Tuesday, December 13, 2005

What strategy?

I read this post by Idris, and it got me thinking about the low level of strategy in our sport. Take the idea that it is truly better (and I actually agree with this) to let a team learn how to flow and adjust on it's own, with little structure. This, to me, is a horrible indictment of the place our sport is at currently. In other, more established sports, you learn the fundamentals, how to make the x different cuts required in your sport, the right way to throw a ball.

Maybe I place football on too high a pedestal, but I'd love to have just a fraction of the strategy from football in Ultimate. How hard would it be to have a standard set of cuts, have players run the cuts, perhaps changing them based on a defensive 'hot read'? Is it simply a physics question? Does the disc move too slowly in the air to run a curl or post? I'd like to think we could just add on to the progression idea. I know player x is going to make cut y. I see how the defense is playing him, and know that he is going to switch to cut z. I can then make the throw to z before the player has even turned.

I know this happens, to a degree, at higher levels, but I don't think it's actual codified in the offense. There is so little consensus in our sport about the best way to do things, that we actually teach new players to try different (probably incorrect) things out because we're worried that we're wrong or we simply know that the next team a player is on will likely have a completely different philosophy. I'm not talking about the difference between the west coast offense and the run and shoot. At their core, those offenses are much more similar than what we do team to team, even if we're running the 'same offense.'

Teaching players to play dynamically makes your team better right now, and probably makes your players better over the next few years. But until we get to the point where we can teach players the right way to play, we're going to be stuck teaching them to play 15 different (non-optimal) ways and we're never going to have real strategy.

Strategy in our sport right now is limited to "create 2 on 1's", "get a step on your guy and jack it", "break the mark to get the defense on the wrong side of the field," etc. Maybe I'm silly to think that we can or should move to, "the defense is in an X coverage so I went to the Y hotroute to beat it."

What is the most advanced Ultimate offense out there? Does it work? and how 'advanced' is it really? What are the limitations in our sport that keep us from having more precise strategy?

11 comments:

Anonymous said...

I agree with your claim that ultimate strategy is undeveloped, but I don't think we should use football as our model for two main reasons.

1. Football is highly specialized. WRs virtually never throw the ball. QBs virtually never run pass patterns. Defensive players rarely do either of those things.

2. Football has a break between plays to call another play. So a WR needs to know only one route, and maybe one or two alterations to that route per play. The QB needs to have one set of reads per play. In ultimate, a single player could easily touch the disc 5 times before scoring, each time in a different position on the field, with the other players also in different positions.

I foresee ultimate strategy as more like soccer or basketball. More fluid sports, with less specialization than football. Unfortunately, I know little about strategy in either sport.

Anonymous said...

Interesting.

About the football comparisons though: an incompletion in football does not lead to a change in possession (except for interceptions, loss on downs). So if you miss that hot read and throw an incompletion, its often not that big a deal, especially if its first or second down. And completion percentages in football are what? Isn't 60% a very good completion percentage? In ultimte that would be awful. I guess what I'm saying is that the consequence of both players not seeing the same read in ultimate are much more severe than they are in football. A turn in ultimate is usually a big deal (except in big wind) whereas in football an incompletion isn't a big deal.

I love the idea of trying an incredibly structured offense. But the risks associated with it seem too great.

Maybe the physics is the reason for the difference. Imagine NFL football played with a disc. I think there would be a lot more d's and interceptions with the slow arriving disc as opposed that rifled football.

The speed with which the football arrives limits the D's recovery time, and therefore hooks and curls are much more effective, even though the receivers often don't "go to".

The length of time an ultimate player is allowed to hold the disc allows for much more spatial separation between receiver and defender to develop.

gcooke said...

Yep. Lots of good questions and comments here.

I think Jeff and Anon both addressed the problems with the football comparison, and I agree with their points. Ultimate seems to require an internal acknowledgement of when the team is moving from scripted movements to improvisational thinking (this seems pretty similar to basketball). 6TM had a woman from Godiva this year, and we spent a bit of time talking about the highly structured, and successful, Godiva O. I commented that you would think that teams would have learned, after all these years, where the cuts were coming from (the cuts are based on stack position). My friend responded "well, that's when Teens and Molly would free-lance".

I am not sure I agree with the assertion that there is a "right" way to play. I am also not sure that I agree that "we" are worried about teaching the wrong things or that the next team will do something different. I think I felt more this way when I began coaching and wasn't as confident about what I believe to be "basic" principles, or at least what is fundamentally sound in the current state of the game (my post on fundamentals goes into my opinions on this at least from a college point of view). This is not to say that I know it all. Far from it. I am just not that worried that I am wrong. One of the nice things about sports is that you have clear feedback. If things are right, you win sometimes.

