Some clarity in NHL’s Sportvision use for goal, video review

A day after I posed some questions I had about how the new Sportvision technology would be used in video review of goals, the Washington Post gained some clarity.

WaPo wrote an overarching view on how the technology would impact the NHL and its players on and off the ice, but we are treated just to a small nugget of what it means for goal review:

“There may be times when we’ll see it and it hits off a goalie’s skate and crosses the line, and for whatever reason the camera angles didn’t get in,” Adams said. “I think it’ll aid in making that decision.”

In other words, the officials will just do their job like they normally do, then defer to the tracking system – like they currently do video replay – if the situation calls for it. For instance, did the puck get deflected into the goal from above the crossbar?

That’s Sportvision CEO Hank Adams speaking to the post in the quote above with a paragraph by Post beat writer Alex Prewitt, but the brief item doesn’t fully give a rundown of scenarios that are presented.

The NHL’s War Room has been known to call a rink and get the timekeeper to blow the horn in the middle of play to get the refs on the phone because they saw on their feed there was a good goal. The St. Louis Dispatch’s Jeremy Rutherford noted this procedural change as much in November:

Beginning with the Blues’ game in Boston, when a goal goes undetected by on-ice officials but is confirmed by the NHL’s situation room in Toronto, a horn will sound to stop play and award the score.

That situation unfolded in the Blues’ 4-3 win over Nashville last Thursday. Defenseman Carl Gunnarsson scored with 7 minutes, 15 seconds remaining in the first period. The goal was not seen by referees, or the players for that matter, and play continued until 5:28 remained in the period.

During the 1:47 of elapsed time, Blues goal judge Jim Kehm contacted Toronto to point out the goal, and once it was confirmed and a stoppage occurred, the horn sounded and Gunnarsson was awarded the goal. With the change, the game will be halted immediately after the goal is detected.

“I think if they know for certainty that the puck went into the net, they want to stop play and award the goal because what happens if there’s a penalty or something else happens, that stays on the clock?” Blues general manager Doug Armstrong said. “So once they know the goal is in, I think it’s smart: stop the play, award the goal, and continue play again rather than going like we did the other day with Gunny for over two minutes.”

You would think whomever is tasked with watching a certain game in the War Room that the Sportvision feed of the game showing the puck’s track will be automatically turned on. I’d imagine the guys in Toronto are watching standard replay anyway before they call the arena to stop play. This actually might speed things up.

Moving on to the other issue about high-stick calls on a puck deflected into a goal, it’s still not clear if the data can transmit height of the puck. The War Room is still stuck with the same camera angles unless everybody gets a GoPro, so while the Sportvision tracking should help in added information illustrating a deflection point, there is still some human judgment calls involved.

Man, could you imagine if this was around when we had to worry about crease violations?

Advertisements

What’s the role of NHL tracking chips in video review?

It’s a great day for analytics and general “wow” data in the NHL.

The league’s first game, albeit the All-Star Game, completed Sunday night with the Sportsvision chips in players jerseys and pucks. Think of this as a tactical version of FoxTrax from the ’90s—used only in replays on air that can show the route of the puck with a line drawing its path, speed of players and more. It’s not a glowing puck that’s on TV forever that sends off a comet tail when Shea Weber rips it for a one-timer.

A tremendous amount of information can be gleamed from the chips and cameras used in the arena, Sportsvision CEO Hank Adams told the Washington Post:

“We know where the puck is, if it’s over the blue line we know it’s over the blue line and for how long, very precisely,” Adams said. “When it gets into who’s in possession of the puck, that again has to be a human involved in it, because we don’t have electronics on the stick. There’s only so much you can use this stuff to replace. A human being’s going to have to get involved and sometimes make those judgment calls, but it can certainly aid and automate the collection of that data.”

So, it knows if the puck is over the goal line then. Can refs and the NHL War Room use Sportsvision technology to rule a good goal or determine a no-goal call?

The technology shows deflections, too. What comes to mind is a goal knocked in above the crossbar. If the chips reveal this much information, can it reveal the height? The opening of an NHL goal post is 48 inches tall. Factor in the thickness of the crossbar, which is 2 3/8 inches around on the outside. Can it tell where on the puck did a stick that could be above the crossbar? In other words, if part of the puck was above the crossbar  but the player’s stick deflected the bottom part of the puck that was at the height of the crossbar, can Sportsvision pick all that information up and be relayed in enough time for a decision in the NHL War Room?

As you read from Adams above, sticks are not tracked but you could have enough information to deduce if a puck was knocked in with a high stick.

If a chip can be placed in a player’s jersey, I would think there could be a chip or some sort of material taped onto a player’s stick that wouldn’t interfere with grip, shooting or balance that the infrared cameras can pick up and determine deflections. I really think if the NHL and Sportsvision has all the rest of this down, they’re not too far off in creating other solutions.

