Monday, December 05, 2005

How much is a yard worth?

I was reading through a bunch of the old stuff on my blog recently. It’s fun to see how some of my ideas have changed even in the short time (1 year) that I’ve been blogging. Anyway, in one of my first posts I said I was going to go through all of Billy Berrou’s stuff and pick out the good stuff. Jim beat me to the punch with this post. In Frank’s response to Jim he says something that’s interesting to me “First, take yardage completely out of the equation. It does not belong. Yardage is irrelevant…It's always about position.” In a sense, this is totally insane – you have to move forward (gain yardage) in order to score. However, it is true that it is oftentimes worthwhile to trade yardage to maintain possession and improve field position.

Anyway – here a couple of things I’m pretty sure I think about the value of a yard (relative to the value of field position and maintaining possession) –

1) The value of yardage decreases as the level of play increases.

2) The value of yardage decreases as conditions improve.

3) The value of yardage increases when a team is going downwind.

Here’s something I think I think about the value of yardage.

1) The value of yardage decreases the closer a team gets to the end zone.

Here’s something I’ll claim I think if people think it makes sense – otherwise I definitely don’t think it.

1) The value of yardage decreases when a team is going upwind.


Mackey said...

Intersting thoughts about yardage and upwind/downwind.

I wonder if yardage wouldn't be less important going downwind, since the possibility of the long strike to the endzone is greater from more points on the field with the wind at your back than otherwise. Upwind I'm not so sure about. Position would certainly take on greater importance, as a trap on the sideline would be much more of a threat than in good conditions. I suppose from the viewpoint that, it's more important to take care of the disc and play safe than to gain the extra yard on every play, the yard is devalued by upwind conditions.

But still...all other things being equal, would you opt for better position more often than making a 5-10 yard pass if you're going upwind? I feel like the level of play generally has a much stronger effect on this sort of decision--whether or not you have confidence in your throws, your receiver, etc--than upwind/downwind.

But I'm only a college sophomore. Perhaps I'm missing something.

Jeff said...

"First, take yardage completely out of the equation. It does not belong. Yardage is irrelevant...It's always about position."

I see a problem in this (though maybe I just misunderstand it). Postion is not independent of yardage. Does a throw from the middle of the field at one endline to the middle of the field at the other endline result in an increase in position? I would say yes, but I would say that it's an increase in position due solely to an increase in yardage.

Could field position be defined better as the ability to maintain possession? When you're trapped on the sideline, your chance of maintaining possession is lower than if you have the disc in the middle of the field. In fact, as you get closer to either endzone, your ability to maintain possession would go down as it becomes more likely to throw the disc out the back of the endzone.

A solely possession based offense would attempt to keep the disc near the center of the field (horizontally and vertically) while throwing to any cutter that gets open. Which I think is close to/the same as Billy Berrou's O. If all throwers will complete a set of throws at 100%, then it is simply a matter of a cutter getting open in a place where the thrower can throw to at 100%. Placing yourself in the middle of the field, with six players around you provides the thrower with six options to throw to. On the sideline/endline, it is harder to space those six cutters such that any one can get open.

It seems like a general rule would be "The value of yardage decreases, as the probability of scoring increases." In the above possession-based offense, as long as throwers constantly throw 100% passes, the chance of scoring is 100%, and yardage truly doesn't matter. Eventually (essentially by chance), a 100% throw will be caught in the endzone.

But my general rule would make invalid the upwind/downwind assertions in the post, which I agreed with when I read them. Hmmm....

Tarr said...

Actual answer: the value of yardage decreases the farther DOWNWIND you are. This applies no matter which way you are going.

Think about it. The 20 yards between the upwind goal and the brick mark are gold. If you're going upwind, you're willing to risk a turnover in order to cover those last 20 yards and score. Conversely, if you're going downwind and you're smart, you're willing to risk a turnover to get the disc out of there and downfield. (This is why the flat zone is so effective going upwind - you can trick the other team into being more patient than they should be.)

But on the downwind half of the field, a yard just isn't worth very much. The downwind team can score fairly easily in one pass from anywhere around there. And the upwind team hasn't increased its chance of scoring too much if they move 10 yards forward. The downwind team should just bide its time and look for a good scoring strike, while the upwind team should be working to get that one big throw that starts a fast break and a real scoring opportunity.

Interestingly, the preceeding dynamic also applies when one team is better than the other. Just imagine that the superior team is always going downwind. The value of a yard is very great near the endzone the weaker team is attacking, but not very valueable near the endzone that the stronger team is attacking.

aj said...

Interesting points.

Mackey: I was thinking yardage is more valuable (relative to maintaining possession/improving field position) going downwind because of the importance of protecting the upwind goal.

About the upwind question – good question - I’m not totally sure I agree with my assessment in the upwind scenario.

Jeff: Good comments. You askCould field position be defined better as the ability to maintain possession? I think I want to define improving field position as trading an immediate lower percentage chance to gain yardage for a delayed greater percentage chance of gaining yardage. In order for field position to actually be “improved” the delayed chance must be sufficiently higher percentage to make up for the field position changing throw(s). Of course, this is really just another way of invoking The Rule (A player should always choose the pass that yields that highest percentage chance that his/her team will score this goal).

