Team rankings are difficult, especially in a sport with as little results as football.
But yet at all levels of the game, computer rankings are used to unbiasedly compare teams. These rankings are even used by the SDHSAA with their point system. Many other states also incorporate rankings into their state tournament pairings (see Iowa).
Inspired by such rankings as BCMooreRankings.com, I have decided to incorporate computer rankings into a power poll for South Dakota football, spending the last 4 months researching ranking systems and mathematical formulas (I encourage you to go check out BCMoore so that you can have an understanding of what goes into these formulas).
And with this studying, a computer ranking was developed. I painstakingly ran the system through multiple years of South Dakota football, and even expanded it to other states.
The results?
The system did really well.
In 2021 for example, the power ranking system correctly predicted 78% of the playoff results, out performing the power points system used by the state to seed teams. It even predicted Beresford’s run to the playoffs despite being a #14 seed. This was due to the fact that Beresford faced a tough schedule of teams that also had tough schedules, deflating the value of their opponents and therefore deflating the value of the Watchdogs in the eyes of the power point system.
There were other examples similar to the previously described situation, and the system did a great job of finding these anomalies. It also correctly predicted all the state champions in that same year.
It should be noted that upsets do happen.
The system will miss on some games, but upsets are what make sports interesting. Last week, it incorrectly predicted the winner on just one 11-Man game (West Central vs Canton) and three 9-Man games (Faith vs Timber Lake, Corsica-Stickney vs Avon, Florence/Henry vs Deubrook Area). For the other 56 games, it correctly predicted the winner. Therefore, it’s safe to say the system has done very well.
So how does it do so well?
The only stats used in this ranking is the final score, and the home/away schedule of the competitions. The rankings don’t take into account class, reputation, past championships or results from a previous season. This is to allow the system to be as unbiased as possible.
The rankings then compare the results of the games, connecting teams through their opponents, and forecasts each game on the schedule. After games, the system readjusts a teams rankings to try and keep the squared error of the forecast, compared to the actual result, as low as possible.
I also set limits to how much a team can win by, or be forecasted to win by. This is to prevent blowouts. After all, what is the difference between beating a team by 50 or 60? Therefore in classes that have a mercy rule, games are capped at a 50 point victory.
For classes with no mercy rule, no victory above the standard deviation of a final score is used. This varies on a year-to-year basis, and currently sits around a 40-point win in 2022. In other words, once you beat a team by 40, it doesn’t affect the rankings in any way by beating them by more.
Each week for the rest of the year, we will go through the weekly results and how they changed the rankings so that it is as transparent as possible. Even though the ranking does rank every team in the state, I will work to present and talk only about the Top 5, as I don’t see a reason to show struggling teams.
The formula is a little too complicated for me to put on here, so I’ll leave it off, instead showing a Rank+ score that is calculated from the final ranks. A 100 ranking is class-average, with any point above 100 equal to 1% increase in rank.
Again, this score is listed as Rank+.
I choose to do this because the actual raw ranking results allow one to predict scores between teams, and I am not comfortable publicizing that yet. But if there is enough public interest I might post week-by-week of select games.
One thing to remember is that as the season continues, more games are played, and more data is compiled. This is important to note, because strength of schedule and your results in that schedule are important right now. If you play a bad game, the system is going to place you lower. But as more weeks are played and the team continues to show good results, the system will continue to raise their rank.
I will break down the first public rankings here, and go through each class.
11AAA
| Name | Record | Rank+ |
| Jefferson Cavaliers | 5-0 | 172 |
| O’Gorman Knights | 4-1 | 162 |
| Harrisburg Tigers | 4-1 | 140 |
| Lincoln Patriots | 2-3 | 113 |
| Brandon Valley Lynx | 3-2 | 106 |
There isn’t much difference between these rankings and the media polls, though I think we will see Lincoln drop to #5 and Brandon Valley move to #4 in those rankings.
Game of the Week: O’Gorman @ Jefferson
Last Matchup: October 21st, 2021 (Howard Wood). O’Gorman 49, Jefferson 6
I don’t think there has been an 11-man game as anticipated as this one. It’s a de facto regular season championship, with the winner most likely securing the top spot for the remainder of the regular season.
