Scoring This Season

[JamesB]

Last season the Canucks set a franchise record for fewest goals scored (182) breaking the record set the previous season and finishing 29th in the league in both goals scored and overall points. It was a tough season for fans. There was not much going on -- just one genuine rookie having a nice first season (Stecher) and a few fairly young guys all taking a nice step forward (Horvat, Baertschi, and Granlund). But, let's face it, there have not been many years in Canuck history with less to cheer about as far as on-ice results are concerned. The best news may have been the 2017 draft, which does look like a great draft.

 

So what about this year? Win or lose, it would be great to score more. There are several things we are hoping for. Maybe Travis Green's systems lead to more scoring than Willie's, maybe some vets have bounce back years (Sedins, Ericksson), maybe we see breakthrough years for some young guys (Boeser, Virtanen) or big steps forward for some other guys like Horvat, Baertschi, Granlund, etc. We can hope for that.

 

But I decided to put optimism and home bias aside to do a mechanical prejection for goals scored based on previous performance and age. Now that we have a very good idea of who will be on the team, it is possible to do this.

 

1. Question. The question I am asking is as follows. If all the Canuck players followed normal progression based on previous performance, how much would they score next year. Of course we are all hoping that most guys have better than normal progression or that coaching will make a big difference. But my question is basically to just set the baseline.

 

2. Method.

 

A. The method is pretty simple. It could be made more complex but that would be unlikely to change the results much. First I form a list of 13 forwards and 7 defencemen who look like they will be the first 13 forwards and 7 D-man on the depth chart. I assume that they will share the games equally, which means 75 games for each forward and 70 games for each D-man. Of course there could be trades and moves up and down from Utica for injuries and other reasons. Therefore, the average number of games played will probably be less than that and other guys will come into the mix. But that should not change the numbers much. As a simple first step approximation, this approach should not be far off and should give a good baseline. 

 

B. I add up each player's goals scored over the past two seasons and divide by NHL games played to get goals per game (GPG). (Using two years is better than one as it smooths out some random variations, whereas going back three years probably goes too far back.) This works for every player except Boeser who spent almost all of the past two years in the NCAA. For him I take his NCAA goal scoring record and apply the standard league adjustment factor, and take into account his goals at the end of last season with the Canucks.

 

C. I then multiply GPG  by expected games played (75 for forwards and 70 for Ds) to get expected goals.

 

D. I do an adjustment for age-related change. This is a simplification but it is pretty close to "best practice" research on the subject. For forwards aged 24 or less right now, I assume a 5% increase in GPG. For forwards between 25 and 31 right now, I assume no change. For forwards over 31 right now I assume a 5% decline. For D's I add one year to each age threshold. This is perhaps a slight underestimate of typical age-related change but 5% is a round number and 10% would be too much.

 

 

3. Players. At this stage, the top 7 D are pretty much set: Tanev, Edler, Stecher, Hutton, Gudranson, Del Zotto, and Wiercioch. The top 13 forwards are not entirely clear but there is not a lot of doubt. I am assuming the forward depth chart goes as follows: Horvat, Henrik, Daniel, Baertschi, Granlund, Vanek, Gagner, Eriksson, Sutter, Dorsett, Boeser, Virtanen, Burmistrov. The order does not matter. And if it is Archie instead of Dorsett or Burmi, that won't make much difference to the calculation. 

 

4. Results. If I knew how to post the entire spreadsheet I would. But the bottom line is that this method yields 223 expected goals, up from 182. Last year the median number of goals scored by NHL teams was 226, so this is very close and would put the Canucks right about the middle of the pack. Last year the Canucks gave up 243 goals. If they gave up the same number this year, the goal differential would be -20. Last year that would have been good enough for 21st place in goal differential, which would probably also mean 21st place in expected overall league standing  -- a big improvement in performance which is, coincidentally, my prediction for where the Canucks will finish this year in the prediction thread.

