Thursday, April 20, 2017

The Sabres Fired Tim Murray (Part 2)

Part 1 primarily focused on drafting (and, obviously, how poorly the Sabres did with little to no improvement) under Murray's tenure at GM. Now I want to look at the trades that he made over his three years and change.

Ryan Miller Trades

  • Sabes --> Blues: Ryan Miller, Steve Ott
  • Blues --> Sabres: Jaro Halak, Chris Stewart, William Carrier, 1st round pick, 3rd round pick
And then, a week later:
  • Sabres --> Capitals: Jaro Halak, 3rd round pick
  • Capitals --> Sabres: Michal Neuvirth, Rostislav Klesla
Essentially, Murray turned a franchise goalie (who happens to be one of the best American goalies ever) into a shittier goalie who was gone after 29 games, a pair of forwards who will never contribute to anything meaningful in Buffalo, Carrier (who scored 8 points in 41 games this season), and draft picks that were flipped in other trades. 

Grade: Not Great!

And then we have two distinct types of trades that seem to be the product of a confused front office. The Murray-era Sabres participated in 20 trades after the Miller trade. Here's how I'd split them up:
  • Dumping players for picks/pros (6 trades)
  • Spending picks/prospects to acquire players (8 trades)
  • My Trash For Your Trash (5 trades)
The Trash trades are largely irrelevant, of course, but why is a franchise dumping players for picks while also, at the same time, using picks and prospects to add players to their current team? Tanking, at least when maximized, is stockpiling picks and young prospects and then being as bad as possibly to maximize the value of those picks. If you can keep guys in juniors or the minors - and maybe even let them serve in leadership roles and make deep playoff runs in those leagues - then everything should come together and you win a bunch of championships like the Blackhawks and Penguins. 

But adding guys like Ryan O'Reilly and Evander Kane seems to run counter to everything that the Tim Murray "Tanking" Sabres seemed to be doing. That's why they ended up with Alex Nylander last summer instead of Auston Matthews or Patrik Laine. It's why they're going to end up with Urho Vaakanainen or Ukko-Pekka Luukkonen or Miro Heiskanen instead of Nolan Patrick or Nico Hischier this summer.

The Sabres, effectively, wasted three years while Tim Murray was in charge. They don't have any good defensemen, they might be able to roll three decent lines next season, and they have not even sniffed the playoffs since 2011. 

The Sabres Fired Tim Murray

Tim Murray had been relieved from his position of General Manager by the Buffalo Sabres. In the coming months, we will likely learn how much of that decision was made with Jack Eichel's thoughts in mind - but that is not our concern right now.

This morning, I want to take a look back at Murray's three years in charge of the Sabres to see how fucking awful he was. Let's start in the middle of the 2013-14 season, when he got hired:

2013-14 Season

We should begin on January 9th, 2014. The Sabres fired Darcy Regier, the longest-serving and winningest GM in franchise history, and turned the reigns over to Murray. They really bottomed out over the final months of the season, finishing in last place in the Northeast Division (RIP).

But, when you're a new GM, you almost would rather start with a last-place finish so you can take advantage of the better draft position. So Murray also traded franchise cornerstone goalie Ryan Miller to St. Louis for Jaro Halak, some shitty players, and some draft picks.

2014-15 Season

We'll kick off Murray's first full season with his offseason moves. The Sabres picked second in the 2014 Draft, missing out on Aaron Ekblad (he went first to the Panthers). Buffalo selected Sam Reinhart with the number two pick.

I think, in hindsight, Sabres fans would rather have taken Leon Draisaitl (who went 3rd), William Nylander (8th), Dylan Larkin (15th), or David Pastrnak (25th). But it's early in Reinhart's career and he'll probably turn out to be perfectly a decent player. I do not recognize the names of any of Buffalo's later-round picks, but, again, it's early.

The 14-15 season was a drag, and the Sabres finished with just two more points than they had in 13-14. They were in last place in the division again, and this time they were last place in the whole league. Draft picks!

Perhaps the most important piece of the 14-15 Sabres' season was the acquisition of Evander Kane. Kane is now just 25 years old, and he figures to be a key piece of the Sabres rebuild that may or may not be happening.

