Why video analysis is crucial to analyzing pro hockey
Posted by Dean Holden at May 3rd, 2013
by David Staples, 30 April 2013
Craig Ramsay on watching a game live: “Your perception of what you thought you saw is often quite off.”
Craig Ramsay has been involved in pro hockey for more than 40 years now. He’s an NHL lifer, his latest job being an asst. coach with the Florida Panthers. In his long career, Ramsay both played for Hall of Fame coach Roger Neilson and also coached with him. As readers of the The Cult of Hockey know, I use video anaylsis of games to study the Edmonton Oilers, employing a method first developed by Neilson in the 1960s and 1970s.
A year ago I had a long talk with Ramsay about what he had learned from Neilson. I’ve finally found some time to transcribe that interview. If you’re interested in scoring chance and video analysis of hockey, Ramsay has much to say, including one key notion: if you want to understand the NHL game and fairly and accurately rate hockey players, it’s a mistake to go only by your initial, eyewitness take on the game.
Neilson coached Ramsay in Peterborough of the Ontario Hockey Association in the late 1960s and later from 1979-81 with the Buffalo Sabres, where Ramsay was an outstanding checking forward.
“I think he started trying to start at that time coming up with scoring chances for and against, to try to have some record to see how we played on a given night,” Ramsay said of Neilson in Peterborough. “You can lose a game two-to-one and play horribly, you can lose two-to-one and play great. Even then he was trying to come up with this concept of where to give up shots from, and where they were dangerous. We were trying to be good defensively without the puck.”
A fellow by the name of Frank Gurney did stats work for Neilson in Peterborough.
“Roger always tried to break the game down to the details in our end and their end. High third man, backchecking, more towards where you would stop in your own zone and who would cover what area. We tried to play not so much man-on-man. In those days the two forwards were supposed to stick with the two defencemen and the other two guys played down by themselves, three-on-three down low. He tried to help get where he had more of a five man unit in the zone defensively.”
When Ramsay got to the NHL, he found his old school pro coaches didn’t do the same kind of detailed work as Neilson had done in junior. “I was surprised. We just worked on basic breakouts, around the boards, hit the centre and go down the ice. With Roger we would do d-zone coverage drills. We would work on concepts for backchecking, and at the NHL level at that time that was pretty uncommon.”
Eventually, with the advent of video, Neilson took his work to the NHL where he became known as Captain Video. “Roger started to break down the opposition and break down into chances for and against for the team and for you individually.”
What did Ramsay think of that as a player?
“It can be a bit unnerving. You think you played a pretty good game and then you were minus two in scoring chances. You could look at them, it was much more difficult to look at them then. You had to go through the VHS tape, which was very time consuming.”
Every player had a differing take on the scoring chance data, Ramsay says.
“Some people took it very seriously and some blew it off. You talk to some players and some players like to just go play and some take everything seriously. There are players who can be adversely affected by statistics and information. I don’t think that it’s a good idea all the time for all the players.”
The way it worked under Neilson’s grading system was the only players who got a mark on a scoring chance, a plus or a minus, were the ones who had an impact on the scoring chance play. In this way, Neilson solved the then decades-old riddle of how to get a better plus-minus stat. The Roth-Irvin plus-minus system for goals — the NHL’s official plus-minus starting in 1967 — awards plus and minus marks to all players on the ice, no matter if they had something to do with the goal or not. About 40 per cent of the time, players do nothing to earn their plus or minus, often leading to misleading results, an issue that also confounds modern Corsi evaluation of individual players.
As Ramsay explains Neilson’s system: ”If you’re on the ice and you have nothing to do with the play, you don’t get a plus. On a given offensive play, there might be four people who made a really good pass, and they would each get a plus. Some coaches will say, ‘No, you can’t do that. You only get two assists, therefore that third guy doesn’t get anything.’ And that’s flawed. He might have made the best of the whole bunch, the defenceman who made that perfect breakout pass.”
And on a negative play, Neilson would only count up the players who made mistakes?
“Correct. Somebody who was directly involved with the play. Just like a goal and two assists, we want to have three people who at fault. You look at it and there’s not three people. There’s one guy who fell down and gave up the puck. Or one guy threw it the middle and got picked off. So that’s one guy. You just do it like that.”
Neilson also started at one time also broke it down into A, B and C chances.
Neilson went over the chances each night after the game. “Roger would do it and have it for you by the next morning. The man never slept. He slept about two hours at a time, maybe three. He would nap. He would be up at all times.”
Neilson would give the players the time of each scoring chance, so they could review their play. “Very few people did it,” Ramsay says of the players. “I wouldn’t do it either most of the time unless there was something that I was concerned about. I’d go ask him and he’d find it a lot faster. There was only one tape. It wasn’t like you had access to it unless you went into his office and asked to see it.
“For instance, I took a good shot in Philadelphia, beat the goalie and hit the post, and the next day on the scoring chances I had none. I said to him, ‘Hey, I hit the post man. It was just about in.” And he said to me, ‘No, you were outside the triangle, so it was no chance.’ But I was outside of his scoring chance area so it didn’t count.”
Neilson counted all goals as chances, Ramsay says, but he did not count missed shots as scoring chances. “If somebody throws a puck out and you’re 15 feet from the net in the slot and you slap it over the net, that’s not a chance. The reality is that’s a chance in my book. But in Roger’s book, and in many coaches book, that’s not a chance because it couldn’t go in, so it depends on your own particular definition of chances.”
A chance from the slot indicates a defensive break down, Ramsay says, so he favours counting some missed shots as chances. “Somebody had a flawed coverage, so you’re tricking yourself sometimes.”
