thanks for the info. It makes me want to join one of the top tier worlds

Posted by strikeout26 on 5/31/2018 7:51:00 AM (view original):
As far as "wait time", how do you differentiate between the worlds that are quick to hit the "go public" button and the worlds who are willing to wait it out as a private world?

I do not differentiate.

Posted by strikeout26 on 5/31/2018 8:01:00 AM (view original):

Posted by mchales_army on 5/30/2018 10:08:00 AM (view original):
This ranking takes a look at turnover, wait time between seasons, and several factors that determine a "parity score".

The first set of numbers represent the raw data for each world and then the second set is that world's "score" for each category.

Category scores are determined by that world's distance from the "baseline world" in each category.

A baseline world is defined as the world with the best raw data in that particular category.

Below is a legend for the abbreviations in the chart:


TO = Turnover. How many total owners have been replaced over the last 4 seasons.

Baseline - 3 - BBWAA

Wait = Wait time between seasons. This number equals how many days longer than 360 the world took to complete 4 seasons.

Baseline - 4 - The MLB/Cooperstown

PO = Playoffs. How many different franchises have appeared in the playoffs over the last 4 seasons.

Baseline - 27 - Happy Jack/Boggs

WS = World Series. How many different teams have made it to the World Series over the last 4 seasons.

Baseline - 8 - 10 tied

100W = How many 100 win + teams during the last 4 seasons.

Baseline - 2 - Cooperstown

100L = How many 100 loss + teams during the last 4 seasons.

Baseline - 4 - Cooperstown/Addicted Users

110W = How many 110 win + teams during the last 4 seasons.

Baseline - 0 - 20 tied

110L = How many 110 loss + teams during the last 4 seasons.

Baseline - 0 - 21 tied

Most W = Most wins by any one team, during any single season, over the last 4 seasons.

Baseline - 101 - Capra

Most L = Most losses by any one team, during any single season, over the last 4 seasons.

Baseline - 101 - Cooperstown

Real Life MLB (2014-2017)

Playoffs	WS	100 W	100 L	110 W	110 L	MOST W	MOST L
21	7	5	1	0		104	102

EDIT: It should be noted that in the "Most W and Most L" column, the actual highest and lowest number for each world was dropped so that one outlier in a skewed season didn't negatively impact a world too badly. That is why you might see a world scored with one instance of having had a team win or lose 110+ but their "Most W" or "Most L" might show a number below that.

I thought this to be a good revision, since looking at the last rankings I thought it may have skewed the numbers too strongly.
Conversely if a world's "Most W" or "Most L" category STILL shows 120+ that means that there was a team not reflected who had an even more absurd figure. Which is pretty telling IMO.

Also, not sure I would drop outliers here. There is a reason that a world has 130 game winners. I think that data is very relevant in your rankings.

I don't necessarily disagree, but in the last rankings there were a few worlds who had a huge number for the LOSS column from a one and done owner and their next largest number in that category was significantly lower.

There is definitely something to be said for the world that failed to replace the absentee or whatever, but I just felt it skewed a score too much if it was an outlier.

You'll notice that even with dropping the extreme there are several worlds with 120+ scores in Most W or Most L categories. Those worlds definitely earned their scores.

This is great data. Fun to see the differentiation across all leagues. Thank you!

Posted by mchales_army on 6/1/2018 9:52:00 AM (view original):

Posted by strikeout26 on 5/31/2018 8:01:00 AM (view original):

Posted by mchales_army on 5/30/2018 10:08:00 AM (view original):
This ranking takes a look at turnover, wait time between seasons, and several factors that determine a "parity score".

The first set of numbers represent the raw data for each world and then the second set is that world's "score" for each category.

Category scores are determined by that world's distance from the "baseline world" in each category.

A baseline world is defined as the world with the best raw data in that particular category.

Below is a legend for the abbreviations in the chart:


TO = Turnover. How many total owners have been replaced over the last 4 seasons.

Baseline - 3 - BBWAA

Wait = Wait time between seasons. This number equals how many days longer than 360 the world took to complete 4 seasons.

Baseline - 4 - The MLB/Cooperstown

PO = Playoffs. How many different franchises have appeared in the playoffs over the last 4 seasons.

Baseline - 27 - Happy Jack/Boggs

WS = World Series. How many different teams have made it to the World Series over the last 4 seasons.

Baseline - 8 - 10 tied

100W = How many 100 win + teams during the last 4 seasons.

Baseline - 2 - Cooperstown

100L = How many 100 loss + teams during the last 4 seasons.

Baseline - 4 - Cooperstown/Addicted Users

110W = How many 110 win + teams during the last 4 seasons.

Baseline - 0 - 20 tied

110L = How many 110 loss + teams during the last 4 seasons.

Baseline - 0 - 21 tied

Most W = Most wins by any one team, during any single season, over the last 4 seasons.

