zeke's Official Top-20 Center Rankings

[zeke]

Production is not linear with ice time

do you know this?

Primary stats are much more important than the liklihood of producing primary stats (corsi) - they definitely shouldn't have equal weighting.

And how do we decide what the "primary", i.e the most relevant, stats are?

 

[zeke]

That is demonstrably inaccurate.

please demonstrate.

do you think, say, that mcdavid's and scheifele's relatively poor PP production means they lack a certain type of offensive skill?

 

[Midnight Judges]

please demonstrate.

do you think, say, that mcdavid's and scheifele's relatively poor PP production means they lack a certain type of offensive skill?

Yes, absolutely.

Firstly PP stats are generally going to favor players with a great shot (Ovechkin, Malkin, Stamkos) - which McDavid clearly does not have. McDavid generates shots and assists from skating and movement which brings me to #2: McDavid doesn't have Nick Backstrom's quick and sneaky hands. Backstrom is better than McDavid at generating assists with little or no movement. He does this with head fakes, amazing hands, and just lulling guys to sleep and then bam the puck is on Ovie's blade having gone through 2 defenders.

 

[Volica]

I love that you brought this up, to be honest, because it's a hilarious example in misremembered history - even recent history. The truth is, the last 100 years has seen oldskool hockey people MOCK anyone who thought points told you the majority of what you needed to know about a player.

Last 100 years:

Statsguy: "This guy has way more points he's clearly a better player"

Oldskool: "LOL you stupid nerd, you can't judge a player by points - it doesn't tell you anything about defense or clutch or grit or intangibles or compeition or role or usage or all that important stuff. watch the games"

Now:

Statsguy: "Wow look we have all these numbers which tell us so much about how a player impacts all-around play on the ice, offensve and defesnive, and can adjust for the role they're used in,and all sorts of great other stuff."

Oldskool: "LOL you stupid nerd, the only thing that matters is POINTS! watch the games"

it's just too rich.

The issue is a mix, honestly, and you're not being completely honest about it.

Essentially what the argument you're making is:

Then:
Nerds: This guy has a bunch of points, he's an awesome player!
Old Timer: There's a lot more to being an awesome player than just points.

Now:

Nerds: This guy's expected goals forward, in addition to his WAR and shot differential are through the roof!!! This guy's an awesome player!
Old Timer: Yeah, but the guy doesn't score at an elite level, and his on ice play show visible holes.
Nerds: What do you know!?!?! How could you calculate someone's on ice play when clearly this player I've seen play twice when he plays my team has stats that are incredible!
Old Timer: I watch him nightly?

 

[Varcus]

It honestly doesn’t matter what formula you or anyone uses even points. No list is perfect and never will be.

The nice thing about advanced stats it shows you things the eye test can miss. Does that mean it’s perfect no. But it also doesn’t mean it’s completely wrong either.

I also find it funny that people are upset about Kadri. I’d think a centre who can score 30 goals and plays a shut down role. Might just be a little underrated. I personally would rank him in the 20 to 25 range based on the last two seasons. I honestly would be shocked if he can put up similar type numbers again if he does I’d have no issues of putting him in the top 15. But I don’t think it’s going to happen.

 

[Frank Drebin]

do you know this?



And how do we decide what the "primary", i.e the most relevant, stats are?

Goals and assists - the things that actually count on the scoreboard.

If you score more goals than your opponent, you will win every game.

Outplaying (corsi) may tilt the odds in your favor, but it is far from a guarantee. Plus the effects of corsi by an individual player are miniscule over the course of a season. The difference between a 55% corsi player and a 50% corsi player is something like 6 goals a season playing 25 minutes a game, all shots being equal.

 

[leafsfan2point0]

to be flat out honest, I literally chose a method I thought would work against the leafs centers, especially matthews....and it seems like that was a mistake. If I had valued pure offense more, malkin, matthews, and kuz would all rank better. it was an intentional anti-leaf bias that caused most of my headaches here, hilariously.

interestingly, though, kadri would still rank well even if I did that.

not to worry, i'm working on a new ranking that avoids the most blatant flaws in this one - i.e. I'm going to give each player full credit for his actual production and possession, and not just his rank.

to be sure, that'll still leave me with a subjective choice as to how to weight the different elements, but I think what I'll do is give a range of options to show what happens if I weight one more than the other, and let people choose for themselves what they think best reflects reality.

Great work! No matter what you do, there will be people crying about where the leafs are ranked. Especially Matthews

 

[Frank Drebin]

The issue is a mix, honestly, and you're not being completely honest about it.

