Thursday, March 15, 2007

300/500

So, I'm sitting here in Dallas Airport waiting for my plane back to my Boston, and I'm asking myself, when will Ultimate become a game of numbers. But first some background.

I got in Monday night around 6:20, went to my hotel, and then found a local movie theater and went to see the movie 300. I had read a lot about, including a lot of conflicting reviews. In particular that a lot of it was over the top, exaggerated, etc. Hello, it's based on a graphic novel, or more fundamentally, a comic book. If comic books aren't caricatures of real people (as a long time collector back in the day, I know of where I speak), then nothing is. Exaggerated prose, over the top plots, heroic behavior? C'mon. My understanding is that the director attempted to replicate a number of panels from the book. Given the number of unusual perspective shots during the movie (not having read the book), I would say he succeeded. There were a number of times where I said to myself "That has to be a shot from the book, nobody would think of doing it that way otherwise".

And regardless of its authenticity, whether it is something like The Perfect Storm where they tried to make up what might have happened given that they already know the final results, who knows. At a very high level, however, the movie was accurate. Except for the fact that 300 Spartans required like 600-800 supporting cast of slaves, cooks, etc. to get by. Those aren't shown of course. But then they never really show the Spartans eating. I won't quibble.

This wasn't really related to numbers and Ultimate, but I wanted to share my review. What got me thinking about numbers was what I did last night. The people I was working with got their hands on tickets to the Dallas Stars/Philadelphia Flyers hockey game. I attended only my second hockey game ever, the other being some Boston Bruins game in the nosebleed seats at the new FleetCenter (TD Banknorth Garden now) back in the day, probably at least 10 years ago. Unbeknownst to me, Mike Modano was on the verge of scoring his 500th goal, which would make him one of only 14 people in NHL history to reach that milestone. And more impressively, he would be doing it with the same team, which only one other person had ever done.

He started the game sitting at 499, and considering that the Flyers were sitting at the bottom of the Eastern Conference, there was a good chance that the Stars would be scoring some goals in this game. The first period ended with the Stars up 1-0 on a goal by Lehtinen. There was definitely a huge air of anticipation in the crowd every time #9 got onto the ice. I had forgotten how short the shifts in hockey are, as it is probably the most intensely athletic sport of the sort-of big 4, with even the best players rarely playing more than half the game because of the non-stop effort required. I'm not saying it requires the best athletes, only the most constant effort.

I never watch hockey, and I had to admit that Modano found himself in the middle of the play a lot. He had pretty impressive stick work, in particular in often deftly stealing the puck from Flyer players. He finally scored sometime in the second period (I think) off of a redirect. The crowd, including myself, went wild. You have to figure this is the equivalent of being there when somebody hits their 500th home run or scores their 3000th point, and there have only been a handful of those games also. They eventually won 3-1 (or it was that score when we left with 1 minute left when the Flyers pulled their goalie, who knows about garbage time goals). Some other observations about hockey and/or the game. You can never take your eyes off the ice, because it is non-stop and you don't want to miss anything. Those guys work HARD. Somebody got nailed in front of the Flyer goal in the third period, ending up with a good amount of blood on the ice right in front of the goal. There was a fight at one point where the Dallas guy got some really solid hits on the Flyer. UFC without the elbows or feet.

So, how does this thread relate to Ultimate? The Ultimate History Handbook made an initial stab at it. They created an appendix of various numeric achievements, like most nationals won in different divisions, most championships won at each level, etc. It was interesting being on the end of some email threads prior to the release as Joe Siedler was doing research and fact-checking, sending out the appendix in its nascent form to see if people could add things, make corrections, etc. However, trying to do it 10 years from now might have been difficult, as a lot of the institutional memory would no longer be available, at least for the early days. I was even fortunate to have been listed in at least one category, like most consecutive nationals with the same team or something.

My point is, what is going to required for Ultimate to truly become a game of statistics and numbers. There were the RUFUS stats back in the '90s, but does anyone keep official stats anymore, even at nationals, more than goals thrown and caught? I know Jim has done some deconstruction of video to gather stats, but there is definitely nothing approaching the level of stats that Earth Atomizer and then Boston Ultimate kept for a few years in the early '90s. And what would really count towards the numbers? Only the championship series? Do you include sectionals and even regionals, where teams and individuals can run up the numbers on some teams? Or do you only count games against 'rated' teams? And where would this be stored and who would do the 'note-taking'? It would be nice to someday be able to measure someone's results, just to get a feel for the evolution of the game, MVP type discussions, etc. Of course, the positions in Ultimate aren't quite as defined as some of the 'big' sports.

