Bo 'Saviour, Hero, God' Horvat Discussion - Part V (MOD WARNING POST 537)

[CanaFan]

QoC isn't really comparable across teams, so you can't just do a league-wide ranking. If I'm reading the tables right at stats.hockeyanalysis.com, Couturier faces the toughest competition on the Flyers, as well as receiving the most defensive zone starts, and the least offensive zone starts on the team. He is absolutely getting the toughest minutes out of any forward on the Flyers. On the Canucks, Bo is... not. I don't think there's any one recipe to guarantee the development of a prospect, but for Bo, I'd lean toward not throwing him to the wolves next year, and continuing to let him develop an offensive game.

But what defines "toughest" minutes from an offensive standpoint? Is it Couturier facing the other team's top offensive forwards or is it Giroux facing the other team's top defensive forwards and/or D pair? I understand that possession time is the standard metric to gauge QoC, but I'm not sure that is a wholly intuitive approach.

And forgive my ignorance, but why can't QoC be compared across teams? In my albeit limited understanding if Couturier is truly playing "hard minutes" (high Corsi opposition) then isn't his relative score comparable to the minutes that Horvat plays? I mean they play in the same league and against the same players, albeit in different amounts. It isn't like the Corsi scores for the opposition are derived only from games against the Flyers or the Canucks, it is against all teams in the league. Thus the proportion of time against "tough" (high Corsi) opposition should be comparable across teams, no?

Just curious tbh.

 

[Verviticus]

what's wrong with replacement level? its a term that communicates a specific thing thats a very useful point of reference

 

[rho]

But what defines "toughest" minutes from an offensive standpoint? Is it Couturier facing the other team's top offensive forwards or is it Giroux facing the other team's top defensive forwards and/or D pair? I understand that possession time is the standard metric to gauge QoC, but I'm not sure that is a wholly intuitive approach.

Well, in this case I'm defining toughest as greatest QoCRel, most d-zone starts, and least o-zone starts. You can argue about which of these should be weighed the most, but in Couturier's case, the argument is moot, because he ranks 1st in all three categories on the Flyers (the evidence suggests, though, that the zone start effect tends to drown out the qualcomp effect).

Edited to add: I misread your question. I think it makes sense to change the definition of what constitutes tough offensive minutes vs. tough defensive minutes, at least in terms of quality of competition. Either way, though, I'd say that getting more d-zone starts and less o-zone starts makes it tough to play both offence and defence.

And forgive my ignorance, but why can't QoC be compared across teams? In my albeit limited understanding if Couturier is truly playing "hard minutes" (high Corsi opposition) then isn't his relative score comparable to the minutes that Horvat plays? I mean they play in the same league and against the same players, albeit in different amounts. It isn't like the Corsi scores for the opposition are derived only from games against the Flyers or the Canucks, it is against all teams in the league. Thus the proportion of time against "tough" (high Corsi) opposition should be comparable across teams, no?

Just curious tbh.

Maybe that wasn't the right wording. Does an absolute ranking via QoC make logical sense? Maybe. The teams that face the Kings and Blues of the league most often will have an upper hand so-to-speak in accumulating the "best" QoC. But, maybe you say that that's only fair -- you face the best, it should be reflected in your QoC stat. The thing is, though, that while that might have an effect on your raw Corsi, it doesn't really affect your CorsiRel, which is typically what we use to measure a player's strength (possession-wise anyway). If you want to move to raw Corsi as your metric, then maybe raw QoC makes sense. Of course that gives you all sorts of weird results, like Henrik Sedin being a worse possession player than over half the Kings roster.

I'll just add that historically, quality of competition metrics were calculated to figure out who was facing 1st liners vs. 2nd liners vs. 3rd liners etc. For that, you definitely can't use QoC and have to use QoCRel.

 

[BlackAces]

what's wrong with replacement level? its a term that communicates a specific thing thats a very useful point of reference

Because around these parts it's becoming a buzzword. I also think each persons definition of the term varies, which leads to arguing about semantics.