Essentially what the argument you're making is:

Then:
Nerds: This guy has a bunch of points, he's an awesome player!
Old Timer: There's a lot more to being an awesome player than just points.

Now:

Nerds: This guy's expected goals forward, in addition to his WAR and shot differential are through the roof!!! This guy's an awesome player!
Old Timer: Yeah, but the guy doesn't score at an elite level, and his on ice play show visible holes.
Nerds: What do you know!?!?! How could you calculate someone's on ice play when clearly this player I've seen play twice when he plays my team has stats that are incredible!
Old Timer: I watch him nightly?

We've gone from completely relying on stats, to completely relying on a wider range of stats.

 

[zeke]

Yes, absolutely.

Firstly PP stats are generally going to favor players with a great shot (Ovechkin, Malkin, Stamkos) - which McDavid clearly does not have. McDavid generates shots and assists from skating and movement which brings me to #2: McDavid doesn't have Nick Backstrom's quick and sneaky hands. Backstrom is better than McDavid at generating assists with little or no movement. He does this with head fakes, amazing hands, and just lulling guys to sleep and then bam the puck is on Ovie's blade having gone through 2 defenders.

I like the theory. well, at least the 1st part. I won't dismiss it. But let's test it out a bit.

Here's the top-20 PP p/60 guys over the last 2yrs (min 200minutes):

1.Kucherov 7.63
2.Stamkos 7.51
3.Marner 7.49
4.Rielly 7.39
5.Hall 7.38
6.Wheeler 7.26
7.Malkin 7.11
8.Kessel 7.10
9.Hedman 6.89
10.Pastrnak 6.72
11.Backstrom 6.70
12.Palmieri 6.65
13.Marchand 6.64
14.Giroux 6.60
15.Nylander 6.59
16.Butcher 6.46
17.Seguin 6.45
18.Crosby 6.44
19.Kadri 6.33
20.Matthews 6.29


starts off sensibly - some great shooters at the top there...but then Marner? Rielly? Wheeler? etc.? not exactly snipeshows there.

more interesting seems to be the abundance of PP partners on this list - Kucherov/stamkos/Hedman, Marner/Rielly/Kadri, Nylander/Matthews, Malkin/Kessel/Crosby, Hall/Palmieri/Butcher - maybe that's a bigger influence than style?

as for McDavid not having Backstrom's hands or guile.....really?

 

[zeke]

The issue is a mix, honestly, and you're not being completely honest about it.

Essentially what the argument you're making is:

Then:
Nerds: This guy has a bunch of points, he's an awesome player!
Old Timer: There's a lot more to being an awesome player than just points.

Now:

Nerds: This guy's expected goals forward, in addition to his WAR and shot differential are through the roof!!! This guy's an awesome player!
Old Timer: Yeah, but the guy doesn't score at an elite level, and his on ice play show visible holes.
Nerds: What do you know!?!?! How could you calculate someone's on ice play when clearly this player I've seen play twice when he plays my team has stats that are incredible!
Old Timer: I watch him nightly?

I appreciate you phrasing it in the most favorable way possible - but make no mistake, pretty much all the objections in this thread have come down to me ranking a player out of whack with one single stat - his point totals.

This isn't about complaining about me not considering a player's all-around game - in fact it's precisely the opposite - people are complaining I'm putting too much weight on factors other than point totals.

But I do like how you give your old timer the credit for watching games, but not the nerd. That's convenient.

 

[zeke]

Goals and assists - the things that actually count on the scoreboard.

If you score more goals than your opponent, you will win every game.

Outplaying (corsi) may tilt the odds in your favor, but it is far from a guarantee. Plus the effects of corsi by an individual player are miniscule over the course of a season. The difference between a 55% corsi player and a 50% corsi player is something like 6 goals a season playing 25 minutes a game, all shots being equal.

why look at goals at all, then?

shouldn't we just look at wins and losses? that's all that actually matters, no?

 

[Frank Drebin]

people are complaining I'm putting too much weight on factors other than point totals.

Exactly, and i think I've given reasons why.

Did you ever calculate the effect of a positive corsi player over the course of an entire season? It's miniscule.

 

[Frank Drebin]

why look at goals at all, then?

shouldn't we just look at wins and losses? that's all that actually matters, no?

Goals and assists have a direct link to wins and losses. That can't be said about any other metric.

 

[zeke]

Goals and assists have a direct link to wins and losses. That can't be said about any other metric.

assists have no direct link to wins and losses.

only goals.

 

[Frank Drebin]

assists have no direct link to wins and losses.

only goals.

Assists directly lead to goals. Goals directly lead to wins.