Each coach has a slightly different definition of a scoring chance. “Most of them are about the same.”
Ramsay says a team has to be careful how it presents scoring chance data to players. For example, if a team is constantly losing while outchancing the other teams, that might be bad for a goalie’s confidence. “If you keep posting those things, your poor goalie is going to go, ‘Hey, wait a sec. Every game is my fault.’ That’s not what you’re trying to accomplish. You have to be careful how you use any information. You want as much as you can so you can make informed decisions, but some of them you keep among the coaching staff, some you post, some you have a meeting where you go over these things. Or you bring people individually or as a line. You don’t have to do everything in that very public nature.”
As a coach, Ramsay doesn’t use scoring chance data as much as Neilson did, but he does rely on video.
“Coaches who think they saw it all, and I’m certainly not one of them any more, you didn’t. Your perception of what you thought you saw is often quite off.
“Sometime you’re surprised when you back over a video. We struggle as a coaching group, you’re on the bench and you’re emotional with the game and you think you see something, you know who made the mistake and you maybe in no uncertain terms say something to that particular player, then you go and look at the video and it’s like, ‘OH, OH, that’s not what happened. This poor guy got stuck trying to save the day and couldn’t get the job done but it wasn’t his fault.’ So you’ve got to be careful as a coach during a game how you react to plays and situations that happened, because you find out looking at the video that things aren’t quite as you perceived them to be. But even looking at video, things happen so much different than when you watch them, you slow it down and think, ‘He could have done this and couldn’t he have seen that player over there.’ But you look at it five or six times to try to decide who made the mistake. What about the player who is out there and has a tenth of a second to make that same decision.
“A goal against ends up because the defenceman didn’t do a great job on a three-on-two but if you really look at it, some guy in the offensive zone got the puck and threw a blind backhand to the front of the net that was tipped away, now it’s three on two. The defenceman was looking for a different play, now he’s struggling to get back, and it’s all his fault. Well, the mistake was made long before the puck entered the net and it was probably made in the offensive zone, not in the defensive zone.”
Ramsay says most NHL teams now do scoring chance analysis. “The video systems now, you just come in after a period and hit ‘chances against’ or ‘chances for; and they show up. And they are basically just using Roger’s theories. Not every play at the net is a chance, it will pull up whatever your guy pulls up or sometimes there is a built-in system.
“Kevin (Dineen) does that just about every period. He comes in and he looks at the chances. We don’t have that much time. So the video guy will throw it up. We have three computers that we can look at. Kevin puts his up on the big screen and Gord Murphy and I would look at plays that we want to see on a computer. So we can do it bang, right away.
Players don’t see the chances between periods, not unless the coaches pull them in. “You’ll go and get a player and bring them in, some players like Mike Weaver, if there’s something that happened on the penalty kill, he might just come over. He wants to see it. Overall, players, in the limited time they have for most of them, they’re just trying to get some rest.
Only once this year did they put it on a computer screen for the whole team. “Overall, you have such limited time between periods that even with the intermission.”
What’s the use of the coach seeing it?
“We can tell them what to do differently. We don’t have to show them.
“You can draw it up on the board.
“So you look at them and, yes, man, you’re talking about the quality of the National Hockey League and how good these players are, you might want to adjust the penalty kill and personnel by one big step. So you want to see that. Now I got to move them one step farther out, over, sideways. But I’ve got to move them one step.”
Players now are more likely to review video information. “Coaches do it, but players will get their individual shifts, they can pull up ever shift for the game if they want to. They can look up anything they want.”
Quite a few players do this. “It’s funny to have a player sitting and looking at his own shifts: Is he really able to critique his own play. So we have individual meetings. We have team meetings. We have all those things. But players absolutely do look at the game themselves and try to critique their own play. They can be overly critical. They can be under-ly critical. The problem you have is if the player is if overly critical of himself. That’s not a good thing. All of this access to information, makes the game, the coaching…when I played we had one coach trying to do everything. How could he possibly? You couldn’t do the kinds of things we do now with three coaches, a video coach, as many coaches as some teams have. So you can overwhelm people and you can inhibit their play, so being critical all the time is a fault. It’s not good. But saying, ‘Oh man, we were great, all the time,’ that’s not good either. So you always walk a fine line of how you pass on the information that you get from watching the video.”
Back to Neilson and what he was like, Ramsay says, “He was a guy that wasn’t a great player. He had played some Junior B goaltending. He was really from baseball. He loved baseball first and foremost, and baseball has always been much more of stats-oriented sport.
“That was his favourite past time was actually baseball. He just was so focused. He had the biggest paper route. He was a teacher. He had four or five things going at a time. His mind was always working at something. I played four years of junior and sometimes we’d sit on the bus for a trip back after a game and he would want to go over our d-zone coverage and our penalty killing, and what can we do differently, and I’d think, ‘Not much, the guy took a fricken’ slapper from 60 feet and it went in, and there’s not a damn thing I can do about that.’ But I actually coached a game of junior for him because he was suspended and he felt I was the most qualified guy to do it since I had played four years for him and I knew what he wanted so I actually coached one game.”
“Roger, he wanted you to be good defensively but did encourage offensive gifts and offensive play.”
“I was really taught to think about the game and try to analyze the game, because, hell, I wasn’t that great a player. I was a little guy that wasn’t that good. I had to figure out ways to play against people who were better than me. So Roger really helped me to look at the game, analyze the game, and try to come up with ideas or concepts of where the puck was going to go so I could be there before the play happened and not react to it, but influence how it happened.”