Unlike NCAA basketball or football, you don’t hear a ton about star ratings and recruiting class rankings when it comes to hockey. Over the past three years Neutral Zone has taken on the task of evaluating players in every hockey market in North America and coming up with a comprehensive ranking/rating system. As we head into the 2019 Frozen Four, we decided to analyze the past three seasons recruiting class rankings and breakdown different strategies and patterns in the data.
Our study focuses on 40 teams in NCAA D1, these are all teams that were top 25 in a major recruiting category from 2016-2018; those categories meaning star rating average, recruiting class rankings or total points.
NCAA D1 Recruiting Class Rankings Study
Neutral Zone created an algorithm that ranks NCAA team recruiting classes based mostly on the individual players in the class’ star rating but it also weighs the age of the player, the league they come out of and position. Each has a different weighted average based on the past 5 seasons trend. For example, success rates of a player coming out of the USHL are different than a player coming out of the CCHL so there is a weight assigned to each league for example. 18-year-old freshman have different stats than 20-year-old freshman and so there is a weight for age as well.
Below (Chart #1) breaks down 40 teams and their recruiting class rankings over the past 3 seasons. Then it takes a cumulative average and that is the order the chart is in from best class average to the least.
Chart #1 Recruiting Class Rankings (2016-2018) – 3 Year Average
What this graphic shows us is a few patterns; first the blue-chip programs (BU, North Dakota, Minnesota, Michigan) have had top 20 classes every year where as other teams have major fluctuations like Penn State who went from 3rd best class in ’16 to 23rd best in ’17 to 57 in ’18. Now some of this is explainable; in ’16 they brought in a big class and therefore they didn’t need as many recruits in ’17 or ’18 so it’s not as if they aren’t getting good players anymore, they just didn’t need as many.
Some interest storylines here are teams on the rise like Denver and UMass Amherst; both teams who are both Frozen Four teams. UMass had the 45th ranked recruiting class in 2016, then jumped to #2 class overall in 2017 and followed that up with #16 class in 2018. Denver was even more impressive going from 53rd ranked class in 2016 to #6 class in 2017 and then #2 class in 2018.
Some coaching changes have made some impacts in these rankings as well; from 2017 to 2018 Nebraska Omaha went 57th recruiting class (3rd worst in the league) to #9 and that was with two new assistant coaches at the helm. Tony Granato & Company have seen Wisconsin go from a 30th rank recruiting classes when they took over to back to back top 10 classes.
Also noteworthy is the recruiting track record of assistant coach Ben Barr, he was with Western Michigan in 2016 where he landed a top 10 recruiting class then went to UMass Amherst and took them from #45 the previous year to a #2 recruiting class and then followed up with #16 class this past year. Coach Carvel and Jared DeMichiel have a major role in that as well but still noteworthy.
Given some coaching changes and more relevant data we broke down the previous 2 years average and re-ranked instead of the 3-year average.
Chart #2 Recruiting Class Rankings (2017-2018) 2 Year Avg
In Chart #2 we can see teams like Denver go from #16 to #3, UMass Amherst go from #17 to #6; Quinnipiac #26 to #10 and UCONN from #27 to #17. These are teams who have picked up their recruiting the past two season and were weighted down by a lower ranked class in 2016.
Chart #3 Recruiting Class Points (2016-2018) 3 Year Average
|36||Alaska – Fair||717.71||259.11||601.69||1578.51||526.17|
The algorithm we use at NZ turns star ratings along with other factors into points. A 5 star player out of NTDP carries the maximum amount of points (over 400 pts) while a 3.5 star prospect out of the NCDC may carry only 75 points. So a recruiting class is ranked based on points which is the accumulation of all the players in the class’ points’. Therefore, this is a measurement of both quantity and quality.
Chart #3 is very similar to Chart #1 in that it shows similar rankings; what is interesting about this is looking at the point differences and the averages. So if you look at a consistent team like Minnesota State you will see a range of between 700-930 over three year period or UNH which is between 675-705 as opposed to a team like Boston University who had the best recruiting class recorded in 2016 with a 1744.14 point total but this past year were 922.74; a difference of over 800 points!
Chart #4 Recruiting Class Points (2017-2018) 2 Year Average
|Rank||School||2017 Pts||2018 Pts||Total||2-Avg|
Same as Chart #2, this graphic just shows the 2-year average as opposed to the 3-year average for the most relevant data. You can see the same jumps here in teams like UMass-Amherst, Denver, UConn and Quinnipiac. We refer to that as “recruiting momentum” and as you can see 3 of the 4 in highest recruiting momentum were in the NCAA tournament and 2 of the 4 are in the Frozen Four!
Chart #5 Recruiting Class Star Rating Average (2016-2018)
For some analysts or coaches this may be the most important metric of all because this measures purely quality of the players. As noted earlier a class brining in 12 players is likely to be much higher than a class brining in 4 players and in hockey recruiting with NHL early signings and other factors classes aren’t the same size every year and most of the year-to-year variability is a result of that. However, this metric ignores that and looks only at the star ratings of the incoming players and this is where we see some interesting patterns.
