Breaking Down: The Start, Then The Stop

With the Avalanche set to return to the realm of active NHL teams, this is a good time to reflect on the first fifth of the season and gain some insight into what has been working and what has not. Depending on how one looks at the standings they were tied for 1st in the West Division before the hiatus and are now either 3rd (points) or 2nd (percentage). In a normal season we are around the point where teams have generally settled into where the standings will end up. This isn’t a normal season so who knows.

I’m going to focus on shot rates in different score situations to get the big picture. I don’t like percentage stats, which combine things that happen at opposite ends of the rink, because they don’t seem to relate as well as many think. So for that we’ll look at rates for and against separately. I also don’t have faith in how score effects are algorithmed out by those that try to do so. We’ll look at the raw rates in three different score situations: leading, trailing and tied, and again separately. At the team level we want to see what they are doing differently and at the player level we want to see who is doing their job well and perhaps who is gaining trust from the staff.

The Team

The Avs are playing slightly less at 5v5 this season than last, around a minute and a half, which points at an increased rate of penalty calls. The percentage of time lead/trail/tie is pretty similar with the lost minute and a half coming almost entirely out of time spent leading.

What is different are the shot rates. Colorado was a lot more high-event last year and the difference is completely from suppressing the opponent’s shot rates, both quantity and quality. It’s very impressive. With the lead they’ve gone from allowing 59.99 shot attempts against per hour to less than 43. While tied they are allowing 15 fewer attempts per hour. There’s a little bit of sample size noise and bad opponents (hello Anaheim & San Jose) as a driver here but slashing those numbers so much is a large change.

In a nutshell, the Avalanche are even more high-event penalty-wise, they are playing a little less with leads and have suppressed opponent shot & quality rates extremely well. This is pretty sustainable stuff and adding in a little bit of bad shooting luck in certain situations means that there’s a fair amount of upside still. Scary.

The Players

This is a bit complicated so I don’t blame you if your eyes glaze over. I’ll run through what I’m looking at and why but the conclusions are the important part. First off, we’re focusing on the 18 skaters that have played in at least half the games. There are 9 others who have started for Colorado this year but realistically haven’t had much of an effect. I’ve taken the on-ice shot attempt and expected goal rates, both for and against, and via some math converted them to plus or minus the team average for that category. There’s also time on ice as a percentage of what’s available in each score situation. What I was hoping to gain was insight into what the coaching staff does in those situations, part of the “trusted player” deal, and how each guy influenced play. We’ll start with all situations then go into lead/trail/tie.

All Situations

See, that’s not so bad, just a friendly rainbow vomit chart. For reference blue is good, red is bad and the more intense the color the further from team average. The players are racked by percentage of TOI available.

General observations are that the guys that play a lot drive a lot of offense and are good shot quantity suppressors. The fellows that don’t play a lot don’t drive offense but suppress quantity and especially quality of shots. I think the staff are reasonably happy with this setup. I’m not much of a fan despite it’s effectiveness. You basically have guys that are supposed to score and then guys that just clog or whatever. It’s too specialized and injuries or the way hockey is played in the playoffs expose problems like lack of depth production from the bottom 6.

Individually, one of the first things that stands out is that Kadri and Burakovsky’s numbers are similar and generally suppressive. The main criticism of both guys has been a lack of offense so far but they are holding their own defensively. Their usual linemate, Brandon Saad, has poorer numbers so whatever he’s been doing away from those two it’s not good. Matt Calvert is playing far too much with how little he generates in the offensive zone and no real suppression of the opponents. Tyson Jost has been really good defensively, it remains to be seen if he can gradually learn to drive some offense in this role.

Play while leading

Having a lead, which the Avs do 44% of the time, means that the staff don’t have to play the 3HM quite as much, evidenced by MacKinnon trailing Jost in TOI. Their shots against and quality against is pretty disturbing and being high-event in general makes it look like they’re completely reactive. It also means some guys that aren’t as trusted can get some time like the 3rd pair of Graves/Timmins playing more than Toews and Byram. The bottom of the chart is highlighted by a bunch of guys that suppress the opponent well but it comes at the expense of generating zilch offensively. Lots of this points to sitting on a lead, which by the eye test hasn’t happened much, so it’s more likely they sit on BIG leads to get these numbers.


The Avs are good so they don’t spend a lot of time trailing, around 8:28 per game at 5v5. Right off the bat I’m like why is Matt Calvert playing 40% of the time when they need to score a goal? Especially with an expected goals for number like that. This is one of the cases where the “trusted player” regime hurts the team. Now he’s only played 6 games and the TOI is a little skewed since they were trailing more when he was in the lineup (which is another topic unto itself) but that’s not helpful usage at all. Bellemare is the same situation. I get it, those guys bring energy, but they don’t bring offense and there’s no way they should be playing more than your 2nd line, even with their warts. Donskoi’s numbers are very troubling despite being one of the 2 forwards that has scored a goal at 5v5 while trailing. The other is Mikko and take a look at his attempts for/against, it’s incredible. Playing while behind is not something the Avs have to deal with a lot but both the staff and the players need to be better here.


