ADVERTISEMENT

WrestleStat Rankings - Week 2 2019

andegre

Well-Known Member
May 18, 2004
4,185
906
1
Here are the updated rankings after a big tournament weekend. All results should be entered, so we aren't missing any tournaments. If you notice any matches that might be missing, send an email to support@wrestlestat.com.

Wrestler Rankings: https://www.wrestlestat.com/rankings/starters

Dual Rankings: https://www.wrestlestat.com/rankings/dual

Tournament Rankings: https://www.wrestlestat.com/rankings/tournament

Don't forget to check out the Tourney Projection Tool so you can simulate the results at the National Tournament.

https://www.wrestlestat.com/tourney/projection
 
You've got to give it a break. RBY has 1 match. A fall over a guy who was 30something to 10something last year at 141 and is currently ranked #83. What is the computer supposed to do with that? Make him top 10????
I didn't mean to create a firestorm, I just don't understand rankings that ignore reality. Do you think there are 67 better 133's? If UCF plays a football game on August 31 and Alabama doesn't open their season until the following week, do the Knights get the higher ranking? It's silly and doesn't tell me anything.

Re-read my initial post... I didn't tell people who like WrestleStat to steer clear of it, just said it wasn't for me, at least until there is enough data to make it align with reality.
 
I didn't mean to create a firestorm, I just don't understand rankings that ignore reality. Do you think there are 67 better 133's? If UCF plays a football game on August 31 and Alabama doesn't open their season until the following week, do the Knights get the higher ranking? It's silly and doesn't tell me anything.

Re-read my initial post... I didn't tell people who like WrestleStat to steer clear of it, just said it wasn't for me, at least until there is enough data to make it align with reality.
Not a firestorm. I re-read your post, and the parting shot wasn't necessary, and even in the post I'm quoting, you take another shot.

The "reality" is that this mathematical model works best the more matches a wrestler gets, so it's not "fiction". My suggestion would be to learn more about the MO of WrestleStat.
 
Last edited:
I didn't mean to create a firestorm, I just don't understand rankings that ignore reality. Do you think there are 67 better 133's? If UCF plays a football game on August 31 and Alabama doesn't open their season until the following week, do the Knights get the higher ranking? It's silly and doesn't tell me anything.

Re-read my initial post... I didn't tell people who like WrestleStat to steer clear of it, just said it wasn't for me, at least until there is enough data to make it align with reality.
The rankings HAVE to ignore reality because it's all math/algorithm-based. The only thing determined by human/manually is the original algorithm and the weighting that it places on matches. This is all a completely automated process for me, otherwise I would not have gotten past the first season (2014).

As to your analogy/comparison, we KNOW Alabama is good based on their previous seasons, whereas [with RBY I'm assuming], he's had 1 college match in his career to use for his ranking. He could very well be ranked the same at the end of the season (although I doubt it), or he could finish the season with a National Championship. Either way, that ranking will be MUCH more accurate once he gets more matches under his belt.
 
You've got to give it a break. RBY has 1 match. A fall over a guy who was 30something to 10something last year at 141 and is currently ranked #83. What is the computer supposed to do with that? Make him top 10????

So how does Mr. Verkleeren wind up at No. 11 after just one bout?
 
  • Like
Reactions: Antaeus
I find wrestlestat very useful. Just have to use your head and realize that it takes a few matches to have enough data on new college wrestlers.
 
  • Like
Reactions: danoftw and ccdiver
The rankings HAVE to ignore reality because it's all math/algorithm-based. The only thing determined by human/manually is the original algorithm and the weighting that it places on matches. This is all a completely automated process for me, otherwise I would not have gotten past the first season (2014).

As to your analogy/comparison, we KNOW Alabama is good based on their previous seasons, whereas [with RBY I'm assuming], he's had 1 college match in his career to use for his ranking. He could very well be ranked the same at the end of the season (although I doubt it), or he could finish the season with a National Championship. Either way, that ranking will be MUCH more accurate once he gets more matches under his belt.
:) I figured somebody would challenge the analogy that way; however, it holds, because all college football teams have substantial numbers of incoming freshmen every season that will potentially alter the teams. "We KNOW Alabama is good based on their previous seasons." ... yes, against teams whose compositions have changed dramatically, just as Alabama's will have.

You're understandably passionate about what you do, and it will have merit and validity in a few weeks, which is when it would logically first be published. You want it to be "scientific," wait until you have measurable evidence. For right now, your introductory clause speaks for itself, because it validates what I said to launch the conversation... "The rankings HAVE to ignore reality." At this point of the season, we're in complete agreement on that much at least.
 