Heck, we’re really at the point where if the technology can advance far enough, you wouldn’t need a linesman to call offside or icing. Though that would never happen because one, what would whistle or sound off when a computer-detected offside or icing happens? An icing horn? Stop. it. now. And you do have a union for officials that certainly would fight for job protection.

Anyway, back to the task at hand: Should the NHL use its new Sportvision technology in determining good goals?

Yes. If you’re willing to go to TV replay to get the truth and run into issues of available camera angles, why not use new puck tracking technology to determine a good goal?

The Lightning saw Eric Brewer’s best and worst

Count me as one of the fans that is disappointed to see Eric Brewer be shipped out of Tampa to Anaheim.

I liked the leadership he offered and what seemed to be a steady influence, at the same time there was always the feeling that there could be more. Now, he’s in the Bygone Bolts category joining Nate Thompson on the Ducks.

Once Jon Cooper came in and managed with minutes in tandem with associate coach Rick Bowness, Brewer’s game was more in control and seemed fresher. Those mistakes that would creep up, especially under Guy Boucher, were minimized.

Somehow this season, it went off the rails. When a veteran player like Brewer gets scratched he either can play worse because of his attitude or motivated to do better. I would want to believe that the more you play Brewer the better he would be this year but he’s only played six fewer games than Garrison, Carle and and Stralman this year.

His regular stat sheet appeared pretty good for this year compared to previous years but the advanced statistics show he was a train wreck this year.

Coaching staff and the front office told Tampa area media that Brewer said he wanted to play but he was controlled in how he was being scratched. Between the salary constraints and his horrible stats for this year, it was justified to move him now.

If his performance was better than what we’ll see below, then it would have been an attitude problem. Here’s a snippet of how Barry Trotz explained how well defenseman Jack Hillen was handling his 19 games being a healthy scratch this season and what happens if he wasn’t a good teammate:

“What happens if you don’t have the good attitude is you start bringing other players down,” Trotz had said, “and when you start bringing other players down around you, it’s time to move that player out, because there’s nothing good that can happen.”

Speaking of the Caps, Brooks Orpik is a lot like a now more expensive Eric Brewer—both heralded for intangibles, but in advanced stats, they aren’t all that great. Orpik’s tradeoff that keeps him going is his jarring hits and physical play. Brewer lacked that grit and jabs after the whistle that could have elevated his game and value that made have played in his favor to hang around.

Horribly advanced

Brewer’s Fenwick, which accounts for all shots directed to the net that weren’t blocked, was the worse in his career topping the year he was traded from the Blues to the Bolts. He gave up about 12 more shots than he directed, according to stats by War On Ice. For five-on-five play, his offensive zone starts were about even this season, at 2 percent, a total swing from the -4 percent last season, meaning he had more defensive zone starts.

Revisiting that 2011-12 season, Brewer’s offensive zone starts percentage was a whopping -20 percent, explaining why that year was so atrocious (and why Boucher screwed himself).

Play with the charts on War on Ice and you’ll see a lot of red for Brewer buried with this year at the bottom of the charts.

Here’s just one of them that shows he was doing pretty well last season and then fell off a cliff.

This chart shows how many shots he was either giving up or getting on net compared to the time on ice of his competition. Basically, he was giving up a lot more in a shorter time up against his competition. Just play around with the charts on War On Ice and none of them look great, but comparatively, he was a lot better last season than this. Somehow getting bumped down on the depth chart is affecting him more than what it should. If anything it ought to have a positive effect on his stats, similar to how Mike Green, an offensive defenseman, is doing better this year seeing weaker competition.

Courtesy of War On Ice

Courtesy of War On Ice

Traditional stat pack

Measuring other areas of his game, the Bolts also saw Brewer’s best.

During Brewer’s St. Louis days, he played on a lean team and if you look at his stat sheet, it looked horrible for a defenseman if you take a peek  at plus/minus. Folks, during his days in Tampa, he enjoyed the most number of plus seasons with any franchise he played for. Out of the parts of five seasons he spent with the Bolts, only one season, 2011-12, he was a minus player.

Look at that roster and you understand why. Brett Clark was a minus-26 and the rest of the cast was spare parts from Breden Mikkleson, Mike Commodore, Brian Lee and Bruno Gervais to Pavel Kubina, Matt Gilroy and Marc-Andre Bergeron who had 24 points in 43 games, enough to lead the defense corps in scoring in an 82-game season. That should tell you how bad that year was.

When Brewer left Tampa this year, he was plus-5 and had four assists in 17 games. That’s better than the minus players Radko Gudas and Mark Barberio (minus-2 each) and the goal and assist Gudas put up. Barberio is scoreless in his seven games.

