Day-of Filming for Worlds

This post was prompted by an article that Lou wrote for Skyd regarding a conversation that he and Bob Krier had.  The gist of the conversation was Bob asking Lou what his setup was, Lou explaining that he didn't really have a great solution yet, but was interested in Hudl and thought it might work if you could get multiple teams to pay for it.  Bob said thanks and that was that.  Aside from being a little hurt that Bob wouldn't email such a question to me (after all he did see me tirelessly work through film this whole summer) I realized that I have a lot to say about the topic.  Not everything I have is best practice by any means, but it is something that I have spent a lot of time with.  So I figured I would put some of those thoughts down here.  It will be long-form, and probably in multiple parts, but since the words/minute rating on this blog is more like words/month I don't think that is a problem.

As you may have guessed from previous posts, I have put some time into film work during my coaching career.  It was something that I picked up in 2003 when I was working at the USOC Training Center in Colorado Springs.  Most of our job back then was about finding ways to quickly get coaches quantitative feedback for their athletes during training and competition.  We were figuring things out as we went along, and were having to mesh data collection with video to meet demanding coaches' desires.  It was stressful, but I learned a little about video production and a lot about the value of live feedback and film for athletes.  On the pre-Olympic level (many of the athletes involved were right on the cusp of making the trip to Athens) the margin for error was slim and any advantage a coach could find was worth it.

Dealing with film was difficult back then, so when I left the USOC I stopped for a while.  Eventually I picked it back up in 2007 during Emory's first and only Regional Championship.  The finals were against FUEL (University of Florida) and because I had two other head coaches for our Semis I was able to leave early and film FUEL's semifinal game.  This was basically me with a tripod, a slight elevation and a camera that shot in 480 and output MPEG-2.  That night I was able to hook it up to the hotel TV through component cables and watch.  Watching that film I realized that FUEL's defense was checking in with the thrower too much and losing sight of the offender.  But our straight stack cutting system typically left those cutters stationary as they waited for things to clear out.  So we knew we could exploit FUEL's defensive strategy with breakmark cutting when the defender wasn't looking.  While that wasn't the only thing that allowed us to win that game, it was the first time that I really felt like the film played a direct impact in a result (there were 2 other adjustments on defense that led to two breaks).  From then on I have tried to figure out how to best do game tape.

When I applied for the U23 coaching position last fall I started thinking about film again, and explained in the interview that film was going to be a large part of what I wanted to do.  Once I got the position it became increasingly more important that I figure out a way to deliver on that promise.  I dug out my old camera from 2007 and found it useless.  The resolution was too low and the format was terrible.  I decided to get a better camera (Panasonic V-201) a few months in advance to see what I could accomplish.  This camera isn't really the solution to my long term problems, but it worked in a pinch.  I knew that I wanted 1080p so that I could actually distinguish players from far away.  I also wanted a decent frame rate and aperture speed.  After talking to a film producing friend of mine I was pointed towards this Panasonic and a few other options.  The V-201 won because it was cheapest while still having decent features.

Getting the V-201 up an running was difficult at first.  There were two main formats for video recording: .mp4 and AVCHD. I had good experience with .mp4 from work in the classroom and knew how to do decent editing with MPEG-Streamclip.  The downside with .mp4 work in the past was that a camera designed to be plugged into a TV has a different interlacing order than a computer focused camera.  This meant that when tracking moving objects there was a strange line-splitting effect that occurred.  In order to get around this the film had to be processed/deinterlaced, which could take a while in streamclip.   I knew that at Training Camp I wanted to be able to view film with a turn around of less than 1 hour, so this type of codec wasn't going to work.

In came AVCHD, as a new format it promised good resolution and the capability to go directly to a computer.  Unfortunately it came with its own headaches.  The biggest one was that the files were large.  I could easily get into the GB range while filming 45 minutes.  But I guess that is what you get for a higher resolution. That would have been a real bummer if I had to do any processing.  Fortunately this format could be read directly by a computer, so I wouldn't need to do any processing to view it.  Unfortunately it wasn't readable by the basic version of quicktime.  So I had to upgrade to the pro version for $30.  It wasn't that big a deal and it allowed me to plug the camera directly into the computer and open the files straight in quicktime.  That part was a success.

Having successfully figure that out we went to Buffalo and filmed a practice.  I had Jason Simpson stand atop a 20 foot ladder leaning against a goalpost so that we got a good angle and decent amount of field without panning too much.  We were able to directly view all of the film on a large flatscreen in a common area, and used it that night in front of the whole team.  It worked as well as we had hoped.  While I don't think we did anything that amazing, it served the purpose of helping get people on the same page quickly.  It reminded me of working at the USOC where we would literally have a weightlifter finish a C&J then walk 2 steps and watch it again from two camera angles, in slo-mo and with force plate and bar velocity data coupled.  It seemed like overkill at the time, but in reality it allowed a coach to point out that the athlete was generating force unevenly (between their feet) and also a bit late compared to the bar velocity.