Be better to talk chess than football. Tool doesn't work as well for team sports. Too few competitions, really risky using previous season's results given graduations/ incoming athletes, etc.
 
I prefer biased human beings. I really believe Nebraska was the best CFB team in 1994 no matter what the unbiased NYT computer algorithm said, which only took into account results, not the pollers fondness for the Nebby coach :rolleyes:

Keep up the good work Greg. Anyone who wants to understand the ELO can read your site, use google or remain in the dark and complain.
 
Come on guys - every year we have all of this wrestlestat bashng from folks either new to the tool or too stubborn to adjust expectations. In full disclosure in my first year I did some ranting too.

Most other rating services are subjective based upon educated assessment - of you want that, go to Flo or intermat or whatever....

Wrestlestat is based on hard Data and analytics, and like all software it has a hard dependency on getting fed good data.

That means for the 1st half of the season all the new true freshman like RBY have no history, thus the program ranks them accordingly. RS Freshman like Verk have a limited college resume but enough to generate some analysis and a ranking.

The dual compare is super fun albeit provacative, but it gives you what the data calculates and nothing more. I love looking back over the histories of common opponents etc...

In the second half of the season and during the conference tourneys there is nothing better and then you realize the power of an objective tool based upon analytics as opposed to opinion.

So in the end extend a BIG THANK YOU to the wrestlestat developers as we enjoy for free what is a boat load of work in areas few of us could play.

In the mean time, let's all sit back and watch RBY climb the rankings over the next 3 months.

Come Feb-March when we all try to predict RBYs tournament performance and his points potential - wrestlestat will be your best friend with a wealth of common opponent info and the ELO chess analysis at your fingertips.
 
Come on guys - every year we have all of this wrestlestat bashng from folks either new to the tool or too stubborn to adjust expectations. In full disclosure in my first year I did some ranting too.

Most other rating services are subjective based upon educated assessment - of you want that, go to Flo or intermat or whatever....

Wrestlestat is based on hard Data and analytics, and like all software it has a hard dependency on getting fed good data.

That means for the 1st half of the season all the new true freshman like RBY have no history, thus the program ranks them accordingly. RS Freshman like Verk have a limited college resume but enough to generate some analysis and a ranking.

The dual compare is super fun albeit provacative, but it gives you what the data calculates and nothing more. I love looking back over the histories of common opponents etc...

In the second half of the season and during the conference tourneys there is nothing better and then you realize the power of an objective tool based upon analytics as opposed to opinion.

So in the end extend a BIG THANK YOU to the wrestlestat developers as we enjoy for free what is a boat load of work in areas few of us could play.

In the mean time, let's all sit back and watch RBY climb the rankings over the next 3 months.

Come Feb-March when we all try to predict RBYs tournament performance and his points potential - wrestlestat will be your best friend with a wealth of common opponent info and the ELO chess analysis at your fingertips.
Thank you for that response!

And you sparked a new idea that I think people will LOVE...

You mentioned going back through the dual comparisons to see how the algorithm/prediction engine worked out......well, how about I just build that into the so it's right there for you? I may even be able to take it a step further and show prediction numbers on the schedule page (for duals), and maybe even the wrestlers match results page as well. Let me check to see if the architecture will allow me to do that fast enough...should definitely be feasible on the dual comparison/box score pages though.

Oh, and you mentioned waiting until the end of the season to project RBY's points at Nationals....why wait, you can do that now with the Tourney Projection Tool :) https://www.wrestlestat.com/tourney/projection
 
You've got to give it a break. RBY has 1 match. A fall over a guy who was 30something to 10something last year at 141 and is currently ranked #83. What is the computer supposed to do with that? Make him top 10????
Exactly the point! "The computer" shouldn't be doing anything until early January. If it doesn't know any better than to make this prediction, how can it possibly be putting out valid individual rankings... https://webcache.googleusercontent....an/39949/rooney-tim+&cd=1&hl=en&ct=clnk&gl=us
 
Exactly the point! "The computer" shouldn't be doing anything until early January. If it doesn't know any better than to make this prediction, how can it possibly be putting out valid individual rankings... https://webcache.googleusercontent....an/39949/rooney-tim+&cd=1&hl=en&ct=clnk&gl=us

If you don't like it, why not just ignore it? For me, wrestlestat is the best tool on the web to quickly get my stats/numbers fix. Unlike you, I could care less about subjective rankings. Means little, except for bragging rights (temporary). Most years we aren't even number 1 until after nationals, thats how good subjective rankings are.