This year, DoG took high level team stats, ie., who was in every point, O or D, and what happened that point (score or scored on). Naturally, Jim saw an opportunity to do some analysis, and that is where he got this
stuff from. There were some other interesting results in terms of plus/minus (like hockey), etc. Nothing necessarily worth repeating in this forum.

Discuss.

28 Comments:

Blogger Whitey said...

Unfortunately I think this is still a ways off. I think before we start really seeing stats it we will first have to see coaches. Coaches are the people who really want to see stats for their players, and add the kind of infustracture to keep accurate stats during a game.

I wouldn't be suprised to start seeing stats coming out of the college game in the next 2-5 years, as it seems like most of the bigger college programs are moving towards having coaches, and I believe the UPA is making a push for having coaches. It will probably take longer for the club level to adopt coaches, but once they have stats will probably come soon afterwards as the the use of stats will have already developed at the college level.

I think with the increase in cellphone technology and the adoption rate of these new technologies, it probably won't be too long before there is a port of Ulti-Stats or some other similar program available for your Blackberry. Which should make keeping and publishing stats easy.

12:42 PM, March 15, 2007  
Blogger parinella said...

There are teams that are taking detailed stats now.

The big hurdles to getting useful information from ultimate stats:
1. Sample size. There just aren't enough games that matter in a year. And most of the games in a year are really pre-season games.
2. Aggregation. Zone games are counted with man games are counted with blowout wins are counted with games played in 25 mph wind and rain. A bad zone handler might get more passes in a point that a top receiver would get in a game.
3. Team context. Basketball offense and baseball defense suffer from this as well, while baseball offense doesn't, for the most part. No matter the quality of the team, someone has to score the goals and make the passes. Surround a player with good teammates, and his counting stats (touches, goals) will go down, although his efficiency might go up. Thus, it's almost impossible to compare stat lines between teams, or even between O and D on the same team.
4. "Yeah, but." Almost everything we would track now has a big "yeah, but..." attached to it that questions the significance of the stat. "A throws a lot of goals." "Yeah, but he forces it into the endzone whenever he's close." "B has a high completion percentage." "Yeah, but he never breaks the mark or moves it downfield." "C gets a lot of blocks." "Yeah, but that's because he baits and gets burned 3/4 of the time." In each case, there probably isn't an existing stat that wuold counter the stat in question (you could come up with them: for A, completion percentage on goal throws vs all throws; for B, % breaking the mark or yards per pass; for C, times scored on).

I guess a good next step for offense would be to track something akin to slugging percentage (completion percentage is akin to batting average), and extend everything to receivers, too.

If only there were some group devoted to improving frisbee stats....

1:14 PM, March 15, 2007  
Blogger Alex de Frondeville said...

I'm not sure that stats necessarily require a coach. Back in the day, we took stats on Earth Atomizer in the early '90s, and this was well before coaches. Granted, Jim did much of the load, with Dennis McCarthy and I also major contributors. But everybody took stats on the sideline. We passed around a notebook that people tracked every pass in and what kind (backhand, forehand, other, and long passes). After the tournament, we would breakdown the notebook and put the results into spreadsheets. I think it is more of a motivational thing than anything else.

Individual stats for team use is one thing. I'm also interested to see how long (if ever) before we get to 'universal' stats (and at what level) so that we can start making comparisons between players of different eras. I imagine that will take longer than 2-5 years.

1:15 PM, March 15, 2007  
Blogger Whitey said...

I know stats don't necessarily require a coach, I just think that a dedicated coach is the only person who is going to show enough interest in keeping track of the data accurately every game, and doing good analysis of the data.

There will obviously be teams that keep stats on their own, but for it to become a "game of numbers" I think it will take more then one or two teams keeping stats. It will have to be a majority of teams (which I don't think will happen without having either coaches or TD's who have stat keepers at each field).

With the current structure of Ultimate (there being only Sectionals, Regionals, and Nationals run by the UPA and the rest of the tournaments run by individual or teams) I think coaches are probably more likely to be the first phase of stat keeping. Maybe someday we will have the volunteer base to be able to keep stats at every field at every major tournament, but that is definitely way more then 2-5 years out.

1:27 PM, March 15, 2007  
Blogger Alex de Frondeville said...

I don't see a coach being able to do stats. He/she may use them, but tracking stats really requires full attention from someone, and a coach will miss the forest for the trees if they are doing stats. That's when the injured guy really makes a difference...

1:35 PM, March 15, 2007  
Blogger The Cruise said...