 

[Alflives*]

Well, in this case I'm defining toughest as greatest QoCRel, most d-zone starts, and least o-zone starts. You can argue about which of these should be weighed the most, but in Couturier's case, the argument is moot, because he ranks 1st in all three categories on the Flyers (the evidence suggests, though, that the zone start effect tends to drown out the qualcomp effect).

Edited to add: I misread your question. I think it makes sense to change the definition of what constitutes tough offensive minutes vs. tough defensive minutes, at least in terms of quality of competition. Either way, though, I'd say that getting more d-zone starts and less o-zone starts makes it tough to play both offence and defence.



Maybe that wasn't the right wording. Does an absolute ranking via QoC make logical sense? Maybe. The teams that face the Kings and Blues of the league most often will have an upper hand so-to-speak in accumulating the "best" QoC. But, maybe you say that that's only fair -- you face the best, it should be reflected in your QoC stat. The thing is, though, that while that might have an effect on your raw Corsi, it doesn't really affect your CorsiRel, which is typically what we use to measure a player's strength (possession-wise anyway). If you want to move to raw Corsi as your metric, then maybe raw QoC makes sense. Of course that gives you all sorts of weird results, like Henrik Sedin being a worse possession player than over half the Kings roster.

I'll just add that historically, quality of competition metrics were calculated to figure out who was facing 1st liners vs. 2nd liners vs. 3rd liners etc. For that, you definitely can't use QoC and have to use QoCRel.

You could try just watching the guys play too.

 

[Verviticus]

Because around these parts it's becoming a buzzword. I also think each persons definition of the term varies, which leads to arguing about semantics.

so question the improper uses. the definition of the term can't vary

 

[BlackAces]

You could try just watching the guys play too.

If I'm watching multiple players on multiple teams so I can win a QoC debate, I better be getting a paycheck for it

 

[TheWanderer]

But what defines "toughest" minutes from an offensive standpoint? Is it Couturier facing the other team's top offensive forwards or is it Giroux facing the other team's top defensive forwards and/or D pair? I understand that possession time is the standard metric to gauge QoC, but I'm not sure that is a wholly intuitive approach.

And forgive my ignorance, but why can't QoC be compared across teams? In my albeit limited understanding if Couturier is truly playing "hard minutes" (high Corsi opposition) then isn't his relative score comparable to the minutes that Horvat plays? I mean they play in the same league and against the same players, albeit in different amounts. It isn't like the Corsi scores for the opposition are derived only from games against the Flyers or the Canucks, it is against all teams in the league. Thus the proportion of time against "tough" (high Corsi) opposition should be comparable across teams, no?

Just curious tbh.

You forgot about the part where people can say "not comparable" if saying so supports their own narrative and nobody else's.

 

[Alflives*]

If I'm watching multiple players on multiple teams so I can win a QoC debate, I better be getting a paycheck for it

You young guns have it way over me with the stats stuff. I simply watch, and think that guy is clearly better than that other guy. Like Bo. He's clearly better as a number 2 Center than Bonino. Bonino can't skate. That's a visual thing. Bo almost always has his big a@s on the defensive side of the puck. That's a visual thing. I understand the value of these new 'metrics', but think the 'eye' test is important too. I do accept it's fun for you guys to argue about this stuff though. Ultimately, Bo's a#se is bigger than Bonino's. That's not arguable.

 

[CanaFan]

You could try just watching the guys play too.

Well ya but I honestly don't watch many games outside of the Canucks (kids do that) so I don't have much chance to catch Couturier play, let alone enough to determine how "tough" his minutes are. In this case QoC stats are better than nothing.

 

[Alflives*]

Well ya but I honestly don't watch many games outside of the Canucks (kids do that) so I don't have much chance to catch Couturier play, let alone enough to determine how "tough" his minutes are. In this case QoC stats are better than nothing.