I agree that the disc could be delivered more often right when a person cuts, but this does vary based on the particular scenario.

Evan said:

Strategy in our sport right now is limited to "create 2 on 1's", "get a step on your guy and jack it", "break the mark to get the defense on the wrong side of the field," etc. Maybe I'm silly to think that we can or should move to, "the defense is in an X coverage so I went to the Y hotroute to beat it."

I don't mean to get picky, but I think there are times where "get a step and jack it" is the same as " D is in X so I go Y". Perhaps in both cases the D player is fronting and the cutter did a good fake and was able to cut deep.

I guess my opinion is that it is optimal, as things stand right now, to have a time for structure (off a pull or a stopped disc), but recognize when you need to react an adjust (when the pull play breaks down, or when to do a fast break).


It is interesting contemplating what an "advanced " system is. It is a very good question. Is is whatever wins? The most efficient from stats point of view? Or based on "elevated" principles, like the Triangle?

-George

Idris said...

mike o'dowd talked about 10+ cuts that chicago would run as audibles from anywhere at anytime. essentially one-man mini plays. teams of course have audibles for a cut in, away, break, open, throw to space, etc. but these were even more complex cuts... various flares, curls, out and in, double moves, etc (this was the 80's btw). it never caught on with jam, but i always thought there could be place for it.

a post wondering when ultimate would begin to take on new, more advance strategies reminded of a recent santa cruz grad who put together an alumni team for the vegas tournament.

though the team has been around for 20+ years, he wondered if this was the first time anybody had ever thought to put together an alumni team for a spring event.

only recently with the advent of the interenet, blogs, and of course "the book", has there been a readily available source of information about current high level strategies. what is new and what has been tried. i'm always weary of saying "why doesn't or hasn't anybody..." and the like. because more often than not, they have, i just hadn't heard of it.

its tough for the people who are willing and equiped to contribute to the progression of the sport to move past the wheel... because they must first invent it on their own. and that can take time.

but as jim alluded to, this is becoming less the case.

just as with RSD, when the college kids post asking a question that has been addressed yearly since the creation of RSD... often startegical questions get asked and addressd multiple times... looks like the old "what about football?" strategy talk is coming back.

post topic idea!

wood said...

I didn't mean for everyone to get stuck on the football thing. I'm not saying we should take strategic elements from football, just that I see football as being a highly strategic sport, as opposed to Ultimate currently.

Clearly Ultimate will require different strategies. It is not even my suggestion that we should be taking strategy from any particular sport, rather that we should be taking a step forward in our level of strategy (if that makes sense).

I went to the football analogy because I don't know what we should really be attempting to do strategically, I just wanted to show a level of strategic complexity.

I do think we can create rules for situations inside an offense, beyond just, look deep/look break/look dump.

On Rival, a couple of years ago, we ran a spread offense with set cuts downfield. It worked fairly well if the downfield players actually ran the cuts, until the opponent figured out what we were doing. I always felt like it should have been fairly easy to develop a set of rules allowing for different cuts downfield. A very simple example is, if you're fronted, cut deep, if you're backed, cut in. If you can integrate that with what the other 3 guys downfield are doing (and even better, what the other 6 people as a whole are doing), then all the better.

You should also be able to come up with rules based on the location of the disc and the location of individual cutters in relation to the disc. Is there any reason you can't have a rule set that includes rules for thrower and receiver for a stopped disc, the disc in the middle/break/open side, trapped, redzone, etc.

I think, intuitively, this is what we do as Ultimate players. I wonder if there's a way to make the learning curve less steep, so that you don't have to be a vet to know what to do in most situations.

_dusty_ said...

I like to treat the offense off of a stoppage like football, where you call a specific play and everyone knows exactly what they are supposed to do. The thrower knows exactly who he should be looking to throw to and knows what to look for in the defense.

In flow, it is more like basketball. The coach provides the general framework of the offense and you just kinda react to how the defense is playing everyone. People should have a pretty well defined role and should recognize when it is their turn to cut.

I agree about the completion percentage as well. In football, if your primary and secondary receivers are covered, you can throw the ball out of bounds and try another play. In ultimate, you need to maintain possession, so you've gotta have some improvisation if your first few cuts get shut down.

Football has the added advantage of getting to set up before every play as well. Even in the hurry up O, the receivers get to set up wherever they need to be before the ball is snapped. In ultimate, if your play is busted and you dump it, people are spread out everywhere so you can't just call another play because you may only have 3 people in the stack, one guy cutting deep and another clearing back to the stack on the break side.