2015-16 Season

The Edmonton Oilers won the draft lottery and were gifted Connor McDavid. The Sabres ended up with Jack Eichel. Eichel is great (certainly a better 2nd overall pick than Reinhart), but McDavid was the best first overall pick in a decade. Here's how Murray handled the lottery results:

Not great. I can't imagine that Eichel was excited to get the prom invitation only after the cheerleader with the cute dimples rejected our boy Timmy.

Buffalo traded the 21st pick in the draft for goalie Robin Lehner. Colin White, who was selected with that pick, may or may not haunt the Sabres - but Lehner appears to be their goalie of the future.

That summer, Murray also hired head coach Dan Bylsma, who had won a Stanley Cup with the Pittsburgh Penguinis after they successfully rebuilt after a tank. The team brought in Ryan O'Reilly (who eventually would drunk-drive his car into a Tim Horton's) and made a huge leap from the previous season.

Well, a huge leap in terms of points - from 54 to 81. However, that "huge" jump turned out to be just enough to push them from 8th place in the division to 7th.

2016-17 Season

Boy, this story of Murray's tenure is wrapping up quickly. The Sabres, obviously regretting their 2014 selection of Reinhart and wishing they'd have taken William Nylander, drafted the younger Nylander brother (Alex) seventh overall.

Murray also gave an enormous contract (seven years, $42 million) to Kyle Okposo in the summer. Okposo scored 19 goals and 45 points, which is suboptimal for a guy making $6 million per year.

The Sabres finished with fewer points than the previous season, and didn't have a single player break the 30-goal or 60-point marks. Again, they finished in last in the division.

RIP In Peace

Murray was at the helm for three and a half seasons. In their best season, Buffalo finished second-to-last in their division. In all the rest, they finished last. They used their three top-ten picks under Murray to draft the wrong guy by choice, the wrong guy because of the lottery, and a middle-of-the-road prospect because they overachieved despite not competing. That's why Tim Murray got canned, RIP.

Looking Forward

Here's who Buffalo has under contract moving into next season:

O'Reilly, Okposo, Kane, Moulson, Eichel, and Reinhart are six solid forwards. Zemgus Girgenssons was an All Star. There is a lot of youth in the pipeline, especially Nylander. The defense kind of loses it a bit, but the Sabres should be able to piece together a decent team.

We'll see who Terry Pegula hires to fill the shoes of Murray and Bylsma, but the right GM and coach could push the Sabres into the playoff picture in the weirdly-not-that-good Atlantic Division next year.

Friday, April 14, 2017

Where Should Steve Mason Look To Play Next Season?

Usually, I get inspired to write my blogs by a random tweet on Twitter. Today, that tweet happens to be (sort of) my own.

Well, it starts with this conversation between O&BP's Mikey D and the only Scottish hockey fan in the world, Berke. They were looking forward to next season's Flyers goaltending situation. It's admittedly pretty murky, and there are a ton of options on the table:
And then I jumped into the Corsica machine:
But now I want to jump in even further, because those rankings for Mason are quite telling.  He's quite solid against low- and mid-danger scoring chances. Among goalies who have played at least 4,000 minutes in the past four years (that's about two dozen games per year, at a minimum), here's where Mason ranks:

  • Low-danger save percentage: 0.9845% (3rd behind Lundqvist and John Gibson)
  • Mid-danger save percentage: 0.9342% (5th behind Talbot, Khudobin, Gibson, and Frederik Andersen)
  • High-danger save percentage: 0.7972% (37th out of 47 eligible goalies)
You expect save percentage to decline as shot quality improves, but Mason's curve is harsher than most goalies. 

It's a very minor difference, but over the 996 high-danger chances Mason has faced in these four seasons, his 79.72% against Price's 85.49% amounts to a difference of 57.5 goals. The Flyers' goal differentials the last four years (though Mason only played 61, 51, 54, and 58 games) were +1, -19, -4, and -17. You always hate to compare your guy to the best in the world, but Carey Price would have singlehandedly erased the Flyers net goal differential over those four years. 

We've always known that Mason is in a tier below the elite level of NHL goalies, and it's almost all because he struggles against high-danger chances. 

