The Quizbowl Resource Center

DumbJaques wrote:I'm sure there exists a team in the Northeast that could supply someone who knows how to do stats, but really, it's not very hard. If you have 15 minutes to spend on IM/the phone, any number of people (including myself) would happily teach you everything there is to know about doing stats.

In fact, there is a fantastic guide to keeping stats right here.

Waterworks wrote:Brandeis would not mind hosting the tournament for New England, as long as we didn't have to do stats (or if we received extensive, extensive help with stats).

User was warned for having a signature without a name.

People have taught themselves how to do stats. It requires 1) downloading SQBS, 2) having someone mess around with it for half an hour.

Stats becomes really easy when you have scorekeepers who follow directions, and is super easy for college tournaments, because you don't have team with 10 players who rotates 3X a match.

I'll take this opportunity to say that online stats are awesome as no scorekeepers are required and it means moderators don't have to go back to the war room for score sheets. That was one kick ass thing Duke did for their Early Collegiate Novice site. Other TD's take note!

Zip Zap Rap Pants wrote:I'll take this opportunity to say that online stats are awesome as no scorekeepers are required and it means moderators don't have to go back to the war room for score sheets. That was one kick ass thing Duke did for their Early Collegiate Novice site. Other TD's take note!

You mean online scoresheets? The U of I has done this several times. It hinges on having a stable wireless connection that every room can connect to, and you better have a back-up plan.

That being said, it is the only major solution I have found for the problem of scorekeepers (I can name names) who cannot account for negs/powers in the total score.

PS: You can learn the basics in 30 minute or less, but you will learn more by doing it for more tournaments.

Zip Zap Rap Pants wrote:I'll take this opportunity to say that online stats are awesome as no scorekeepers are required and it means moderators don't have to go back to the war room for score sheets. That was one kick ass thing Duke did for their Early Collegiate Novice site. Other TD's take note!

online stat sheets are awesome, if you have two laptops, and a scorekeeper in every single room. If any of your moderator has to scorekeep, it is tough. But since your saying moderators don't have to go back to the war-room, I will tell you that a lot of moderators out there would rather walk back to the war-room than have to flip to excel after every question.

nalin wrote:

Zip Zap Rap Pants wrote:I'll take this opportunity to say that online stats are awesome as no scorekeepers are required and it means moderators don't have to go back to the war room for score sheets. That was one kick ass thing Duke did for their Early Collegiate Novice site. Other TD's take note!

online stat sheets are awesome, if you have two laptops, and a scorekeeper in every single room. If any of your moderator has to scorekeep, it is tough. But since your saying moderators don't have to go back to the war-room, I will tell you that a lot of moderators out there would rather walk back to the war-room than have to flip to excel after every question.

Psst, moderators need to suck it up, clicking a couple of times after each bonus doesn't seem like a big price to pay for saving so many resources and so much time. Besides we all know there are at least a few moderators at most tournaments who have to keep score on a score sheet as well, which is more of a pain. Granted some time is lost in the process of doing all that data entry, but that's less than the time to get a new scoresheet and also probably means less time going over disputed scores at halftime/full time (in my experience scores seem to be more accurate when recorded by the same person rewarding the points/are less likely to be questioned).

Zip Zap Rap Pants wrote:Psst, moderators need to suck it up, clicking a couple of times after each bonus doesn't seem like a big price to pay for saving so many resources and so much time. Besides we all know there are at least a few moderators at most tournaments who have to keep score on a score sheet as well, which is more of a pain. Granted some time is lost in the process of doing all that data entry, but that's less than the time to get a new scoresheet and also probably means less time going over disputed scores at halftime/full time (in my experience scores seem to be more accurate when recorded by the same person rewarding the points/are less likely to be questioned).

I prefer the use of electronic scoresheets because numbers input into Excel are much easier to read than the numbers written down by scorekeepers.

Also, moderators who keep score are more likely to make math errors, because they are trying to do 2-3 things at once.

From my experience, scorekeepers should not hesitate if they think they messed something up. Such errors should be resolved in the game room, and not by someone entering stats who has no clue what happened in the game. Just from last weekend, I lost count of the number of scoresheets with various errors.

Zip Zap Rap Pants wrote:
Psst, moderators need to suck it up, clicking a couple of times after each bonus doesn't seem like a big price to pay for saving so many resources and so much time.

This would take more time than the current standard, even if it does save paper.

Besides we all know there are at least a few moderators at most tournaments who have to keep score on a score sheet as well, which is more of a pain. Granted some time is lost in the process of doing all that data entry, but that's less than the time to get a new scoresheet and also probably means less time going over disputed scores at halftime/full time (in my experience scores seem to be more accurate when recorded by the same person rewarding the points/are less likely to be questioned).

No, the delay from having to go to a different program to put in a score change, and then having to email scoresheets (if and only if you have everyone on a dependable wireless connection) in order to do any progress on SQBS in the time it takes a round to happen. Also most score discrepancies in my experience are from someone writing down the wrong number or not writing down a number at all, so yeah this wouldn't help with that either. Basically doing these and the packet on a single monitor would be more hassle than its worth.

Have you actually witnessed this being done? In your experience did it take more time? In mine at least, it was the tournament with the fastest transition from one round to the next that I've ever seen.

Zip Zap Rap Pants wrote:Have you actually witnessed this being done? In your experience did it take more time? In mine at least, it was the tournament with the fastest transition from one round to the next that I've ever seen.

We use these in practice from time to time. I don't trust our wireless network enough to even attempt this at a tournament. Also this takes away every tournament director's favorite method of getting everyone on the same page: holding the staff until everyone is there. I can't imagine blindly placing my trust in a tournament functioning on the reliability of a) one person to be doing 3 different things at once constantly or b) a college wireless network actually functioning for an entire tournament.

Using Avram's gimranov stats software has worked wonders for UCLA's tournaments, we've been able to run tournaments with no dedicated stats person since stats can be entered from each individual computer, and teams have loved it because it means stats are available online immediately. We still use paper scoresheets, but at the end of each game, each moderator inputs their stats online into system. This depends heavily on an internet connection.

Golran wrote:This depends heavily on a reliable, strong internet connection.

Fixed.

At UIUC, if we're about to use the excel scoresheets, then people use it to get used to it. Alt-tabbing is quicker than clicking, by the way.

The 2009 NSC can show what happens when a connection screws up.

For running paperless packets, we send out the packets for round X+1 during round X. This is usually done via email, or by file transfering in an IM.

PS: Could someone siphon this discussion off to a distinct thread?

Golran wrote:Using Avram's gimranov stats software has worked wonders for UCLA's tournaments, we've been able to run tournaments with no dedicated stats person since stats can be entered from each individual computer, and teams have loved it because it means stats are available online immediately. We still use paper scoresheets, but at the end of each game, each moderator inputs their stats online into system. This depends heavily on an internet connection.

Did moderators usually enter the stats on paper after the round or half? That would save time too, of course.

I guess one of the major problems with everything being done online is the possibility of the wrong round being read, which happened at Duke because no one had to get a password off of a paper scoresheet. I suppose the TD could easily fix this by emailing everyone the password to the next round as soon as all rooms send in stats. And yeah I suppose this all does depend on really consistent connections and squirrels not chewing on your internet cables.

What worked best for us was to have one "bracket master" with the paper packets, and have that one person (usually the fastest reader in the bracket hand out packets to the other readers in their bracket, and enter the scoresheets of other readers. So for a 24 team tournament with 4 brackets of 6, we had 4 people entering stats and handing out packets to their part of the tournament.

Golran wrote:Using Avram's gimranov stats software has worked wonders for UCLA's tournaments, we've been able to run tournaments with no dedicated stats person since stats can be entered from each individual computer, and teams have loved it because it means stats are available online immediately. We still use paper scoresheets, but at the end of each game, each moderator inputs their stats online into system. This depends heavily on an internet connection.

Interested parties may want to consult http://code.google.com/p/qbsql/ for more information. And "Gimranov[a]" is my wife's maiden name; the software has the horrible name of QBSQL, mainly because it was approximately the second thing I ever wrote employing MySQL, and I showed bad judgment. Maybe the 1.0 release will come with a new name. WaSpS?

As the main developer of JAGUARS (http://jaguars.sourceforge.net/), I have been considering how to export data for a completed game to a file. I was seeking feedback from statkeepers on what type of file format would be the easiest to work with. I've been thinking about using some sort of XML, but I'd like feedback from more advanced programmers on all the details.

rjaguar3 wrote:As the main developer of JAGUARS (http://jaguars.sourceforge.net/), I have been considering how to export data for a completed game to a file. I was seeking feedback from statkeepers on what type of file format would be the easiest to work with. I've been thinking about using some sort of XML, but I'd like feedback from more advanced programmers on all the details.

The only export format currently in use in the quizbowl world is the SQBS data format. It's newline-delimited and rather opaque, but it does work well enough. I built support for reading and writing SQBS data files into my stats software, QBSQL, and NAQT built an SQBS importer for their stats database.

Look at functions.php:504 in QBSQL to see how the SQBS format works. If the player and team numbers are stable, it should be possible to combine fragments of SQBS data files rather easily and make a valid SQBS-formatted tournament file for the whole event.

If you go that route, QBSQL would work properly with JAGUARS without any modification, so the game results could be uploaded and incorporated rather easily.

The program looks like a good start. There are clearly some features that need to be added (scoresheet display, editing/correction), but it works cleanly for me.

I'll add that it's not necessary for the purpose of keeping the stats that are currently kept to transmit entire scoresheets. On the contrary, only the following data are needed:
Each player's name
Each player's +10's
Each player's -5's
(Each player's +15's or other powers if applicable)
Each player's TUH
Each team's score.

That's something like 186 bytes at absolute maximum that can easily be transmitted by any variety of means with any kind of internet connection. Obviously, more elegant solutions are possible in some cases, but let's keep our eye on what we really need here.

M

I'll add that it's not necessary for the purpose of keeping the stats that are currently kept to transmit entire scoresheets.

After bogging down QBTPS in file format recommendations, I wholeheartedly agree here. The SQBS fragment for a game is quite compact, probably near-minimal for a format encoded as text. It is simply around 20 pieces of newline-delimited data (numbers) with no tags, descriptors, nothing. And the data itself is simple enough that support for any new format could easily be added to existing stats software.