Including outliers is also fundamentally flawed, as is using data from seasons that don’t reflect how good players are today. Both of which you did in your post. None of the other players have a goal season outlier like Hyman in the data. I was showing how much of an effect the outlier has. And even if we use Hyman’s numbers from 22-23 to today, he still falls behind the top 8 in points per 82 and the rest of the post applies.
Who was the biggest omission from the Four Nations Cup rosters?
- Thread starter Took a pill in Sbisa
- Start date
Including outliers is also fundamentally flawed, as is using data from seasons that don’t reflect how good players are today. Both of which you did in your post. None of the other players have a goal season outlier like Hyman in the data. I was showing how much of an effect the outlier has. And even if we use Hyman’s numbers from 22-23 to today, he still falls behind the top 8 in points per 82 and the rest of the post applies.
No, deciding based on your own bias what seasons to include for what players is flawed. You have chosen to remove a single players best season, and to compare that against the complete seasons of all other players. Especially when for the player you are "picking and choosing" you place weight on the last 26 games, ignore the last 27-106 games, and then include the last 106 to 185 games. Also no mention of postseason play from you either. That is a hugely flawed methodology.
It’s not a methodology but ok. I’m not going to bother with you if you can’t see how removing an outlier might give some insight. It doesn’t mean I’m using it as the only deciding factor
Using your methodology with Sam Reinhart is a 39-45-84 player. Guess he's borderline for Canada too.
Remove outliers sure, but do it for all in the sample. Otherwise you are just introducing a bias.
Last edited:
Beerz
Registered User
- Jun 28, 2011
- 37,132
- 13,548
Tage is a choir boyReinhart was absolutely borderline going into the year because his season was a huge outlier (and this is my bad because I forgot to mention Reinhart’s outlier). Fewer people would have had him on the team going into last year than Hyman, and if he was considered a consistent 50 goal player, he would have been selected initially instead of Marchand. The fact that he’s continued his production is a big reason he was taken.
Hyman’s slow start hurt his case, and if he was continuing his production from last year like Reinhart has done he almost certainly would have been taken. So in some ways I agree with you that Hyman lost his spot due to this season. Where I disagree was that Hyman was an obvious pick that only lost his spot due to a bad month. This isn’t like if Marner had a slow start after being a 100 point pace player for years. Hyman was a borderline player who needed to show his year wasn’t an outlier to be a lock and he didn’t. He still would have been a fine pick, but it’s not egregious if he’s only going to be the player he’s been outside of last year.
The difference with Reinhart as well is that he’s also an elite defensive player, and he’s been even better than Hyman at Hyman’s PP role, which reduces one of Hyman’s reasons for being there. If Reinhart wasn’t the obvious net front on PP1, Hyman probably would have been taken in part to play that role.
I can see your point from these two paragraphs. I disagree strongly with it but it at least makes sense.
What I take issue with is deciding which seasons to compare between a player and a sample of players:
To compare 1 players current 26 game season, and their seasons from 3&4 years ago to a sample of players last 3 seasons in just not apples to apples.
And again the team decided that its best player doesn't need his most regular linemate for a short tournament due to a slow month. Its just a dumb decision.
My guess is that by February team Canada will realize they have made a mistake, and find a way to have Hyman on the team.
Beerz
Registered User
- Jun 28, 2011
- 37,132
- 13,548
I agree it’s not apples to apples but I don’t believe determining the best players today can ever truly be an apples to apples comparison because we need to find some balance between variance in smaller samples and players actually improving or regressing, as well as both positive and negative outlier years. I didn’t intend the numbers to be the only factor to consider, but I was trying to show how the extended timeline you used for Hyman in Edmonton and missed games can make the numbers look better for Hyman relative to other players, particularly ones like Konecny and Hagel, who made jumps in 22-23, and Stone who has missed time. And even in my data, Jarvis gets underrated because he made the jump last year and has been a 30+ goal scorer since then.
I can see your point that Hyman should have probably been taken for chemistry in a short tournament, even if it doesn’t end up working out, but it’s possible they also load the top lines and that chemistry becomes moot. I also imagine he picks it up again and he’ll probably end up with 35-40 goals on the year. I just don’t think he’s so much better than the other names taken that he’s a huge miss here.
I'd have Bedard on team Canada. I think the kid still believes he can be the best in the world and I don't know how serious the veterans will take this tournament. I know he's not the most complete player but he's better offensively than some guys and he did well for Canada in the past.
