Weekend Upsets with WrestleStat | Week 12 2020
- Thread starter andegre
- Start date
Love the website! It has become my go to for wrestling info. By the way, I noticed the RPI match limit is down to 15 this year (instead of 17 like last year). It has to be at the current weight and cannot include FFT or medical forfeits. Not sure how hard that would be to code.
https://ncaaorg.s3.amazonaws.com/ch.../2019-20D1MWR_QualifierAllocationCriteria.pdf
https://ncaaorg.s3.amazonaws.com/ch.../2019-20D1MWR_QualifierAllocationCriteria.pdf
All 3 of those are factored in, so I just need to swap out 17 => 15. Thank you!
Often sitting at the dual, I'll punch up your site. I especially like it when the opponent is not listed in the PSU program and then I can quickly select the new opponent our guy is facing.
Thank you for your work!
Thank you for your work!
No really, if you guys have ANY requests on what doesn't work well, is hard to see (other than the red-on-black), or anything you'd like to see added, PLEASE just let me know! You all are my bosses!
I’m giving you a raise, and the employee of the month parking spot!
I was looking at the PSU-tOSU and see it has Meredith over Heinselman? I will take it
but they have one common win and Meredith is overall 1-5 and Heinselman is 7-3 with other common opponents. I was confused how that one worked?
Just a slight suggestion.
I don't have a specific example to cite right now but they're have been many. I would like to see the algorithm tweaked so that recency is a much stronger influence. I.e. whether it be common opponents or even head to head; what happened last week, last month should have a much stronger influence over the rankings.
There have been times when two close wrestler's in the rankings meet, there I is an upset, but they do not swap the rankings or the projected outcome of the next meeting due to a longer history of common opponents and maybe head to heads. Once you get out a full calendar year, matches before that time should be de-weighted significantly, likewise anything within the past month or two should have the most weighting.
To be clear, every upset does not mean the wrestler's should swap in the rankings, but that match should mean more than those that occurred last season.
I know this might be an impossible request (and feel free to tell me as much) but you have really done an amazing job and shocked me before with what the site can do.
With allocation stuff in the coming weeks, is there a way on the RPI table (or somewhere else on the site) for their RPI record and winning percentage to show up at the current weight only.
Just to use an example: Thomas Penola of Purdue who has bumped up to HWT for the team: https://www.wrestlestat.com/wrestler/57105/penola-thomas/profile
He is 16-9 on the year which is his correct record against college competition and shows.
The 14-9 in paranthesis is what I assume is being used to calculate his and other's RPI but that includes matches at 197 & 285.
If I understand correctly, his record to calculate RPI (and his opponents) at 285 should be 8-7 since he is 9-7 overall at 285 but it would remove the MFFT match which I believe is currently being done in your formula to get to 14-9.
However, for winning percentage, his winning percentage at 285 would be 9-7 (56.25%) since all of his matches at the weight are Division I matches and FFT and MFF do count for winning percentage.
Here is the allocation criteria if you read it differently than I do: https://ncaaorg.s3.amazonaws.com/ch.../2019-20D1MWR_QualifierAllocationCriteria.pdf
With allocation stuff in the coming weeks, is there a way on the RPI table (or somewhere else on the site) for their RPI record and winning percentage to show up at the current weight only.
Just to use an example: Thomas Penola of Purdue who has bumped up to HWT for the team: https://www.wrestlestat.com/wrestler/57105/penola-thomas/profile
He is 16-9 on the year which is his correct record against college competition and shows.
The 14-9 in paranthesis is what I assume is being used to calculate his and other's RPI but that includes matches at 197 & 285.
If I understand correctly, his record to calculate RPI (and his opponents) at 285 should be 8-7 since he is 9-7 overall at 285 but it would remove the MFFT match which I believe is currently being done in your formula to get to 14-9.
However, for winning percentage, his winning percentage at 285 would be 9-7 (56.25%) since all of his matches at the weight are Division I matches and FFT and MFF do count for winning percentage.
Here is the allocation criteria if you read it differently than I do: https://ncaaorg.s3.amazonaws.com/ch.../2019-20D1MWR_QualifierAllocationCriteria.pdf
Good catch! I swear I added that in, or maybe it was just missed on this page....I'll check it out and get it fixed. It will be the column in the RPI Rankings page that will change (shows 14-9 right now).
https://www.wrestlestat.com/rankings/rpi
Just curious why Billy Janzer's win over Rocky Jordan didn't show up on any of the list. Seems like a pretty significant upset even if the rankings choose to continue ranking Jordan higher.
It's because the losing wrestler needs to be ranked in the top 20; Jordan was ranked #30 last week.