11AA
| Name | Record | Rank+ |
| Pierre Governors | 5-0 | 224 |
| Tea Area Titans | 5-0 | 180 |
| Yankton Bucks | 3-2 | 161 |
| Brookings Bobcats | 4-1 | 114 |
| Aberdeen Central Golden Eagles | 2-3 | 90 |
Again, not too much of a shock here. Tea and Yankton was a very good game that didn’t give either team too much separation (although Tea did gain some Rank+), and Pierre holds the top spot pretty securely. Aberdeen is coming off of back to back wins, so #5 is entirely appropriate. Now, you may ask how Aberdeen has a Rank+ below 100. Remember, Rank+ is an average of everyone in the class, and Pierre and Tea Area lift that average really high.
Game of the Week: Yankton @ Brookings
Last Matchup: October 15th, 2021 (Brookings). Brookings 36, Yankton 20
This week’s matchup is not only going to be big for these rankings, but also playoff seeding. Yankton has had back-to-back weeks of tough games, and the Bobcats are going to be a tough matchup as well. Brookings has won 3 straight games, and took a 36-20 win against the Bucks last year. In fact, Yankton has not beaten Brookings since August 26th, 2016 in a 29-28 thriller.
11A
| Name | Record | Rank+ |
| Dell Rapids Quarriers | 5-0 | 164 |
| West Central Trojans | 5-0 | 161 |
| Beresford Watchdogs | 4-1 | 154 |
| SFC Chargers | 3-2 | 143 |
| Canton C-Hawks | 2-3 | 142 |
11A has been a lot of fun this year, with a ton of exciting matchups between the top teams in the class. With a Rank+ this close, it means that any of these teams could potentially beat anyone else on a given night (with obvious favorites). Dell Rapids and West Central are the consensus top two anywhere you look, while the rest of the top 5 is up for debate. Beresford, Canton, and SFC are #3, #4, #5 respectfully in the SDHSAA’s power point system, and all three are receiving votes in the media, along with Dakota Valley and Lennox.
Now, the Panthers do have a better record than both Canton and SFC, but remember, as this is important for later, winning games isn’t the be-all-end-all in these rankings. (Note: DV is justtttt out of the top 5. With how close these rankings are, one great or bad performance from any of these teams is going to change the top 5 pretty quick!)
You have to play well against the teams who have struggled this season, and have to stick with teams who are favorites against you. Canton has done an excellent job of this, losing by only 1 to Beresford, and 6 to Dell Rapids. A 22-point loss last week against West Central was their lowest performance according to the computer, but good performances against SFC and Lennox helped eliminate the negative.
Dakota Valley, on the other hand, didn’t perform as well against expectation versus SFC. They have also had close games against some teams that haven’t performed great so far this year. It’s no fault to the Panthers, who just keeping winning games, but the computer also looks at how well your opponents have done, and well, they haven’t played particularly great or at a level that would compare to top 5 competition.
If you have a close game against a team that struggles the rest of the season, it will lower your ranking, because that’s what the results have shown. You don’t have as much control of the schedule as some teams would like, and you can’t help other teams not performing well, but as the season goes on, teams like Dakota Valley will have more opportunities to perform well in situations that they can control, and outplay the results of games they didn’t perform well in.
In particular, those last two games on the schedule, #3 Beresford and #1 Dell Rapids, are going to be big indicators of not only where the Panthers should be in the computer rankings, but also how they might perform in the playoffs.
Game of the Week: Dell Rapids @ Beresford
Last Matchup: October 6th, 2006 (Beresford). Dell Rapids 35, Beresford 0
Anytime two of the top teams in the class battle it out, it’s going to be fun. The computers calling for basically a coin flip of a game, with Dell Rapids (obviously) as the close favorite. I’m excited to see this one play out.