I really don't pay attention to wrestlestat rankings that much either, but just use the data and like to see how guys climb in the rankings over time. Shows progress and you can see the why of the move. Being able to compare apples to apples is the fun part and what gives a much better understanding of where each team/individual is at any particular point in relation to the rest of the country.

No need to rip on something that you don't understand/agree with just because you don't like its purpose.
 
i just looked at tom's weekly rankings review thread and only 1 site has RBY ranked other than wrestlestat, and that's trackwrestling who has him at 25. why aren't all of the rest of the services put up to the same standard by ccdiver?
 
  • Like
Reactions: dbldoofus
  • Like
Reactions: dbldoofus
If you don't like it, don't open it until January. Personally, I enjoy using it as a tool and knowing that there are inaccuracies early on. This arguing is going no where.
We can always install an internet filter on your computer, @ccdiver.

Some reasons I love Wrestlestat: it's a labor of love, is very transparent about what it is and isn't, and is more responsive to fan feedback than any commecial project I know about (though Willie's Flo Arena inquiry was nice!).
 
We can always install an internet filter on your computer, @ccdiver.

Some reasons I love Wrestlestat: it's a labor of love, is very transparent about what it is and isn't, and is more responsive to fan feedback than any commecial project I know about (though Willie's Flo Arena inquiry was nice!).
Why you quoting me? I like WrestleStat and was supporting it o_O
 
Last edited:
Because he has data from last year when he was competing as a redshirt.

Okay, explain to me then how DeSanto gets ranked ahead of Pletcher.

I understand the rankings will get more accurate as the season progresses, but nothing has happened this year that should overturn a full season of data from last year.
 
I didn't mean to create a firestorm, I just don't understand rankings that ignore reality. Do you think there are 67 better 133's? If UCF plays a football game on August 31 and Alabama doesn't open their season until the following week, do the Knights get the higher ranking? It's silly and doesn't tell me anything.

Re-read my initial post... I didn't tell people who like WrestleStat to steer clear of it, just said it wasn't for me, at least until there is enough data to make it align with reality.
I can see your argument but I love Wrestlestat. I’d really encourage you to take a second look at their site. The tools there are thorough and quite helpful. Everything from team rosters to team/wrestler comparisons and yes even rankings. Check out the rest of the site and skip the rankings.... I’m sure you’ll see what I mean. Btw, I like that they aren’t doing their ranking by opinion like everyone else out there. They go by the numbers and it’s just another way to look at things. I agree it’s tough when you have freshman or new wrestlers entering the picture but that works itself out in due time. I’ll be honest, I think their year end rankings right before Nationals we’re quite impressive.
 
Okay, explain to me then how DeSanto gets ranked ahead of Pletcher.

I understand the rankings will get more accurate as the season progresses, but nothing has happened this year that should overturn a full season of data from last year.

I'm certainly no expert on wrestlestats but to answer your question, to me it seems Desanto has better wins than Pletcher does for the ranking to come out that way. It's not going off of NCAA placements it's going off of a formula that takes into account wins and the better person you beat the more points you get.

If Desanto is ranked ahead of Pletcher than he probably ended the season higher than Pletcher last year
 
Okay, explain to me then how DeSanto gets ranked ahead of Pletcher.

I understand the rankings will get more accurate as the season progresses, but nothing has happened this year that should overturn a full season of data from last year.
I'll give this a go...

-- In March, at NCAA's, DeSanto was #4, Pletcher was #8, so the mathematical model already had DeSanto higher.
-- Their overall win % is within a percent or two, so that's a minimal impact
-- Their common opponents, particularly results against Micic won't separate them much
-- I suspect the biggest factor is Bonus Point Wins, which shows a decided advantage for DeSanto
-- Note: they are VERY, VERY close already, so even using the WrestleStat model, this could be called a toss-up
 
I'll give this a go...

-- In March, at NCAA's, DeSanto was #4, Pletcher was #8, so the mathematical model already had DeSanto higher.
-- Their overall win % is within a percent or two, so that's a minimal impact
-- Their common opponents, particularly results against Micic won't separate them much
-- I suspect the biggest factor is Bonus Point Wins, which shows a decided advantage for DeSanto
-- Note: they are VERY, VERY close already, so even using the WrestleStat model, this could be called a toss-up
A "like" from andegre :). Does this mean I was close? ;)
 
I'll give this a go...