Brewer was also averaging 22.7 shifts per game and 17:50 in ice time this season, about 20 seconds above his average last season with one fewer shift. Both of those numbers are way down from the 20 minutes and 27 shifts he averaged in the lockout shortened 2012-13—also the final season of Boucher.

Lightning top league standings, have more work to do

The Lightning are settling in to where those offseason expectations put them in—atop the NHL standings. At least for the night.

Scoring, goaltending and for the most part, solid team defense is clicking. It looks like the second period needs some cleaning up, especially when it comes to penalties and killing them successfully.

Here’s a look at team stats for the Lightning in order of worst to first:

  • Penalty Kill Percentage: 76.1 percent—25th in league
  • Power Play Goals Allowed: 11—24th in league (tied with four other teams)
  • Shots per game: 30.1—16th in league
  • Goals against per game: 2.67—14th in league
  • Face-off winning percentage: 50.6 percent—13th place
  • Leading after second period win percentage: 88.9 percent—12th place
  • Winning percentage when outshooting opponent: 70 percent—9th place
  • Leading after first period win percentage: 83.3 percent—7th place
  • Shots against per game: 27.4—6th in league
  • Scoring first winning percentage: 80 percent—5th place
  • 5v5 goals for/against ratio: 1.38—4th in league
  • Trailing winning percentage: 60 percent—4th place
  • Power Play Percentage: 26.4 percent—3rd in league
  • Goals per game: 3.80—2nd in league
  • Power Play Goals: 14—Tied for 2nd in league
  • Winning percentage when being outshot: 75 percent—1st place

The takeaway for me is that the Bolts need to close the door a bit more, especially late in the second and into the third. They allowed Detroit to creep back in but as the stats show, they were able to overcome that deficit and win.

Not every team is going to be perfect and some of these stats offset the others, allowing a team to cover up for its mistakes and shortcomings. Take a look at the penalty kill. It needs huge improvement at 25th in the league, but I was surprised that the Lightning only managed to score one shorthanded goal with how aggressive the forwards are.

Brian Boyle popped in that shorty on Thursday night against the Flames where the Lightning earned several other short-handed opportunities and a couple breakaways thanks to the drop-pass used by the Flames breakout. However, after Boyle scored the shorty, they allowed a power play goal 20 seconds afterward. Buzzkill.

Short handed goals are a bonus when killing penalties and at this point, the unit needs to not be aggressive to the point where they’re looking to strip a puck and go the other way with it. Instead, they need to still limit space but focus on getting a stick on the puck enough to clear it and change.

The Lightning are taking 9.8 penalty minutes per game, which is the eighth lowest in the league, so it’s not like the team is on a parade to the box. That’s 66 penalties in 15 games. For comparison, Winnipeg has the most at 90, which includes three misconducts

Here is how the Lighting penalties per period break down after I went through each box score for the season (NHL doesn’t keep track):

  • 1st — 17
  • 2nd  — 36
  • 3rd  — 13
  • OT — 0

Remember that some penalties offset like roughing and the five fighting majors the Lightning have on record. Three of the fights occurred in the first period and two in the second period, according to HockeyFights.com. Maybe it’s the long change, but couple the a long change with a penalty kill and you can see where things can get hairy. Seven of the goals allowed on the PK also occurred in the second period by my check and then three in the third period and only one in the first.

If the Lightning want to be ultra stingy, it’s clear that second-period penalties need to be cut down along with snuffing out power plays in the second period. (Maybe no coincidence that the second period typically has more goals scored in the league.)

These numbers also bleed another way in that refs put their whistles away in the third period sometimes (and too many in overtime) and the first period seems to be standard.

At 15 games in, the Lightning are showing what they’re capable of and hopefully the team continues to do its homework for the rest of the season to become an elite first-place team and not a San Jose Sharks/Washington Capitals first-place team.

Lightning’s year-end stats: What to improve upon for 2014-15

Embed from Getty Images

A four-game sweep in the playoffs means that the Tampa Bay Lightning have some evaluations and work to do for next season.

Lightning general manager Steve Yzerman wouldn’t divulge what he wants to fix and improve upon other than grit, but he did tell reporters Thursday that he’ll look over stats with head coach Jon Cooper:

“Our plan is next week is to sit down and kind of go over our team. Look at where we were statically throughout the season, our goals against, goals for, power play, penalty killing–kind of go over everything and decide what we really need to do. I have some thoughts, I prefer to keep those to myself for the time being. We do have to improve in certain areas. Over the course of time, change our team a little bit. I prefer not to say that at this time.”