 

The main improvement just comes from better personnel. The vast majority of goal-scoring comes from forwards. Last year, on any given night, the Canucks had 4 or 5 replacement level forwards on the ice instead of the normal 1 or 2. Guys like Skille, Megna, Chaput, Gaunce played a lot of games and scored very little as those 4 guys had 13 goals in total, despite getting a lot of minutes. And Shore, Crammerosa, Labate, and Molino played a combined 42 games and had a combined 0 goals.

 

This year's replacements: Vanek, Gagner, Boeser, and, yes, Virtanen, have much higher expected goal totals. Eriksson is projected to recover to somewhere between last year's very poor 11 goals and his 30 from the previous season (and is projected to hit 20). And guys like Horvat, Granny, and Baertschi should continue to show modest (5%) improvement.

 

I also think that this baseline is fairly conservative, because I think the Canucks will beat this baseline due in large part to improved coaching, especially on the PP. Taking coaching into account, I think the Canucks will beat the league median in goals scored. Goaltending is a big question mark and I have no more idea than anyone else (and less than many) about how goaltending will pan out. My best guess is that it will be similar to last year as I think Nilsson looks pretty good. If they have above average goaltending, the Canucks could be in the playoff conversation.

 

5. Individual Results

Here are some of the interesting results from the individual baseline projections, athough there are no big surprises because the method is based on past performance.  Based on league adjustment factors and age-related improvement I am projecting Boeser for 20 goals. In fact, the Canucks should have a large group of guys threatening the 20 goal plateau, including Eriksson, Boeser, Baertschi, Granlund, Daniel, and Vanek. Horvat should not be far off. .And Sutter and Gagner are both projected to reach or beat 15. 

But no-one is projected to threaten the 25 goal threshold. On D, the projected top scorers are Edler and Del Zotto, each projected to score 7 goals.

 

This is extremely broadly distributed scoring. From that point of view it is a very interesting experiment. It is an extreme version of "scoring by committee". Maybe it will be a substitute for having one or two "franchise" players. (By the numbers, the Canucks will have no franchise players or even elite players next year in terms of scoring as those terms are typically defined. But they will have a lot of good players.) 

 

Thanks for reading this. I would of course be interested in comments. If you don't like the method, please try to suggest feasible improvements or, better yet, do the calculations.

 

 

 

[Davathor]

I think the Canucks are going to score a lot more than people expect... and it's not just regular player progression. It's a better coach, better matchup, better puck moving, better linemates... Boeser, a matured JV, Vanek and Gags, a healthy Gudbranson... there's a lot of reason our scoring will go up - I'm going to predict 20th -16th in scoring, but a bottom 5 finish due to the worst goaltending tandem in the NHL

 

#reality

[Fanuck]

Reasonable logic James.   Wondering how other teams' improvements impact us - if we're hoping to be better aren't other teams expecting to as well, making it somewhat of a wash? (Yes,  I know some teams might drop down as well) Also,  does it really help us long-term by finishing 21st or thereabouts? 

[DJSkingz]

Good read. Thanks for writing it. +1

[blackhawksdown]

Someone get this man a cookie!

[Alflives]

Is there an award for reading the entire OP? 

well thought out, and interesting.

 

 

 

[6string]

Excellent article, Canucks finish in the bottom five but win the lottery!

[PistolPete13]

Good post James. Your logic is sound and now that you have a spreadsheet you can easily tweak it for factors such as better power play, coaching, etc. 

The X factor is going to be better productivity due to our not spending large  chunks of each game trying to protect one goal leads. 

I swear that Willie D was convinced that if the Canucks scored first, the surest way to  victory was to sit back and defend.

So far it seems that TG would rather try to push the pace and attack more when we have the lead.

Under this more offensive approach guys like Megna and Chaput are expendable. So much ice time was allotted to such “trusted” non- offensive types, no wonder we couldn’t score.

Ive been a fan of this team since day one and I have witnessed many inept offensive years, so it was a real body blow to hear that these past two years have been the worst ever.