St. Cloud is a great example here of a team that in the recruiting class rankings and points were ranked 20th and 24th respectively, but when it came quality of prospects in star rating data they are ranked #10. Ohio State is another team with a major jump here going from #15 and #17 in recruiting class ranking and point totals to #7 in average star rating. These would signify both of these teams have brought in high quality players but not as many players as their competitors. Also worth noting both of these teams made the NCAA tournament this season.
The best way to read this is to take a look at the average and then compare that to the individual years data to see the variability and if teams are consistent or not in their talent. Quinnipiac is a good example of a consistent recruiting team (NCAA tournament team) who has had average star ratings of 3.7, 3.81, 3.81. So that is a program that knows the kind of players they want and can get and stick to the formula. Other programs have shown major spikes in their recruiting quality such as Colorado College who went from 3.58 in 2016, 3.50 in 2017 and then a major jump to 4.0 in 2018. Again a coaching change occurred that year when Leon Hayward came into the program as an assistant and landed some major talent upgrades. UConn is another good story who was recruiting a 3.65 caliber player in 2016 then jumped to 3.85 in 2017 and then 3.98 in 2018. Joe Pereria and Brandon Buckley took UConn to another level in talent bringing in 10 NHL picks over those 3 classes including Tage Thompson, Maxim Letunov, Adam Huska, Ruslan Iskhakov and Jachym Kondelik.
Chart #6 Recruiting Class Retention (2016-2018) – Players Leaving Early
|Rank||School||Departures||Total Rates||Points Lost|
NCAA hockey is unlike football where players have to play a minimum of 3 seasons; in hockey they can leave after 1 season like basketball. There are 12 of the 40 teams who have experienced a player in the 2016 and/or 2017 class leave early. We have put together the number of players, the accumulated star rating and the points lost from that class. This is an important graphic as it can help explain why some of the blue-chip programs like Boston University and Minnesota, Michigan, North Dakota and Wisconsin who are all top ranked recruiting programs are not in the NCAA tournament every year. A lot has to do with the very players who give them high recruiting class rankings leave after one or two seasons and therefore teams with better retention get 3-4 years out of their players. A great example of this are the teams you don’t see on this list like Providence, Ohio State, Notre Dame, Harvard and Northeastern; again, all teams in the NCAA tournament.
BU had the toughest impact losing 5-star Clayton Keller, Brady Tkachuk and Kiefer Bellows off the 2016 and 2017 recruiting classes. Minnesota lost Mittelstadt, North Dakota lost Tyson Jost and Wisconsin lost Trent Frederic; all 5-star talents who leave major holes.
Chart #7 Non-Recruiting Metrics – Team Age Average– Youngest to Oldest
*Info was gathered by College Hockey News (www.collegehockeynews.com)
The MOST unique aspect of college hockey is the age discrepancy of its players. Unlike football or basketball where all freshman are 18 or 19 years old; hockey can range from as young as 17 in the first semester to 21. Therefore, you have 18-year-old players going against 24-year-old players regularly. With that being said, teams are built differently; teams with high level prospects out of the NTDP or USHL are likely going to take them at 18 or 19 where as teams building from the NAHL, OJHL, CCHL are more likely to take older, more experienced players at 20 and 21.
There isn’t an absolute pattern in regards to older teams perform better or younger teams perform better, the statistics are mixed. However, there are some interesting patterns; the four youngest teams in college hockey were BU, BC, Wisconsin and Michigan. These four teams have been top 10 recruiting classes over the past 2-3 years and have some of the most NHL draft picks on their rosters. However, their team averages are under 21 years old which means they are bringing in players young and they aren’t seeing all of them over 4 years. None of those programs made the NCAA tournament and none of those programs, despite their talent level, had winning records.
However, that doesn’t suggest older is better; of the 6 oldest teams in the country only Minnesota State made it to the NCAA tournament and lost in the first round. In fact, if we split the 40 teams in half; the youngest 20 teams had 8 teams in the NCAA tournament and the oldest teams had 6 teams in the NCAA tournament.
On it’s own the age of the players doesn’t tell us enough to be statistically significant but used along side other metrics above can help paint a picture and give more context behind the recruiting class rankings.
Chart #8 Non-Recruiting Metrics – Team Size Average – Heaviest to Lightest
*Info was gathered by College Hockey News (www.collegehockeynews.com)
The size factor is an interesting data point in how coaches build their rosters. Some teams with big sheets like UNH are built more on speed and are 3rd smallest team among the 40 here whereas teams like UConn and North Dakota want to play a big, heavy style. There isn’t a real pattern here either to say bigger teams are better as 4 teams from 1-10 heaviest were in the NCAA tournament, 4 teams from 11-20 heaviest were in the NCAA tournament, 3 teams from 21-30 were in the NCAA tournament and 3 teams from 31-40 heaviest were in the NCAA tournament. Again, looking at this along with age along with recruiting class rankings will give a clearer picture of recruiting strategy and roster building. For example, a team like Maine who is big and physical may take 3.75 star prospect who fits their style of play and can be heavy and hard to play against as opposed to take a 4.0 star player who is small, skilled and doesn’t like the heavy lifting. Style of play matters to the coaches recruiting these players and Cornell is a great example of a team who doesn’t have the highest recruiting class rankings but they target a very specific player and they are one of the best program over the past 3 season in Pairwise rankings without the high volume of 4-star and NHL draft pick players.