So, looking forward, from Mason's point of view and not a Flyers point of view, which destinations make sense? Here's our checklist, based on what Mason has said and what we can infer about his statistical history:
  • Team without a number one goalie
  • Defense that doesn't allow a lot of high-danger chances
That's it! That's the whole list. First, let's narrow our list down from 30 teams by eliminating the franchises who have cornerstones already. That knocks out:
  • Metro: WSH, PIT, CBJ. NYR, NJD
  • Atlantic: MTL, OTT, BOS, TOR, TBL, 
  • Central: CHI, MIN, NSH
  • Pacific: ANA, SJ, LA, ARI
We also have a bunch of Maybe guys. Is Jake Allen that guy for St. Louis? Is Brian Elliott that guy for Calgary? Is Cam Talbot that guy for Edmonton? Is Winnipeg going to pay Connor Hellebuyck this summer and make him the number one guy? Is Semyon Varlamov the guy in Colorado? Is Jimmy Howard that guy for Detroit? Is the Greiss/Halak tandem going to work in New York? I mean... maybe. 

But that list removed 17 teams from the list (and maybe 25, we'll see), leaving us with seven contenders for Mason: 
  • Hurricanes: have $6 million committed to two goalies next season
  • Panthers: have $5.3 million committed to Roberto Luongo until 2022
  • Sabres: will likely re-sign RFA Robin Lehner this summer
  • Stars: probably not going to commit more money to goalies because they're paying $10.4 million already
  • Canucks: Jacob Markstrom is signed at $3.6 million for three more years
Barring something wonky, I would expect the 'Canes, Panthers, and Stars to duck out of this race. So, between Buffalo and Vancouver, who would Mason prefer to play in front of? Let's compare the biggest thing that matters for Mase - high-danger chances against. 

Now, considering the Sabres, and Canucks finished 17 and 25 points out of the playoffs respectively, we shouldn't get our hopes up too high that any of them do anything good on ice. But we might surprise ourselves. For reference, league-average High Danger Chances Against Per Sixty was 6.41 last season, and that made up an average of 21.5% of total scoring chances.
Let's start with the three goalies that Buffalo used this year:
  • Robin Lehner (59 games played): 5.85 HDSA/60, 17.6%
  • Anders Nilsson (26 games): 7.13 HDSA/60, 21.6%
  • Linus Ullmark (1 game): 5.98 HDSA/60, 17.9%
Wait, Buffalo is a better-than-average team in terms of denying high-danger chances?

Now for Vancouver's carousel:
  • Ryan Miller (54 games played): 7.00 HDSA/60, 21.8%
  • Jacob Markstrom (25 games): 6.45 HDSA/60, 22.2%
  • Richard Bachman (5 games): 9.31 HDSA/60, 27.3%
That would make an easy decision for Future Sabre Steve Mason. 

Tuesday, April 4, 2017

Applying Playing Styles/Clustering To The Flyers

If you are a hockey fan, you should read Ryan's post from today. Stop reading this, open that link in a new window, and read it from start to finish. It's probably going to end up being a huge piece for hockey's internet community, if for no other reason than it creates a SHIT TON of possibilities for expansion.

The nuts and bolts is simple: like EA Sports NHL, most players fall into four major categories:

  • Forwards: Playmaker, Shooter, Balanced, and Dependent
  • Defensemen: All-Around, Volume Shooter, Puck-Mover, Defensive
There are obviously some ways to nit-pick those groups (and perhaps dividing them even further is next up on Ryan's to-do list), but according to his notes there are pretty even cuts for those eight groups. 

