far as to say that they "suck," at least in a public forum, but I agree that there is more work to be done in making the ratings accurate. We welcome suggestions from statisticians and interested players who are browsing the database.

The biggest issue to be tackled is properly handling teams that play very few games, particularly in regions where the competition might be both (1) weak, and (2) only local.

I, personally, am interested in pursuing more complex stats like Pat Matthews' "Points Created" (from way back when), when we have development time.

Re: NAQT Statistics Database

Posted: Thu Aug 21, 2008 5:50 pm
by ClemsonQB
One problem with these stats is that sometimes they aren't correct at all. Neither Dorman A or Dorman B played at the Spartanburg Viking Tournament, yet they have stats from those tournaments. "Dorman B" lost twice at that tournament to teams that would score less than 50 points to the real Dorman B.

Re: NAQT Statistics Database

Posted: Thu Aug 21, 2008 6:02 pm
by rhentzel
ClemsonQB wrote:One problem with these stats is that sometimes they aren't correct at all. Neither Dorman A or Dorman B played at the Spartanburg Viking Tournament, yet they have stats from those tournaments. "Dorman B" lost twice at that tournament to teams that would score less than 50 points to the real Dorman B.
That's certainly true; that's probably the most extreme example (that I know of) in these stats of the general question about when teams are "the same." A school might field three (or two, whatever) of its best players for a given tournament. Should those "A team" stats count toward the overall rating and record of its "true" A team (assuming that such an entity has a clear and uncontroversial existence)?

NAQT has basically decided that whatever teams schools field will be considered to be their A, B, and C teams, regardless of composition. If they aren't named A, B, and C, we'll rank them in order of finish. That is, if Charter Death Star finishes higher than any other Charter team, we'll merge its stats in with Charter A from other tournaments. Without a rule like that, I think, it'll become pretty subjective about whether or not teams are "the same" unless we take the approach that only teams with the exact same players playing the exact same proportion of the time are the "same team."

That said, if I remember correctly, this was a weekend on which Dorman sent teams to multiple tournaments; if that's case, we would probably be willing to relabel Dorman A and B as C an D (or whatever is appropriate) since the tournaments were taking place at the same time. Drop us an e-mail (or have your coach drop us an e-mail) with the details.

This is the sort of thing that we are looking to clean up, particularly for teams as prominent as Dorman.

Re: NAQT Statistics Database

Posted: Thu Aug 21, 2008 6:09 pm
by cdcarter
R, Is it possible to get any sort of API or better way of accessing the data programmatically besides just parsing the pages?

Re: NAQT Statistics Database

Posted: Thu Aug 21, 2008 6:13 pm
by ClemsonQB
At the Spartan High tournament, "Dorman A" consisted of four sophomores that compose the JV A Team, or Dorman C. "Dorman B" was similarly the normal Dorman D. The actual A and B team members read at this tournament to help out an inexperienced local club.

Re: NAQT Statistics Database

Posted: Thu Aug 21, 2008 6:15 pm
by Gonzagapuma1
Yea, we have the same problem. Our B/C Team went 0-5 at 2007 WJ. Our A Team had already played the set at Princeton.

Re: NAQT Statistics Database

Posted: Thu Aug 21, 2008 6:40 pm
by JackGlerum
Very cool.

Re: NAQT Statistics Database

Posted: Thu Aug 21, 2008 6:57 pm
by Sen. Estes Kefauver (D-TN)
I appreciate Joel Knight's top notch performance for Grand Rapids Catholic Central HS.

Re: NAQT Statistics Database

Posted: Sat Aug 23, 2008 9:51 pm
by wowitsquinthaha
I think LOT Academy will surprise everyone and win HSNCT this year.

Re: NAQT Statistics Database

Posted: Wed Sep 24, 2008 7:14 pm
by master15625
Deesy Does It wrote:I appreciate Joel Knight's top notch performance for Grand Rapids Catholic Central HS.
I can't tell if you are purposefully misstating, but Joel Knight played at Detroit Catholic Central HS, right?

Re: NAQT Statistics Database

Posted: Thu Sep 25, 2008 11:55 am
by Maxwell Sniffingwell
master15625 wrote:
Deesy Does It wrote:I appreciate Joel Knight's top notch performance for Grand Rapids Catholic Central HS.
I can't tell if you are purposefully misstating, but Joel Knight played at Detroit Catholic Central HS, right?
Ees choke. Ees fonny, ja?

Re: NAQT Statistics Database

Posted: Thu Sep 25, 2008 12:04 pm
by Stained Diviner
I sent an email a month ago with some corrections and additions, and nothing happened.

Re: NAQT Statistics Database

Posted: Sat Nov 08, 2008 7:15 am
by mastila
can you please tell me the result when A Team had played the set at Princeton?

Re: NAQT Statistics Database

Posted: Wed Feb 04, 2009 12:26 pm
by Charbroil
Is NAQT going to update this as the '08-09 season goes on, or is the database only of past completed seasons?

Re: NAQT Statistics Database

Posted: Wed Feb 04, 2009 3:16 pm
by Byko
Charbroil wrote:Is NAQT going to update this as the '08-09 season goes on, or is the database only of past completed seasons?
I've seen them make updates as the season goes along for tournaments that they get data back from. As you can see, though, there are several events that they don't get any response back from.

Mostly, it's a pretty good repository. When something gets mis-entered or appears to be mis-entered (see TOMCAT from this year, where I can't really figure out anything as to what's going on), it becomes difficult.