-- In March, at NCAA's, DeSanto was #4, Pletcher was #8, so the mathematical model already had DeSanto higher.
-- Their overall win % is within a percent or two, so that's a minimal impact
-- Their common opponents, particularly results against Micic won't separate them much
-- I suspect the biggest factor is Bonus Point Wins, which shows a decided advantage for DeSanto
-- Note: they are VERY, VERY close already, so even using the WrestleStat model, this could be called a toss-up

Well okay then, but using the mathematical model to support the mathematical model is a bit like constructing a circular proof.

In real life, Pletcher entered NCAAs seeded #3, DeSanto #7. Pletcher finished an AA at #4, DeSanto went 2-2. That mathematical model needs a fudge factor two, IMO.

But, no big deal ... even though I shake my head at some of the rankings, I do enjoy looking at dual comparisons.
 
I'm certainly no expert on wrestlestats but to answer your question, to me it seems Desanto has better wins than Pletcher does for the ranking to come out that way. It's not going off of NCAA placements it's going off of a formula that takes into account wins and the better person you beat the more points you get.

If Desanto is ranked ahead of Pletcher than he probably ended the season higher than Pletcher last year

Pletcher beat 6 of the other 7 AAs (Micic, Wilson, Delvecchio, Brock, Bridges and Parker) ... also beat ranked guys like Terao, McKee, Meyers and Naser. He had 4 losses, all to AAs

DeSanto had wins over 2 AAs (Micic and Bridges) ... also beat ranked guys like Muller, Terao and Forys. He had 9 losses, 7 were to non-AAs.

The only thing I can guess is that the algorithm places a lot of weight on bonus points. That's fine, but I'm still scratching my head on this one.
 
Last edited:
  • Like
Reactions: andegre
I'm certainly no expert on wrestlestats but to answer your question, to me it seems Desanto has better wins than Pletcher does for the ranking to come out that way. It's not going off of NCAA placements it's going off of a formula that takes into account wins and the better person you beat the more points you get.

If Desanto is ranked ahead of Pletcher than he probably ended the season higher than Pletcher last year
There is a rank history link on Wrestlestat, which also has ELO history, by week.

At the end of last year, Desanto was 1454, Pletcher was 1418. As Roar states in another message, not much has happened this year to change those. Desanto currently at 1457, Pletcher at 1424.
 
  • Like
Reactions: andegre
Pletcher beat 6 of the other 7 AAs (Micic, Wilson, Delvecchio, Brock, Bridges and Parker) ... also beat ranked guys like Terao, McKee, Meyers and Naser. He had 4 losses, all to AAs

DeSanto had wins over 2 AAs (Micic and Bridges) ... also beat ranked guys like Muller, Terao and Forys. He had 9 losses, 7 were to non-AAs.

The only thing I can guess is that the algorithm places a lot of weight on bonus points. That's fine, but I'm still scratching my head on this one.
It's "body of work", so AA really doesn't matter. DeSanto doesn't have 9 losses, he has 7;
-- 3 against Jack Mueller (24-7 last year, before NCAA's), whom he also MD'd at NCAA's
-- 1 vs Delvecchio
-- 1 vs Micic, whom he also MD'd
-- 1 vs Tucker
-- 1 vs Parker

The Parker loss is the only marginally bad loss, and he was 20-8 last year.

Guessing you're including the two MFor's, which don't count. The more I get into this, the more I see how close these two guys are. And I agree that the Bonus Point wins are what separates the two.
 
  • Like
Reactions: andegre
It's "body of work", so AA really doesn't matter. DeSanto doesn't have 9 losses, he has 7;
-- 3 against Jack Mueller (24-7 last year, before NCAA's), whom he also MD'd at NCAA's
-- 1 vs Delvecchio
-- 1 vs Micic, whom he also MD'd
-- 1 vs Tucker
-- 1 vs Parker

The Parker loss is the only marginally bad loss, and he was 20-8 last year.

Guessing you're including the two MFor's, which don't count. The more I get into this, the more I see how close these two guys are. And I agree that the Bonus Point wins are what separates the two.
Just so people are aware, win type is not built into the chess version of ELO. Wrestlestat, if I'm not mistaken, built that in, and ran scenarios to see what various multipliers for win types would do.
 
  • Like
Reactions: andegre
ADVERTISEMENT