I’m currently working on a browser-based scouting system to use at Nats. Assuming no technical difficulties, how many other teams would be interested in scouting? How many teams would want to get data from it?
Alternatively, does anyone already have a reliable system in place that we could help with and/or get data from?
The scouting program we use would allow sharing data fairly easily (all the data is stored in individual files for teams/matches/regional - there would just have to be some agreement on how to label matches based on field at the Championships, perhaps calling each field a ‘regional’ for the purposes of the program).
I don’t know yet who on our team will be running the laptop, but I’m sure we could work something out among anyone at Atlanta who is using our program. I do know we’re on Archimedes, so it’s a matter of making sure there’s coverage of all the fields and coordination for exchanging data.
Depending on what you’re collecting, we could possibly come up with an exchange format, like a CSV file, that could be imported back and forth between the systems…
Calling each field a regional is what I thought of as well. I downloaded a copy of your team’s prog and looked at the match form. Slightly different data, but I’d say half the data is the same. Heck, I might remove some of my data that isn’t the same.
Assuming it’d help, I’ll probably post a description once a get a couple more details finalized… Just today, I realized I was neglecting the match breakdowns (refer to: Reflections on 2007 thread).
If you’d like to make a standard format in the next day or two I could make MARRS comply with it. We will be running it on Newton.
Yeah my team will contribute we’re on Curie
we are using team 768’s system
We’ve used this successfully in two regionals:
It was amazing the data it gave us, especially in Boston once we got used to it.
We are upgrading with a potential output summary sheet that gives analysis on the teams and some other cool features. We would be happy to share that info or the database with you if you want.
I haven’t designed that many file formats myself, but I think I can come up with something usable. Unless people say otherwise, I’m targeting one “physical” file, preferably well-labeled (INI files come to mind). If anyone else has a program I should keep in mind, speak up. (/me watches as everyone and their brother comments…)
As for syncing data, that depends on what we can rely on. If we can’t anything, then a person going around with a laptop or usb key will probably work. If we can get a network, we could potentially have a MySQL server somewhere, and write a program/script that goes of every hour. Since the soonest the divisions mix is finals, we could also keep thursday/friday syncing to “just meet at the end of the day”. The real test will be if we can get the data to all the teams, if they care…
Thoughts?
I think at the end of the day (and I’ve been through quite a many of these days going to 3 competitions, and one pre-season scrimmage) the top detail I personally (not my team per say, I won’t speak for them) want to know from a scouting report is how many tubes a team scored.
If that was publically accessible as a hard data final number, and very accurate, I think we will see this data will help in scouting immensely.
Just my $0.02.
/me thinks I will join my scouting team this year more than in past years and try to lead this effort by keeping everyone on the same page as to what the ultimate goal is the data they collect is going towards.
I sense some long nights of hard number crunching. Oh wells.
(I should of started caring more about this weeks ago… lol) :-x
My thought is likely CSV or an XML format would probably be the easiest to use between the systems, as opposed to an INI. Both are fairly easily read or written with almost any software language. (CSV might be the easiest, since in theory it would also be readable directly in Excel.)
I could probably have some form of input/output ready by later this evening, depending on how my workday goes. 
An XML format would probably be the most readable but CSV would be the quickest to implement.
We need to decide what information we are going to have consistent becuase MARRS stores data on a match by match basis and I would have to implement an export/import option to just export the vital stats.
My vote is CSV.
(That stat was mentioned elsewhere.) So (a)there should be a computer or network where anyone can view the data collected, and (b)We need to make sure that the program used to share the data can take averages? I believe MARRS is web-based, so not too much of a problem on (a). Come to think of it, sundial might be a good option, but I can’t imagine how the heck we “plug into it”. As for (b), the programmers can speak for themselves. Worst case scenario, I’ve got the half-designed web app.
Readability will be an issue. From what I’ve seen, any program that uses whatever format will need to handle extra info and missing info. Or everyone arbitrarily decides to use the same program and re-design the scouting system around the new program. That’s why I prefer a well-labeled file format; so the programmers know that the data isn’t getting mixed up. If that’s a non-issue, then I’ll start laying out the columns for CSV.
As for what data is the same, I thought I was going to be stuck with that since the post went “If you’d like to make a standard format…”? All I need is the relevant forms. I’ve got 768’s already, just need to make sure I get the right one for MARRS.
My gut feeling is that CSV is probably the easiest way to go with a week left before the Championships - it’s the route I was thinking of hammering out this evening if needed. It’s quick to layout and has the benefit of being readable in Excel if necessary. Plus, it’s easy enough to write code to either ignore extra data or add in ‘missing’ data to make it valid during the import/export process.
I had actually sent a PM to the maintainer of Sundial to see if there was a possibility for some kind of basic ‘feed’ (for lack of a better word) to retrieve schedule information for import, but didn’t get a response so I’d imagine that’s not going to be viable for the moment. The only question for using a web app for data exchange is the question of networking. 
I’m open to suggestions on the ‘how’ of getting the data passed around, though I’m more busying myself with format questions at the moment.
we are willing to use it
I guess unless someone complains quickly, I’ll work on CSV.
magic8ball, can you get me an html dump of the match form? You guys did a real good job with designing a hard-to-mess with login system, but I locked myself out…
EDIT: And the pit form, while you’re at it.
sweet…
we would use it probably to get infomation
Are you talking about the data collection form? If so I’ll need to upload that. Currently I am at work and can’t do that :).
I was thinking about just exporting the data after it is compiled (IE total spoilers, total ringers, matches recorded). But if you would like to do data for every match I’d be willing to get you a SQL dump (just a few rows).
I’ll PM you with a username and password for the MARRS Demo that has unlimited access.
Actually, I took a look at the dump generated by mysqldump that was kindly passed to me by our partner in crime here. 
I think part of the question is what do you need in the CSV to generate your stats, to make your program as effective as possible with the imported data? (Fair’s fair, right? :))
Insofar as the 768 scouting program, my preference would also include a little more of the match data such as the scoring on the rack, but that’s the basic gist of what we’d need for our purposes. Some of the fields, such as what I believe are strategy rating fields in the dump, aren’t perhaps as portable since they’re a bit more subjective.
(I feel for you about being at work, as I’m at my job too…)
(Sorry, I’m on Spring break and jobless…)
tjcasser, as I seem to be missing part of the picture on your app, I’ll e-mail you an HTML dump of the forms provided in MARRS, unless magic8ball prefers an alternative… That should give you an idea of what MARRS likes to have. I’d love to be able to develop the format myself, but I seem to be missing part of the picture…
No one has offered to cover Galileo. My team is in Archimedes, so I probably shouldn’t volunteer them for that unless I have to. Anyone in Galileo able to help?
I think we need Curie too?