Tuesday, November 10, 2009

You haven't left the GOAT Herd yet... Have You?


If you have had the pleasure of playing or watching GOAT or Furious George, you'd be hard pressed to dispute that:
  • Both teams are good at 'frisbee'
  • Both teams are very good elite teams at the UPA level
  • Both teams had disappointing UPA runs this fall
Usually when a team I support or am involved with does well, I think I must come across as a cynic devil's advocate. I don't mean to be, I just simply don't like it when people make too much of a win and make incorrect inferences from them. It's the complete opposite when said team doesn't do well. I get to be honestly positive when people are going over the top negative.

The wheels fell off for GOAT at nationals. Furious lost the last spot to finals at regionals versus a team they beat at the same tournament the day before. Hence, we have been subjected to "what's wrong with GOAT/Monkey" questions ever since.

These teams are elite teams. They played elite teams and won and lost. Heading into each game, there were certain probabilities that GOAT/Furious would win or lose. I think people would be shocked at how evenly likely teams at this level can win or lose and how arbitrary the outcomes can be.

A good example: Dave Berri of Wages explains Randomness and how people are fooled. (original concept from Nassim Taleb's book):

"At the start of each football game a coin is tossed to determine who will receive the opening kick-off. Let’s imagine if instead of just one team calling heads or tails the fans in attendance were also asked to make a call. And let’s further imagine that if you make the correct call, you get to stay. But if you are wrong, you have to leave.

Okay, now let’s do a bit more imagining. Let’s say 80,000 fans are in attendance – and since fans know it is a fair coin (equally likely to be heads or tails) — about 40,000 make the wrong call. So these fans exit the building. After they are gone, let’s imagine we play the same game again. This time, about 20,000 fans are incorrect and they depart. And then we play it again, and again, and again… After three tosses we are left with about 10,000 fans. After seven tosses there are about 625 fans. After twelve tosses we should still have about 40 people left in the stands.

Now what have these 40 people learned? These people have just called a coin flip correctly twelve consecutive times. Clearly these people are incredible at this game.

If we play the game one more time, though, we should expect about 20 more to depart. What will these departing fans have learned? Well, clearly they just didn’t match-up with the 20 who got the 13th call correctly. And they better go home and figure out why that particular match-up didn’t work if they ever wish to see another football game.

The Point?

Makes you realize how good Fury and Riot are in the women's division compared to the rest of the field.

People often have problems understanding randomness. And what it says is relevant to how people see the results of a team at UPA finals or tournament finishes in general.

As for our Canadian teams
  • Furious proved they are able to win out of their ultra tough region, but they simply didn't win the game that counted.
  • GOAT shouldn't over-react to their recent slip.. unless there is a clear team in their team/opposition that leads them to conclude their chances to win have decreased.
So don't throw out the Furious or GOAT foam finger yet. You could replay Finals again this weekend and have completely different results on the open side.

This shouldn't stop teams from doing things in the off season to improve their probabilities and odds for next year. Training, studying video, strategy, recruitment, and all the fun stuff!


H.C. said...

I am not sure how the randomness scenario is relevant here. If you play the "Stadium Game" a second time the odds of those same 20 people with 12 correct coin toss calls in a row would be astronomical.

In the case of the UPA Nationals, most of the same teams return year after year to the finals after "tossing their coins" through multiple games at Sectionals and Regionals. The random component seems really small to me.

westwell said...

H.C., I think you're taking the "coin toss" metaphor a little too much at face value (pun intended?) I think the analogy Steve used is a valuable one. When referring to Sectionals and Regionals, instead of using a coin toss analogy, let's use a die. When GOAT matches up against Replicants at Regionals, for example, GOAT will win whenever the die rolls on any number other than one; thus, they will win if the die lands on the numbers two through six. When GOAT plays Thruway at Sectionals, the die we use to roll now has 100 sides, numbered 1-100, and GOAT will only lose when the die lands on one.

When referring to the UPA Championships, the teams in question have a much stronger parity, and thus the coin toss analogy becomes more relevant. GOAT beat Revolver at ECC just a few months prior to Sarasota, but at the championships GOAT didn't make quarters and Revolver made it to the finals. Maybe the coin is slightly favoured to land on heads, but it becomes more of a 55/45 scenario rather than 99/1.

In reference to Fury and Riot, even at the Championships, it seemed unimaginable that they wouldn't meet in the finals. Previous years (and this year's tournaments) indicated their sheer strength over all other women's teams. Fury has only lost to Riot this year, Riot had lost to 3 teams all season before UPAs. Their dominance is so apparent throughout North America that even at the championships they seldom flip coins; they prefer rolling dice.

T1000 said...

Well, you're right, H.C., but you're talking about the odds of 20people correctly calling a fair coin flip more than 24 times in a row. Obviously, that would be a ridiculous analogy for sports.

Conversely, though, is it reasonable that a "better" team should never lose to a "worse" team? Of course not; upsets happen all the time. Here, you can make a choice in worldview; either the "better" team simply was not as good this time, or random chance adjusted their averages this time. Whichever view you choose depends on whether you're inclined to determinism or fatalism. Regardless, the two descriptions are equivalent with regard to the statistics.