Bo is better than Couturier. Bo skates better, and has far more skill - at speed. Plus, Bo's hands are far softer too. Bo has the heavier rear end, which makes him more solid on his skates. I see Couturier as a 'lumberer'. Bonino is even better than him.

 

[CanaFan]

Maybe that wasn't the right wording. Does an absolute ranking via QoC make logical sense? Maybe. The teams that face the Kings and Blues of the league most often will have an upper hand so-to-speak in accumulating the "best" QoC. But, maybe you say that that's only fair -- you face the best, it should be reflected in your QoC stat. The thing is, though, that while that might have an effect on your raw Corsi, it doesn't really affect your CorsiRel, which is typically what we use to measure a player's strength (possession-wise anyway). If you want to move to raw Corsi as your metric, then maybe raw QoC makes sense. Of course that gives you all sorts of weird results, like Henrik Sedin being a worse possession player than over half the Kings roster.

I'll just add that historically, quality of competition metrics were calculated to figure out who was facing 1st liners vs. 2nd liners vs. 3rd liners etc. For that, you definitely can't use QoC and have to use QoCRel.

Thx for the explanation, tbh I'm not exactly sure how BTN calculates its QoC estimates, as there is also a Corsi rel QoC (which I understand is a measure of possession adjusted for QoC) so I assume the simple QoC is the most appropriate but clearly I may have to read up on the subject further.

 

[CanaFan]

Bo is better than Couturier. Bo skates better, and has far more skill - at speed. Plus, Bo's hands are far softer too. Bo has the heavier rear end, which makes him more solid on his skates. I see Couturier as a 'lumberer'. Bonino is even better than him.

Ya, whenever I do catch the Flyers (not often) I'm never overly impressed with Couturier. Doesn't show anything special in the offensive zone but does seem to be working his butt off. Hard to believe he scored 96 points in his pre-draft year but then that just shows the limits of stat scouting I guess. I do like how Bo "looks" offensively so I am also confident he can progress further than SC in that regard.

 

[bossram]

Agreed. I think people aren't used to seeing rookies have a lot of success, so its easy to hype him up.

At the end of the day, he is currently ~replacement level (which is very good for a 19 year old), but lets not act like hes become a super star.

I think Sean Couturier is probably a good comparable in terms of what I am expecting offense wise. I don't think Bo is going to crack .5 ppg for a couple seasons

Agreed on almost all counts. I think his two-way game will develop earlier than consistent offense. I do think he's better than replacement level now though. He has got better as the season progressed. I think he's legit NHLer. Which at 19 is very good.

It's laughable though that some people think he would easily hit 30G/50P with better linemates/PP time. The production is almost entirely driven by inflated %s.

What has me optimistic is Bo's SAT differentials are vastly improved. If he can even be a break-even player - at 19 with his two-way reputation - that bodes well for the future.

 

[CanaFan]

It's laughable though that some people think he would easily hit 30G/50P with better linemates/PP time. The production is almost entirely driven by inflated %s.

Haven't seen anyone say Bo could "easily hit 30G" but you don't think 50 pts is potentially attainable with better linemates and PP time? His personal sh% has nothing to do with assists which are definitely lagging right now. And while I haven't looked up Hansen and Dorsett's sh% but I'd guess they are on the low side. Kenins is of course high but what has he actually factored in for Bo, like 2-3 assists? How much better would he be with someone like Vrbata on the receiving end of his passes instead of Hansen?

He could see his sh% regress to a more sustainable 15-17% and still reach 20 goals based on his style of play. But his ability to hit 50pts (~30A) is really only possible with better wingers and some PP time.

*edit: Kenins at 14%, Hansen at 12%, and Dorsett at (ugh) 6%. Doesn't seem his assists are overly % driven.

 

[AirMinivan]

Bo is better than Couturier. Bo skates better, and has far more skill - at speed. Plus, Bo's hands are far softer too. Bo has the heavier rear end, which makes him more solid on his skates. I see Couturier as a 'lumberer'. Bonino is even better than him.

