I've been interested in creating a transparent and highly automated ranking system for awhile now, and I finally took the time to do it.  My intention was for it to be purely data-driven.  And much like Ranger's rankings, you gotta beat top guys if you want to be ranked at the top. I also wanted to use the data collected to try to come up with some fun conversation starters and notifications, so I added features like "This Weeks Top Ranked Matches" and "Biggest Climbers". (Biggest Climbers will be a better feature when I have a longer history of rankings). I have some other ideas to add as well.  

A couple of notes:

1. I'm not trying to replace Ranger's rankings. As far as I'm concerned, his rankings are the gold standard.  This just offers a slightly different perspective and hopefully creates some fun conversation.  And anyways, my rankings and his aren't that far off in most cases. 
2. I decided to rank more than one wrestler per weight class.  So you'll see some of the more deep teams with 2-3 guys in a weight class.  I think that's pretty cool and highlights which programshave the most depth.  It's not surprising that the teams with the most ranked back-up wrestlers are Union, Paducah, and Ryle. 
3. I really wanted to highlight the importance of wrestling tough competition. I hope the tables in the back of the report show that records don't matter.  You gotta beat a top guy if you want to move up.  And if you're a coach, go get some tough competition!
4. Wrestlers are removed after 45 days of inactivity.  If they return, they'll get added back in.
5. Weights are based on most recent 10 matches, defaulting to the lower weight.  I have the ability to override the weight, so I did that in cases where I know they're going a different weight at regionals.  
6. If you look at the tables in the back, you'll see a lot of "CO"s in the table. That's common opponents. So a green CO means you beat someone that the other guy lost to. This includes out of state competition, which helps rank the guys on teams that regularly leave the state.  
7. As I said, you gotta beat guys to move up the rankings.  There are exceptions around the very top.  If you're a returning state placer, you'll get some benefit of the doubt.  If you're Jordyn Raney, you'll get ranked at the top even with just one ranked win.  

If you see any mistakes, or features you'd like to add, let me know.  I hope to see this report used to highlight exciting recent events like ranked matchups, and eventually it can also highlight exciting upcoming events as well.

rankings-report-022025.pdf

Quite interesting but not bad. 

LOVE IT!  Those H2H grids are amazing.

58 minutes ago, gator1 said:

I've been interested in creating a transparent and highly automated ranking system for awhile now, and I finally took the time to do it.  My intention was for it to be purely data-driven.  And much like Ranger's rankings, you gotta beat top guys if you want to be ranked at the top. I also wanted to use the data collected to try to come up with some fun conversation starters and notifications, so I added features like "This Weeks Top Ranked Matches" and "Biggest Climbers". (Biggest Climbers will be a better feature when I have a longer history of rankings). I have some other ideas to add as well.  

A couple of notes:

1. I'm not trying to replace Ranger's rankings. As far as I'm concerned, his rankings are the gold standard.  This just offers a slightly different perspective and hopefully creates some fun conversation.  And anyways, my rankings and his aren't that far off in most cases. 
2. I decided to rank more than one wrestler per weight class.  So you'll see some of the more deep teams with 2-3 guys in a weight class.  I think that's pretty cool and highlights which programshave the most depth.  It's not surprising that the teams with the most ranked back-up wrestlers are Union, Paducah, and Ryle. 
3. I really wanted to highlight the importance of wrestling tough competition. I hope the tables in the back of the report show that records don't matter.  You gotta beat a top guy if you want to move up.  And if you're a coach, go get some tough competition!
4. Wrestlers are removed after 45 days of inactivity.  If they return, they'll get added back in.
5. Weights are based on most recent 10 matches, defaulting to the lower weight.  I have the ability to override the weight, so I did that in cases where I know they're going a different weight at regionals.  
6. If you look at the tables in the back, you'll see a lot of "CO"s in the table. That's common opponents. So a green CO means you beat someone that the other guy lost to. This includes out of state competition, which helps rank the guys on teams that regularly leave the state.  
7. As I said, you gotta beat guys to move up the rankings.  There are exceptions around the very top.  If you're a returning state placer, you'll get some benefit of the doubt.  If you're Jordyn Raney, you'll get ranked at the top even with just one ranked win.  

If you see any mistakes, or features you'd like to add, let me know.  I hope to see this report used to highlight exciting recent events like ranked matchups, and eventually it can also highlight exciting upcoming events as well.

 

rankings-report-022025.pdf

Can you legitimately see Ricketts not winning state?

Sure, the head to head stuff may hurt Union to some affect but you have to take overall strength of schedule into some account.

That is awesome

22 minutes ago, Pépitedepoulet said:

Can you legitimately see Ricketts not winning state?

Yeah, it's hard to imagine Rickett's losing.  He's a great wrestler, and so is McCoart.  The only reason I placed McCoart higher is because he pinned #4 Ben Hall, while Ricketts won by decision(7-2).  I know a wins a win, but to me it's a two man race and I just hope they meet in the finals next week.

I'm not predicting a McCoart win, by the way.  I think it will be a great match.  I have no idea who will win. 

37 minutes ago, Ranger123 said:

LOVE IT!  Those H2H grids are amazing.

Thanks!  That matrix is what I've been visualizing in my head for years, and I was curious how it would look in real life.  It really highlights the upsets too.  Like what is that red cell doing in that giant sea of green?  Sometimes it's a huge upset, other times it's a loss to a guy that maybe should have been ranked higher. 

Where is Caleb Barnes (Johnson Central) at 157?  

26 minutes ago, wrestling fan said:

Where is Caleb Barnes (Johnson Central) at 157?  

 

That's a good question!  I noticed when I was working on the code that he was showing up at 144 along with his brother.  I meant to go back and figure out what was wrong with the code, but I forgot to do that. 

He'll drop in at #12 for now, although he could go up quite a bit if he beats Leader at regionals this weekend. 

That is one of the difficulties with any automated system is the human input error.  There are TONS of wrong names and results in track.  And trying to automate weight class when guys bump all around is a major challenge.

And circling back to the Ricketts / McCoart discussion.  Ricketts was the only ranking that stood out to me at first glance.  I do feel like he is a favorite, but wouldn't consider that ranking to be way out of line.

13 minutes ago, Ranger123 said:

That is one of the difficulties with any automated system is the human input error.  There are TONS of wrong names and results in track.  And trying to automate weight class when guys bump all around is a major challenge.

As an old boss used to say, it doesn't matter how good a process is.  Garbage in gives garbage out.  Of course in the case of Barnes I think it was my own error. 

I've worked in ways to override weights and match results, but we'll see.  I won't be surprised at all if I run into unexpected problems like the ones you're talking about. 

Very interesting and awesome! The cream rises to the top.

Thought it was just an oversight.  I like this ranking system.  Thanks for putting it together.