Personally I like "Atlas"... seems like the easiest to do monthly, and can be done relatively quickly for the BOE/BOG

Yeah I understand your point and affirm the validity of it. Again, this is more of a PSA to let GMs know who vote for the simplest one that they it will not be an accurate representation of the rankings, relative to the other options.

If you are interested in reading up on some of the "underlying metrics", I can point you to some articles that explain what it is and it's relevance. As for the reading of the stats, literally speaking, GMs will simply have to go to Corisca and just copy the stat under the specific column and transfer that data into a spreadsheet, ideally, pre-coded with the adjusted values. To give sort of a reference, it took me about 1/2 hour to track all of the 18 skaters (whomever was top-12 and top-6) and the two goaltenders (whomever was top-2) of @BoltsPoint21 's team. Again, that 1/2 hour doesn't incorporate the time it took me to adjust (about 10 minutes).

Just to specify more, this will only have to be done in my suggested increments (two weeks or one month; I'd lean towards one month, as it is a better sample size and isn't as time consuming as a two-week sample). The increments are really only relevant to see how well teams are doing, so if GMs want to do it every two months, or three months, or once at the midterm of the season, the trade deadline, and at the end with the Mint Model, that's fine.

The fixed dates are really only to give an update as to where teams stand. GMs will need to weigh how frequent they want to have updated rankings and how much work they want to put into getting said rankings. If GMs are fine with only knowing where they stand at the midpoint of the season and the trade deadline, then it makes perfect sense to go "all-in" with the Mint Model and bang out the 1hr or so it takes to get those rankings.

This is a list of questions I would ask myself before voting one way or another:

1.) How frequent do I want updated rankings?
2.) How accurate do I want the rankings to be?
3.) How much time am I willing or able to allot to these rankings?
4.) Do I agree that the complexity of the models is correlated with accuracy?

As always, reach out to me with any questions or suggestions on how we can make this the best of all worlds.

I'd refer you to the post above.

Since people are sharing their thoughts, I might as well add mine.

In the overly complicated example you provided, you keep referring to it being done using TBL. That's pretty simple in my mind since Bolts really hasn't changed any of his roster from that of the real life team.
So basically you can pull up an entire team at once. But what about those in game who have players from sometimes 15 or more teams spread out through the real world.
Now you're having to do this research player by player which naturally takes far longer to do.
On top of that, it's asking people to engage in an activity that they may absolutely no interest in, which increases the chance for errors or just for people to say F it, I'm not doing this.
Granted, it took "you" about 1/2 hour because you have more knowledge of this which inherently makes it easier for you.
Add to that, unless something has changed you're in high school and have far more free time than many of the others who have real life jobs, families, commitments etc.
If I read things correctly, not only will GMs have to do all the research on every player ont heir team, they will also have to create a spreadsheet with calculations built in that they may not have any understanding of, which again just screams errors waiting to happen.

No matter what the system looks like int he end, there will no doubt be teams who feel they should be ranked higher than they end up at. I mean we all think our team is good in some ways and better than others.
Remember, these are just my own personal thoughts and do not reflect that of any other GM or BOG member.
We the BOG continue to have behind the scenes discussion about this, but it's not a simple concept to hammer out.
Our drop dead date for this is Oct 2nd. The night before the NHL season begins for real.

You're stuck with protecting him. Sorry.

Teams were required to protect any contracted players with no move clauses (NMCs) with one of the team's slots for protected players, unless the contract expired on July 1, 2017, in which case the NMC was considered void for the draft.[9][10] Players whose NMCs had limited no trade clauses had to still be protected, and any players with NMCs were able to waive the clause and become eligible for the expansion draft.[9]

You can always trade him

True, so WHOEVER has him at the time of expansion draft will need to protect him.

@rangersandislesfan there is also a buyout window before expansion (it's in the rules) so you could just buy him out as well

I can assure you that 31 GMs aren't going to track all this every month, and I can double assure you that the BOG and Commissioner do not have the time to run these for every team. Also, there are GMs that have already said to us that they don't feel this is the best way for the rankings to be made. It gives the GMs who are more familiar with these advanced analytics an advantage, which is not something we are going to allow to happen. It will also assuredly lead to errors being made and skewing with these stats anyways. We will find something that ranks the teams 1-31 in the most efficient, fair and agreed upon method we can find. It will most likely take some of these analytics into consideration, but not to the extent that you are suggesting.

I was trying to figure out if i should be trying to move him or not.

Thanks for letting me know, i actually forgot about that.

Let me know what you would throw in to dump him

Would you give me anything for him 50% retained.