Modano has played for two teams, The NHL Minnesota NorthStars and some other team that doesn't even really count because they are in the south.

Speed of the game and ammount of downtime is one of the biggest factors of sports stats. Baseball has such defined beginnings and endings of play time, making it the most stat filled game, then football, then basketball, then hockey, then soccer.

The fact that the top teams only play about 10 tough games a year, makes the discrepancy in stat importance obvious. Writing down and remembering 5 scores for an elite player against a team that never practiced together carries as much weight as Cube fucking around and getting a triple double in pickup bball.

2:09 PM, March 15, 2007  
Blogger Alex de Frondeville said...

From ESPN.com

After the Minnesota North Stars made him the second American-born player to be selected first overall in the entry draft in 1988, Modano came to the NHL with high expectations. A sleek player who honed his skills with the rough-and-tumble Western Hockey League's Prince Albert Raiders, Modano has -- by almost every measurement -- met or exceeded whatever lofty expectations he and his country have set on his shoulders.

The 36-year-old has captained a Stanley Cup-winning team; played on three U.S. Olympic teams, earning a silver medal in 2002; and played in six All-Star Games. He twice led all playoff performers in assists, all while playing his entire career (at least thus far) with the same franchise.

8:32 PM, March 15, 2007  
Blogger dusty.rhodes said...

Good heavens, a hockey semantics discussion?

The Minnesota Northstars moved to Dallas to become the Stars in 1993 depriving a true hockey fan base of their team and transplanting them to damn Texas. How do I know? Well, I was born in Minnesota and actually have a Northstars Campbell Conference Champions t-shirt from 90-91, back when Modano was but a pup.

11:04 PM, March 15, 2007  
Blogger _dusty_ said...

At UGA I keep the stats during the game as well as calling lines and coaching during play. We don't track touches, so it's a little less intensive, but it mostly requires attention at the start and end of points and after a turnover (ie the downtime). We track points played on O/D, breaks while on O/D, goals thrown, goals caught, 'hockey assists', throwing turnovers, receiving turnovers, and takeaways.

With some handy spreadsheets, we can calculate interesting stuff like percentage of time the team was broken while a particular player was on offense or percentage of defensive points that result in a break for a given player. Some things like turns/point or goals/point are obviously going to be skewed by percentage of points played on offense, but you start to notice trends of certain players having a high number of turnovers despite not being one of our deep throwers or handlers. I've also messed around with a Fantasy Average (goals thrown+goals caught+hockey assists+D's-turns)/(points played), but I'm not convinced whether it's a good measure of anything or not. Our top players are around .275-.290 while a few of our defensive handlers who don't force many D's but throw lots of hucks after we get the disc are actually in the negative.

2:00 AM, March 16, 2007  
Blogger parinella said...

I was born in Minnesota and actually have a Northstars Campbell Conference Champions t-shirt from 90-91

That would be the year that they made the playoffs with about a .500 record but caught fire in the playoffs, only to be knocked out by the Pens. Some buddies of mine (Lance and Kweder and another guy, I think) decided on the spur of the moment (probably were at a bar) to drive to Minnesota to catch the final game. They said everyone was nice to them.

9:51 AM, March 16, 2007  
Anonymous Anonymous said...

For statistics to be comparable, there needs to be a league where the competition is comparable. You hit on this point in mentioning where the statistics would or should be taken.

Ultimate has a series. Which is a system that gives every team a chance to qualify for the championships.

Because the series is an equalizer of opportunity, it gives the highest possible variety of skill. The importance of a player to a team is hard to measure when playing varied competition.
A second string player of an elite team may come out as a stud in the statistics after the sectionals tournament but only play a couple points at the championships.

Taking statistics at the actual championship tournaments seems like a good start to keep statistics comparable from year to year.
Or, maybe you could add weight to the statistics as they go further into the series.

Comparing statistics of players in a league can still be questionable. In the NCAA comparing the statistics of players in different conferences can be difficult. An A-10 player getting 20/10/5 a game isn't the same as a Big East player with the same stats.

Did the high detail statistics that were taken when you were in college useful in any way? Did they help with strategy, or with helping people see some tendencies that they had but perhaps refused to notice?

11:19 AM, March 16, 2007  
Blogger dusty.rhodes said...

Statistics that I kept while coaching a college men's team (finished tied for 4th or 5th at regionals):

D Points, O Points, Score, Goals Caught, Goals Thrown, Goals Involved In, Turnovers Thrown (TT), Intended Receiver of a Turnover (IRT), Turnovers Involved In (TII), Ds, and Possession Saving Play (PSP).