 

[Sidney the Kidney]

Or maybe he really is just that good? Why is than not an option? Why have you disregarded that as an option?

Just to save you time, telling me kopitar should be higher because everyone else would say so isn't an argument that is going to fly with me; it is also logical fallacy, so any one convinced by that line of argumentation isn't worth convincing.

I'm not saying you are wrong in saying Kopi>staal, but, you aren't really providing any good arguments

Not providing good arguments for what? Kopitar just won the Selke and scored over 90 points as well. I don't need to have some mathematical formula tell me that he should be ranked ahead of someone who has never come close to matching those offensive totals, while not being better defensively, either.

As for coming up with a different methodology, like I said earlier in the thread, I don't know enough about these various metrics and calculations in order to do that. All I said was that *something* must be weighted wrong in these calculations to come up with J. Staal being better than Kopitar.

Blindly adhering to these numbers just because there's no inherent bias in them is just as bad as blindly ignoring all numbers available to you. Both can result in bad takes. I mean, logically, do you truly believe Jordan Staal is a better player than Kopitar and would easily take him on your team over the latter?

It's like seeing a doctor doing plastic surgery on a patient and instead mangling her face. I'm not a doctor, so I can't "do better" nor can I offer suggestions about what, technically speaking, he did wrong. But that doesn't mean I can't say "this doesn't look right..." when I see the end result. And that's what I'm saying about that metric that results in Staal being ranked ahead of Kopitar. I don't know what metric is weighted too highly, or how to fix it, but "it doesn't look right".

 

[zeke]

Here's a list.

Here are the top goals for percentage centers in hockey, all situations, over the last 2yrs (min 1000minutes), because only goals are 'real', and because goals against are every bit as real as goals for, and are the only things that directly impact wins and losses:

1.Malkin
2.Stamkos
3.Crosby
4.Backstrom
5.Bergeron
6.Wennberg
7.Dubois
8.Kuznetsov
9.Matthews
10.Thornton
11.Bozak
12.Turris
13.Scheifele
14.Zetterberg
15.McDavid
16.Barzal
17.Schenn
18.Barkov
19.Kadri
20.Johansen

good list? these are the centers who outscored the other team by the most while on the ice over the last 2yrs. The only thing that actually directly links to winning games.

(and lmao at Kadri making the top-20 again. trust me that wasn't planned.)

 

[zeke]

Assists directly lead to goals. Goals directly lead to wins.

you don't beat a team by out-assisting them. there is no direct link. you can win games with fewer assists.

 

[Frank Drebin]

you don't beat a team by out-assisting them. there is no direct link. you can win games with fewer assists.

You cannot register an assist without a goal.

 

[Midnight Judges]

...as for McDavid not having Backstrom's hands or guile.....really?

I think the more salient point is that McDavid generates points from speed and movement, and the PP mitigates that. Backstrom is one of the slowest C's in the league, but he kills it on the PP. Admittedly, a big part of that is having Ovechkin (this is reflected mostly in Backstrom's secondary PP assists because he passes to the D who passes to Ovie for the primary assist). But the other part is Backstrom is extremely deceptive. And no, I don't think McDavid matches Backstrom is this very limited context.

 

[CherryToke]

Proof that Toews is no worse than he was in the Hawks glory days, he just has much worse line mates now.

 

[Marshy]

Zeke I never would have believed Kadri was a top 20 C in all these types of lists using so many different statistical methodolgies....and points.

Thanks for convincing me!

 

[leaffaninvancouver]

Here's a list.

Here are the top goals for percentage centers in hockey, all situations, over the last 2yrs (min 1000minutes), because only goals are 'real', and because goals against are every bit as real as goals for, and are the only things that directly impact wins and losses:

1.Malkin
2.Stamkos
3.Crosby
4.Backstrom
5.Bergeron
6.Wennberg
7.Dubois
8.Kuznetsov
9.Matthews
10.Thornton
11.Bozak
12.Turris
13.Scheifele
14.Zetterberg
15.McDavid
16.Barzal
17.Schenn
18.Barkov
19.Kadri
20.Johansen

good list? these are the centers who outscored the other team by the most while on the ice over the last 2yrs. The only thing that actually directly links to winning games.

(and lmao at Kadri making the top-20 again. trust me that wasn't planned.)

I think Bozak being there is even funnier.

 

[Turin]

possibly.

then again he plays and has always played in a very cushy role, behind the best center of his generation, and is a negative possession player even in that sheltered role.

Malkin is not a negative possession player. If he is, the term possession no longer has any value.

 

[Marshy]

I think Bozak being there is even funnier.

Holy Christ we had 3 top 20 centers last year too?