Curls and outs are effective in football because the QB is timing his throw to arrive just after the receiver turns around. If you watch a good QB/WR combo, you'll notice the ball is on the way to the WR before he even turns around. If the disc is delivered early enough, curls are very effective in ultimate. It takes a lot of faith in your receiver to make the throw before he has cut for it, though. It can be very effective as well, since the defense isn't prepared for the throw either.

Anonymous said...

In 1995, UGA took the region using a pretty simple offensive strategy with an avergae team. Rick Whitcomb came out to Lake Herrick and analyzed our play and players and developed an offense that cycled its attack against field space. He always said he wanted to use it with Chain, but his teamates were too dumb to figure it out...

Four players were assigned designated areas to attack:

1 = Open side, mid to deep lane cut
2 = Break mark, anywhere
3 = Swing out to the open side flat
4 = Dump

The cycle was determined by the count, so:
stall 1 / stall 2 / stall 3 = look for #1,
stall 3 / stall 4 / stall 5 = look for #2,
stall 6 / stall 7 / stall 8 = look for #3,
stall 8 / stall 9 / stall 10 = look for #4.

There was a little overlap to allow for setup, etc.

We then paired two strikers to the #1 and #2 cutters for the continuation and the leftout guy was the designated scorer for the strikers.

Sometimes, we would miss a cut as that # had the disk before a reset. Didn't usually matter because we worked hard on the dump. If the #4 (dump) got stuck with the disk he would call out a set play involving the strikers or the scorer (typically Will Deaver) and reset the offense.

What we found was for the college level, this thing worked like butter. Noone got in each other's way. Even if it didn't score immediately, we cycled until the combinations clicked.

Further, we didn't cut each other off and everyone was one the same page. We also found out we liked each other when we weren't frustrated over a lack of flow.

Its simple and easy to figure out in retrospect, but it won the tourney in Baton Rouge with no wind. We were basically unstoppable against the other mediocre teams in the South that year. We threw in a few set plays off the pull or stop, a couple of endzone plays, and viola the region was ours.

Unfortunately, big wind at Nationals revealed our offense to be pretty one-sided and we 0'fered.

At the time and still, it is a pretty simple formula that allows for a lot of freedom and could work as a varied offensive set to throw out there.

Brawner

Edward Lee said...

What if the stall count in elite ultimate were lowered to 5? Would this force offenses to craft more precise strategies? Or would this just lead to swilly huckfests?

Anonymous said...

What are the limitations in our sport that keep us from having more precise strategy?

No one has really attacked this question yet. Here are a few of (I think, leading) candidates:

1. Ultimate is not a pro sport. Virtually everyone who plays and coaches has a job or school that demands a bunch of their time. The time it would take to learn a relatively complex playbook or system or whatever you want to call it is something that most of the players on a given team either won't have at their disposal or won't be willing to put in.

2. Ultimate really hasn't been around that long. Check back in 50 years.

3. Heretofore, few players have grown up playing the game. This means the knowledge base on which a complex strategic system would have to be built just isn't there for many players. Compare this to sports like football, basketball, baseball, and soccer, which kids learn as toddlers. By the time they get to the college level, no one has to teach them the basic rules, how to throw/catch/kick/hit/shoot the ball, etc. They've also probably already been exposed to a lot of the basic strategies. And, they watched people who were really good playing on TV. Again, check back in 50 years (assuming ultimate continues to infiltrate primary and intermediate schools, youth camps, ESPN, etc.)

Martin said...

I think the point that few have been playing the sport since they were young is a really important one, however there is another part of that . . . the coaching that they recieve as young players. Just playing at a young age is great, because it gives a player time to develop and experiment with complex strategies in their head if they so choose, but it is being coached when you are young that gives you the proper foundation to implement complex strategies. How many pro athletes, when asked about their major influences, refer back thier little league/pop warner/AAU coaches as the people that told them the one gold nugget that they still carry to this day?

Youth coaches are becoming slowly more prevalent, so in the next few decades (maybe less than 5) we can start to see better strategies developed through a proper foundation of how to do the basics of the game.

In the meantime it certainly doesn't mean that the persuit of strategy is doomed (especially at the college level where the average experience is so small), it just means you have to figure out how to find the balance of teaching and strategizing. Oh, to be a professional coach and have developed players drop in your lap.

B-ran said...

I just want to put out their, I have put together zone and man defense for open field and redzone, designed to stop and create turnovers. While at the same time created a attack system on offense that will ensure movement, even if it has to be slow at first. I havn't tried these out unfortunately, but the plays are laid out as a foundation and their is still a sense of being aware of each other and the ever shifting scheme of the game. I am still working on these things and will continue to work on strategy, this is my goal.
B-ran