Better days are ahead.

 

[Nuxfanabroad]

1 hour ago, Fanuck said:

 Also,  does it really help us long-term by finishing 21st or thereabouts? 

It's a good question. If EP, OJ, Lind & Gads come thru, we might be fine, even finishing in the dreaded "middle ground" territory.

 

If development is just as important(or even, almost...) as draft position, then we can just cheer for the team's climbing progress from here on out.

[CanadianRugby]

2 hours ago, JamesB said:

4. Results. But the bottom line is that this method yields 223 expected goals, up from 182. Last year the median number of goals scored by NHL teams was 226, so this is very close and would put the Canucks right about the middle of the pack. Last year the Canucks gave up 243 goals. If they gave up the same number this year, the goal differential would be -20. Last year that would have been good enough for 21st place in goal differential, which would probably also mean 21st place in expected overall league standing  -- a big improvement in performance which is, coincidentally, my prediction for where the Canucks will finish this year in the prediction thread.

Well goal scoring can't really get any worse, we're already beating out own records for fewest goals scored.  So it can only go up you'd think.  However even playing WD's conservative system and with Miller in net, Vancouver gave up the 6th most goals in the league.  If they open up to get more scoring they'll expose Markstrom and get lit up even more than last year.  So even if Boeser plays well and some of the additions (and Eriksson) add scoring, the goal differential won't be pretty.  

 

That said, I hope Horvat has a great year.  I bet a jersey that he'll have more points than that dirty Slovenian Kopitar.  

[JeremyCuddles]

One of the biggest things people gotta keep in mind is injuries. We had the second most man games lost due to injury. Meaning last year we had, less than ideal players playing much more than expected. Our depth is better this year, health should hopefully be better, potential bounce back seasons, and players progressing.

 

I could see this team being a playoff team IF we stay healthy. It's a bold statement but last years team was a bubble team til injuries took over, and Hansen and Burr being traded. Scoring should be much better, even if we sustain a bunch of injuries, as I am much more comfortable with our depth. I really do think a healthy Vancouver can turn a lot of heads. Of course, I could be going full Homerism. 

[khay]

Nice post!

 

I think the system change and new quality additions will lead to more goals as the OP said. WD had poor personnel to work with, and he did not have a very good offensive system. Green's system looks decent thus far in the preseason games.

 

Last season was indeed horrible. Nothing against guys like Skille, Chaput, and Megna, but each team should have at most one of these replacement level players, not three playing major roles! Heck, Megna saw some time on the first line, Chaput centered a second or third line at some point too!

 

[DrJockitch]

This is the rare pre-season where it still looks like everything is up in the air.  Not sure how this team is going to do. 

I feel that the balance between players lost and players gained via free agency really don't make us a better team.  A better scoring team but softer, less defensively sound and goaltending is unproven.

Brock and Jake to me are the biggest potential arbiters of success for this squad that doesn't stand 6'6" or taller.  While the free agents aren't going to make us a lot better, these two might.  This is probably how it should be.

Like what I see from TG, not sure he has the horses to get it done this year.  WD took the coaching change year and got us to the playoffs, then the wheels fell off the bus.

Certainly an exciting year because-no idea what to expect, could be scraping into playoffs or 29-30 in the league and either potential seems reasonable (Colorado is holding on to 31st and that is one race we definitely can't win).

[JamesB]

11 hours ago, Fanuck said:

Reasonable logic James.   Wondering how other teams' improvements impact us - if we're hoping to be better aren't other teams expecting to as well, making it somewhat of a wash? (Yes,  I know some teams might drop down as well) Also,  does it really help us long-term by finishing 21st or thereabouts? 

 

11 hours ago, PistolPete13 said:

Good post James. Your logic is sound and now that you have a spreadsheet you can easily tweak it for factors such as better power play, coaching, etc. 

The X factor is going to be better productivity due to our not spending large  chunks of each game trying to protect one goal leads. 