Tarr: I think I agree with you. I’m not sure I completely agree with this statement though: If you're going upwind, you're willing to risk a turnover in order to cover those last 20 yards and score. I should rephrase – I do agree with you, but I think The Rule still applies. In other words, if throwing a break mark swing will make scoring this goal more likely you should still do it. Perhaps this goes without saying.

aj said...

A couple of other (vaguely related things) I meant to post about Billy’s offense – 1) field position is considerably more important (relative to yardage) in goaltimate, I’d expect his offense to be much more effective in that game. 2) An additional problem with the offense is that it relies on an assumption about the defense that is more true in goaltimate/basketball than it is in ultimate. Good defenders in basketball and goaltimate generally default to playing goal side defense rather than ball side defense. In goaltimate/basketball if someone makes an aggressive ball side defensive play the slightly lower percentage backdoor pass = a goal in goaltimate or a lay-up (a dunk if you play with Chain) in basketball. This means that the pass is almost assuredly worth the risk. In ultimate everyone is going to play ball side defense on handlers and a backdoor pass gets you five yards.

A while ago I actually designed a “motion” goaltimate offense based on the old “flex” basketball offense. We’ve never actually tried to implement it, but I honestly think it develops too slowly to work. Additionally, it’s easier/more amusing to pick defenders off the big yellow hoop rather than your teammates. If I ever get truly desperate for posts maybe I’ll put it up here.

parinella said...

P=f(x,y,wind, flow, opp, sun, temp, etc.)

V = dP/dx

V_x = DP/Dx = dP/dx + dP/dy dy/dx + dP/d(wind) d(wind)/dx + ... (the little d's here are partials)

Evaluate at each point for x=0:70, y=0:40 and draw a series of contour maps. The areas with the steepest curves are where yardage is the most valuable.

BK said...

I think there are several different, but valid, tradeoffs that you can make relative to yardage, and they should be considered separately.

1) when deciding whether to take a shot upfield to gain yardage, considering the yardage you'd gain vs. the risk of the throw
2) deciding how much of your time with the disc you should spend looking for shots upfield compared to moving the disc to a position where shots upfield may be more appealing. In this case you're trading off the yardage gain vs. the increased chance to score from a better position.

I think Tarr was talking about 1 and Crazy Frank talks about 2. I think CFrank's right that most players and coaches overvalue the time spent looking upfield, and undervalue the position of the disc on the field and the orientation of the receiver. Having played a lot of soccer, I can sympathize with his beliefs that soccer coaches have found that the best players to move the ball forward are not those that just received a pass coming back, away from the goal. That its often better to get the ball to someone with better field vision to spend the time looking for the upfield shot. This also can increase your expected completion percentage of the shots since the defense overvalues stopping yardage gains versus stopping improvements in position, which lead to high percentage passes to unguarded or momentarily unduarded players to make long plays with their momentum.

There is also the tradeoff of taking a goal shot vs. the risk of the throw. This is a special case of 1, where instead of just some arbitrary set of yards, its _all_ the remaining yards. This should be considered separately since all other passes include uncertainty. It might be possible to structure a team strategy around the idea of a shooting percentage. You could record it for all scrimmages, and scimmage with a specific target percentage. This percentage would indicate how free players should feel when taking goal chances, knowing that they need to factor in the current team success rate when considering their own shot options. It might be possible to determine an optimal rate for the team, for individuals, and for O\D teams.

This last tradeoff fits into Idris' post about B teams. If your expected scoring percentage from close to the goal is low, your team is better lowering your big shot threshold, or your expected completion percentage at which you'll take the big goal shot.

heacox said...

I agree with BK here, and it seems to go along with some of the stong points I was able to pick out of Billy's writings (granted, this came from the "discussions" in Jim's blog a few months ago, not a reading of the primary material).

In soccer there is something called "play the way you face," which is what BK is talking about. Despite playing soccer for years, I had completely forgotten about this concept until Drew Levine reminder me of it during a Rippers game last year.

Basically, the idea is that if you receive the ball while facing the goal you are defending, it is better to find space in front of you and then pass the ball to a player moving up field, than to try and turn around and take the ball upfield yourself. I recently had the opportunity to see this in action when I caught some of the Manchester/Portsmith (I think) game a couple weekends ago.

Anyway, Billy does seem to promote something similar where the best person to make the upfield pass is not the guy who just came underneath but a person behind the disc, who is in position facing the endzone the offense is attacking.

Hence, you trade a yardage throw for a position throw (fine), but gain by getting the disc to someone who doesn't have to pivot 180 degrees to throw upfield.

I think the detractors of this strategy would point to the extra throw that resets the disc as liable to increase the chance making an unforced error, but I thinnk the strategy has merit.

parinella said...

Heacox, I don't think that the problem is that there is an extra throw, but that it doesn't come up that often in regular flow. Even if you go out of your way to look for this pass, you won't have a lot of them unless you structure your offense around this idea.

heacox said...

Thanks, Jim--that makes sense.

So, is the question then, is it worth structuring your offense like this? I'm actually wondering now if we're beginning to get into the territory of AJ's most recent post about reads and throw priority. What happens if the dish to the handler becomes the priority throw for most of the cutters?

Also, could you institute such a guideline in certain areas of the field, like "any come-back cutters receiving the disc outside of the redzone should make the dish to the handler their priority." However, part of me thinks this reset throw might be even more important inside the redzone.

But as I'm currently hibernating from ultimate, this is all conjecture.