What is parity?

Submitted by Will Leckie on Wed, 2016-01-06 10:25

Forums:

How is parity defined?  

Is it simply an attempt to balance team salary across all teams in the league?  Point differential across a number of games among all teams?  Parity in the overall won/loss records in the standings?  Something else?

How do the game masters decide which roster moves to make?

How do we know how close (or far) we are to acheiving "parity"?

((And honestly, my post has nothing at all to do with Monday night's debacle - that's all on our team, nothing at all to do with the league or how the game masters manage it.  These are curious questions that I've had all season)).

I took a very quick look at point differentials in each game for this season's Parity League and compared to previous seasons of Thursday IAD/SEL.  On average across a large number of games, the point differentials are much closer in SEL.  I haven't actually calculated all the point differentials and looked at the mean and variance to quantitatively compare the two leagues, but just eyeballing it's clear that average point differential is much higher in Parity League.  I can think of a few possible reasons for this:

  • The mindset of players and the level of organization of the captains in Parity League is probably different than in SEL; even the competitive players in Parity probably approach the game a little more lightheartedly than they likely would their SEL games.
  • In Parity League the "cross section" or variance of skill and experience among players in the league is quite a bit bigger than in SEL, so you could make the argument that the teams in SEL are inherently better balanced and that it follows that games are closer in SEL.  You could then make the argument that in Parity League since the variance of skill among players is larger, the point differentials in games will naturally be larger regardless of how well balanced the team salaries are.
  • In Parity League the teams are shuffled every three weeks (?) so chemistry/familiarity among players on a given team can suffer.  In the last round of trades, some teams had literally half of their rosters turn over.  Chemistry/familiarity is further compounded by the larger variance in skill/experience among players in Parity League.  It's been a couple of seasons since I played in SEL but as I recall, the teams were only rebalanced once per session, and even then there were only a handful of players moved.

There are probably other factors at play as well, but these are the most obvious to me.  

Of course, I'm not expecting compare Parity to SEL in an apples-to-apples sense and find that Parity has more parity than SEL.  And I'm certainly not complaining about the way the Parity League is run, or the competitiveness of the games.  

I'm simply thinking about how we define parity, and how much we can control it via trades, versus how much inherent variance there is in our level of play.

Mehmet Karman

Wed, 2016-01-06 11:34

I sent a message to other team captains/coordinators on this - my idea of parity is based on the SBF (Spence Balance Factor).  For explanation see postscript.

Essentially I think the most important aspect of parity is the expectation of close games, such that going into a game you don't know who is going to win as both sides have comparable talent levels and whichever team plays better will take the game.  I would call a league unbalanced or lacking parity if team A goes into a game with team B expecting to lose due to talent disparity.  Therefore won/loss isn't as vital in my opinion, assuming no blowout wins or losses.  An odd hypothetical, but I think if a league like ours had the top and bottom teams with the corresponding records of 9-0 and 0-9 but with +/-'s of +18 and -18 that could still be defined as achieving parity.

In the Parity league, there is an assumption that equal salaries will mean roughly equal talent at least on a macro team sense, not comparing individual to individual.  Through the trades required by salary cap compliance in theory teams get balanced. The coordinators' control over trades to acheive salary cap (and those extra trades unrelated to salary cap) tinker with the overall balance, sometimes effectively, sometimes not.

Stealing Steve's chart from a week ago, and adding Wk8:

Here is the average point disparity per week (score difference for each game)

Wk 1 = 5.5 (4,7,6,5)

Wk 2 = 6.75 (13,6,5,3)

Wk 3 = 5.25 (2,6,2,11)

Wk 4 = 3.75 (2,3,7,3)

Wk 5 = 3.25 (1,2,9,1)

Wk 6 = 3.25 (4,6,0,3)

Wk 7 = 9.25 (11,12,8,6)

Wk 8 = 7.25  (5,3,20,1)

Wk 4, 5 and 6 (after the first round of trades) look pretty good.  Trades after week 6, somewhat less so.  Wk 8 was actually better than Wk 7 other than the outlier game Will alluded to (my sympathies, sometime's everything seems to go wrong, but for my team everything went wrong last week against Geoff and Owen as well, coincidence?)

Taking a look at this season of the SEL league (shout out to the old guys and my masters friends for having a great season on Game of Throws!) for example had the following SBF for their most recent week of games:

Wk 7 = 11.67 (11, 11, 13)

Back in Wk 5 there was a more balanced night of games (4,5,4) = 4.33

The Wk 7 number (11.67) is higher than anything in the parity league, so far.