Chart #10 Recruiting Class Rankings (2016-2018) Weighted & Adjusted Rankings vs. (2017-2019) Pairwise Ranking
The last study was to take all the recruiting data points from Star Rating Average to Points Average (discounting the teams who lost players early) to Recruiting Class Ranking Average and then comparing that to the Pairwise Ranking Average over the past 3 years.
This isn’t an entirely fair metric because in 2016-2017 season we are only accounting for the freshman class; in the 2017-2018 season only the freshman and sophomore class and 2018-2019 season freshman, sophomore and junior class. However, with that in mind it still gives an interesting look at teams success on the ice versus their recruiting success.
If we look at the top 20 teams in recruiting class rankings, they represent 16 of the top 20 teams in average pairwise rankings or 80%. Also to note, 2 of the teams not in the top 20 were ranked 21 and 22. If we look at the average pairwise ranking for a team in the top 20 recruiting classes the average is 13.1. Compare that to the average team ranked 21-40 in recruiting class rankings and its 27.9, a multiple more than 2x. If we look at the teams in the NCAA tournament this year we see 11 of them were in the 1-20 recruiting class rankings and only 3 were outside of the top 20. The other two teams were Bowling Green and AIC who did not make this study based on the recruiting data.
A Closer Look
There isn’t a one-size fits all recruiting strategy. We have taken 4 different programs who have different pairwise rankings and looked at their recruiting trends, their age, their star rating averages, retention, etc.
- Top 3 in: Recruiting Class Ranking (2nd), Average Star Rating (2nd) and Total Points (1st)
- Youngest Team in College hockey with average age 20.5 years old
- Worst Retention Rate in College Hockey losing over 315 pts per year in early departures
- Most 5 Star Prospects signed since 2016
- #15 Pairwise Ranking Average from 2016-2019
Comments: BU has been able to land some of the biggest names in the country since 2016 including Clayton Keller, Brady Tkachuk, Joel Farabee, Joel Oettinger, Kieffer Bellows and Dante Fabbro. However, they lost Keller, Tkachuk, Bellows and Farabee after their freshman seasons so BU has a high level coming in but they also have a high level going out. Retention is their biggest hurdle.
- Top 6 in: Recruiting Class Ranking (5th), Total Points (5th), Average Star Rating (6th)
- 11th ranked team in youth with average age of 21.33 years old
- Perfect Retention Rate from the 2016-2018 Recruiting Classes
- Have landed the #1 rated age out junior player the past two years: Jack Dugan (’18), Scott Conway (’17).
- #7 Pairwise Ranking Average from 2016-2019
Comments: Providence has built a balance roster with both veteran junior players and younger talent. They have landed the last two #1 ranked age-out junior players who are both top 5 scorers on the team as freshman and sophomore. They haven’t lost anyone early off the 2016-2018 recruiting classes and have maintained consistent recruiting class rankings going from #6 to #18 to #5.
- Ranks 7th in Star Rating Average, 15th in Recruiting Class Rankings, 17th in Total Points
- 7th Oldest Team with an average age of 22.08 years old; 9th heaviest team in college hockey
- Perfect Retention Rate from the 2016-2018 Recruiting Classes
- Landed 3 of the Top 10 ‘97s in 2017 Class
- #6 Pairwise Ranking Average from 2016-2019
Comments: Ohio State is an interesting team because they rank in the top 20 in all recruiting categories but only top 10 in one (Star Rating Average). They are much older than both BU and Providence with an average age just over 22 years old and they’ve done a good job recruiting age out junior players and retaining the talent they bring in.
- Ranked 21st Recruiting Class Rankings, 19th in Points, 25th Star Rating Average; 28th Overall
- 12th Oldest Team with an average age of 21.92; 9th lightest team in college hockey
- 8th Worst Retention Rate losing about 100 pts per year
- Landed #9 Ranked Recruiting Class in 2017
- #31 Pairwise Ranking Average (2016-2019)
Comments: UVM is an older team, a bit light and have an overall recruiting class ranking of 28th which is close to their pairwise average ranking of 31. They tend to recruit older, smaller players and were able to pull off a big class in 2017 which was a Top 10 but their 2016 class was 15th and their 2018 class was 43rd. What really hurt this team was losing Ross Colton, 4.75 star prospect, after his sophomore year.
Photo Credit: Dan Hickling/Hickling Images