11B
| Name | Record | Rank+ |
| Elk Point-Jefferson Huskies | 5-0 | 210 |
| Winner Warriors | 5-0 | 176 |
| Roncalli Cavaliers | 5-0 | 169 |
| Mount Vernon/Plankinton Titans | 6-0 | 157 |
| McCook Central/Montrose Fighting Cougars | 5-1 | 151 |
EP-J has frankly dominated this year, outscoring opponents 251-7. When you win games constantly by mercy rule, your going to be the top ranked team. Winner is also being Winner, and cruising to a lot of victories themselves.
After the Warriors, how about Roncalli?! I hinted last week that I thought the Cavaliers were underrated, and that’s because I have had access to these rankings the whole year (haha), and they were in my top 5 before they reached the top 5 for the media! Seriously though, they have put together a great season and have a big game this week against STM. If they win that matchup in Aberdeen, they have an excellent chance at the #1 seed in the playoffs.
MV/P and B-E/E are the other pair of undefeated teams. While MV/P is #4, B-E/E comes in at #6 just behind the 1-loss MC/M. The Fighting Cougars only loss is to EP-J, so there is a note, but I think the divergent point was September 9th.
On that day B-E/E won 20-2 against Parker, while MC/M beat WWSSC 40-0, who is the 12th rated team in the class according to our Power Rankings, and is 9th in the SDHSAA’s Power Points. That win was huge for the Fighting Cougars season, and the Seahawks haven’t gotten a game that big yet, but that also figures into strength of schedule.
The good news for B-E/E fans?
They have two really big games these next two weeks…including:
Game of the Week: Winner @ B-E/E
Last Matchup: November 12th, 2021 (Vermillion). Winner 40, B-E/E 8
Coming off a bye week, this game also means A LOT for the Seahawks community. Winner has beaten them in 2021 and 2020 for the state championship. B-E/E won the 2019 championship matchup, however, so there is definitely some history.
The Warriors also have a really tough schedule coming up (West Central, MV-P) while B-E/E will have a chance to boost their ranking the week after this against MC/M.
Although I’m sure that neither team is looking past this week. This is the game both have had circled on their calendar.
9AA
| Name | Record | Rank+ |
| Wall Eagles | 6-0 | 220 |
| Bon Homme Cavaliers | 4-1 | 187 |
| Hamlin Chargers | 6-0 | 181 |
| Hanson Beavers | 4-1 | 177 |
| Parkston Trojans | 5-1 | 164 |
Okay, now it gets really interesting. Two undefeated teams were left out of the top 5, including the defending champions, Howard, along with Elkton-Lake Benton, the team that beat Bon Homme.
Speaking of Bon Homme, the Cavaliers haven’t gotten a single vote in the media polls. Is the media underranking the Cavs so far this season? According to our rankings, this is the case. I’m actually super excited about this difference, because it’s a great learning moment.
I don’t think any of the other four teams in this ranking are really controversial, and Wall is pretty firmly at #1 after dominating (309-30 point differential) this year. So let’s take a look at the other three.
Bon Homme beat Platte-Geddes 42-6 on Friday. Before that, P-G took down Kimball/White Lake 32-0, and Viborg-Hurley 35-18. Viborg-Hurley also beat Canistota by 6. Why does any of that matter?
Remember, the computer takes all the game scores from 9-man. All the games are connected so your opponent’s results and their opponent’s results are all connected. This helps the computer get a gauge of your entire body of work, not just one game.
Howard beat Canistota by just 1, and the next week won by 3 against Kimball/White Lake. From the eye test, I thought their win 51-0 over F/M/FA was notable, but when I later looked, the Phoenix haven’t had a difficult schedule, and with their best win against Avon, who dropped to 3-3, it doesn’t help (Everything is connected!).
Irene-Wakonda is a great connector, as they have faced a bunch of tough teams, and the Tigers won by 34 against I-W, but it wasn’t enough to put them into the top 5, especially when teams like Hanson and Gregory won by more.
If Howard picks up a big win against Viborg-Hurley, that is surely going to help them, and a deep playoff run is crucial too. Now, you may also wonder because Howard beat Hanson in week 1. And that is a fair point, but I want to point out that if you expanding the whole rankings to show all the teams in a class, the difference between Hanson and Howard is a about a touchdown. These classes that are stacked at the top don’t take much to change.
