Jump to content
Ching

Rank the Rankers

Recommended Posts

Looks like Andre has some work to do on his algorithms.

 

 

Sent from my iPad using Tapatalk

I’m not so sure. Ashnault deserved that year long #1 ranking? Name one wrestler who beat him this year!

 

Exactly

 

 

Sent from my iPhone using Tapatalk

Share this post


Link to post
Share on other sites

Looks like Andre has some work to do on his algorithms.

 

 

Sent from my iPad using Tapatalk

WrestleStats is the best tool in this sport for us fans the last few years. Let’s not bust on Andegre for all the time he puts into it.

 

Without wrestlestats, good luck looking up results, comparing duals, etc.

 

His rankings are a lifetime achievement Eli rating system not a 1 yr glance.

 

 

 

 

Sent from my iPhone using Tapatalk

Edited by pish6969

Share this post


Link to post
Share on other sites

To the OP?

 

What do the points represent? Good topic to see what service provides the best rankings. I’m sure it changes every year but would be interesting to track

 

 

Sent from my iPhone using Tapatalk

Edited by pish6969

Share this post


Link to post
Share on other sites

Are you doing a victory lap for beating flo? Not bad for a secret formula.  

LOL at the secret part...given the lack of secrecy of it.

 

That said, no, no victory laps for beating Flo here.

 

For what it's worth, based on my bracket projections (which are slightly different than the simply ordinal numbers used across the board here, as they account for the actual matchups, something taking the top eight in order doesn't do), I had 555 points, which doesn't move me up at all in these rankings. I will say, however, that I had 58 of 80 All-Americans, one more than the seeds got, and 15 finalists, versus 12 based on the actual seeds.

Share this post


Link to post
Share on other sites

This is really cool and I appreciate you compiling this information.

 

I've/we've know since about week 6 that the WrestleStat rankings needed a pretty good tweak. We've got some notes on where to start tweaking, as well as some other ideas that I have to make it look more accurate.

 

Having this information definitely helps, and gives me a good way to "weight" the different versions/tweaks that I'll be applying this summer. It will allow me to run through the rankings HUNDREDS of times until my number gets to the top of your list :).

 

Last summer I did exactly ZERO updates to the rankings process, all while announcing that I WAS going to make some tweaks. Seeing as how this is a side project and I have a real job, with a family and two kids under the age of 5, my time is limited. Last summer was spent almost ENTIRELY on integrating with TakedownLIVE's system (it's the icon that will show takedowns, escapes, etc for individual match results....if you're not sure what I'm talking about, check out https://www.wrestlestat.com/wrestler/14443/butler-nathan/profile).

 

So, I have a [LONG] list of updates/tweaks/additions that I'm going to work on this summer. The rankings will be near the top of the list....priority #1 is to re-design the site with a different dark-themed style that will make it easier on the eyes, IMO.

 

BUT, one thing that seems to still be lost on some of the people that criticize my rankings is the fact that these are CAREER based, just like the real Elo rating system that this is based on. Prime example, Tomasello #1, Spencer Lee #2.

 

Lastly, one quick question, Ching, did you happen to use the "Starters" version of my rankings, or the regular that includes injured, redshirting, etc? (Crossing my fingers that you used the "regular" version, which would mean my numbers would move up :)).

Share this post


Link to post
Share on other sites

This is really cool and I appreciate you compiling this information.

 

I've/we've know since about week 6 that the WrestleStat rankings needed a pretty good tweak. We've got some notes on where to start tweaking, as well as some other ideas that I have to make it look more accurate.

 

Having this information definitely helps, and gives me a good way to "weight" the different versions/tweaks that I'll be applying this summer. It will allow me to run through the rankings HUNDREDS of times until my number gets to the top of your list :).

 

Last summer I did exactly ZERO updates to the rankings process, all while announcing that I WAS going to make some tweaks. Seeing as how this is a side project and I have a real job, with a family and two kids under the age of 5, my time is limited. Last summer was spent almost ENTIRELY on integrating with TakedownLIVE's system (it's the icon that will show takedowns, escapes, etc for individual match results....if you're not sure what I'm talking about, check out https://www.wrestlestat.com/wrestler/14443/butler-nathan/profile).

 

So, I have a [LONG] list of updates/tweaks/additions that I'm going to work on this summer. The rankings will be near the top of the list....priority #1 is to re-design the site with a different dark-themed style that will make it easier on the eyes, IMO.

 

BUT, one thing that seems to still be lost on some of the people that criticize my rankings is the fact that these are CAREER based, just like the real Elo rating system that this is based on. Prime example, Tomasello #1, Spencer Lee #2.

 

Lastly, one quick question, Ching, did you happen to use the "Starters" version of my rankings, or the regular that includes injured, redshirting, etc? (Crossing my fingers that you used the "regular" version, which would mean my numbers would move up :)).

 

I did use the "Starters" version. I made sure every top eight were actually competing. Looking at individual weights, 125 you were the clear best, 184 you were tied for second, and 149/157 third.

 

I still haven't dug into the data, but I think most of the underperformance (for all the lower rankings services) are a handful of individuals:

Forys, Jack, Walsh, and Massa hurt

 

I'm looking forward to the updates.

Share this post


Link to post
Share on other sites

It has been shown many times that all the predictors are about the same, which is what your figures show.  If you break it down to predicting AA's and predicting exact placements you would get the following from a long time ago on TOM:

 

I analyzed two factors; Correctly predicting the AA's and the placers.  Guessing who AA's is easier; V-62, JB-57, Fuhr-56, Ron G-62, and Bryce 56.  The team averaged 73%. 

Correctly predicting the placers was much more difficult; V-16, JB-19, Fuhr-18, Ron G-20, and Bryce-17.  The team averaged 23%. 

133 and 174 were the easiest weights to predict AA's, but 174 was also the hardest weight to predict the placings. In general there wasn't much variance among the weight classes.

The seeding committee received comparable scores: 59 AA's and 17 placers
.

 

TWT's JensenS had the best score with 61/22, while FLO's CP (VA elsewhere) scored 61/20.  ZAP/VAK had a poor tourney (must be all that Iowegian bias) 53/16

 

http://forum.theopenmat.com/index.php/topic,26858.msg352572.html#msg352572

Share this post


Link to post
Share on other sites

Very cool stats... Thanks for putting in the time Ching.

 

I think it's really impressive that SHP's rankings scored as competitively as they did given that they are entirely computer generated. I would have thought having a human element and the ability use common sense would be a huge advantage.

Share this post


Link to post
Share on other sites

Not sure if anyone cares, but WrestleStat has gone through quite a few tweaks and iterations to the algorithm. Latest version has WrestleStat with a final number of 554. I went back and filled it out for 2017 and it ALSO came up with 554.

 

WrestleStat algorithm is getting better, we'll continue to work on improving it this offseason.

 

By the way, if anyone knows of a way to get past historical rankings from any of these services, please let me know as I'd like to see how the rest of the field did for last season as well to see where WrestleStat stands.

 

Edit: I ran through 2017's to see how the "seeds" fared using that scoring model....540.

Edited by andegre

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...

×
×
  • Create New...