Before we jump into applying this to a Flyers context, I want to note that there are only 54 Playmakers, 69 Shooters, 41 Volume Shooters, and 11 All-Arounds. I was curious who those All-Around guys were, so here's the list:
  • Obviously: Doughty, Karlsson, Keith, Pietrangelo, Letang, Hedman
  • Okay That Makes Sense:  Tyson Barrie, TJ Brodie, Hampus Lindholm,Keith Yandle
  • Wait What: Derrick Pouliot, Jakub Nakladal, Jamie McBain, Shea Theodore, Tyler Wotherspoon
We're going to let ourselves get sidetracked really early on in this piece because how the fuck are those five guys among the best statistical defensemen in the league?
  • Pouliot: Played 34, 22, and 10 games the last three years with Pittsburgh. Used very sparingly and sheltered heavily. 
  • Nakladal: Played 27 games with Calgary last year and 3 games with Carolina this year. He kind of stinks, I think. 
  • McBain: Played 348 games in 8 seasons. Scored 92 points in his first five seasons but just 18 since (including 0 in just 3 games this year). 
  • Theodore: Played 50 games for Anaheim between last year and this year, scored 8 points each season. He should almost certainly have more this year, though, because he's shooting just 1.9%. He's probably the best of this bunch.
  • Wotherspoon: This was just a fake name that I made up. 
That little subquest means we're going to allow ourselves the ability to tweak who the model says qualifies for each category. The most obvious, from the Flyers fan's perspective, is Ivan Provorov. He's technically a Defensive Defenseman (the worst of the four categories), but we're going to move him to Puck Mover because he's been (1) stapled to Andrew MacDonald and (2) faced with the toughest competition of the team's blueline. Fair? Good. 

Here's who the Flyers will be icing next season:

Playmaker: Giroux
Shooter: Voracek, Simmonds, Konecny
Balanced: Filppula, Couturier, Schenn, Raffl, Weise, Bellemare, Lindblom
Dependent: Read, Cousins, Laughton, Leier, Vecchione

All-Around: Nobody
Volume Shooter: Gudas, Gostisbehere, Del Zotto
Puck-Mover: Provorov, Sanheim, Myers
Defensive: Manning, MacDonald, Morin

A few notes:
  • Considering Schenn is near the top of the league leaderboard for power play goals and Lindblom is near the top of the league in Sweden, I wouldn't be surprised them to classify themselves more as Shooters than Balanced next season
  • I was tempted to move Provorov straight into the All-Around (read: Elite) category, but keeping him as a Puck Mover seems like the safe move
  • Sanheim and Myers, if they develop into anything like what they should be as professionals, should both fall in the Shooter or Puck Mover category. Based on the qualifications listed in Ryan's post, I think Puck Mover might make more sense for both of them. 
  • Morin could be a Gudas-type Shooter, but given that he's only just making his NHL debut tonight, it's tough to say for certain. We'll take the safe route there too. 
Really, the most valuable piece of Ryan's work published today is the way the different player archetypes interact when paired together: 


Playmakers are the most valuable thing you can have when constructing an NHL roster. All-Around defensemen are best when paired with Shooters - not when paired with other All-Around defensemen. 

The best the Flyers can do with their lineup is something like this on offense:

Playmaker - Shooter - Shooter (55.1% xG)
Balanced - Balanced - Balanced (51.5% xG)
Balanced - Balanced - Shooter (50.9% xG)
Balanced - Balanced - Dependent (45.4% xG)

I'm going to say something really crazy here. This is a point to Dave Hakstol for pairing Jake Voracek and Travis Konecny, at different times, with Pierre-Edouard Bellemare and Chris Vandevelde. If he believed that PEB and CVV were Balanced and not Dependent (which would be stupid, but stay with me), then it's better to put them with a shooter than it is to put them with a Dependent (read, in this context, as "someone even worse than they are"). 

If and/or when Jake Voracek earns his contract and crosses the threshold from Shooter to Playmaker (less shots, more shot assists, more overall expected goals), then you can change the second player on the first line to green and change that line's xG to 55.4%. It's not a huge difference. What would be a huge difference would be replacing that Dependent on the fourth line with a Balanced player or a Shooter. That could be Cousins/Laughton/Leier making a leap, or it could be a rookie developing quicker than expected, or it could be another Dale Weise-type free agent signing. 

And on defense:

Volume Shooter - Volume Shooter (54.5% xG)
Puck Mover - Volume Shooter (51.1% xG)
Puck Mover - Volume Shooter (51.5% xG)

If and/or when Ivan Provorov establishes himself in the All-Around category, then change that first D-slot to green and change the first pair's xG to 56.5%. 

