Laxfanatic2022 wrote: ↑Wed Apr 26, 2023 9:00 pm
Ezra White wrote: ↑Wed Apr 26, 2023 5:46 pm
Laxfanatic2022 wrote: ↑Wed Apr 26, 2023 12:23 am
...
faircornell wrote: ↑Sun Apr 23, 2023 11:59 pm
This is one of those random pieces of analysis where I believe that my calculations are correct...
Both Cornell and Princeton have strong contenders for both Ivy POY and First Team All-Ivy goaltender. Some statistics going into next Saturday's game:
POY: CJ Kirst vs Coulter Mackesy:
Mackesy:
vs Ivy Goals: 24
vs Ivy Assists: 11
vs Ivy Points: 35
Kirst:
vs Ivy Goals: 27
vs Ivy Assists: 5
vs Ivy Points: 32
First Team Goalie: Chayse Ierlan vs Michael Gianforcaro
Gianforcaro:
vs Ivy saves: 76
vs Ivy Goals Allowed: 43
Save % vs Ivy: 64%
Ierlan:
vs Ivy saves: 65
vs Ivy Goals Allowed: 41
Save % vs Ivy: 61%
These numbers are vey close. I believe that the All-Ivy team is picked based on the regular season (I am happy to be corrected). This coming game will most likely tip the scales one way or another.
...
It’s a weird argument here because Mackesy has to go up against the best defenseman in the country in Gavin Adler. Let’s not forget that Adler would win IVY POY if Kirst wasn’t producing at such an insane level. I expect the matchup to be great but Mackesy finds goals off ball a ton while Kirst is way more of a dodging threat. Really looking forward to seeing the game Saturday. Final score prediction Cornell 15 Princeton 11
Here's a way of possibly estimating the "
Adler Factor."
- Identify each Ivy team's top scorer
- Look up their goals, assists, and total points. Do this for both "all games" and "Ivy games."
- If they've played Cornell already, look up their goals, assists, and total points for the Cornell game.
- Adjust both their "all games" and "Ivy games" values by subtracting the Cornell values for goals, assists, and total points. These values will be the totals scored in games against teams other than Cornell.
- Divide these totals by the number of non-Cornell games played. These will then be each top-scorer's average goals, assists, and points scored per game against teams other than Cornell. Call these the "adjusted averages."
- Divide the corresponding values for goals, assists, and points for the Cornell game by the averages computed in the previous step.
- While you're at it, total the number of goals, assists, and total points by the top scorers in their games against Cornell. Divide these totals by the number of games Cornell has played against other Ivies. For the record, as of today Cornell has played 5 games against other Ivies and had 11, 6, and 17 goals, assists, and points scored by the other team's top scorers. The top scorers have averaged 2.2 goals, 1.2 assists, and 3.4 points against Cornell.
- Divide each top scorer's goals, assists, and total points in their game against Cornell by the corresponding values for their adjusted averages. This gives index numbers for each top scorer. Each index number represents the percent of their adjusted season averages the top scorer achieved in their game against Cornell.
- For example, Penn's Sam Handley scored 3 goals and 1 assist against Cornell but averaged 1.9 goals, 2.0 assists, or 3.9 total points against all other teams; he also averaged 2.0 goals, 2.6 assists, and 4.6 total points against other Ivy teams. So his index numbers for goals, assists, and points are 158%, 50%, and 102.6% compared to his performance against all other teams and compared to his performance against just other Ivy teams, 150%, 385, and 87% against Cornell.
- Overall, when they played against Cornell, the top scorers on other Ivy teams scored 59.2% of the average number of goals they scored against other Ivy teams, 64.9% of the number of assists, and 55.2% of the total number of points. The corresponding averages are 83.8%, 72.4%, and 79.8% for the top scorers' performance against Cornell relative to their performance in all other games.
- Coulter Mackesy hasn't played Cornell yet, so we can't compute his index number. But we can repeat this exercise, removing Princeton from the totals the same way we removed Cornell. Then we compare his performance against the other Ivies versus the other top scorers'. This is left as an exercise for the reader.
Calling this an "Adler Factor" implies quite a few assumptions:
- Adler always guards the other team's top scorer. Besides slides, picks, etc., sometimes this is impossible. For example, Yale's Matt Brandau and Chris Lyons are tied with 52 total points during the season. Adler probably tried, but he couldn't guard them both simultaneously. I do believe he gave Brandau most of his attention, and Brandau scored only a goal and an assist against Cornell. Meanwhile, Lyons scored 3 goals. Brandau was held to 44% of his Ivy-only scoring, while Lyons achieved 63%. If Adler was primarily responsible for Brandau's subpar performance, and Adler achieves the same thing with Mackesy, we'd expect Mackesy to score about 4 points, most likely with 3 goals and 1 assist.
- Characteristics of opposing players and their teams' styles of play might cause Adler to be assigned to mark someone other than the other team's "top scorer."
- Game situations and coaching decisions affect these statistics. For example, in a 10-point blowout, coaches might decide to eat up the clock or clear the bench, in which case perhaps only data for the 1st 3 quarters should be used.
- The team might play zone or use a 5 + 1 defensive formation, in which case Adler would not necessarily be covering the other team's top scorer.
- A team's top scorer might change as the season progresses, due to injuries, younger players becoming acclimated to the college game, etc. A coach studying films from the first two games of the season might assign Adler to player X and then, at halftime, change the assignment to player Y.
This is an absurd analysis that doesn’t tell the whole story. Stats don’t tell the story as guys score and assist when they get matchup changes, emo, etc. Just watch the tape Adler has manhandled every opponents #1 and any points that come 99% of the time are when Adler is sliding or isn’t on the guy. He’s only responsible for giving up 1 point total this year to the guy he’s covered in the Ivy League which is INSANE just watch the tapes. Teams are actually game planning around him to get matchup changes now. I’m sure Princeton will have a scheme to get Mackesy free from Adler. Looking forward to a dog fight
Of course it doesn't tell the whole story. Statistics, and, more generally, quantitative analyses, never do.
Unfortunately, none of the ESPN+ videos to which I have access show 100% of Adler's coverage in any given game. Even if they did, there would be other things outside the camera's view that influence Adler's play. We'd still have to judge how much of a result is due to Adler.
Remember SOT? The coaches decided to have JT51 go the the sideline, along with his marking long-stick opponent, and then have the rest of the attack play 5 v 5. Things would have been very different if Jeff had the green light to beat his marker, although either strategy might have resulted in Cornell winning.
I was just trying to put a number on what we might expect from Mackesy.