If you are resigned to random chance, then all your strategy, nutrition, focus, and hours of preparation amount to weighting a coin in a flip, unique among many. But the final outcome was subject to chance.

If you are a determinist, then the outcome is determined by your will and skill in each game. Sometimes, you are in "the zone," and sometimes, you are not. Your statistics result from your average ability to enact your will over your opponents'. And statistically, you are just priodically unable to do so. Again, this is something we all have experience with, when you win or lose unexpectedly. Why did it happen? Why that time?

The effects of randomness are still noticeable, but they are recognized and described in different ways.

RJ said...

wow, this is one of the most lame conversations about ultimate frisbee I've heard in my life. Even if I heard it 100 times, and given a mild propensity to change my opinion based on weather, and my last bowel movement, I'd still rather slam my head on my desk then read these comments again....

Maybe Goat lost because their studs are getting older and are giving ground to better, younger defenders

Maybe Goat lost because of the level of competition in their own backyards make it impossible to train and prepare at an appropriate level

Maybe Goat lost because ultimate is gaining ground in the united states as a legit sport and not in canada

maybe Goat lost because they brought in players from sweden, England and Ottawa to play major roles in the UPAs instead of concentrating on their core of young talent like Mark Lloyd and Malcolm Johnston

Maybe all the vets that left over the last couple years like Jammer, Sawoda, graham Fletcher really hurt their depth more than they realized before.

Or maybe they just can't play in big game scenarios, and lose at crunch time.

I don't know... maybe.

T1000 said...

Really? I think probability is fascinating. I might even read the comments again.

Jeff said...
This comment has been removed by the author.
Jeff said...

This was a severe wake up call to GOAT. While we keep in mind Steve's 'randomness' analogy about anyone can beat anyone on a given day, that's only to not mistake a poor showing as an inability to succeed. In reality the idea that we have a ton of stuff to work on is much more fresh in our minds then any sort of luck. The parity at UPAs is also something to be acknowledged as a 14th seed like Streetgang never had the ability to take down higher seeds like Ironside and as well GOAT, although at this point we can't claim to be much of a higher seed. But in reality - GOAT's goal is not to hold an 8th seed but instead to be among the elite and, on the final Sunday in Florida, win the metaphorical coin toss.

In that sense this complete crash may be beneficial as it will allow us to fix some bad habits that might not be fixed with continual quarterfinals or better performances in Sarasota.

The only part of RJ's post I would agree with is that this was one of the lamest frisbee conversations.

n said...

I'd be interested in knowing a) whether the heat made any difference, and b) what goat thinks of the ironside comment in this blog (http://mittechultimate.blogspot.com/2009/11/nationals.html) about their prep for game 1.

Sport Management Steven said...

Scotty- You captured my example exactly and improved it. I think the teaching profession is going to suit you well!!

RJ- Your post was hilarious (I mean that sincerely). This post was to get people away from assumptions, many of the same ones you just made. I realize that you might be right on some, but player stats are a better indicator of who/what did or did not work. (If GOAT sends me video, I will break it down)

For such a boring post, it illicited some quick response. :)

Hodge said...

"maybe Goat lost because they brought in players from sweden, England and Ottawa to play major roles in the UPAs instead of concentrating on their core of young talent like Mark Lloyd and Malcolm Johnston"

While GOAT brought in Derek Alexander from Ottawa, it's not like they 'brought in' the guys from England and Sweden. They happen to be in Toronto for the time being, so why wouldn't they have them play with the team, obviously they are good enough to make the team. The way you make it sound, the guys came to the tournament to play with GOAT, but they've been playing with them most of the year.

jhaig said...

"I realize that you might be right on some, but player stats are a better indicator of who/what did or did not work. (If GOAT sends me video, I will break it down)"

I will agree Steve that if you have the right metrics that stats can tell a more accurate story then what people think they saw, but I think that you will agree that we are a long long way from being that advanced with our frisbee stats.

I also got a laugh at this:

"So don't throw out the Furious or GOAT foam finger yet. You could replay Finals again this weekend and have completely different results on the open side."

When a week or so ago you had this in your women's prediction for Capitals:

"And if they don't make semis, at the very least, I suggest they change their merger system or personnel. I really like how these ladies have stuck to their program and system for several years (others could learn from this), but they must take the next step."

Sport Management Steven said...

Who taught Haig how to quote? This could be trouble! :)

-Yes, we're a ways off from having robust stats in ultimate. However some of the better thinkers of the game have been collected their team stats for years and breaking them down in great detail, so when a critical number of elite teams do it, we might be able to share data of past years and determine some elite level averages and a better understanding.

-The comment about Capitals was made partly to entice comments. However, I did mean it. When you engage in a harvest strategy every fall, and see players in each city miss out on fall ultimate as a direct result of efforts focused on the combo team.. then this combo team better be improving every year.

2005- 6th
2007- 5th
2008- 9th

If the goal is to win it all, they didn't seem to be narrowing the gap with Riot or Fury. (Losing 15-3 to Fury on day one in 2008).

Making it to semis was good progress. Beat Riot this year on day two was also huge.