Rulez?

So what are the league mechanics this year?

Is there a salary cap?

Is there a salary floor?

How often is the cap/floor enforced?

Do GMs operate in a free market and negotiate their own trades?

What metrics determine salary?

We have great league mechanics. The BEST league mechanics. Way better than anyone else's league mechanics. 

The salary cap is going to be sky-high. It's going to be huge. You'll be amazed by the salary cap. It's going to be great.

I can't give you details, however. I have no idea what I'm talking about.

We're going to make GM's great again!

I want a t-shirt with "Make GMs great again" ... and then repurpose it and sell it to Owen and his buddies for profit... eh, Grandmaster? ;)

Can we make the soccer players pay for a wall so their balls don't fly over to our side?

How is your team name not  "China!"?

If I can somehow get all the asians on my team, you'll see this happen. Realizing that all asians are not all Chinese... #buttheyalllookthesameanyways

Shouldn't be allowed to be a GM!

That will be included in everybody's feedback form at the end of the season. I'll even try to work it in as the default text so as not to waste anybody's time.

But in 8 years, I'll be a GM again regardless ;)

meow?  Does Kevin know about these forums?

Kevin Hughes's picture

I do I just don't check it often enough (a notification feature would be excellent ...)

The stats and salaries are the same as last year with a slight difference in how to handle absent players. We're looking into the best way to handle this still and keep things reasonably simple. (I want to go with an average for game played salary but we need to evaluate that still)

For reference here is the salary values and subsequent calculation if you care https://github.com/kevinhughes27/parity-server/blob/master/lib/calc_sala...

Because of the way salaries work right now its hard to predict what they will be. On week 2, 4, and 6 we're looking at the average salary and putting the cap and floor at +/- 2% of the average. 

We're aiming for free market but the league realizes that isn't always perfect and sometimes intervenes - sort of like a socialist governement :p

Thanks!  Oh no, it's in Javascript!  My eyes!

I'm sure you've got this under control... but it's Friday afternoon and these user stories are tedious.

How about:

team component = (team's salary increase this week -  team's average salary increase) / num players;

missing player delta = missing player delta average + team component;

Kevin Hughes's picture

This might actually be the best idea anyone has put forward (disclaimer its also probably simplest for me so maybe thats why I like it). By including the team average you pretty much isolate the missing player from affecting the overall team salary. The team overall is accurate and we can base salary caps on that.

A counter example would be a really strong team whose weakest is missing for the first 2 weeks they would get assigned a posible higher salary than they deserve but at the same time if the team was going to win anyways it would pull their salary up.

I am going to consider this and probably implement.

Also whats wrong with JS lol? 

Kevin Hughes's picture

I tried this out here: https://github.com/kevinhughes27/parity-server/pull/64

Results - it didn't seem to change much. If you have actual ideas please comment on Github :)

Without trying to drum a discussion that may have already been had (since I'm a newbie)... I've always wondered why salaries aren't kept zero-sum? Maybe someone can explain why/how this isn't a good idea? I might be missing something totally obvious and in that case feel free to mock me for my ignorance...

- Sum weekly total salary gained in the league

- Divide by # of active players that week to produce average

- Subtract that average from each individual active player's raw salary gain [repeat cumulatively]

Visually: http://imgur.com/a/i1ljO

Pros: 1) Manager salary caps can be kept static throughout the season. 2) Inactive players "Player N" salaries are static and don't need to be fluffed for balance. 3) Tempered relative gain rate for league's most elite players "Player H". 4) An explorable option to calculate two averages to isolate and compute salary change by gender, if desired [this removes M/F from the same vacuum and allows for closer comparison within gender]. 

Cons: 1) A persistently very-low-scoring player "Player M" could have their salary conceivably approach zero. 2) Lower scoring players could potentially feel demoralized by having a decreasing salary number, rather than seeing just very small relative gains at current [although, week 4's max/min delta is 447k already]. 

Con 1 and 2 trumped all cons. "Parity" is a gimmick more than a goal, and hurt feelings aren't worth it.

Plus, we are all worth hundreds of thousands of fake dollars this way.

Some of us are worth millions #expensive #leadingscorer #checkthenumbers

I figured this was the reason, and agree that it's likely not worth it to make the numbers easier to work with

There are a number of ways that salaries could've been managed for missing players.  It all comes out to none of them will be perfect and parity would possibly be achieved.  Maybe.

 

I tried to come up with a salary based model to propose to the league  that would be most fair but then I started thinking of female models and I got nothing accomplished after.

Kevin Hughes's picture

I would still love some feedback on this idea https://github.com/kevinhughes27/parity-server/pull/64 (its not just code there is a graph and a full description) I think it helps.