Weekend Upsets with WrestleStat - Week 3

[andegre]

Trying something new, let me know if you guys like and whether I should continue doing these...

https://wrestlestat.wordpress.com/2016/11/23/weekend-upsets-with-wrestlestat-week-3/

 

[Cali_Nittany]

Trying something new, let me know if you guys like and whether I should continue doing these...

https://wrestlestat.wordpress.com/2016/11/23/weekend-upsets-with-wrestlestat-week-3/

I missed a couple of the upsets so I like seeing it all in one place. Well done.

 

[andegre]

I missed a couple of the upsets so I like seeing it all in one place. Well done.

Thank you, what would make this even better is if I could get the video clips of the matches.....

 

[Lion-around]

Trying something new, let me know if you guys like and whether I should continue doing these...

https://wrestlestat.wordpress.com/2016/11/23/weekend-upsets-with-wrestlestat-week-3/

Didn't Cenzo have some upsets?

 

[andegre]

Didn't Cenzo have some upsets?

Beat #7 Chad Walsh of Rider (which I'm sure you already knew about) but it doesn't fit the "upset" criteria...

Basically, criteria is as follows:

Losing wrestler ranked in the top 20
Difference in losing wrestlers ranking to winning wrestlers ranking is greater than 5

Obviously could use some fine-tuning, because IMO, if #5 beats a #1, I'd consider that an upset...and it would NOT fall under the current criteria.

 

[RoarLions1]

How much impact does a forfeit (W or L) have on your rankings? Forget how Elo handles this...though my thought is it doesn't, so has to be something you added to the algorithm.

 

[andegre]

How much impact does a forfeit (W or L) have on your rankings? Forget how Elo handles this...though my thought is it doesn't, so has to be something you added to the algorithm.

Correct, the rankings process ignores forfeits and medical forfeits. Otherwise, every other match result type (Dec, MD, TF, etc) are used in calculating the new ELO rating.

As far as this report, I ignore these: MFOR, FOR, DQ, INJ, and DEF

 

[Cali_Nittany]

Bump. Andegre does great work, this should get more feedback.

 

[Cali_Nittany]

Beat #7 Chad Walsh of Rider (which I'm sure you already knew about) but it doesn't fit the "upset" criteria...

Basically, criteria is as follows:

Losing wrestler ranked in the top 20
Difference in losing wrestlers ranking to winning wrestlers ranking is greater than 5

Obviously could use some fine-tuning, because IMO, if #5 beats a #1, I'd consider that an upset...and it would NOT fall under the current criteria.

Rather than basing it on rankings, is it possible to base it on some kind of point system that would distinguish the tiers within the weight class. For example, #1 Zain losing to #2 Sorenson would definitely be an upset vs Heil losing to McKenna which wouldn't be that much of an upset.

 

[andegre]

Rather than basing it on rankings, is it possible to base it on some kind of point system that would distinguish the tiers within the weight class. For example, #1 Zain losing to #2 Sorenson would definitely be an upset vs Heil losing to McKenna which wouldn't be that much of an upset.

Yeah, I'm playing around in my test environment right now with this exact approach...