Not to sound snarky, but this is a perfect example of why the "eye test" is horribly subjective.

The "eye test" is usually influenced by whim and personal bias, such being the case in your post. Also, a lot of people really don't have a keen eye for what is actually going on in a game, and tend to focus on huge plays/mistakes while ignoring the little things. This is something most viewers are guilty of, including myself. The "eye test" from the perspective of Mike Babcock is invaluable and irreplaceable. The "eye test" from some spaz on HF is rarely worth a nickel. That being said, there ARE some very informed and intelligent posters on here too.

Also, unless you have a very keen eye, the eye test also doesn't account for other metrics that you don't notice like who these players are playing against, and coaching decisions that affect the chances of a player to have an impact.

As a result, statistics are generally a better means of making arguments between a group of people over which we have no quality control.

When someone whose name I don't recognize says "Oh player X is better than player Y because I watch them play," I always take what they say with a grain of salt. I don't know how qualified they are to be making an objective and accurate analysis, so I will just assume they aren't.

Advanced stats are the most factual and measurable thing we have to work with here. I for one greatly appreciate and enjoy their use.

 

[bossram]

Haven't seen anyone say Bo could "easily hit 30G" but you don't think 50 pts is potentially attainable with better linemates and PP time? His personal sh% has nothing to do with assists which are definitely lagging right now. And while I haven't looked up Hansen and Dorsett's sh% but I'd guess they are on the low side. Kenins is of course high but what has he actually factored in for Bo, like 2-3 assists? How much better would he be with someone like Vrbata on the receiving end of his passes instead of Hansen?

He could see his sh% regress to a more sustainable 15-17% and still reach 20 goals based on his style of play. But his ability to hit 50pts (~30A) is really only possible with better wingers and some PP time.

*edit: Kenins at 14%, Hansen at 12%, and Dorsett at (ugh) 6%. Doesn't seem his assists are overly % driven.

Bo's on-ice SH% is still at like 13%. That's not really sustainable for anyone either. It probably comes down to about 8-9%.

Dorsett sinks any line he was on. And while the Kenins - Horvat - Hansen line also had inflated %s, they were also legitimately good. Controlling ~54% of SATs while together (albeit in a limited sample).

A 15-17 SH% still isn't sustainable for anyone really. Over the last 3 seasons at all situations with players >2000 minutes played (for solid sample) the highest personal SH% is Palat at 13%. Stammer is at 12.3%. I don't think Bo will be as prolific a shooter as Stamkos in his career.

Rather than talking about production which is unlikely to continue (for now), what is really encouraging about Horvat are his SAT differentials. He started the year at basically replacement level, but his underlying numbers have also turned a corner since the all-star break to where he is now regularly >50%. For a 19 y.o. to even break even is quite impressive and bodes well for his future value as a two-way player.

 

[Verviticus]

Not to sound snarky, but this is a perfect example of why the "eye test" is horribly subjective.

Advanced stats are the most factual and measurable thing we have to work with here. I for one greatly appreciate and enjoy their use.

yeah

http://nhlnumbers.com/2015/3/1/the-limits-of-observation

 

[Alflives*]

Not to sound snarky, but this is a perfect example of why the "eye test" is horribly subjective.

The "eye test" is usually influenced by whim and personal bias, such being the case in your post. Also, a lot of people really don't have a keen eye for what is actually going on in a game, and tend to focus on huge plays/mistakes while ignoring the little things. This is something most viewers are guilty of, including myself. The "eye test" from the perspective of Mike Babcock is invaluable and irreplaceable. The "eye test" from some spaz on HF is rarely worth a nickel. That being said, there ARE some very informed and intelligent posters on here too.

Also, unless you have a very keen eye, the eye test also doesn't account for other metrics that you don't notice like who these players are playing against, and coaching decisions that affect the chances of a player to have an impact.

As a result, statistics are generally a better means of making arguments between a group of people over which we have no quality control.