I swear that Willie D was convinced that if the Canucks scored first, the surest way to  victory was to sit back and defend.

So far it seems that TG would rather try to push the pace and attack more when we have the lead.

Under this more offensive approach guys like Megna and Chaput are expendable. So much ice time was allotted to such “trusted” non- offensive types, no wonder we couldn’t score.

Ive been a fan of this team since day one and I have witnessed many inept offensive years, so it was a real body blow to hear that these past two years have been the worst ever.

Better days are ahead.

 

 

7 hours ago, khay said:

Nice post!

 

I think the system change and new quality additions will lead to more goals as the OP said. WD had poor personnel to work with, and he did not have a very good offensive system. Green's system looks decent thus far in the preseason games.

 

Last season was indeed horrible. Nothing against guys like Skille, Chaput, and Megna, but each team should have at most one of these replacement level players, not three playing major roles! Heck, Megna saw some time on the first line, Chaput centered a second or third line at some point too!

 

 

4 hours ago, DrJockitch said:

This is the rare pre-season where it still looks like everything is up in the air.  Not sure how this team is going to do. 

I feel that the balance between players lost and players gained via free agency really don't make us a better team.  A better scoring team but softer, less defensively sound and goaltending is unproven.

Brock and Jake to me are the biggest potential arbiters of success for this squad that doesn't stand 6'6" or taller.  While the free agents aren't going to make us a lot better, these two might.  This is probably how it should be.

Like what I see from TG, not sure he has the horses to get it done this year.  WD took the coaching change year and got us to the playoffs, then the wheels fell off the bus.

Certainly an exciting year because-no idea what to expect, could be scraping into playoffs or 29-30 in the league and either potential seems reasonable (Colorado is holding on to 31st and that is one race we definitely can't win).

Thanks for the comments, guys. I think the bolded points above are particularly important. First, as @Fanuck says, next year we will face a by now familiar "tanking" or "no man's land" problem. If the team does as my estimates suggest and finishes in 20 or 21 range, we don't into the placeoffs and we are also very unlikely to get a high draft pick (very low chance of a lottery pick), probably picking 11th or 12th. Also, we would probably be close enough to the playoff race at the deadline that the team would not make deadline deals of vets for draft picks or prospects. If, on the other hand, we had another really bad year (like last year) we could hope to have a similar deadline and draft and add another 5 or 6 high quality prospects, including at least one who looks like a potential elite level player (like Pettersson). That would give us a great prospect pool.  Personally I am kind of torn. It would be nice to see the Canucks start an ascent from the bottom, but it would also be great to have another big in infusion to the prospect pool and maybe get another franchise or elite level player.

 

I agree with @PistolPete13. WD played Magna and Chaput and Skille so much because they could be "trusted". Yes, they could be trusted not to score. I would much rather see guys who have a chance to develop into impact players, even if they make more defensive mistakes. But, in fairness to Willie, he had a tough hand to play because, as @khay says, the Canucks had 3 or 4 or 5 replacement level players on ice a lot of the time when good times would have maybe 1.

 

Finally @DrJockitch makes a fair point. This year's team is certain to score more. But team could be even softer than last year (although that would not be easy to achieve). However, wIth Guddy back, and (we hope) with Dorsett healthy and Virtanen and maybe Archie or Pedan in the mix if necessary, the team could actually be tougher than last year. Defense could also be suspect, partly because of more offensively oriented personnel and partly because of more offensively oriented coaching but I don't actually see defence being any worse. I agree that goaltending is a big question mark and we won't have a good idea of how that goes until we actually season the season in progress. Goaltending is in the general the hardest position to project in any case.

[kingofsurrey]

Last season the Canucks had better luck scoring at the Roxy.

 

 

 

Scoring NHL goals can only go up this year for the Nucks...

[bloodycanuckleheads]

You are getting a higher goal-total than you should expect (basically, you have to take your 223 goals and lower it by a fair amount in order to get your final number).