The unique part of these statistics was tracking which receivers were constantly the target of throws that would end as turnovers instead of completions. TII = (TT + IRT) and GII = (GC + GT) This particular stat did not deliniate who was at fault for any given turnover (drop v point block v d or whatever) and, when evaluated with the other stats gave a rather clear picture of which throwers were forcing throws and which players were having throws forced to them.

As it turned out, our best player (both thrower and receiver, Dan Patisteas who now plays for Slow White) was involved in nearly 70% of our turnovers at the first tournament. Showing him and the team this statistic alone was enough to illustrate my off-the-cuff remark that we were forcing the disc to particular players.

Two other key measurements that I felt painted an accurate picture of the overall effectiveness of individuals were:

Lineups:
ratio of scores for/against a particular lineup on d or o vs ratio of scores for/against a particular player on d or o.

Players:
(GII + D)/(TII + PSP)= efficiency (ranged from .1 to over 3)

The First gave me a better sense of which parts within a whole were interchangeable and which parts clearly altered the chance of our team scoring a given point. This really helped with strategic subbing.

The Second was used to balance the efficiency ratings for particular players with what I determined through tape review to be ideal ratings for each of type 4 types of player (possession handler, hucking handler, possession receiver, big-play receiver). I tweaked these ideals a little to accomodate combo-type players, and was able to use numbers in discussions with players and the team regading what I expected from them. Reaching certain goal percentages would nearly guarantee wins while others would severely decrease the chances of winning a given game.

I would providemore of the direct info, but those excel files have been corrupted. I have the stat sheets, but I'm not sure that I want to go back and hand-enter everything.

1:35 PM, March 16, 2007  
Blogger Alex de Frondeville said...

So there are two separate threads here. One is the value/applicability of individual (and team less so) stats for historical purposes, which is the far more difficult one to establish I think. In terms of which

The second one is the value of stats within an individual team. Dusty, sounds like you got real value from yours.

Back in the day on Earth, the stats were far more individual. We ONLY tracked people and passes, and it was only on offense and not who was explicitly in every point. So we couldn't derive team related stats. The stuff you are doing sounds far more useful nowadays. Of course, back in the day, it was all about lauding ourselves... :)

6:21 PM, March 16, 2007  
Blogger parinella said...

Well, it was also about convincing others to stop sucking.

Dusty, I'm surprised that your best player would have been involved in 70% of the TOs, and also that given that fact, that it was because people were forcing throws to him. In the first place, I would expect fewer turnovers because he would be more open than others and that he would be able to save more bad passes than others. After the fact, I would have surmised that the reason for the high percentage was that he was asked/forced to carry more of the burden of creating offense through his throws, and that he was a last-second bail more often, when of course you would expect a lower completion percentage.

12:27 PM, March 18, 2007  
Blogger Coach Lou said...

I wrote about a couple thoughts on ultimate stats on my blog.

3:10 AM, March 19, 2007  
Anonymous Anonymous said...

Speaking of hockey and ultimate, and Modano's first trip to the finals with the Minnesota North Stars in 90-91, here's a clip showing that hockey players can bring the heckle game as well -- Kevin Stevens/Brian Trottier giving it to Brian Bellows. Close the door before turning up the sound, 'cuz this is not for kids or prudish co-workers
http://video.google.com/videoplay?docid=7087099771269561926

11:39 AM, March 19, 2007  
Blogger Alex de Frondeville said...

Wow! Now THAT was an awesome clip. Obviously in ultimate we don't go anywhere near that kind of stuff, except maybe sotto voce... :)

9:57 AM, March 20, 2007  
Blogger dusty.rhodes said...

Jim:

The thing about Dan and that team was that he was asked to carry the burden and he wanted the burden. That guy never backed down from a challenge.

Things that led to turnovers being forced to him:
1. Throwing into double/triple covereage
2. Throwing to him 1-2 out of every 3 passes for every point on O and D for a full 2 day tournament
3. Throwing to him when he was standing still
4. Hucking to him when you can't actually throw that far to begin with

The assumption was that forcing a pass to Dan, no matter what else was occurring, was always the best play.

Things that led to him throwing turnovers:
1. Players only cutting deep when he got the disc
2. Getting the disc every 1-2 out of every 3 passes for every O and D point for a full 2 day tournament

I guess one way to put it is that when you have the disc you are in complete control of what where it goes. When you don't have the disc, any idiot on your team can just throw it to you.

Teaching one player to make better decisions/throws is pretty easy. Teaching a whole team to make better decisions/throws is a bit harder.

