I don't mean you don't look at results or that it's not an effective way of comparing two players but what I do mean is that two players with the same production aren't going to be identical in terms of offensive skill. I know Jfresh has microstat breakdown of different metrics but I'm guessing you or others (including myself) aren't comparing each metric individually and assigning values to each metric (i.e, making our own models). Determining who is the better puckhandler or skater isn't as simple as looking at a single metric.

I mean the best way to compare overall impact is to see our their offense and defense has helped their team overall. Off and Def GAR and WAR displays the approximate quantitative impact a player has had on their team. Also puck handling and skating is difficult to measure, and doesn’t really matter. On ice production is what matters most. Using those skills to convert them into quantitative results. Those can be measured

The models do provide a way to compare offensive defensive, and overall impact and they try to standardize things but it's important to recognize they don't capture everything and there does have to be a determination about what's more important. If it's production, does that mean the 4% EV OFF Hughes has on Pettersson is more significant than the 6% EV DEF that Pettersson has on Hughes? Idk what Jfresh has to say about that since I don't know how he calculates the overall WAR (which takes into account other factors as well). How much does Pettersson's edge in finishing boost his impact.

I would say the best players typically have the most impact but when looking at players who do have similar high impact, I do think it's important to look at the qualities (the skills) that enable them to get those results. Skating and puck handling are just examples but I don’t think just because measuring these qualities are difficult that they don't matter. They'll contribute directly to things like zone entries and exit, and offensive production as well as defense.

There needs to be more data imo to even effectively use these models to determine which of these two different players (different play styles and utilization) who had similar production last year is better. Saying Pettersson is better because of better defense seems like a lot is being glossed over. That to me just makes him the more complete player.

What are your thoughts on point shares? I'm not too familiar with what they are looking at that but it seems that if production is the main criteria to say X is better than y, defense relevant in terms of point share. I do recognize that you weren't saying only production, and you're looking at overall impact on both sides of the ice. So, basically to you the best of two players with similar production is the more complete rather than the more skilled (not saying Hughes is more skilled) since you're not too concerned about them. I get that you can't really measure a lot of skills which is why the eye test matters (which has it's own issues) but really if they aren't on the same team the best player is the more complete player with the production and not the more skilled (not saying Hughes is more skilled).

Now we may settle this, gentlemen.

@GMBL @Headhighsauce @littlejerryseinfeld

I feel as if this card confirms some things but overall still leaves more questions than answers. Hughes has better 5v5 offense, Petterson is better on the PP and with 5v5 defense. I think of Pettersson as the better shooter and Hughes the better playmaker, but last season it's been quite the opposite. Hughes is better at things that speed, skating, and puck handling come in handy, not sure why but interesting to see. What I can say for sure is that I'm looking forward to see what they each do next year.

I think it confirms what I had been saying, that the difference between them is very small, so Hughes gets the edge due to age and contract

Sure, if the question is who would you rather have? or who is more valuable?. If we're talking about the better player though, contract is irrelevant, and maybe someone gives the edge to Hughes in the future due to age but Pettersson now like the OP did.

> If it's production, does that mean the 4% EV OFF Hughes has on Pettersson is more significant than the 6% EV DEF that Pettersson has on Hughes

It almost certainly is. per EH, the 100th percentile for EVD is 9.1 GAR. For EVO it's 25.6. Offense is much more valuable than defense (which, duh. you can't win if you don't score goals) in these models.

FWIW EH has Hughes at a total 22/23 WAR of 3.6 and Petterson at 4.4 with the biggest difference being defense and penalties. xWAR is 3.7 to 3.7.

different models will value different things differently though.

Certainly as far as individuals go, offense is generally speaking than defense, but yeah I was talking more about the models which as you said do reflect that but there is the question of which metrics are being valued more over others in each model (as you mentioned).

The models do have a larger emphasis on 5v5 offense which is a good standard for comparing players, to try and isolate the players from the situational impacts. With these two who had the same overall production essentially, the way I see it (until there's more data to separate them) they essentially got to the same place while excelling at different things.

Advanced analytics will only tell so much of the story. Hughes is noticeable on the ice 100% of the time. He'll be a top 5 player in the league in maybe 2 years, if not in this upcoming season.
And I'm a Rangers fan.

Pettersson is the better player, Hughes easily is more valuable on his contract, unless Pettersson loses his mind and signs for less than 9 which would never happen

imo the absolute most definitive statement you can say on this topic is that Pettersson and Hughes are statistically equal, certainly within the error bars. I don’t think it’s really possible to say which is the better player right now as they’re on as equal footing as there can be in the NHL.

Their offense is pretty equal with Hughes having a SLIGHT edge, but Pettersson’s defense is still significantly better

And yet, still basically between the error bars, mostly due to penalty drawing and powerplay.

Jacks defense metrics also blew Pettersson away last season and the season before lol

Hughes had a better EVD in both 20/21 and 21/22. Pettersson’s outlier season this year has him ahead on 3 year EVD WAR, jacks outlier has him ahead on 3 year xEVD

Just to further expand on this, if you look at Evolving-Hockey's 3 year playercard:

Hughes: offense 92, defense 59, overall 95
Pettersson: offense 90, defense 63, overall 95

This argument seems very silly as we're having a conversation about which of these two players who have roughly equal advanced and counting stats, is better. Theyre just about as on-par as you can get.

Pettersson was playing injured for the first half of last year

Jack Hughes played 49 games. So it evens out just fine

That’s, not how it works…

That actually is how a cumulative stat like WAR works.

Petteys primary linemate shot 27% and gave pettersson 23 assists
Hughes primary linemate shot 5% and gave hughe 1 assist

Both Haula and Kuzmenko had the same amount of xgoals, Kuzmenko simply did way more with them, inflating petterssons stats.

Not even considering the garbage goaltending in the Pacific

We've been over this a thousand times but that offensive difference imo is much bigger than you think, and I am nearly certain it will show this year

The difference in war is literally because of the shooting differences by teammates.

Kuzmenko, Horvat, Miller, were ELITE finishers over expectation and made up nearly 2/3 of Petterssons assists.
Haula was massively below expectation, Bratt was above, Dougie above, Hischier way below, and the rest hovering around

Offensive GAR in this scenario is skewed by ridiculous shooting rates. Again, these models are not be all end all, you need to understand the context behind them. In general these metrics are accurate, but a stat where a player gets 20 some assists to a guy shooting 27% will be skewed and make that player look better than he is

How can the offense be so much bigger when the evidence simply does not support that?. Take away 15 assists to Kuzmenko and Pettersson is still a 90 point player with elite defense. That doesn’t seem like a big gap in offense

Make all the excuses you want. Pettersson didn’t put up 102 points with elite defense by luck