As for E-LB, they are right there with with Parkston and Howard. The computer literally has matchups between many of these top teams as literal toss-ups, with situations that could feature a 2-point conversion as a game changer.
Finally back to the Cavaliers. Despite their loss to E-LB, who again, has a very close ranking to Bon Homme, the Cavaliers have shutout victories over K/WL (46-0) and V-H (20-0), and a solid win over P-G (42-6). Those three have played a lot of these ranked teams, and are therefore, sort of connectors between all these teams. It’s these other data points that allows teams to cancel out loss or bad victories, as long as teams perform well against teams that play a lot of other tough teams.
That’s exactly what Bon Homme has done. I hope that gives a good explanation, and I thank you if your still with me.
Game of the Week: Howard @ Viborg-Hurley
Last Matchup: October 12th, 2018 in Howard. Viborg-Hurley 54, Howard 52
A big victory here will give Howard a chance to strengthen the resume.
9A
| Name | Record | Rank+ |
| Gregory Gorillas | 5-0 | 192 |
| Lyman Raiders | 4-1 | 178 |
| Harding County/Bison Buffaloes | 5-1 | 176 |
| Philip Scotties | 4-1 | 157 |
| Wolsey-Wessington Warbirds | 3-2 | 155 |
Well, if I haven’t worn you out yet. Let’s get to this ranking. Undefeated Warner and Alcester-Hudson have been left out, but neithers S.OS. has been great. Meanwhile, W-W lost to Gregory and Parkston, but with both games close against great teams, along with the Warbirds big win against De Smet, they have done enough to reach the Top 5.
9AA has proven to be the strongest class of the three, and if Warner can take down their #3 Hamlin, that’s going to give them a great game for the computer. The Monarchs, again, though, are so close to the top 5 that I really should just put the top 7 or 8 in these rankings, and in next few weeks if it interest everyone, I can expand the rankings before the playoffs.
For Alcester-Hudson, they are in the same boat as EP-J, they have played some close/close-ish games against teams that haven’t played well this season. However, if they just keep winning and make a good playoff run, I suspect we will see their name in the Top 5.
Game of the Week: Hamlin @ Warner
Last Matchup: October 15th, 2021 (Hayti). Warner 35, Hamlin 24
It only features one 9A team, but I’m curious where a win would move Warner.
9B
| Name | Record | Rank+ |
| Herreid/Selby Area Wolverines | 5-0 | 198 |
| Kadoka Area Kougars | 3-2 | 176 |
| Hitchcock-Tulare Patriots | 6-0 | 168 |
| De Smet Bulldogs | 3-2 | 162 |
| Sully Buttes Chargers | 5-1 | 124 |
Last but not least, everything in these rankings is pretty straightforward. Kadoka Area has a loss to Wall, who is by and far considered the best 9-man team according to the computer. The Kougars also lost to Lyman, but they are again considered a top two team in their class.
Corsica-Stickney has the best record of teams left out, but it’s worth noting, that their wins are against team’s with a combined record of 5-22. Record is a decent gauge of where teams are at, as wins obviously means that team performed to a certain floor, so the Jaguars are going to need another good win to get into the top 5. Their final matchup against Sully Buttes will be that important game.
Game of the Week: Alcester-Hudson @ Avon
Last Matchup: October 28th, 2021 (Avon). Avon 30, Alcester-Hudson 14
I think it’s a fun game that is going to be a connector in the future.
Love it. Hate it. There you have it, the first public computer rankings.
Each week, I’ll take a look at the class-by-class results and we can see how the rankings reevaluate teams. We’ll post the composite rankings that combine the Power Rankings, Media Rankings and the state’s Power Points.
I’ve also been working on historical seasons. I will be posting these at the end of the season!
Thank you for taking the time to read this article and support our journey. We now have over 2,500 followers on Twitter and have seen a lot of growth on Instagram.
If you have any questions or want to learn more, you can email me at digitalzsports@gmail.com