Team rankings

[Nimrod]

Okay I got in. Curious about the methodology. I can tell someone put some work into this. Good job on that. I having taken a deep dive into this. What I like about the usclubrankings.com is that they put emphasis on playing in certain tournaments which basically takes care of strength of schedule. When you play 20 games against to 20 team but have a winning % of 750, and another team that is playing less competitive teams and have a winning % of 950 that team will be ranked higher.

i like the effort also. Problem with putting emphasis on certain tournaments is that you have to make a subjective judgment about which tournaments are stronger. Plus, just playing in those tournaments doesn't make a team better than another. I get why usclubrankings does that, but the goal of a good computer model is to prove which tournaments are better, not to pre-assume anything.

Here are the 2019 North Carolina rankings that I spoke about earlier.

Notice that the #9 team is only 13-10-2 while the #30 team is 30-12. Further, the #9 team is rated 5 runs better than the #30 team, so these computer rankings are projecting that this 13-10-2 team would ordinarily beat the 30-12 team by a score of 8-3, 9-4, etc., even though they might not have played each other.

That's evidence of a model that can figure out strength of schedule - provided you feed it with enough scores. This website tracked down virtually every score of every game played in North Carolina over spring/summer. Over 170 teams were ranked in this age division. This shows just the top 30.

Now, understand that the computer model doesn't give extra points for playing in certain tournaments. It figures out what tournaments (and teams) are the strongest through comparison of scores for all the teams. You'll find mediocre teams ranked in the top 10 early in the season just because they won a small tournament with teams that are isolated. But over time, those teams will venture out and get tested, and the proof will be in the thousands of scores that are played every year. This ranking system will put every team in the general neighborhood of how good they are among the 170 teams statewide. Key - Getting the scores, having a good model.

 

[radness]

Interesting.
Any guestimations on how accurate ?

 

[_MIA_]

Honest question...why does it matter? Feel the same about the player ratings they put out. If TB orgs/coaches/parents were banned from posting about it on FB they wouldn't even exist probably..

 

[BT3100]

What age bracket?

18U

 

[SBmom8992]

How can there be rankings without teams playing 2020?

Saying this about travel in calif.
Calling it 'California Rust'
( opposite California gold rush)

Its the teams who have not played tournaments in 10 months.
Or VERY FEW !
While have read on dfp of all the tournys elsewhere being played for months.

From a state that is use to playing year round...and hasnt...

How would California Rust fit into rankings?!

Just like there being Gatorade Softball Player of the Year in 2020 when not a single game was played in most states. Makes no sense


Sent from my iPhone using Tapatalk

 

[Nimrod]

Interesting.
Any guestimations on how accurate ?

I think they're highly accurate as far as placing teams at their level. As far as fine-tuning whether you should be #51 vs. #63, or #4 vs. #7, you can't eliminate all those debates. It's just like college football rankings done by Sagarin, Massey, etc. You're going to see teams you think should be higher or lower, but you're not going to see anything ridiculous, or if it does look ridiculous, you can look at the scores and the teams they've played more closely and then you'll say, 'OK, that makes sense now.' Computer rankings should be the best mathematical explanation for the scores. If you have enough scores, and teams don't get isolated, it works pretty well. But most times, you can't get enough scores, or the computer model is weak.

 

[Nimrod]

Honest question...why does it matter? Feel the same about the player ratings they put out. If TB orgs/coaches/parents were banned from posting about it on FB they wouldn't even exist probably..

Great question. I agree that rankings could do harm. There's far more to youth softball than which team is better than the other, and rankings could encourage people to take it too seriously.

But, there can be value in rankings if they educate coaches and parents about the level they're playing. Most parents don't have a realistic sense of playing levels, and I believe the N.C. rankings provided that.

Player rankings are another thing. That's too subjective.

 

[Rick M]

Honest question...why does it matter? Feel the same about the player ratings they put out. If TB orgs/coaches/parents were banned from posting about it on FB they wouldn't even exist probably..

This is true. And a person with a real entrepreneurial sense would use the location data on a user’s browser to determine which rankings they’re shown.

Viewing the site from Virginia? Oh, look how many Virginia teams are in the top 25! We’re playing in the best softball market around. FB share!


Sent from my iPhone using Tapatalk

 

[_MIA_]

But, there can be value in rankings if they educate coaches and parents about the level they're playing. Most parents don't have a realistic sense of playing levels, and I believe the N.C. rankings provided that.

Player rankings are another thing. That's too subjective.

Maybe if you are new to the TB scene and are looking for guidance but in that case how good the team is wouldn't necessarily be the top priority anyway since if you are new that typically (nowadays at least) means you are young. Other than that, one would think playing 30+ games against teams in your state for a few years would give a pretty good indication where your team stacks up against other teams in the area or just where the teams stack up in general

 

[benjey]

I like these rankings....... because they rank DD's team really high

In all seriousness, while I appreciate the approach they took here using data instead of the typical reputation rankings, I think it is next to impossible to produce accurate rankings due to the sheer number of teams and games making accurate/complete data collection a herculean task.