I'm not a betting man, but based on past performance I would think the game tomorrow between Game of Thrones and Astronutts in the SEL league is trending towards a certain result, and that might illustrate a lack of parity in that particular league.  To be clear I have no idea how rosters work in that league and am not offering any critique, just observing the data.

Mehmet

P.S.

What the hell is SBF you may ask?  Here's where I demonstrate my age - in the early ages before Parity coordinators in competitive leagues were tasked with making balanced teams where they had absolute power over all trades.  They were judged, in part, on a made up stat thought up by a friend of mine, Colin Spencer James - read below for more.  It's no longer calculated in leaguerunner (Zuluru). 

 

 From an old FAQ -  The SBF is the "Spence Balance Factor". It was suggested during discussions about Thursday Indoor 2003 as a way to see how balanced a particular tier was.The Leaguerunner implementation of the SBF is a straight average of the point differentials (as a positive number) of all games played. In the case of the League SBF, this is all games in the league. For the Team SBF, this is all games played by that team.Assuming that this measurement has actual statistical validity (something that has not yet been proven), these numbers can be used to measure the 'closeness' of game scores for a tier/division or a team. Tiers with lower SBF values generally have closer games, and thus are more balanced. Teams whose SBF is much higher than the SBF for the tier they're in may be good candidates for moving up or down (depending on their win record) to another tier.

 

James King

Thu, 2016-01-07 10:34

My quick, non-statistical, admittedly biased, two cents regarding Mehmet's statement that GoT vs Astronutts is "trending toward a certain result": untrue. As Keates indicated, Seb put in place a system on GoT that maximizes his players' strengths and has been highly effective in the short-term.  In contrast, my take is that TK and Jess have focused on the long-term by choosing to develop individual player skills while ensuring that everyone touches the disc lots over the course of a game.  Had TK and Jess decided to focus on winning, they could have simply implemented a strategy to share the disc between their high end players, which would be highly effective, but unpopular. To that end, if one of Astronutts' high end players decides to "take over the game" tonight, with all due respect to the members of GoT, good luck trying to stop them. 

For me, parity is about having all members of a team contribute to close and spirited games against opponents of a similar caliber. Being a betting man, I'm willing to pit my lucky quarter against Mehmet's that over the course of the last three games, the Astronutts will achieve that version of parity.

Jared Cohen

Thu, 2016-01-07 14:17

As a player on GoT, my strongest objection to this discussion is the notion that we play with any kind of system, let alone some fancy system that Seb has allegedly put in place. The closest thing we have to a system is to maintain possession long enough to find Ashlin, sprinting open to the endzone.

Beyond that, our success has largely been due to ritualistic animal sacrifice and clean living.  With eight grandmasters age players, any game without a broken hip is a victory.  But I'm with James.  Astronutts are far more skilled than their record would indicate, and I wouldn't bet against them even though I'm playing against them.

 

Christopher Keates

Thu, 2016-01-07 14:58

http://www.ocua.ca/zuluru/teams/view/team:9302

Ariel Untiveros

Thu, 2016-01-07 15:44

I don't care what anyone says, Ken Maclean looks (from a distance) and runs like someone in his mid 20s

Sebastien Belanger

Fri, 2016-01-08 11:04

Honestly, the success of our team is due primarily to an emphasis on disciplin. We have some basic strategies on defense and offense but system has very little to do with it. Also, disciplin is only possible if everyone feels included in the game play. We pass to whoever is open, even if that person is Stan ;-).

I do agree that I put an emphasis on winning, but player development is a huge part of captaining a team. Both can be achieved and if you don't believe me, watch Justine coach up everyone around her. I also try not to kick scoreboards ;-)

Mehmet Karman

Fri, 2016-01-08 17:19

 

I certainly shouldn't have neglected the Ashlin factor in my shout out, as I was fortunate enough to play with her last summer and she just scored 10 goals against my team in Parity a few days ago.  Of the 30 points (a current season SEL record) GoT scored last night I'm assuming she was involved in quite a few.  Let's definitely credit her, Kenny Mac's ageless legs (and charm) and Jared's advanced hip health before we credit Seb, it should just be a rule!

Leaving the player development vision vs. win at all costs to the point of scoreboard sacrifice method debate aside the score from last night's game (30-18) wouldn't seem to fit into most people's version of Parity*.  The two leagues may have different goals but I'm hoping after the redraft in the Parity league that's coming up we'll have less blowouts, but just as much if not more heckling.

Mehmet

*Then again my record in a certain fantasy football league season which Mr. King also participates in can't really be associated with the word "parity", perhaps even "competency" despite my best efforts.

 

 

Christopher Keates

Sun, 2016-01-10 22:23

"Nearing Competency."