Let’s pause for a second and check in on Yzerman’s areas of review that he identified. The critical areas for improvement are highlighted in red:

  • Goals scored per game: 2.83 (9th place)
  • Goals against per game: 2.55 (11th place)
  • 5-on-5 goals for/against per game: 1.16 (7th place)
  • Power play: 18.5 percent effectiveness (13th place)
  • Penalty kill: 80.7 percent effectiveness (23rd place)
  • Shots for per game: 29.8 (18th place)
  • Shots against per game: 29.2 (13th place)
  • Winning when scoring first: 73.8 percent (12th place)
  • Winning when trailing first: 37.5 percent (5th place)
  • Winning when leading after first period: 72 percent (21st place)
  • Winning when leading after second period: 84.8 percent (16th place)
  • Winning when outshooting opponent: 57.8 percent (7th place)
  • Winning when being outshot: 58.8 percent (10th place)
  • Faceoff win percentage: 49.2 percent (19th place)

What  does this mean?

The most disturbing stats above involve losing the game after the Lightning take a first period lead and a second period lead. The team gave up the first goal more often than not, showing why they were able to be fifth in the league when trailing first. The Lightning were more comfortable winning when they were losing and not keeping a lead.

Some of that can be explained in the shots per game stat. That’s now shots against, that’s shots for. Martin St. Louis was leading the team, and finished with 204 when he reached New York. Still, that was only good enough to be 61st in the league with other teams having multiple leaders rank higher. Tyler Johnson’s team leading 181 shots was only good enough for 92nd in the league with Alex Kiilorn next at 107th with 173 shots.

There were 10 defenseman who had more shots than Johnson this season. Victor Hedman had the 18th most shots for defencemen in the league. LA, Washington and Phoenix had two defenseman reach rank higher than Hedman. That’s not to place the blame on Hedman, who ranked 9th in goals by a defenseman with 13. Other defencemen need to get shots through, instead they were opting to pass the puck instead to a forward, especially Radko Gudas with his heavy shot.

Throughout the season, Bobby “The Chief” Taylor would gripe, gasp and groan over the players, especially young ones, passing up an opportunity to shoot in favor of a perfect chance, a perfect shot. They were waiting for the Grade A scoring chance on the first try and sometimes when they hesitated, chance A disappears and so does B and C. When that happens, they end up over-passing the puck and now now scoring chance presents itself.

Instead, put the puck on net, and then you’ll get rebounds and deflections that would produce multiple B, C chances that end up in the net.

That has to be nipped in the bud immediately. When the Washington Capitals came up reloaded with then-young talent Alex Ovechkin, Niklas Backstrom, Alex Semin and Mike Green, the team was called the “Comeback Caps” because they could somehow claw out a win when trailing in a game, especially late. The theme evolved that while the Caps were a comeback team, players kept saying how they had difficulty closing out a team in a game and not allowing them to come back in. The team would relax with a lead or try too much offensively instead of going into a defensive shutdown mode that dominant teams were able to do.  Hire a sports psychologist–do whatever you need to do, but learn the mental tools to fix this.

It’s exciting hockey when they’re able to comeback, but it’s even more exciting when you see those wins pile up because you didn’t let teams come back into the game.

Getting back to basics, let’s look at face-offs. The play starts at the face-off dots. You lose the puck, you lose possession, thus you give up goals.

Only Valtteri Filppula and Tyler Johnson really shined, with Filppula checking in at No. 19 for most face-offs won in the league. Tyler Johnson led all rookies in face-offs, wins and percentages and won the 34th most faceoffs in the league for any player, signaling some promise for the future. Tom Pyatt led the team in percentage, but only took 101 face-offs. Only him, Nate Thompson and Filppula were above 50 percent.

A healthy Stamkos should improve this, but it doesn’t take care of the bottom six. Alex Killorn has bounced from wing to center and of all the forwards, his face-off winning percentage at 43 percent is pretty bad for how much he was used.. The team will have to work on winning the face-offs as a team, getting good position in the circle while the center tries to legally tie up his man and use his skates to knock the puck back.

The face-off issues could explain some of the reasons why the penalty kill is so bad in 23rd place, but when it’s that bad, more needs changed than winning the face-off. Somebody is drawing up a new PK in the offseason. That can be guaranteed.

The powerplay was pretty good at 13th, but the mark teams aim for is 20 percent effectiveness. It will be interesting to see where that number trends next year. This team was able to maintain that mark without Steven Stamkos, thanks in part to Martin St. Louis, but then when Marty left, Stamkos was just kind of there caught looking for that perfect pass that rarely came on the PP.

On the positive side, not only did the team scored enough goals to be in the top 10, but also was 11th in goals against average. Room for improvement? Sure, but it’s not something to be overly concerned about when 16 teams make the playoffs and you’re 11th in that category. The goals surrendered this season hasn’t been an issue of how many but instead, when they were surrendered as the stats showed above.