Baseline - 101 - Capra

Most L = Most losses by any one team, during any single season, over the last 4 seasons.

Baseline - 101 - Cooperstown

Real Life MLB (2014-2017)

Playoffs	WS	100 W	100 L	110 W	110 L	MOST W	MOST L
21	7	5	1	0		104	102

EDIT: It should be noted that in the "Most W and Most L" column, the actual highest and lowest number for each world was dropped so that one outlier in a skewed season didn't negatively impact a world too badly. That is why you might see a world scored with one instance of having had a team win or lose 110+ but their "Most W" or "Most L" might show a number below that.

I thought this to be a good revision, since looking at the last rankings I thought it may have skewed the numbers too strongly.
Conversely if a world's "Most W" or "Most L" category STILL shows 120+ that means that there was a team not reflected who had an even more absurd figure. Which is pretty telling IMO.

Also, not sure I would drop outliers here. There is a reason that a world has 130 game winners. I think that data is very relevant in your rankings.

I don't necessarily disagree, but in the last rankings there were a few worlds who had a huge number for the LOSS column from a one and done owner and their next largest number in that category was significantly lower.

There is definitely something to be said for the world that failed to replace the absentee or whatever, but I just felt it skewed a score too much if it was an outlier.

You'll notice that even with dropping the extreme there are several worlds with 120+ scores in Most W or Most L categories. Those worlds definitely earned their scores.

Great Job doing this - Thanks!

I was really excited to see this poll as I feel like I run a pretty decent world. I’m going to have to examine these numbers closer when I have a moment, but I’m definitely surprised to see that I’m essentially the second worst commish in the game based on these rankings... not the best note to be trying to go to sleep on!

Yeah, I appreciate the effort that was put into this, but the method used to rank worlds needs to be evaluated. My world that is solid is ranked below my world with a massive tanking problem and that goes public immediately upon rollover.

Just a Bit Outside has had consistent ownership and quick rollover....considering it was a merged league, it has taken a few seasons for some teams to even out, but seeing our league ranked so low, which is by far the best I have been in with a commish who works hard to make sure folks stay/recruits when needed, seems to suggest a flaw in the method.

Nice collection of raw data but the derived rankings seem to be not a good representation of actual league quality (based on the worlds I am in / got to know).

It isn't a "flaw in the method" at all. As far as "actual league quality" that is certainly in the eye of the beholder.

The rankings are heavily weighted towards parity, which has been sorely lacking.
That is also the reason that I posted rankings for the wait time and turnover alone.

JaBO scores better than most in both of those categories, but had the 5th worst parity score.

There is one team in that world which has won 110 or more for 9 straight seasons. That same team has won 5 of the 10 WS titles, and only 2 other teams have ever made the WS from that NL. From the AL is a team with 9 straight 100 win seasons which has made 4 WS appearances winning 2.
There have only been 5 franchises in 10 seasons who have managed to best one of your 4 juggernauts to make a WS appearance.

On the other side is a team which has never won more than 79 and has lost 100+ for 7 consecutive seasons now.
And another which has lost at least 92 every single season.

Being only 10 seasons removed from a merger is certainly playing a factor I am certain.

Note that this most recent season had two teams face off in the WS who had never been there before, and the one most egregious juggernaut has apparently finally run out of steam as they only won 99 last season.
So it would seem that the balance is starting to shift and maybe more parity is due in the next few seasons.

Again, the quality of a world can be determined many ways.
This particular method is primarily focused on parity, and is only a 4 season snapshot, which may or may not reflect more poorly or more favorably than a world's "typical" season.

I like it. Lots of data to work with.

Posted by strikeout26 on 6/13/2018 5:43:00 AM (view original):
Yeah, I appreciate the effort that was put into this, but the method used to rank worlds needs to be evaluated. My world that is solid is ranked below my world with a massive tanking problem and that goes public immediately upon rollover.

I assume you are referring to Doubleday vs. Gehringer.

Doubleday is a full 20 spots ahead of Gehringer in the parity ranking (likely due, at least in part, to the massive tanking problem you reference).

Where Doubleday loses ground is being bottom 20% in the other two categories.
These are the "modifier" categories.

99 lost days during rollover is a full season+. Only 20 somewhat worlds had a worse score here.
26 departures in 4 seasons is 6.5 per season. That's pretty high.

I don't claim that these rankings are perfect or anything, but I'll stand by them.
Every instance, so far, of someone calling a particular ranking into question has resulted in me going back and checking their raw data.
Without exception it only bolsters my feelings that these rankings are a pretty solid reflection of the state of any particular world, concerning these categories, which I feel are most important.

Also remember, as I mentioned in my previous post, this is a "snapshot" of the last 4 seasons only. This may reflect better or worse than that world's "typical" season.

I think a good way to judge a world's quality is by looking at the rule 5 draft results.

Bump

This is nice. How often does this data get run and posted to the boards?