Here's my best case scenario Flyers lineup for next season. If players develop like they should and Hextall keeps (what we expect to be) the same roster, this could be your opening night lineup:

Konecny - Giroux - Simmonds (55.1% xG)
Lindblom - Filppula -Voracek (55.0% xG)
Raffl - Couturier - Schenn (51.5% xG)
Bellemare - Vecchione - Cousins (51.5% xG)

Provorov - Gudas (56.5% xG)
Sanheim - Morin (51.1% xG)
Gostisbehere - Myers (51.1% xG)

That's four lines and three pairs that can hold their own at even strength, plus the following buckets of special teams players:
  • PP forwards: Giroux, Voracek, Simmonds, Schenn, Filppula, Konecny, Raffl, Couturier, Lindblom
  • PP defensemen: Gostisbehere, Provorov, Sanheim, Myers, Gudas
  • PK fowards: Giroux, Simmonds, Couturier, Filppula, Raffl, Bellemare
  • PK defensemen: Gudas, Morin, Provorov, and then I guess throw one of the offensive guys to the wolves? Idk this isn't my job
Big night tonight - Sammy Morin making his NHL debut against those scumbag shits from New Jersey. I'll see you out there. 

Monday, April 3, 2017

Gonzaga-UNC Preview

North Carolina - Team Profile
  • 84.3 points per game (8th in the country)
  • 43.3 rebounds per game (1st)
  • 18.1 assists per game (3rd)
Using basic per-game stats, UNC looks like a juggernaut. They score a ton, they outrebound literally everybody and they seem to have a balanced offense that can move the ball. But is that assists stat just a product of playing high-pace, high-event basketball?
  • 0.592 assists per field goal made (37th in the country)
  • 1.511 assists per turnover (7th)
Maybe not. They're in the top ten percent in assist rate and even better at maintaining possession. The other way UNC can maintain possession, in addition to not turning the ball over, is controlling the glass. And oh boy do they ever do that:
  • 40.6% offensive rebounding (1st in the country)
  • 76.7% defensive rebounding (32nd)
Generating tons of second-chance opportunities while limiting them for your opponent is almost always a recipe for success. 

But if we are to look for a "flaw" in UNC ("flaw" in quotations because they're one game away from winning the national championship), it's their shooting metrics:
  • 51.1% on two-point shots (99th in the country)
  • 36.2% on three-point shots (113th)
  • 70.6% on free throws (163rd)
  • 0.343 free throw attempts per field goal attempt (209th)
  • 52.0% effective field goal percentage (102nd)
That certainly paints a picture more along the lines of "they score so much because of their pace" - they don't really do anything particularly great, but they attempt (and make) more field goals than just about everybody else. 

Gonzaga - Team Profile
  • 83.1 points per game (11th in the country)
  • 40.5 rebounds per game (6th)
  • 15.4 assists per game (49th)
Gonzaga isn't far behind UNC in the "juggernaut" discussion based on basic stats. They score, they rebounds, and they assist (though all at less impressive rates that UNC). However, let's start with shot rates to see how their offense compares to Carolina's:
  • 58.4 field goals attempted per game (138th in the country)
  • 29.7 field goals made per game (8th)
  • 57.1% effective field goal percentage (6th)
  • 57.1% on two-point shots (5th)
  • 38.1% on three-point shots (44th)
  • 71.9% on free throws (117th)
  • 0.389 free throws attempted per field goal attempt (76th)
The enormous elephant in the room here is strength of schedule - Gonzaga played seven games against "tournament" and/or "power five" teams during their entire regular season. And during their tournament run they beat a 16 seed, an 8 seed, a 4 seed, an 11 seed, and a 7 seed. You can only play the teams in front of you, but the toughest team they've played during this stretch finished 12-6 in the Big 12. 

Four Questions

The way I interpret this matchup, four points of emphasis will determine who we should bet on. We need to look at four matchups (two on each side) to see how these teams have played against teams who exhibit similar traits to their opponent. Let's start with Carolina. 

