By the Numbers, Vol. 3: Dec. 10 through Dec. 15
If you have been following both the “human” polls and these computer rankings, you may notice that there are some big differences in where a few teams fall. Some may have you pleasantly surprised, while others may cause you to…
Access all of Prep Girls Hoops
Continue reading this article and more.
Continue ReadingIf you have been following both the “human” polls and these computer rankings, you may notice that there are some big differences in where a few teams fall. Some may have you pleasantly surprised, while others may cause you to shake your head in disgust. I have found a few reasons as to why this happens.
First, and probably the biggest impact early on, is that on the computer side of things the only way to start the season is to begin with the data from previous seasons. This can be somewhat accurate as teams tend to be similar to how they have been, but there is no way for it to account for improvements, transfers, or graduation. Human polls do this much better and will therefore have a better chance of accuracy early on.
Secondly is something I call recency bias. A human poll tends to treat recent games with more weight and therefore you often see teams that lose in the beginning of the season and win toward the end ranked higher than those that do the opposite. The computer weights all games the same (unless there is a blowout) whether they occur in November or February. This is good in theory, but teams evolve throughout the season so games at the end of the season probably do tell a more accurate story for each team even though every game matters.
Lastly is the head-to-head fallacy. It is quite widely accepted that when Team A beats Team B that they are most likely the better team. This is usually true, but some take it too far. We often use one game as a way to rank the winning team higher no matter what. Human polls, and section seeding, do this on a regular basis. The problem with this is that the teams play 20+ games in a season, and to use one game to judge whether a team is better diminishes the other results.
Both systems can exist and be valuable for teams. The computer ranking system provided here will continue to evolve in order to become as accurate as possible, but it will not treat teams the same way as a human poll. I hope that people can be open to it as I think it can provide some fun and unique information.
Biggest Upset
Greenway 38, Littlefork-Big Falls 35 (2.15% chance of upset)
Littlefork-Big Falls had been mostly competitive in their season to date while Greenway had not. That is why this result was so surprising. With the exception of their win over Wrenshall, the closest Greenway had been to anyone was 25 points, including a 27-point loss to Deer River. The same Deer River team squeaked by Littlefork-Big Falls by three.
Honorable Mention: Pine River-Backus 46, Pequot Lakes 39 (3.12%), Blake 68, Jackson County Central 54 (6.02%)
Biggest Gainers
A – North Woods (last week Class A #95, currently #70)
North Woods started this week with a “productive” loss to a Cherry team that has been blowing teams out of the water. They followed that loss up with a blowout of their own against Fond du Lac Ojibwe.
AA – Lake Crystal-Wellcome Memorial (last week Class AA #65, currently #37)
Three wins this week, including one over another team moving up the rankings in Belle Plaine, have LCWM as the Biggest Gainer in Class AA. I’m not yet ready to anoint them as a top 20 team, although the win over Blue Earth last week means they have potential.
AAA – Princeton (last week Class AAA #36, currently #24)
Princeton went 1-1 on the week with a blowout win over Sauk Rapids-Rice and a loss at Bemidji. The blowout win was the most impressive result for Princeton this season so far, just edging out their opening win over Grand Rapids. The loss at Bemidji does not hurt as much as one would believe, with the Lumberjacks playing a difficult northern schedule so far.
AAAA – Tartan (last week Class AAAA #50, currently #38)
Only one game this week for the Titans but they made the most of it, blowing out St. Paul Highland Park for their first win of the season. The 45-point win rates as the second best outing for Tartan this season, behind the 28-point loss to Hopkins.
Biggest Fallers
A – West Lutheran (last week Class A #63, currently #100)
Two blowout losses to good, but not great, team have West Lutheran in this spot. Both CHOF and Lester Prairie/Holy Trinity beat the Warriors but 35+ points.
AA – Plainview-Elgin-Millville (last week Class AA #45, currently #82)
PEM took three losses this week. In those three losses were two blowouts to good teams in Dover-Eyota and Rochester Lourdes and a seven-point loss to St. Charles. Good news is there should be a couple of good opportunities for wins this week.
AAA – St. Croix Lutheran (last week Class AAA #27, currently #39)
How could a team that goes 2-1 on the week be our Biggest Faller in Class AAA? Some of that has to do with the preseason ranking still factoring in. The other part is that the teams they have played so far this season are not holding up their end. When teams you have played all underperform either by score or flat out losing to teams they shouldn’t, your wins (or losses) do not look as good as they used to.
AAAA – Champlin Park (last week Class AAAA #21, currently #33)
Champlin Park went 0-3 this week with losses to Maple Grove, Minnetonka, and an upstart Spring Lake Park team. The Rebels probably were not predicting a winless week, but they will need to figure it out soon as they have two more conference games this week.
Highest Scoring Games
It was a pretty average week for points with 161 being our lowest high total so far to date. It also appears that the points are racking up in the southern part of the state as no one on the list this week is north of St. Cloud.
- Becker 91, St. Cloud Apollo 70
- Lake Crystal-Wellcome Memorial 89, Cleveland 66
- Marshall 81, Worthington 71
- Park Center 92, Anoka 60
- Austin 75, Mankato West 74
- Norwood Young America 83, Belle Plaine 66
- Southwest Minnesota Christian 84, Murray County Central 65
Predictions for Next Week
WARNING: Most of these predictions still take into account preseason ratings, since many teams have not yet reached eight games. Also, predictions for later in the week will change based on the results of early in the week, but I put them out there because it would be boring to just put out predictions for Monday’s games. A reminder, the PEV stands for the Potential Excitement Value of the game.
Monday
- Minnewaska @ Belgrade-Brooten-Elrosa (Minnewaska 54-53 PEV: 76.3)
Tuesday
- Apple Valley @ Farmington (Farmington 53-49 PEV: 83.0)
- Rochester John Marshall @ Austin (Austin 77-71 PEV: 74.8)
Thursday
- Sauk Centre @ New London-Spicer (New London-Spicer 67-61 PEV: 84.2)
- Duluth Marshall @ Cherry (Cherry 77-66 PEV: 75.1)
Friday
- Eden Prairie @ Prior Lake (Eden Prairie 82-79 PEV: 83.8)
- Chaska @ Buffalo (Buffalo 71-67 PEV: 84.2)
Saturday
- Proctor @ Roseau (Roseau 63-55 PEV: 81.1)