Boychuk + 1st for Cody Ceci

I sorta think that what I am trying to convey is getting lost in transition, so I'll go into more depth so everyone is on the same page. When I say "tracking", this doesn't mean that GMs need to go thru each individual game to get their desired figures. What this means is that GMs need to simply select the timeframe that they want to use within Corsica. For example, if we use two month increments for every update, all I need to do to find my figures is to select October 3 and November 30 and all the data that is pertinent will show up. I also agree that putting this task to the BOEG is unreasonable and too time consuming.

I don't really agree that there is an inherent advantage for the underlying aspect of the rankings for GMs who value them more. The weight of the rankings is universal and this is objective rankings. If GMs value players with "better" underlying metrics than everyone else, then that's not unfair - it's just how they value them. Everyone has access and has the opportunity to access the same information. Simply because certain GMs don't value xGF doesn't mean that the GMs who do have an advantage. As I mentioned above, I can gladly link some articles that explain the underlying metric and it's relevance, all of which is universally accessible.

I agree that there will be some errors, but they can be corrected at the end of the year by having GMs run someone else's numbers to ensure that there is correctness. Any system will have the opportunity for errors, but I agree that my original model has the opportunity for the most.

Again, the one month increment was an example of how often we need the update. The questions that I outlined in the post you cited explain my position.

I don't agree that since I used Bolts' team that it took me any less time. You still have to manually enter in the players name, meaning that I still typed "Stamkos" and "Kucherov" the same way I would have to type "Henrique" and "Simmonds". The 1/2 hour was literally of me typing in the last name of each player and writing down their stats on to a sheet of paper.

I don't agree that simply because I understand the stats means that it's inherently easier. It's literally just writing down data from a spreadsheet. It would take me the same amount of time if this was some sort of soccer game. All I have to do is copy down a number onto another sheet and then aggregate and adjust. No knowledge of the sport is needed to copy down data.

The entire reason that I am not the GM of the Philadelphia Flyers is because I will not have time for this game with any serious interest this year. I am still in high school, but this task literally takes 1/2 hour to copy down data.

I think that your under estimating how easy this task is. It's literally copying down data and then having a coded spreadsheet do the rest of the work. Unless GMs don't know how to use simple functions in a spreadsheet like autosum or multiplying a cell by a specific number, the errors will be minimal.

I don't really think any BOEG member has seriously looked into how my model functions. This isn't a personal attack, it's just what I'm observing. I can gladly tell everyone how I did this, step-by-step. I think that anyone who has major pushback on this, specifically the algorithmic tasks, needs to actually try it out before hand.

Automating stat retrieval might help.

I'm not quite certain if the website allows for that function, but yes, that would be ideal.

Also, I will be timing myself for the Calgary Flames later tonight (spreadsheet is already coded - took about 7-8 mins).

For starters, we should have a general vote on who would want what sort of format we should use instead of it being five people deciding for the whole league. I agree with incorporating some form of analytics to determine who is the best team, but not everyone uses analytics (even though they should), so maybe some percentage can be allocated to everyone submitting their own set of rankings from 1-31 to determine who are the best and worst teams.

For the analytic part, I would be willing to cover a portion of the league that feels they aren't able or don't have enough time to collect statistics. In order to make this easier to track, I would want some sort of roster freeze for a 24-48 hour period in order to collect data.

You can scrape data from any site with sufficient knowledge. However, Corsica seems to have a csv format option. That's one way to import the stats into a spreadsheet, and requires less actual coding knowledge.

Once the data is in a sheet, you can manipulate it any number of ways. Including sending the relevant stats for each player to the respective gm game team, provided rosters are kept up to date on a spreadsheet.

One person could do the bulk of the work if they were so inclined. There would just have to be a middleman and corresponding sheet, adjusting the teams for each player as they were traded.

No, even at his full contract that's too much. Do you see him having value to your team if i retain half?

No he has no value. You are gonna have to dump him

Some people have been interested 50% retained. The NMC i guess changes things. If i'm giving up something to get rid of him i'm gonna have to do it at his full contract.

Also, so there's no confusion: Boychuk is NOT forced to be protected, as his NMC expires at the start of 2019-20. (See: Vermette, Antoine.)

@phillyjabroni:
To @ricochetii 's point, I can put together a spreadsheet that will automate everything. I *might be* willing to download the CSV data from Corsica (3 DL's per month - 5v5, EV, Goaltending) and import the data into the spreadsheet once per month (or whatever time segment we choose). But the GMs would have to provide an accurately spelled roster each month.

I'm going to get a working copy of the sheet going over the next few weeks and I'll share with the group - y'all can help de-bug it