When someone whose name I don't recognize says "Oh player X is better than player Y because I watch them play," I always take what they say with a grain of salt. I don't know how qualified they are to be making an objective and accurate analysis, so I will just assume they aren't.

Advanced stats are the most factual and measurable thing we have to work with here. I for one greatly appreciate and enjoy their use.

I'm happy for you. If those sorts of things make the game more enjoyable for you, that's fine by me. I, however, believe watching the players is more enjoyable than staring at numbers. I wonder if Babcock reads those numbers, or he has someone (maybe someone like yourself?) updating him on those things?

 

[Verviticus]

edit: misread, thought you said valuable instead of "enjoyable"

 

[Wilch]

I'm happy for you. If those sorts of things make the game more enjoyable for you, that's fine by me. I, however, believe watching the players is more enjoyable than staring at numbers. I wonder if Babcock reads those numbers, or he has someone (maybe someone like yourself?) updating him on those things?

He's not saying advanced metrics make the game more enjoyable, he's saying it provides a standardized and objective basis to argue certain points.

The eye test is your word against my mine, and more often than not end up in either a pissing match or agree to disagree.

 

[Brock Boeser]

For QoC, wouldn't it be skewed by coaches that like to play their checking lines against the other teams top forwards and coaches that like to go power vs power and during away games where the home coach tries to find a better match up for certain players?

If there's a website someone can lead me to I'd love to read more and get a grip on it.

 

[Wisp]

Horvat hasn't been 'replacement level' since December.

His experience here is a very good exhibit A for the benifets of letting young guys adjust to NHL pace.

 

[Wisp]

yeah

http://nhlnumbers.com/2015/3/1/the-limits-of-observation

This is a great article but it isn't really going to change anyone's mind. In this setting, By and large the people who poopoo on stats are those who just don't want to be bothered with them, especially when they refute their preconceived notions.

The 'Watch the game' comeback is particularly immflamatory, as it is easier to pretend I don't actually watch the Canucks play then understand the concepts I'm using to try and support my own arguments and eyeball tests.

 

[CanaFan]

Bo's on-ice SH% is still at like 13%. That's not really sustainable for anyone either. It probably comes down to about 8-9%.

But his on-ice SH% includes Bo's personal SH% which we've already agreed is unsustainable and have used to adjust his goal production down. His own SH% doesn't impact his assists, only his teammates and we've already seen that Hansen (12%) and Dorsett (6%) aren't shooting at an unusual level.

A 15-17 SH% still isn't sustainable for anyone really. Over the last 3 seasons at all situations with players >2000 minutes played (for solid sample) the highest personal SH% is Palat at 13%. Stammer is at 12.3%. I don't think Bo will be as prolific a shooter as Stamkos in his career.

Hockey analysis shows over 40 forwards with a SH% between 15-17% this season (in all situations, not just 5v5). Now I adjusted the min. games down to 500 mins (which is roughly 25-50 games, depending on their TOI) because no one has 2000 mins played this season. But it seems that 15-17% is sustainable over fairly long stretches, which is all that matters (not that it is sustainable over entire seasons).

Bo could certainly stay at a high enough level (15-17%) due to his play style and shot selection so that he could reasonably hit ~20 goals in a full 82 game season. His assists are not driven by his personal SH% and there is nothing to suggest they are unsustainable (quite the opposite, they should go up with better players like Vrbata or Kassian instead of Hansen and Dorsett). It certainly isn't a given, but neither is it as unreachable as I think you are suggesting. Bigger minutes, PP TOI, and lineman quality could easily offset a regression in SH%.

Rather than talking about production which is unlikely to continue (for now), what is really encouraging about Horvat are his SAT differentials. He started the year at basically replacement level, but his underlying numbers have also turned a corner since the all-star break to where he is now regularly >50%. For a 19 y.o. to even break even is quite impressive and bodes well for his future value as a two-way player.

I'm not overly familiar with the newer stats like SAT diff's but I'm glad they speak well to Bo's play.