One For The Nerds

[Wrestleknownothing]

You have been warned.

This one is not for everybody.

But if you want to read about the expected points model - have at it.

https://wrestling.guru/2026/02/01/expected-points-friend-or-foe/

As a footnote, I am glad sad to see the site keeps degrading. I now have an ad that is supposed to be in the margin, but instead covers the left side of the content box. Well done.

[Caveira]

2 minutes ago, Wrestleknownothing said:

You have been warned.

This one is not for everybody.

But if you want to read about the expected points model - have at it.

https://wrestling.guru/2026/02/01/expected-points-friend-or-foe/

As a footnote, I am glad sad to see the site keeps degrading. I now have an ad that is supposed to be in the margin, but instead covers the left side of the content box. Well done.

There are no mods.  We’ve been left alone to fend for ourselves like mad Max 

[flyingcement]

2 minutes ago, Wrestleknownothing said:

You have been warned.

This one is not for everybody.

But if you want to read about the expected points model - have at it.

https://wrestling.guru/2026/02/01/expected-points-friend-or-foe/

As a footnote, I am glad sad to see the site keeps degrading. I now have an ad that is supposed to be in the margin, but instead covers the left side of the content box. Well done.

Awesome stuff.  I appreciate the probabilistic analysis you have put forth instead of the other less accurate measures.  Hope life is treating you well.  

[GreenHorn]

Good stuff! I look forward to the next one. 

[Wrestleknownothing]

13 minutes ago, flyingcement said:

Awesome stuff.  I appreciate the probabilistic analysis you have put forth instead of the other less accurate measures.  Hope life is treating you well.  

Never better, my friend.

In 2025 I turned 60, celebrated my 40th anniversary, and welcomed 2 precious grand daughters to the world. That is a damn good year. 

I was also heartened to see your good news. 

[Wrestleknownothing]

16 minutes ago, Caveira said:

There are no mods.  We’ve been left alone to fend for ourselves like mad Max 

Two men enter. One man leaves. 

Fight the good fight. 

[Wrestleknownothing]

6 minutes ago, GreenHorn said:

Good stuff! I look forward to the next one. 

Thank you. I have some ideas. 

[Mr. PeanutButter]

Just now, Wrestleknownothing said:

Never better, my friend.

In 2025 I turned 60, celebrated my 40th anniversary, and welcomed 2 precious grand daughters to the world. That is a damn good year. 

I was also heartened to see your good news. 

Damn dawg that is amazing! Hope life keeps treating you well

[flyingcement]

2 minutes ago, Wrestleknownothing said:

Never better, my friend.

In 2025 I turned 60, celebrated my 40th anniversary, and welcomed 2 precious grand daughters to the world. That is a damn good year. 

I was also heartened to see your good news. 

Great to hear!  And congratulations on all fronts.  

Really appreciate your last comment as well.  

[Wrestleknownothing]

4 minutes ago, Mr. PeanutButter said:

Damn dawg that is amazing! Hope life keeps treating you well

Amen.  Same to you. Lets start with a resurgent Iowa. 

[PortaJohn]

2 minutes ago, flyingcement said:

Great to hear!  And congratulations on all fronts.  

Really appreciate your last comment as well.  

Wait, what was your good news?

[Caveira]

11 minutes ago, Wrestleknownothing said:

Two men enter. One man leaves. 

Fight the good fight. 

 be well boss

[bnwtwg]

1 hour ago, Wrestleknownothing said:

Never better, my friend.

In 2025 I turned 60, celebrated my 40th anniversary, and welcomed 2 precious grand daughters to the world. That is a damn good year. 

I was also heartened to see your good news. 

 

[TampaRyley]

Great to see you post again @Wrestleknownothing! And great to hear all is nice on your end.

I love the analysis. As an often time statistical graph creator myself by profession, this was a fun read.

[GV4444]

Best article so far, but I spent younger years building models.  If you spent 3 months adjusting the 12 years of data, your model's predictive power would be much better (e.g. normalize for every rule change like 3-pt takedown).  Your model truly would be a better predictor than the chalk in all cases.  

Going by the chalk will always overvalue the top 8/undervalue the lower seeds.  The probability distribution of each wrestler over or under performing will always create such results.  A #2 seed can only overachieve by finishing #1, whereas the 16 seed has far more scenarios that can allow overachievement.  Now, factor in the way points are distributed and the point results (chalk vs. model) will skew even further (e.g. point difference between 1 and 2 greater than 17 and 18.

Can't wait to see the model prediction after seeding is done for NCAAs.  

[bnwtwg]

2 hours ago, GV4444 said:

Best article so far, but I spent younger years building models.  If you spent 3 months adjusting the 12 years of data, your model's predictive power would be much better (e.g. normalize for every rule change like 3-pt takedown).  Your model truly would be a better predictor than the chalk in all cases.  

Going by the chalk will always overvalue the top 8/undervalue the lower seeds.  The probability distribution of each wrestler over or under performing will always create such results.  A #2 seed can only overachieve by finishing #1, whereas the 16 seed has far more scenarios that can allow overachievement.  Now, factor in the way points are distributed and the point results (chalk vs. model) will skew even further (e.g. point difference between 1 and 2 greater than 17 and 18.

Can't wait to see the model prediction after seeding is done for NCAAs.  

Must not have checked out previous work regarding the implementation of THREEEEEEEE. Rule number one of academic article critique is to see if it exists. For shame!

 

[MPhillips]

22 minutes ago, bnwtwg said:

Must not have checked out previous work regarding the implementation of THREEEEEEEE. Rule number one of academic article critique is to see if it exists.

For the dummies in the back...

[ionel]

4 hours ago, MPhillips said:

For the dummies in the back...

... say what?  

[GV4444]

12 hours ago, bnwtwg said:

Must not have checked out previous work regarding the implementation of THREEEEEEEE. Rule number one of academic article critique is to see if it exists. For shame!

The previous article doesn't mean the data used for the newer was normalized for the change in TD points, nor does indicate that the data have been normalized for bonus score differences based upon TD point change, a monumental task, which was my point.  There may have been other rule changes during the sample time period that would also cause a data adjustment.  Nor does it impact the primary point of my post that using the model to generate simulations would likely result in 1) highest seeds underperforming more often, while lower seeds overperforming (vs. chalk) and 2) a more accurate prediction than the chalk.

 

[bnwtwg]

Just now, GV4444 said:

The previous article doesn't mean the data used for the newer was normalized for the change in TD points, nor does indicate that the data have been normalized for bonus score differences based upon TD point change, a monumental task, which was my point.  There may have been other rule changes during the sample time period that would also cause a data adjustment.  Nor does it impact the primary point of my post that using the model to generate simulations would likely result in 1) highest seeds underperforming more often, while lower seeds overperforming (vs. chalk) and 2) a more accurate prediction than the chalk.

 

I believe the data model is open-source and WKN strikes me as a team-first player open to peer review