 

Let's see, just off the top of my head...:

 

1.  This system doesn't take injuries into account.  Replacement players tend to be worse than the player they are replacing, a lot worse.  Your system assumes that all our top players stay healthy all season long.

2.  Similar to #1, this system doesn't take player-development into account.  This system assumes that our top-13 forwards in every game are the best 13 players we have.  That's never the case.  We had 25 forwards play for us last year.  So, during a lot of games, we were using our 21st or 23rd best forward.

3.  Last year, our 10 defensemen scored 22 goals total.  Next year, you have us getting 14 from just two players.

4.  You appear to have Daniel's scoring going up a fair bit next year - so your bit about adjusting for age and decline was apparently disingenuous.  In fact, it looks like you have all our old-guys scoring more next year!  Some dramatically more.

5.  Conveniently, you hid all the important numbers (your predictions for every single player that isn't easy to predict).  Presumably, because they are all way too high.

6.  Basing everything on the last two years of goal-scoring is highly-biased.  It vastly over-values declining players.  It over-values rejects from other teams.  Etc...  It also under-values improving players or injured players.

 

Accounting for all this brings your 223 goals down to maybe 200-210, which would put us bottom-10 or bottom-5.

[Industrious1]

Two hundred sounds about right. 

 

A bit up from last year because of some emerging players.  However,  more of a drop from the Sedin's which is slowly being overshadowed by young guys starting to pick up some of the slack.

[JamesB]

3 hours ago, bloodycanuckleheads said:

You are getting a higher goal-total than you should expect (basically, you have to take your 223 goals and lower it by a fair amount in order to get your final number).

 

Let's see, just off the top of my head...:

 

1.  This system doesn't take injuries into account.  Replacement players tend to be worse than the player they are replacing, a lot worse.  Your system assumes that all our top players stay healthy all season long.

2.  Similar to #1, this system doesn't take player-development into account.  This system assumes that our top-13 forwards in every game are the best 13 players we have.  That's never the case.  We had 25 forwards play for us last year.  So, during a lot of games, we were using our 21st or 23rd best forward.

3.  Last year, our 10 defensemen scored 22 goals total.  Next year, you have us getting 14 from just two players.

4.  You appear to have Daniel's scoring going up a fair bit next year - so your bit about adjusting for age and decline was apparently disingenuous.  In fact, it looks like you have all our old-guys scoring more next year!  Some dramatically more.

5.  Conveniently, you hid all the important numbers (your predictions for every single player that isn't easy to predict).  Presumably, because they are all way too high.

6.  Basing everything on the last two years of goal-scoring is highly-biased.  It vastly over-values declining players.  It over-values rejects from other teams.  Etc...  It also under-values improving players or injured players.

 

Accounting for all this brings your 223 goals down to maybe 200-210, which would put us bottom-10 or bottom-5.

Good comments. Thanks for taking the time to read my OP carefully and make thoughtful comments. Here are my responses.

 

1. Injury.  I thought about injury but, to keep things simple, I assumed that things would work out as if the 12 top guys each missed 7 games due to injury (i.e. each plays 75 games). That is 84 man-games. I assumed the 13th  would also play 75 games. I assumed the other 9 man-games would be replaced by someone who would not score at all. I agree that this is optimistic. The actual distribution of play will be much more complicated. The question is, does it change expected scoring? I think the answer is not much. 

 

My model is based on an optimistic injury story. A pessimistic injury outcome would be if the Canucks lose the equivalent of one additional good player for the entire season (or two for half the season, etc.) The marginal top 6 forward is projected for 14 goals. So this would cost an additional 14 goals. That player would be replaced by a replacement player. If the replacement player was Megna or Chaput for the entire year, he might get 5 or 6 goals, for a net loss of 8 or 9 goals. However, I think that the actual replacement player is likely to be Goldobin or Rodin. If Goldobin played a full season he would also project at about 14 goals. So there would actually be no loss of expected goals (although there would probably be more goals against). It might be more reasonable to deduct a 3 or 4 goals due to injury beyond what I originally projected, but not more than that.