Looking at the tie numbers is odd because there’s really two different situations wrapped up in one. There’s the initial game state tie, which is usually pretty fast, and then the late game playing for OT tie where both teams play not to lose.

Everybody looks like pretty much who they are here. The lines rack up exactly 1-4 by TOI, the 3HM and top 4 D are all effective and the bottom 6 is a nightmare. Spotlight on JT Compher, who is bottom of the barrel in 3 of 4 categories and aggregately terrible on both offense and defense.

The Avs forward depth is largely an illusion. Take any one player in the bottom 6 in isolation and they’re all justifiable but together they really don’t mesh well and the lack of a strong center on either line makes doing anything productive difficult. The best you can hope for is clog, suppress and kill time until the top line comes out again. Not much of a bargain for over $15M in cap.

The New Guys

The Avalanche kept their forward corps largely intact over the off-season with the only change being acquiring Brandon Saad and losing Matt Nieto to free agency. The defense has been almost completely made over, adding Devon Toews in a trade, graduating Bowen Byram and Conor Timmins and losing Ian Cole and Erik Johnson (for now). Ryan Graves’ role has changed significantly from last year. Let’s take a look at how they are used and whether they keep up.

Brandon Saad missed a bunch of camp with C19 protocol so it’s no surprise it took a while to integrate into the Avs system. Gradually he’s settled into the 2LW role and 2nd PP unit with a little bit of PK thrown in. After a slow start he blew up with 7 points in 4 games vs SJS & MIN. That’s what he’s here for, he’s not going to drive play and the stats bear that out. I’m fine with that, he’s good at what he does and as long as his linemates complement him this will go well.

The Devon Toews trade looked good at the time but it really paid off as soon as games began. He fit in perfectly with the team’s style right away and this is not really what he was with NYI so the scouts deserve a shoutout. The thing that jumps out from his shot metrics is that he’s consistently but slightly suppressive in both offense and defense. This is similar to Sam Girard and I have a feeling it has to do with playing with all the lines a lot rather than just with Makar/3HM. He is already very trusted by the staff.

Bowen Byram missed camp and the first week of the season from quarantine after a strong World Junior Championship performance. Like Toews, he’s been a revelation on the blueline. He doesn’t look 19 at all and doesn’t play like a teenager either. Usually rookies take a while to acclimate to the speed of the NHL and generally suppress offense at first but that’s not the case here. It’s entirely the opposite. He has an extremely positive effect on offense and his only real wart comes from opponent shot rates while the Avs have the lead, which by the way mirrors the 3HM. The staff have limited his minutes late in games with leads a bit so he’s not entirely trusted yet but there’s no doubt that will come.

I’m surprised Conor Timmins has more or less cemented a spot in the NHL at this point. The Avs seem like they really wanted to open the door for him which is a little contrary to what Joe Sakic said in the off-season. With EJ injured he’s solidly in a bottom pair role and gets 2nd unit PK time. Unlike Byram, the rookie on-ice offense suppression does manifest in his stats but playing mostly with the bottom 6 forwards that’s not all his doing. The trust and usage he gets from the staff is inconsistent but generally low, which is what you would expect at this point. The best thing for Timmy is gaining this experience so let’s see what he does with it.

Ryan Graves isn’t new but his role sure is. He gets a bad rap from a lot of folks but being a benign passenger on the Makar/3HM train isn’t as easy as it’s made out to be. This season he’s been pushed down to a bottom pair role while anchoring the penalty kill. His PK work has been fine like it was last year so that’s a good thing. His numbers 5v5 are pretty ugly but once again they mirror most of the bottom 6 forwards he hangs with now. This is a tough spot for Graves, role is reduced and trust isn’t what it was, even leading to a healthy scratch earlier in the year. I would love to see both Graves and Timmins get a little confidence in their game, along with the more useful bottom 6 guys, and start producing, they’ve got the ability.

The Thousand Yard View

Colorado is one of the best teams in a weak division and barring catastrophe will make the playoffs easily. They have one extremely good forward line, 2 very good defensive pairs and a goalie that looks like he’s jumped up a level. As much as I love delving into the numbers the major obstacle to winning the Stanley Cup this season is going to be health. That doesn’t mean looking at very specific stats is pointless, I like to know why things work or don’t and what conditions influence performance. This is a far cry from 4 years ago when we didn’t need to break down things so minutely to examine the issues. The Avs need tweaks, not excavations. The staff can get away with self-indulgent behavior like trusting ineffective yet experienced players because that can be made up in other areas and the overall talent level is so much higher now. We’re in a good spot, just had a nice long rest and have another 45 (hopefully) games to dial in the weaknesses. Hit it.


Scoring LW, punchy climber for the Ardennes classics, spirit guide

One thought on “Breaking Down: The Start, Then The Stop

%d bloggers like this: