2012 GA State Championships -- What we did, how we did it.

[jcwren]

As Alan mentioned in this post, we did indeed score the 2012 Glock Georgia State Championships with Practiscore. Here's the story:

The match was 4 sessions, 2 on Friday, 2 on Saturday. Each session had a maximum of 10 squads, with a maximum of 8 shooters per squad. Most of the staff shot on Friday, using embedded RO's. Friday morning we had 8 squads with a total of 52 shooters and for the afternoon, 4 squads of 28 shooters. On Saturday, we ran 10 squads with a total of 66 shooters in the morning, and 10 squads with a total of 68 shooters in the afternoon.

Our match hardware consisted of 9 primary NOOKs, 22 back-up NOOKs (stage scoring devices), 2 iPads (a master registration device, and the roving data synchronization device), 2 laptops running EzWinScore (also with Practiscore Importer), and a printer.

The primary goal in the model I explain below is that at all points we'd have back-ups. We always try to setup a match so the data flow is in one direction, once the devices are all synced. That starts out as EZWS -> Practiscore (iPad) -> Stage NOOKs. Once the match starts, its NOOKs -> Practiscore (iPad). At the end of the match, we post directly to Practiscore, and once I get home, export the match into EZWS to post to USPSA and handle any classifiers. Basically, we treat the NOOKs as sacrosanct, and try never to push changes back to them.

Registration was handled online, using Shoot'n'Score It (SSI). Although it's not a highly intuitive user interface, we've had good results using it for our last 4 monthly matches. We do have a problem with people not always entering all the information necessary. For the GA State (GAS) match, we had to track down about 20 people who didn't enter their USPSA numbers, correct divisions, etc. At least if nothing else with SSI, you *do* get people's email addresses.

The match was defined with a maximum of 275 competitors permitted, 40 squads, and a maximum of 8 competitors per squad (yes, that's 360 slots ). The 40 squads comes from having a total of 10 stages (9 actual + chrono) times the 4 sessions.

When registration closed, we had 259 people who had registered, with 28 removing themselves from the match, for a total of 231 registered.

At this point, I took the EzWinScore export from SSI and ran it through the ssi2ezws conversion utility. This brings all the competitors and any defined stages from SSI into EzWinScore. Corrections were made as necessary to the stages (SSI doesn't know about multi-string stages or IPSC vs classic targets). This EZWS database became the master database to be used for competitor check-in, and will never have scores in it (more about that in a bit).

I then used a utility I wrote (but have not yet released) to convert the EzWinScore competitor database and stage definitions to a match file that Practiscore can directly import (the .psc files you may have seen, if you've ever emailed yourself a back-up of a Practiscore match). In this particular instance, I brought over only the stage definitions. I kept a back-up copy of the Practiscore database with only the stages so I could start over without redefining the stages if I had to start over.

After the Friday AM registration closed at 7:15AM, I exported the registration (Match -> Handheld -> Export Registration) from EzWinScore over to Practiscore on the master iPad. We then synced the 9 stage NOOKs to the master iPad, and also synced the scoring iPad.

After this point, as competitors checked in for their session, we made updates simultaneously to the EZWS database and to the master iPad. The reason for this is that Practiscore 1.5.3 and prior don’t handle the import of the PSS.txt file correctly. After the initial import of the PSS.txt file, if you import again, it duplicates the data, rather than updating. Import a third time, and you’ll have three copies of each shooter. The Android version 1.0.6 and prior have the same problem. It’s been fixed in 1.0.8, but I have not yet tested it in iOS 1.5.4 version yet. Because of this problem, we decided it was better to use the [manual change document] (pdf) that Bill Noyes created, and make sure the change was applied to both EzWinScore and the master iPad.

Once the match started, Alan cruised around the range periodically, syncing each stage NOOK to the scoring iPad. We have an Engenius ENH-200 outdoor high-power access point installed that covers the majority of the bays. Unfortunately, there were 2 stages where we didn’t have coverage, and due to time constraints, we also didn’t have a link between the club house (at the top of a hill, and isolated from the bays by terrain and foliage). For the 2 bays the ENH-200 wouldn’t cover, Alan had a pocket access point to allow the iPad and the NOOKs to sync.

After a sync session, the scoring iPad was brought back to clubhouse and immediately synced to the 2nd laptop using Practiscore Exporter 1.0.1. Bill Noyes had his iPad with him, and he’d sync against the scoring iPad also.Bill took up the role of posting the intermediates to Practiscore.com.

The 2nd laptop was not technically necessary. I used it because I had it available. I could have just as well used one laptop, and switched between two different EzWinScore databases (the database with only the registration and stage definitions, and a duplicate of that one which included the scores).

Earlier I mentioned that our normal match data flow never pushes anything back to the NOOKs during a match. Since we had 3 additional registration periods after the initial database creation, we knew we’d have to sync the NOOKs at the start of each session. Since we only had 4 squads for the Friday PM session, we elected just to hand out the back-up NOOKs (which were configured immediately after we deployed the 9 stage NOOKs to their respective stages, and could start breathing again).

At the end of Friday, we effectively had 8 NOOKs with the AM squad’s data, and 4 NOOKs with the PM squad’s data. All of our scores were on the scoring iPad, and also on the 2nd laptop in EZWS. We backed all the various data up to email and USB sticks, then shut down the match.

For the Saturday sessions, after registration ended at 7:15AM, we synced all the NOOKs to the master iPad, and sent them down to the stages. At that point, it was basically lather, rinse, repeat. We did have one major panic and one small panic, which I’ll detail shortly.

When the Saturday PM session started, like Friday, we synced the backup NOOKs to the master iPad, and rotated them in to the bays, picking up the original morning NOOKs as the squads finished.

At the end of the match, I removed all the confirmed no-shows (along with one extra person as a result of a communications error), verified the 10 DQ’s we had, and ran a classification update in EzWinScore. I then copied it to the 2nd laptop, and synced against the scoring iPad to bring in all 1770 scores. We ran a missing score sheet report, WITH ZERO MISSING SCORES. We actually would periodically run the missing score sheet report periodically, and confirm that it was pretty much in line with what we expected, so while we were ecstatic, we weren’t *really* surprised.

One of the two scares we had was a real heart stopper. Alan brought the scoring iPad up, I synced with Practiscore Exporter, and it said (in effect) “Your shooters numbers aren’t sequential. Aborting.” WHAT THE ... ?!? After a few moments of looking at various data, we determined that one of the shooters on the scoring iPad had an EZWS shooter number of -1. Alan did report he had a “1 shooter modified” message when syncing against a particular NOOK. This means that *someone* went into the Edit Shooter/Squads menu on the NOOK, and “did something”. We STRICTLY forbade this, but herding staff can be like herding kittens.

After a few minutes of experimentation, I pin-pointed the bug. It turns out the Android version of PS doesn’t allow a shooter number > 99. When importing it from the PSS.txt file, it’s not a problem. But if you edit the shooter, and it’s greater than 99, it just erases the field, and internally sets it to -1. You can only enter 2 digits into the field. We then determined that we could reset it in the iOS version with no problem (with that fancy little iOS spinner). We exchanged a couple emails with Josh (the Android developer), and he was under the impression he was supposed to limit it to 99. What was actually intended was that the SQUAD number should be limited to 99 (we talked them out of this, too, for other reasons).

We suspect that the reason the shooter was modified was someone moving him to another squad. We had one shooter that jumped squads for some reason we never found out. Rather than using the ‘Move a shooter to this squad’ button, the RO used the ‘Edit shooter’ method, which was the way it had to be done prior to 1.0.6. Again, in spite of being explicitly told NOT to do this, and to contact stats, they took it upon themselves.

The second scare we had was after syncing the scoring iPad to one of the stage NOOKs, it report “9 stages modified”. Alan rejected the sync, and compared all 9 stages in the NOOK against the scoring iPad. No differences could be found, so he synced again, and accepted the changes. After returning to the clubhouse, we compared the stage configuration on the scoring iPad to the actual stage definitions, and could find no change. We never did nail down what caused this. Practiscore uses a bunch of different internal time stamps to determine what’s the most recent copy (an important reason to make sure the clocks are set correctly if you’re EVER using more than 1 device), and we’re wondering if the clock slipped somewhere. We just don’t know on this one. But since the stages were determined to be defined the same, we let that one go.

I can’t say for sure if the EZWS shooter number issue on Android would have killed another match or not. Alan and I have had quite a bit of experience with Practiscore on both Android and iOS. We’ve brought Practiscore into two clubs that shoot weekly, and two clubs that shoot monthly, and I think we’re probably responsible indirectly for two other clubs picking up PS. We’re both very strong advocates of PS, and provide a good deal of feedback to the developers. We’ve got a “suggestion” list that runs to about 3 pages right now.

One of our hot-buttons for major matches is the ability to lock people out of editing shooters and squads. At a major match, you simply CAN NOT have RO’s accidently or deliberately editing shooters or stages. All that needs to go through stats, and have change control forms to track the changes. Anything else is just playing fast and loose, with no audit trail.

For those that are interested (and have actually read this far down), here’s links to the various forms we used for the match.

Change Control Form (pdf)

Score Sheet (pdf)

For future matches, I don’t think I’d use the 8-up score sheet we used this time. I’ve found some on-line printing services that are *cheap* (and I definitely can recommend Carbonless On Demand). Next time I’d rather use a generic summary sheet. Bill cranked one out in PowerPoint(less) while we were waiting for some scores to come up. The rows are in the same order as the Practiscore ‘Review Scores’ screen, making it easier to copy over. I can get 3000 for $127, shipped, and I’ve seen several 25% off coupons go by in the last couple weeks.

While I might make a couple changes if we did all this over again with what we know, I have to say that given the various complaints others have made using Practiscore in large matches, from the competitor’s perspective, this match went flawlessly. From our perspective, with the exception of a couple of RO’s making changes that should have NEVER happened, it went flawlessly.

And I expect that as Practiscore continues to develop and a few of these niggling little problem areas get fixed, if you suggest scoring the match on paper, they’ll look at you as if you had said you intended to shoot a flintlock in open division, minor power factor.

Edited September 12, 2012 by jcwren

[Ronnie j]

Wow , awesome job Chris , my thanks to the smart guys who are making this happen .

[a matt]

Thts easy for you to say JC. . . .

[Will_M]

This was by far the easiest match I've ever participated in as far as logistics go. I was able to register/squad and pay online in a matter of minutes, I had a near "live" update of the scores as they were entered, and after I finished shooting (Saturday AM slot) I had the final results by the time I made it back to Mississippi.