In Buffalo the advantage was clear because I was able to show these 26 players from all across the country what I was seeing on the field and how that shaped what I wanted to do offensively.  It made it easy for me to display the idea of functional space to players in terms of disc movement, and allowed Eli Kerns and I to go back and forth on the cost/benefit of a throw-and-go driven offense.  **I would like to point out that neither of us "won" that conversation, but the film allowed me to see what he was talking about and him to see what I was talking about in such a way that we both moved towards the middle**  From that point on we filmed every practice, and while the players were getting dinner the coaches would often be watching the film in order to figure out what to talk about that night or what to do with the next practice.  At times we would use the film to show players tendencies that we wanted to stress/diminish, but mostly it became a diagnostic tool for the coaches.

There was one hitch with the process, and it was quicktime.  An annoying feature of quicktime's AVCHD compatibility is that it opens a preview window from which you can select any of the video clips that are in the AVCHD file.  But once you select a video and watch it quicktime will kick you out so you have to do the whole thing over again.  It made hunting and pecking for specific video very slow and a momentum-kill in a film session.  If we had even a night to prepare we could have clipped all of the sections and made a unique video, but as it was we were trying to find specific plays in 1-4 minute long clips.

That hitch could have really been a damper for the process if it wasn't for one of the players (I think it was Justin Norden) that pointed out we could use VLC.  While VLC doesn't have some of the scrubbing features that QuickTime does (in quicktime the arrow keys will let you go frame by frame in either direction, in VLC you can only easily go forward) it will queue all of the individual videos in an AVCHD file so you can jump around with ease.

The use of day-of film became more evident in Toronto.  We had no prior knowledge of opposing teams going into the tournament (which is not that uncommon in mixed), so being able to scout opposing teams was vital.  We apparently got a reputation for being the CIA because we would bring a camera to other fields and film opponents, but it was worth it.  We also filmed most of our games so that we could watch the film that night and see if we were falling into bad habits.  All teams fall into bad habits, and having the ability to recognize that is important.  Having film showing a person displaying those bad habits goes a lot farther than just telling them after a point.  We used film often to show that we weren't running the team's end zone offense but rather each person's end zone offense, which explained why scoring from the red zone was taking so long.  Or to show how well we cleared out space in front of the disc as it swung to a sideline.

By the time we made it to power pools to play Canada we had film on all of our opponents, had broken down all of their offenses and all of their defenses.  But here we fell into a trap. Being all club players, the coaching staff felt that we needed to tell our team as much about our opponents as possible to be ready.  If we could use that to produce even one break that could be the difference between a win and a loss.  So the night before the game we spent a lot of time explaining everything that Team Canada was likely to do and how we were going to stop it.

The trap was that we unloaded too much information on our players.  I believe the captains accurately described it as "making the game more about Canada and less about US."  That was a problem that came from having access to so much film.  The well was so deep it was easy to drown in.  In the past, my page of scouting notes could have been finished quickly, but since we had film of everything we could go over too much.

The lesson we learned was that controlling the amount of film was important.  That is easy to see when making instructional film, but we lost sight of it in the day-of film process.  We continued to film everything, and use it for the coaches in bulk, but for the players sparingly.  By the time we played Canada in the finals we were able to refine the message to just a few clips of specific things we wanted to do.

So where does this leave me with day-of film going forward?  With the high school season ahead of me, we don't really scout as much as we do in club/national competition.  So I imagine the film will turn introspective and we will film our own games.  Then that night I'll go through stuff, pull some clips for the nightly team meeting and try to improve our game for the next day.  This film will most likely get kept for later use during rainy days, but at that point it will turn into "stale" film and I should be able to edit and telestrate it for particular messaging/instruction.  That feels like it should be another post.

I think the benefits the coaching staff and players got from day-of film was worth the headache, but it was easy to go overboard.  Players are tired after a day of ultimate, so forcing them to sit through film sessions may not be the right idea.  But that answer might change from team to team.  Getting a system up and running isn't terrible from a technical side.  You should be able to purchase a decent camera and software for under $350.  If done correctly you can then just plug it right into a computer (coupled with an HDMI out or AirPlay you can go to a TV easily).  From that point on you actual get to focus on coaching and can figure out how you want to use that video to convey a message to your team.

I'm sorry if this has been a rambling mess.  I have had so many thoughts about this topic they kept getting stuck together.  One thing that I would like to end with is a huge thanks to Kyle Krumwiede from Orlando.  He helped us with tryouts and traveled to Toronto in order to help the team.  He was my camera man for the entire week and without him we wouldn't have been able to get as much film as we had or been as prepared as we were.  He wasn't formally affiliated with the delegation, but he made a big impact and deserves credit.  USX.