Draft Format Proposal: Auction

I calculated the % difference between a players current salary and a calculated salary based on a curve fit to a plot of current salaries vs. draft position. I also did this with my own metric of player quality instead of salary just to compare.

Some players have played better than their draft position would suggest, some worse - no surprise there. Adding up the %difference values for each team as it was drafted, the teams that have done better so far have the highest sums and the teams that have not done well have the lowest sums (also not really a surprise). So either the GMs that have done well were able to draft good players farther into the draft, or the players they drafted have just played better than their draft position.

What is clear from the data is that the draft order had little or no bearing on GMs ability to draft better players earlier. Lots of players who have been playing like they should have been drafted in the first half of the draft were drafted in the second half, and vice versa. Probably a lot of this is due to GMs not actually knowing all the players that well, which should get better in the second half of the season. I think not much of this is due to the draft format itself. My point is that a change to the draft format would not have much effect on initial team balancing, unless all the GMs have much better analysis/understanding on how each player will affect game outcomes.

The idea from Amos about forcing GMs to draft themselves in certain rounds makes sense - I don't think it would actually require the creation of tiers of players - just agreeing on what draft position the GMs lose to draft themselves. Also drafting genders separately might help ease complexity of decision making, leading to better balance.

There aren't enough Alessandros to go around. We could make pairs though. 

Line 1: Alessandro, Chris, Laura, Melissa, Rob, Steve.

Line 2: Alessandro, Chris, Laura, Melissa, Rob, Steve.

After the break, some nefarious GM would draft Alessandro, Alessandro, Aleks, Alex, Chris, Chris, Chris, Christine, Krys, Jaime, Jamie, Melissa, Melissa. It would be awesome to prevent switches.

Mehmet's strategy from last year would not work in this format, and his team was in the finals. His team was largely criticized as being "weak", but in the second half of the year he won more than he lost and kept every game close.

In year 1, Seb's team was of similar construction, and won.

In year beta, Martin and Glen Ford played fetch for a bunch of games while Hadrian bottomed out his roster so he could give and go with Kinley, but it was Rob's balanced offense and great use of novice players that won.

> but the way it's done now doesn't really jive with how OCUA wants the league to unfold at the beginning of the season.

I said this in the podcast, and I meant it: "parity" as defined by salaries is a gimmick. It will not work perfectly (but it does work pretty well). Your suggestion requires people to agree on "tiers" and it's all based on a subjective valuing of players. The state of the league right now is largely a result of draft results (there are a few notable draft decisions that I would suggest were mistakes). Players taken earlier than they should, or later than they should, will be drafted much closer to where they should next time around because of the knowledge and understanding we have of them after many weeks of play.

This alone fixes most of the parity issues in the league, and it then becomes a game of building a cohesive team with a good on field strategy.

I get what you're trying to do with your suggestion, but you tie GMs hands to specific subjective valuations of players and reduce or eliminate the freedom to game plan in the way that several of the most notably savvy GMs have in the three years this league has run.

Building balanced teams (whatever "balanced teams" are) isn't easy. Strong teams on paper regularly do not mesh, weaker teams regularly play well together and perform better than expected, etc.

We played Kindha this week and won by 1 point, and I'm really trying not to lose.

>  would we get more balanced outcomes if salaries scaled more to performance.

The math would just be "different".

Because the economy of the league is interconnected, small gaps in salary and performance should be seen as fairly significant (but they aren't because 725k looks a lot like 775k). It's an optics thing.

The flip side of this is team effects: strong teams have strong collective outcomes, and for maybe the first time in my life I will unironically say that a rising tide lifts all boats.

Put another way: how many players on Boss or Kaboom are undervalued precisely because those teams have scored few goals? How many players on Basket are overvalued because the team has scored a lot?

Parity and math are fun you guys!

> Sometimes I feel like there should be an extra adjustment for the difference in how much you win/lose.

We did this year 1. Every player got +/- $1000 for the +/- of their game. It didn't do much to the numbers, and because a player's salary is part of the league's salary as a whole, the math was a bit of a wash.

> Like a Boss and SOS where there was a 9 point difference (SOS won the game). Yet, Like a Boss owes $188,070 and SOS has space for $359,053 in fantasy money

I'm not as well informed on the math of the season as I have been in past years, but it looks like SOS is in cap right now (no need for trades), and Boss was left above cap last trade cycle and is still a bit over the cap as a result.

When we look at the cap, in the way that we have it implemented, it's really just 96 player salaries summed, and divided by 12, +/- a small %. This creates an upper (cap) and lower (floor) limit for an acceptable team salary. A team's 12 players are then compared against that gap to see if it's in range (gap = cap - floor).

Every player affects EVERY team's cap. Every team affects EVERY team's cap.

When you mix in effects that we've talked about before (teams that score on few passes v. lots of passes, teams that score few goals relative to league average, games that are short relative to a week's average) and certain strategic and systemic issues come up. The economy has a few flaws that probably can't be fixed, and there are things that can be (marginally) exploited.

Which all comes back to it not being a perfect system. I don't think it ever can be. Coordinators and GMs are the stakeholders of "parity", and need to work together to create it, while simultaneously working towards their own selfish ends (GMs want to win, coordinators want to stay out of managing the league directly because there's a lot of other things to do).

Whatever Parity is, it ultimately needs to be fun. Draft rules should be fair and fun. Trades should be fair, fit the goal of "parity", and be fun. GMs should try to build winning teams that don't exploit too many of the inefficiencies, and have those teams be fun. Coordinators should create a sandbox for GMs to play in that have understood rules, encourage balance, and is fun.

Frisbee is fun you guys!

Part of the issue with this game was that Boss was more than half subs so most of our players weren't affected by our less than awesome game last week.