NAIA XC/TF 2024-25

Shut up.

can't wait to see Cumberlands destroyed on Friday in Knoxville by Milligan for example and other teams :)

TLDR: Used a Network model to evaluate all head to heads for each runner in the NAIA to generate a ranking for them. This system uses margin of victory over an opponent rather than time. Biggest draw back of this method is you can’t factor for bad races, tempoing a race, etc. I have provided a larger write up of the methods at the bottom.
The model has really punished anyone who ran at Gans Creek, as Spring Arbor beating many people there, but then running relatively even with Goshen at Bethel pushed all those teams down. Anyone who didn't run at Louisville avoided this. I might even redo the rankings to exclude the Bethel meet entirely--but it's hard to judge the "effort level" at any given meet.
At this point I don't consider these rankings much more than a novelty but I am open to suggestions if anyone has ideas on accounting for meets to exclude or how to more appropriately rank the individual runners. The model will likely do much better at predicting a nationals scenario as more crossover is generated with the regional races this week and then conference championships.

NAIA Men’s Great Lakes Model Predictions:
Team Score
Rio_Grande 50
Shawnee_State 85
Olivet_Nazarene 115
Campbellsville 146
Indiana_East 165
Indiana_Wesleyan 171
Spring_Arbor 179
Marian_IN 193
Taylor 205
Grace 224
Aquinas 263
Lawrence_Tech 273
Goshen 283
Cornerstone 344
St_Francis_IL 420
Governors_State_University 511
Indiana_Tech 520
Lindsey_Wilson 526
Madonna 534
Ave_Maria 534
Missouri_Baptist 556
Carroll_MT 576
Rocky_Mountain 590
Montana_Tech 660
Athlete TeamName Weighted BestTime TimeInSeconds
Tyler__Jenkins Rio_Grande 147.5 24:33.9 1473.9
Dominic_Malel Campbellsville 173 24:29.9 1469.9
Duncan_Kogei Rio_Grande 227.5 24:43.7 1483.7
Isaiah_Lanoy Shawnee_State 251.5 24:53.8 1493.8
Danuel_Persinger Rio_Grande 294 24:56.7 1496.7
Shawn_Little Shawnee_State 384 25:12.4 1512.4
Eli_Fullerton Indiana_Wesleyan 385 24:14.4 1454.4
Mylan_Nocton Indiana_East 430.5 25:06.0 1506
Flynn_Prather Olivet_Nazarene 453 25:06.7 1506.7
Brandon_Armendarez Olivet_Nazarene 496 25:11.7 1511.7
Kaiden_Lanoy Shawnee_State 533 25:30.8 1530.8
Brandon_Kleber Grace 565 24:49.0 1489
Mathew__Keitany_ Goshen 586.5 24:37.4 1477.4
Lukas_Smithson Campbellsville 631 25:28.0 1528
Full Individual Ranking:

docs.google.com

Mens Great Lakes Rankings

Open link

METHODS:
Use a Networks model to objectively rank each NAIA runners performance by analyzing any meet in which an NAIA runner participated in since 8-31-2024, assuming it was loaded to TFRRS. 154 as of today. This data includes 11,292 runners across all divisions.
Part 1: Match ups
This model generates each unique head to head matchup every runner was involved in during the year.
EXAMPLE:
Meet: Gans Creek 9-21-2024
Runner1: Ethan__Barker vs Runner2: Matt_Schieber
Margin of Victory : 5.5 (Positive Implies Victory for Runner1)
Meet: Seminole Valley 10-05-2024
Runner1: Ethan__Barker vs Runner2: Matt_Schieber
Margin of Victory: -5.6 (Negative Implies Victory for Runner2)
Part 2: Calculated Margin of Victory
Next generate a calculated amargin of victory for each runner against all of his opponents. This margin of victory was weighted so that more recent results hold greater value. While a standard average would show Matt_Schieber over Ethan__Barker by .05 seconds, this model uses a Time Decay Average Margin of Victory of .76 as Matt beat Ethan at their most recent race. This can be interpreted that you would expect Matt to beat Ethan by .76 seconds.
EXAMPLE:
Runner1: Ethan__Barker vs Runner2: Matt_Schieber
Standard Average Margin of Victory: -.05 (Negative Implies Victory for Runner2)
Time Decay Average Margin of Victory: -.76 (Negative Implies Victory for Runner2)

Part 3: Directed Graph
The graph network will have a single node, represented visually as a dot, for each runner in the results database. We now construct an edge, represented visually by a directional line, pointing from the slower runner to the faster with a weight value of “Time Decay Average Margin of Victory”. The weight value can be conceptualized as the thickness of the line. Runners who are relatively similar will be connected by a thin line, while runners with disparity will be connected by a thick line. This results in a graph with 11,292 nodes representing all individual runners, each with approximately 100 to 1500 incoming or outgoing lines connected to runners they have competed against. If the line is pointing into the runner, it indicates they are superior. If the line points to a different runner, that means the original runner is inferior.