UNC against teams whose defenses are strong against two-point shots
  • Relevant Gonzaga stats: 41.2% opponent effective FG% (1st), 39.8% opponent two-point FG% (1st)
  • Teams with similar opp eFG%: Louisville (45.7%/17th, Georgia Tech (46.0%/20th, Virginia (46.1%/24th). Duke (47.5%/41st), Maryland (47.5%/42nd), Kentucky (47.5%/44th), Wisconsin (47.6%/47th)
  • Teams with similar opp 2FG%: Wisconsin (43.1%/11th), Georgia Tech (43.9%/17th), Louisville (44.8%/23rd), Virginia (45.1%/27th), Maryland (45.6%/41st), Oregon (46.1%/55th), Tennessee (46.2%/56th)
So how did UNC fair in their games against Louisville, Ga Tech, UVA, Duke, Maryland, Kentucky, Wisconsin, Oregon, and Tennessee? They played 12 games total against those teams.

As expected, UNC did not fare quite against this group as they did against the rest of the college basketball world. They shot an average of 43.4% from the field - below their average of 46.6% - and they weren't quite as dominant on the offensive glass either - 35.8% compared to 40.6%, even with the decreased shooting percentage.

Point: Gonzaga

UNC against teams who shoot the ball efficiently
  • Relevant Gonzaga stat: 57.1% effective FG% (6th), 57.1% two-point FG% (5th)
  • Teams with similar eFG%: Oregon (55.5%/17th), Duke (54.8%/28th), Butler (54.6%/32nd), Wake Forest (54.1%/38th), Oklahoma State (54.0%/40th), Florida State (53.7%, 46th), Syracuse (53.5%/52nd)
  • Teams with similar 2FG%: Virginia Tech (54.6%/20th), Oregon (54.5%/21st), Butler (54.5%/22nd), Florida State (54.4%/23rd), Indiana (54.1%/26th), Duke (53.6%/32nd), Kentucky (52.9%/42nd), NC State (52.4%/51st)

The Tar Heels fared a bit better here than they did against the teams who are solid on defense, but this is still a bright spot for Gonzaga. UNC allowed these teams to average more than 80 points per game despite the fact that this list includes:

  • Oklahoma State, the 5-seed in the Big 12
  • Indiana, the 11-seed in the Big Ten
  • Syracuse, the 9-seed in the ACC
  • Virgina Tech, the 7-seed in the ACC
  • NC State, the 14-seed in the ACC
I'd say the Gonzaga team total of 76.5 might be an attractive option. Point: Gonzaga

Gonzaga against teams that attempt and allow a lot of field goal attempts
  • Relevant UNC stats: 65.7 FGA for per game (4th), 59.5 FGA against per game (250th), total 125.2 FGA per game
  • WCC teams that push the pace: BYU (60.5 FGA for/64.0 FGA against), Pepperdine (58.3/60.0)
  • Non-conference teams that push the pace: Iowa State (62.5/62.2), Washington (62.9/62.1), Tennessee (59.2/57.1)

UNC is going to grab like 30 offensive rebounds tonight. That 67.9% defensive rebounding rate would be 328th out of 351 teams in the country. And if you factor out Pepperdine it drops to 64.3%, and if you drop BYU to just look at the three power conference schools it drops all the way down to 59.2%. That would be far and away the worst mark in Divison 1, It's fucking atrocious and given how much talk there is about Gonzaga's big men, I almost can't even believe it.

Point: UNC

Gonzaga against teams that rebound well (especially offensively) 
  • Relevant UNC stats: 58.2% rebounding rate (1st), 40.6% offensive rebounding rate (1st)
  • Teams with similar reb rate: St. Mary's (57.2%/2nd), Arizona (55.1%/9th), Xavier (54.4%/18th), BYU (53.5%/31st),  
  • Teams with similar o-reb rate: West Virginia (37.1%/5th), Washington (33.6%/25th), Xavier (32.6%, 37th), St. Mary's (31.3%/51st)
This is the big one. Essentially, if we accept the fact that UNC is going to crush the offensive glass, how goes Gonzaga fare in situations where they give up a lot of second chance points?

All you can do is beat the teams in front of you. Sure, SMC and BYU aren't the ideal competition to gauge a game against UNC, but we know two things about Gonzaga: (1) they're going to get KILLED on the glass tonight and (2) they know that, and they've still got a good chance to win. 

Point: Inconclusive

That's two points for Gonzaga, one point for UNC, a point split between them.