4:32 PM, March 20, 2007  
Blogger parinella said...

Well, if he's catching 1/3 of the passes, that means he's involved in 2/3 of the passes, and 70% isn't far off from that, so removing "forcing" wouldn't really improve your completion percentage. And when you factor in that a lot of those turnovers are on deep passes, it actually does seem to me that forcing it when Dan was involved was indeed the best play, or at least a better than average play.

12:10 PM, March 21, 2007  
Blogger dusty.rhodes said...

Maybe I will go back and plug those numbers in again. I'm hesitant to make more statements without actual numbers in front of me. I should also, at this point, be sure to mention that I know very little about how to use numbers in an acedemic way.

What it comes down to (in the end) is that when they stopped constantly forcing it to him, the team won more games by a larger margin. More players were involved in the offense and the team's turnover/point ratio decreased significantly. This isn't to imply that we stopped forcing it to him at all, just less often. I guessed that if we cut out the worst 5 turnovers per game, we'd improve dramatically.

Perhaps an overlooked advantage to this was simply playing fewer points per game and reducing the length of the points that we did play.

1:30 PM, March 21, 2007  
Anonymous Anonymous said...

Even in sports like baseball which have been around for a hundred plus years, statistics are deceiving. There's always subjective judgement involved in the ruling of an error. Ballparks are not standardized so ding dongs, doubles, and foul ball pop outs vary since home teams have advantages/disadvantages. The most egregious of these have to be Fenway, with monstah and the weird but charming design, and the Oakland Colloseum, with the ridiculously large foul ball territory.

Ultimate, at least in the fall series and in the college series, doesn't have those issues since the field is standardized. But there will always be the judgement call of whether or not someone should have caught that throw. After all, it him him/her in the hands right? ;-)

I wonder if we ever get to a point where MLU takes over for the fall series that there will be true statisticians making those judgement calls...

great topic!

Peter B from cali.

6:54 PM, March 22, 2007  
Anonymous Anonymous said...

http://youtube.com/watch?v=f9_Vn1qUBAU

12:41 PM, March 31, 2007  
Blogger Unknown said...

The Flyers have played their home games on Broad Street since their inception, first at The Spectrum from 1967 until 1996, and then at the Wachovia Center from 1996 to the present. They have had rivalries with several teams over the years, sportsbook, the most heated rival of late being the New Jersey Devils, with whom the Flyers have traded the Atlantic Division title with every season since 1995 and have faced three times in the playoffs, winning once in 2004 and losing twice in 1995 and 2000.
http://www.enterbet.com

4:00 PM, November 16, 2007  
Blogger kimberly sayer said...

Which is a daily email with some sort of special in the greater Boston area, whether 70% off at a restaurant, or spa treatments, or something of the ilk. signed up for one of them,costa rica fishing3 classes of ultimate bootcamp for the price of one ($20) to be used by the end of the year. Basically, it is a 1.5 hour Saturday morning workout at various sites around Boston.
http://www.fishingcostaricaexperts.com

12:38 PM, January 03, 2010  
Anonymous Anonymous said...

The Spectrum, formerly known as the CoreStates Spectrum (1996–1998), First Union Spectrum (1998–2003), and Wachovia Spectrum (since 2003) is a now closed indoor arena in Philadelphia, Pennsylvania.Costa rica tours Opened in the Fall of 1967 as part of what came to be known as the South Philadelphia Sports Complex, it eventually seated 18,136 for basketball and 17,380 for ice hockey, Arena football, indoor soccer, and indoor lacrosse after several expansions of its seating capacity. The building was formally closed on October 31, 2009, after a four-night concert stand by Pearl Jam. It is planned that the arena will eventually be razed to make way for a hotel in the Philly Live! complex.
http://www.kingtours.com

10:50 AM, April 15, 2010  
Anonymous home exercises said...

awesome blog man, the things you have mentioned above are really informative and are

examples of your awesome writing skills and a good blog, marvelous work done star exercise home exercise equipment push ups

11:11 PM, August 24, 2011  
Blogger david thomas said...

hello,
your articles are always informative and cool for learning, they always increase my knowledge just wanted to appreciate you and say thanks for sharing them with us chicken biryani recipe sms hindi friendship sms sms jokes latest sms

12:20 AM, August 27, 2011  
Blogger Mohtasham said...

this is an awesome article must appreciate it, the ways you have explained above is really awesome, exercises to do at home l how to make chicken biryanil latest sms

11:51 PM, June 12, 2012  

Post a Comment

<< Home