Prediction Engine built-in to Comparisons

[andegre]

Now that the prediction engine is being used in the new dual rankings process, I was able to get this implemented into the dual and wrestler comparison pages. Unfortunately, I have a bug in the wrestler comparison version so that isn't working yet. But the dual comparison version works great.

On the dual rankings page, at the top their will be a "WrestleStat Prediction" that will be for the dual. Then for each weight, there will be another "WrestleStat Prediction" that is for that individual match.

This can also be used to figure out which teams beat which teams in the dual rankings process.

Here's an example dual comparison that shows the WrestleStat Prediction:

http://testwrestlestat.azurewebsites.net/compare/dual/60/penn-state/55/ohio-state

Go ahead and try out any combination you want and let me know how you like it:

http://testwrestlestat.azurewebsites.net/compare/dual

 

[obrats]

Hi Greg,

Got an error message trying to run Rutgers vs. Rider.

 

[andegre]

Fixing that now....should be fixed in about 5 minutes.

 

[andegre]

It's fixed now...

Here's Rutgers vs Rider: http://testwrestlestat.azurewebsites.net/compare/dual/66/rutgers/65/rider

 

[Passe24]

Ok so how does it come up the Bo Jordan would lose to Joseph? And better yet Bo Nikal beating Martin 13-5. It's just not realistic to me. You asked for examples and here are some. I am not putting this down I'm just not agreeing with its results.

 

[andegre]

Ok so how does it come up the Bo Jordan would lose to Joseph? And better yet Bo Nikal beating Martin 13-5. It's just not realistic to me. You asked for examples and here are some. I am not putting this down I'm just not agreeing with its results.

No worries, someone else on another board had these exact same questions....here's my response:

Ok, regarding Nickal vs Martin, the FALL that he had is definitely weighing heavily on those results. Also, he's got another Fall vs Brunson (#8) as well as another MD. That's why it's predicting a MD.

For Cenzo and Jordan, it looks like it's weighing his matches too heavily that he had against inferior competition (and ALL matches were won by bonus). Since they were all won by bonus, and Jordan doesn't seem to have as many bonus wins, Cenzo is off-setting, and winning.

This one will take a little more work to get right...

 

[Passe24]

I get it. Just like Ashnault is predicted to beat Koldzik 12-0. I don't see that happening. But Kolodzik has no official match ups to sort threw I guess.