Re: NAQT Statistics Database

Posted: Wed Feb 04, 2009 3:56 pm
by Sen. Estes Kefauver (D-TN)
I know I sent them the data for Mizzou's Fall Academic Tournament but they haven't entered them yet, possibly becuase we didn't run the stats as SQBS. I think it would be kind of cool if they got more on entering stats from tournaments that have their results posted but that are sitting blankly on the NAQT page, but whatever.

Re: NAQT Statistics Database

Posted: Wed Feb 04, 2009 4:11 pm
by rhentzel
Charbroil wrote:Is NAQT going to update this as the '08-09 season goes on, or is the database only of past completed seasons?
This intent is to update it as soon as possible with whatever results we have, but we haven't been able to completely live up to that.

Some of that is because loading results requires database access and the few people at NAQT that have permission to do that are also working, more-or-less continuously, on higher priority tasks (like finishing packets).

The bigger issue is that not all hosts send results, many send incomplete results, and many send results that are in an inconvenient format. For example, a spreadsheet of teams' scores in each round but no indication of which teams were playing. We then have to request a schedule and sort things out manually to enter the data.

I hope that we will become more efficient with this as time goes on and eventually reach a point where every tournament is promptly reported to us and available in the database within a week of it's completion (and hopefully faster). It will take us a while to get there, however.

I currently have quite a few tournaments that will get uploaded as soon as Sectionals are over and IS #83 is completed and shipped.

Re: NAQT Statistics Database

Posted: Wed Feb 04, 2009 4:14 pm
by rhentzel
Byko wrote:
Charbroil wrote:Is NAQT going to update this as the '08-09 season goes on, or is the database only of past completed seasons?
I've seen them make updates as the season goes along for tournaments that they get data back from. As you can see, though, there are several events that they don't get any response back from.

Mostly, it's a pretty good repository. When something gets mis-entered or appears to be mis-entered (see TOMCAT from this year, where I can't really figure out anything as to what's going on), it becomes difficult.
The listed final rankings for TOMCAT teams outside of the top 12 are currently meaningless due to the powermatching format used. I'll be cleaning that up at some point when question production needs are less pressing.

Re: NAQT Statistics Database

Posted: Wed Feb 04, 2009 4:15 pm
by rhentzel
Jeremy Gibbs Free Energy wrote:I know I sent them the data for Mizzou's Fall Academic Tournament but they haven't entered them yet, possibly becuase we didn't run the stats as SQBS. I think it would be kind of cool if they got more on entering stats from tournaments that have their results posted but that are sitting blankly on the NAQT page, but whatever.
Getting sent a SQBS data file is the easiest thing for us to process; I typically upload those immediately when I receive them.

Other formats usually have to manually rearranged into one or another standard format, which takes time. If we have your results, they will get posted eventually. I do apologize for the delay and hope that we become more efficient at keeping the database up to date.

Re: NAQT Statistics Database

Posted: Wed Feb 04, 2009 6:51 pm
by Byko
rhentzel wrote:
Byko wrote:
Charbroil wrote:Is NAQT going to update this as the '08-09 season goes on, or is the database only of past completed seasons?
I've seen them make updates as the season goes along for tournaments that they get data back from. As you can see, though, there are several events that they don't get any response back from.

Mostly, it's a pretty good repository. When something gets mis-entered or appears to be mis-entered (see TOMCAT from this year, where I can't really figure out anything as to what's going on), it becomes difficult.
The listed final rankings for TOMCAT teams outside of the top 12 are currently meaningless due to the powermatching format used. I'll be cleaning that up at some point when question production needs are less pressing.
No problem--I understand. It just confused me to no end to see some teams playing 6 prelim games and some playing 7, which seemed to differ from years past with TOMCAT. Plus seeing Chaska win the tournament playing 3 playoff games while Eden Prairie played 4 and Wayzata going 3-0 in playoffs but finishing third--I had to make sure my head didn't explode. I'm guessing the semifinals are just missing or something (since nobody appears to have played in round 11), but I just thought you'd want to know. But I can be patient with this.

Re: NAQT Statistics Database

Posted: Wed Feb 04, 2009 8:43 pm
by rhentzel
Byko wrote:No problem--I understand. It just confused me to no end to see some teams playing 6 prelim games and some playing 7, which seemed to differ from years past with TOMCAT. Plus seeing Chaska win the tournament playing 3 playoff games while Eden Prairie played 4 and Wayzata going 3-0 in playoffs but finishing third--I had to make sure my head didn't explode. I'm guessing the semifinals are just missing or something (since nobody appears to have played in round 11), but I just thought you'd want to know. But I can be patient with this.
In 2007 there were 64 teams so everybody played six preliminary games. In 2008 (and 2009) there were 80 teams, so some played six and some played seven as part of the powermatched preliminaries.

Some 11th-round games (and some other games, mostly forfeits) are missing from the currently posted stats. I'll have to sort through the scoresheets to identify those and enter them.

Re: NAQT Statistics Database

Posted: Wed Feb 04, 2009 10:21 pm
by Charbroil
Okay, thanks. I'll go see if anyone I know has results for NAQT tournaments that they can send in.

Re: NAQT Statistics Database

Posted: Thu Feb 05, 2009 12:27 am
by BuzzerZen
Speaking of SQBS files, if anyone out there happens to have a specification of the SQBS data format, I'd quite probably attempt to implement an export-to-SQBS-file feature for Taft.