Sebastien Belanger

Mon, 2016-01-11 09:35

So the Rule of giving credit should go something like Ash, Ludwick everyone else then Mike, Seb and Stan in some order at the bottom. Does that sound about right ;-)

Will Leckie

Wed, 2016-01-06 12:10

Two things:

1) Inherent Variance (TM) would be an excellent team name.  I'll happily license the right to use it to any interested party :)

2) This is what I find really interesting about the concept of "parity" and why I really like this league:  we can choose how we define parity and we can try to somehow control parity with roster moves, but I'd posit:

  • The inherent variance in our level of play will limit any quantitative measure of "parity"
  • The lack of a large enough sample size for statistics (three weeks between trades) means that meaningful quantitative decisions for roster moves are difficult if not impossible.
  • These two effects are simply compounded the more often we make roster moves, or the more players involved in a given roster move.

​Again, I'm not complaining at all; I'm praising the Parity for the interesting thought debtes and conversations it gives us.

I think Keates did a fantastic job summarizing the league and all of its idiosyncracies with this quote:

Mostly, the league is meant to be fun. The money and salary are a gimmick, albeit a fun and semi-useful one. All players should be included on the field as much as possible, people and teams should play to maximize fun, learning (the league has a huge skill gap between the most and least experienced players), and winning and not salary efficiency. Yes you should bid for that disk you might not catch, even though we might say "haha you dropped it!"

 

It's a game. The heckling isn't serious. The salary doesn't measure your worth on the field. It's a meta-game abstracted on top of frisbee. When people are worried about playing the meta-game and not the frisbee, it's just less fun for everyone."

(http://www.ocua.ca/comment/33093#comment-33093)

 

PS, Bonus third thing:

3)  "For my team everything went wrong last week against Geoff and Owen as well, coincidence?" - point taken.  Great players will usually beat good players.

Christopher Keates

Wed, 2016-01-06 12:53

I'm gonna reply to you and Mehmet in one big post. Quotes are to reference things I'm addressing.
 
> How do the game masters decide which roster moves to make?
 
Whimsy, mostly.
 
I kid. Lots of reasons. Parity, sometimes. OCUA goals, others. After week 9 it will be almost entirely parity.
 
> How do we know how close (or far) we are to acheiving "parity"?
 
I would argue that on any given week we are pretty far away from "parity" and always will be. We are balancing the game based on metrics that do not fully reflect the value of a player on the field. I say this a lot: salary is a gimmick. It's a fun gimmick, and perhaps a somewhat meaningful gimmick, but it is still a gimmick.
 
The games played on week 7 had "achieved parity" by salary, but if anyone thinks those teams are fair and balanced, they are wrong. Salary is so wrong that it currently suggests I am a better player than Owen Lumley and Geofford Seaborn. This is very wrong. It's like, the wrongest. It is the most wrong. It is basically, peak wrong.
 
It also says I'm better than Alex Bush which is 100% correct and not wrong.
 
RE: why the score gaps are bigger, you have the answer here:
 
> In Parity League the "cross section" or variance of skill and experience among players in the league is quite a bit bigger than in SEL
 
And this is a contributing factor with respect to this:
 
> In Parity League the teams are shuffled every three weeks (?) so chemistry/familiarity among players on a given team can suffer
 
Better players can move more easily and integrate in new teams and systems, because of their (presumably) better understanding of timing, spacing, and strategy. Better players play better with one another. The higher the collective "game sense" on a team, irrespective of other factors (athleticism, for example), the better everyone on the field looks.
 
I actually really like that Mehmet has brought up Game of Throws v. Astronutts in the Thursday SEL league. Sebastien (who is likely reading this) has a team that is collectively older than every other roster. He (in some areas) is giving up a bit in terms of athleticism relative to some of the other rosters. But he has a team that spaces a field very well, and plays intelligently in all areas of the game. Positioning on his roster is good in all facets of the game, and Seb is good at implementing team systems. He has smart players who will do what he wants. He wins consistently.
 
Alex Bush and I agree that he has arguably the strongest team in that league. And yet... in early-season discussions about his team, Seb felt he did not have enough athletes to compete. He is wrong (you're wrong, Seb), and his nearly unbeaten record reflects this.
 
Astronutts, on the other hand, has a big skill, experience and field-awareness gap between the top and bottom of the roster. The sort of systemic and strategic cohesion Seb's team easily sticks to can be hard when the gap in how the game is viewed by all players on the field is so different.
 
SEL this session has been difficult to balance because the cross section and variance that Will talks about is much greater than in past seasons. Everyone is still very good, athletic, experienced, etc. The gap between the top and bottom of the league is broader. Parity is not easy.
 