 

2. I don't see how your point 2 relates to player development. I think it is really the same as point 1 -- the actual distribution of play across players will be much more complicated than I assume as guys move up and down between the Canucks and Utica (possibly for developmental reasons, I guess). But the point is that this movement will not affect expected goals totals much because, due to greatly improved depth, the guys moving up from Utica won't be very different in expected scoring than the guys moving down. If the guys moving up are guys like Goldy, Rodin, and Molino, scoring might actually increase (although defensive play would likely get worse).

 

3. Yes, scoring from the D was terrible last year. Edler had 6 goals in each of the past two seasons, but he missed a lot of games due to injury. Given slightly better assumed health his expected value goes to 7 goals for this year. And adding Del Zotto adds quite a bit to the D as his expected value is also 7 goals next year. The other 5 guys combine for an expected value of only 13 more goals, so 27 for the D overall, which is still pretty low. I think this is quite conservative, not optimistic at all. If anything Stecher and Hutton are likely to score more than this projection given Green's emphasis on more offensive aggressiveness from the D.

 

4. The numbers for Daniel are as follows. In 2015-16 he scored 28 goals, In 2016-17 he scored 15 goals. That is 43 over the two seasons combined and he played all 82 games in both seasons. His goal per game number is 43/164 = 0.26. Applying that to the assumed 75 games yields 19.7 goals and multiplying that by a depreciation factor of 0.95 yields an expected total of about 19 goals to the season ("threatening 20"). So this does assume some bounceback which is what you would expect from ordinary regression to the mean. Every player is like this. These are just mechanical calculations based on a large body of empirical evidence about player performance. There is no particular upward (or downward) bias for Daniel or anyone else.

 

5. I did not "hide" anything. My OP was pretty long as it was, but here are projected goal totals by forward (rounded to whole numbers): Boeser 20, Eriksson 20, Daniel 19, Granlund 19, Baertschi 19, Vanek 18, Horvat 17, Sutter 16, Gagner 15, Henrik 12, Virtanen 8, Burmi 8, Dorsett 5.  And the defence adds 27, for a total of 223. There is nothing unusual about any of those numbers as they are very similar to what those players have done in the recent past. They are bit high because I am assuming that they will play most of the games. In fact, some of these will play less and therefore score less. But other players will play and score to the overall expected value does not change much.

 

6. Basing projections on the past two years does not create bias as long as there is a reasonable age correction. It is true that my age correction is modest (+5% per year for young players and -5% per year for older players) but it is not far off (and I have looked at that data closely in the past). If that correction is too modest is actually underestimates expected goal scoring for the Canucks because they have more young improving guys than older declining guys. But even a much more detailed simulation of the age effect would not change the overall expected value by much -- maybe 2 or 3 goals -- and in a positive direction. 

 

Adding up all these points, there is no reason to believe that 223 is an overestimate. The simple fact is that Benning has built a line-up with a significantly better goal scoring record. As I said before, replacing guys like Skille and Chaput and Megna (and Gauce) with guys like Vanek and Gagner and Boeser and Virtanen has to raise expected scoring a lot. Replacing Tryamkin with Del Zotto also adds a few expected goals.

[EdgarM]

Taking all of the scientific info. aside, I believe the biggest question mark on scoring this year is how the Sedins are going to get deployed. Are they still going to be treated as the the #1 line and #1 PP? If so, things will be similar to last year where the PP suffers and the Sedins are stuck playing in their own zone for much of their shifts. (Last night one of them played for about a minute and a half of the entire PP!)

Baert, Bo and Granny have shown that they can hold their own now and they have much younger support from such players as Vanek, Gagner and Sutter. Not to mention the up and comers such as Virt and Boes.

The Sedins primary skills such as passing and PP creativity have all but diminished. Yes they have their high Hockey IQ but they are mute when they cannot execute.

Sit these two on occasion and put them on the #2 PP and watch the rest of the team blossom.