Part 4: Ranking Algorithm
Next generate a ranking with the PageRank algorithm, using the thickness of the arrow to weight the results. Second create a separate ranking using Eigenvector Centrality. Create a composite ranking to be used as the final ranking by:
FinalRank = PageRank_Rank *50% + Eigenvector_Centrality_Rank *50%.

Part 5:
Order the Runners by FinalRank. Best time is provided only as a reference.

Wow I really do not care lil bro

Man this model was atrocious

Wow you probably just wasted hours of your day just to come up with the worst model I've ever seen in my life.

I upvoted comments saying how bad it is--I already know.

Anyone have a meaningful thought on improving it?

Example: Eli Fullerton is dragged down because Brandon Kleber is ranked as a close competitor, but Brandon was beat by many guys at the IWU invite. Is it better to toss out the early season races entirely?

And I don't feel too bad--I had fun and learned a lot even if the results aren't great. I've been around this site long enough to know there are some genuinely smart people lurking the boards, so if I get one piece of meaningful feedback, it's worth sharing despite the harsh criticism.

Lawrence Tech ranking in this "model" is hilarious.

Don’t listen to them Nick. Shawnee State is going to take Spring Arbor this weekend and prove all of the haters wrong. Most of the trolls on this forum are your own athletes so I would tell Woj to get them to stop bullying you.

The guy posted something that took effort abd time and negative losers post about it. You go do better bum

Apparently, Anthony Robert's was struck down with diarrhea and Drew Hogan is not redshirting anymore. This will affect Goshen's score as Robert's won't be scoring but Drew will clear the field by over 20 seconds. Also, Korry Hamlin from Grace College still has eligibility left and seems lethal coming off his latest marathon. Ben Bird will also be at this meet! This guy doing the rankings is out of his mind, the meet winner is clear, Zach Gaknic had a good workout last week and is excited to be at this meet. His immense joy will be just enough to come out on top!

If I don’t get an award for every time I take an effort filled dump while constipated then he doesn’t get an award then he doesn’t get an award for an effort filled dump model

Nice uneducated response loser.

Very lol guy, I’ll have you know I went to hustlers university so how dare you call me uneducated and a loser. I’ll also have you know I skipped the first grade buddy so I’m a pretty smart fella.

This guy could be pretty legit Macey, I'd back off if I were you. If he really studied at hustlers university he likely is a very high IQ individual. I had to drop out after two months cause even I couldn't handle on everything.

I’ve returned for more abuse 😊
I have rerun the Men’s model by tossing out Bethel Invite as it was a known drag on Spring Arbor, and thus the rest of the Crossroads.
Next I swapped the Final Rank to be 50% of your composite rank from last time, 50% your rank by season best time.

All the individuals ranks are here:

docs.google.com

Great Lakes Rankings

Open link

This method generates a team situation which seems much more plausible to me:
TeamName Score
Rio_Grande 88
Shawnee_State 113
Indiana_Wesleyan 131
Spring_Arbor 136
Olivet_Nazarene 150
Taylor 165
Aquinas 174
Campbellsville 182
Marian_IN 221
Grace 240
Indiana_East 248
Goshen 311
St_Francis_IL 327
Cornerstone 329
Lawrence_Tech 370
Indiana_Tech 463
Rocky_Mountain 470
Carroll_MT 472
Madonna 526
Missouri_Baptist 529
Lindsey_Wilson 564
Ave_Maria 573
Governors_State_University 614
Montana_Tech 616

Compare with Lacctic rankings:
Indiana Wesleyan 96
Spring Arbor 105
Rio Grande 130
Olivet Nazarene 157
Taylor 158
Shawnee State 185
Campbellsville 220
Aquinas 284
Grace 285
Carroll (Mont.) 292
Rocky Mountain 293
St. Francis (Ill.) 298
Indiana East 315
Marian (Ind.) 342
Goshen 369
Cornerstone 384
Lawrence Tech 415
Governors State 502
Madonna 520
Montana Tech 526
Ave Maria 589
Indiana Tech 592
Lindsey Wilson 620
Missouri Baptist 650

Some personal observations on teams:
Rio Grande: The model loves them because they ran fast a big meet. Beating many other high quality runners improves your network rank.

Shawnee State: I can’t quite see why the model loves them, but maybe it indicates they are an underranked team.

Grace: The model hates them and I’m not sure why.

Lawrence Tech: The original composite rank loves this team because they placed 5 in front of UM-Dearborn #1 and #2. Those UM Dearborn runners ran very close to the Aquinas pack at Seminole Valley. When you factor back in season best time, this washes out and they appear to be ranked more appropriately now.

IU East: Similar to Lawrence Tech. They got tied to the success of Rio Grande in the original method. Using best time washes out some of this effect.

Carroll / Rocky Mountain:
They get isolated in an island in the network and can’t make a meaningful jump out to rank against other teams at this time. My gut feeling is they will race close to their current NAIA rank, somewhere near us at Aquinas and Campbellsville.

Looking forward to seeing everyone at the meet. Coach Woj put a lot of effort into making the course as best as it can be, and if the recent HS races are any indication, it should put up some fast times.

Ah come back for more eh???

good girl