In the Parity league, things will always vary. At the moment the winningest team is 5-3 and the losingest (is this a word?) team is 3-5. That's not bad. In short runs, even week to week or 3-game sets to 3-game sets, parity will probably not happen. Winning teams end up above salary caps, and we use salary as a proxy for player quality. Winning teams end up losing good players, sometimes even key players. They then lose for a bit. Sometimes the swings are big, sometimes they are small. 
 
Over 20 weeks, I expect win/loss and +/- to be fairly close, same as last year. But I also expect teams to win close games that they feel were much easier than the score reflects, and lose games by big margins and feel like the game was closer than the score reflects. I expect key pieces to be absent sometimes, and players to miss games and be replaced by good subs who don't immediately mesh well with the team systems, and players to have really off games.
 
Balanced teams, in both short and long runs, is really, really hard. In all the leagues I've run, it's harder to achieve the bigger the gap between the best and worst players on a team or in the league in general. Parity and SEL are both strong leagues where everyone knows what they are doing and everyone is coming from a fairly strong frisbee background, and balance is still hard.
 
I think it was 2012 SEL East, where we somehow managed to make teams where the average +/- per game was about 2, and the winning record was 6-4. There were something like 10 ties over the course of half the season. The final standings are lost in the ether (Zuluru archives), but that was basically a unicorn of a season and I'm not even really sure I saw it in the first place.

Adam MacDonald

Wed, 2016-01-06 13:45

As to Will's comment on drive or mental approach i play both sel and parity currently. my approach mentally is the same to both have fun, work on my weakest aspects, spike every D, and try to win as a team. I would say my sel team has a better defined offensive and defensive strategy that more people buy into but we also have had very few changes so cohesion and buy in time have been a lot longer.

Ps parity has a far better "fanbase" or chirping section than sel does so win parity

Justine Price

Wed, 2016-01-06 14:23

I always want Parity to lead to constant close games, I would love every night to be a total guess as to outcome. Any team could win "any given Sunday", or in our case, Monday. The one-point game is just so much fun! We had the not-coveted late slot on Monday, in frigid conditions, so usually not a great quality of game. But we took it to 17-16 in a super fun, constantly contested game. For me, even at playoffs every game outcome should be unknown. The eighth seeded team should be able to have a one-point game with the first seeded team, and so on. This involves trying to ensure there are good match ups, not just for the Geoffords and Keates, but also for the less skilled players, so they can be regularly part of the action, not just  a couple of touches.

Personally, I find Parity fun. I throw fistfuls of dollars in the air regularly by putting up deep throws to whoever is stretching out the field. Pew! Pew! Pew! Dollars for lots of people up for grabs. I'd be more conservative in SEL (naw, not really, but I may not be so lovingly encouraged. Chirp! Chirp!).

Watching Rob Ives, a highly competitive minded player, grin as he got schooled deep by Khalid this week was just so much fun. Demonstrated respect and fun.

Great comments by all, and yes, I agree with Keates original quote. Nicely put.

 

Alex Davis

Wed, 2016-01-06 14:02

Parity will be achieved when the D-line rises up and the dictatorship of the proletariat begins.

Christopher Keates

Wed, 2016-01-06 14:20

Where is the подобно button, comrade?

Alessandro Cola...

Wed, 2016-01-06 16:30

Alex Davis' stats-based erection couldn't let him not comment in a stats-based frisbee thread.

 

(Of course, my Alex Davis-based erection couldn't let me not comment on it.)

Steve Bisang

Wed, 2016-01-06 15:39

There are a few elite players, and a few on the lower end, and the majority in the middle.

Therefore, there is not as fair of a distribution that you would get in SEL, because SEL was selective wth two divisions. There are not enough players  at either end of the spectrum to distribute evenly amongst all teams, inevitably, some teams are going to be stronger and some weaker.  So parity is moreso achieved, as Keates kind of mentioned, in terms of a team's overall win/loss record.

Kris Bergmann

Thu, 2016-01-07 02:11

Hopefully not toooo off-topic...

Although Parity is right there in the name, I always thought that a major component of this league was to demonstrate the ability of captains to make the most of a disparate group of players. You have a group of people of varying skill levels, althleticism, familiarity with each other, etc. but that's where strategy and leadership come in. I've played on teams that didn't seem to really know what to do with me, but I've also been on teams that learn how to use my strengths and minimize my weaknesses - that's when this league really becomes something special to me.

That, and I think the whole host of Parity captains have made me a better player (in a way that I wasn't achieving through summer league play) by giving me the opportunity to learn from some great players.