After Saturday’s 7-1 loss, Minnesota Wild fans have gone into full blown panic mode. Many drunken howls for the heads of Mike Yeo and Chuck Fletcher can be heard if you put your head out the window late at night and listen hard.
Fans who are upset aren’t in the wrong. Expectations for this season were high and things are going very poorly. The team’s goaltending is a mess (and hasn’t improved since that linked article), they haven’t won back-to-back games since November 16 and 20, and the wins we’re celebrating are victories over the Toronto Maple Leafs, who have the third worst Corsi For in the NHL and dominated the Wild in shots and possession, despite losing the game.
So, what happened?
The Good, the Bad, & the Wild
The common wisdom seems to be that the Wild weren’t winning enough games early in the year. A.611 point percentage from the beginning of the season through November 20 — the back end of their last back-to-back wins — isn’t President’s Trophy material, but it’ll get you into the playoffs, so it wasn’t exactly a disaster.
Overall, in the early going, there were a couple things happening that were good and a couple that were bad.
- The Wild were getting, with the exception of the opening four games, some sub-par goaltending from Darcy Kuemper and Niklas Backstrom.
- The power play wasn’t connecting. It took the team until November 1 to score their first power play goal and it wasn’t exactly as though the dam had broken open at that point. However, I think you can — and probably should — argue that it wasn’t as bad as it seemed. They were taking more shots than most teams and doing a lot of things right. The goals weren’t coming, but systematically it wasn’t the tire fire it was being made out to be. But in the end, they weren’t scoring at all on the power play and that’s what counts and that’s what hurt them.
- The team was dominant in terms of possession. Through November 20 the Wild had the second best Corsi For in the NHL at 54.9%.
- The team was dominant in terms of shots for. It wasn’t until November 28 that a team outshot the Wild. That means they outshot opponents in 21 straight games. They were the only team in the league to do that. By the beginning of November every team had been outshot at least once, except for the Wild.
- The team’s shot suppression was excellent as well.
What’s happening now?
- The Wild’s goaltending is still terrible. Out of all NHL goaltenders who have played at least 300 minutes (60 total), Kuemper ranks 50th and Backstrom ranks 54th in 5-on-5 save percentage (SV%). In adjusted SV%? Kuemper ranks 53rd and Backstrom is 55th.
- They aren’t getting as many shots. Up through November 20 the team was averaging 24.5 shots per game at 5-on-5. Since then, they’re down to 22.6.
- They’re allowing more shots. They’ve been outshot regularly since November 20. From the beginning of the season to November 20 they allowed 18.5 shots per game at 5-on-5, the fewest total shots against at 5-on-5 of any team up to that point. Since then they’ve allowed 23.6. If you look at the bullet point above there’s a glaring problem and it’s the shot ratio. They went from taking 24.5 and allowing 18.5 to taking 22.6 and allowing 23.6.
- Their possession totals have dropped off a cliff. Once at 54.9%, the team is the owner of a 49.9% Corsi For since November 20. That’s a fall from second in the league to 16th. In score adjusted Fenwick, the Wild ranked third at 54.8% up until November 20, since then they rank 20th at 49%. Their record has correspondingly dropped. Up through the 20th they had acquired 61.1% of the available standings points. Since, just 47.4%. 61.1% of the points throughout the season gets you 100 and gets you in the playoffs.
- The power play is better. It’s not really a problem any more. It’s not the best around, but it’s vastly improved.
- The penalty kill has really been decent all year. The team is currently ranked eighth in the NHL on the PK with a 84.2% success rate.
You can add to that pile that the team’s scoring rate has been pretty garbage lately too, as highlighted by Micah Blake McCurdy in the below graph.
The biggest issue here: the possession totals. With how the team played in the early going, it may be a fixable problem as well. Goaltending would be great to fix, but the Wild kind of have the goaltenders they have right now and that’s highly unlikely to change. The shots are going to be fixed with better possession. Power play is functional again, but if you’re scoring at even strength, that’s less of a concern.
Some might call smoke and mirrors on the team’s possession totals after the team started last year looking dominant in possession and then trailed off, ultimately ending the season as a negative possession team. The biggest difference this year is that they aren’t playing low-event hockey. There are a lot of shots happening, increasing the sample size and its reliability.
Why Focus on Possession?
Why is Corsi and/or Fenwick so important? Because it’s the best predictor of future success we currently have. No other set of data from a team is as useful in predicting future success. Scoring chances, goals, save percentage, shots, power play success… none of it statistically functions as a better predicator of future success than what we call possession metrics but are actually just shooting attempts1.
Corsi and Fenwick tend to be more predictive because they create a larger sample size, increasing the accuracy and removing room for error. The team’s drop in possession numbers is worth focusing on because it’s the most correctable of all of these problems and can lead to improvement among the other numbers.
Going back to the start of the season, many fans will remember Michael Russo’s piece on how Zach Parise, once a defender of dump-and-chase, saw a study looking at zone entries that concluded dump-and-chase isn’t the most effective method of entering a zone. Teams tend to gain better possession, take more shots, and score more often when they carry in more often than dump in.
The Wild did just that to start the year. They were carrying. In my game notes I keep, it felt as though I was taking a note each time the team dumped because it was happening so rarely compared to past systems under Yeo.
That strategy seems to have gone to the wind about the time the team starting losing. In recent games the team has sacrificed possession of the puck over and over by dumping the puck in and losing the race to the puck. I don’t know of anyone tracking entry zone data on the Wild, so I can’t display the dip that’s happening on carries versus dumps, but it’s not a gradual thing. It’s very clear that there’s been a shift in focus.
It seems that a combination of a few losses and a four-game losing streak in early November had the team losing faith in the adjustments made to the team’s system that focused on possession, attempting to carry the puck into the offensive zone instead of dumping it being a notable part about that. Or maybe the coaching staff lost faith in those adjustments. Either way, there was a lot more dump and chase happening and a lack of focus on doing the things that had the Wild controlling play early in the season.
Is this a #FireYeo situation? No.
That’s pretty reactionary and won’t solve the team’s problems. There’s enough blame to go around. The players, the goalies, Chuck Fletcher, they all deserve a slice of the pie. Though I do question some of the personnel decision being made — playing Kuemper in back-to-back games when we know that it’s basically always a bad idea, playing Stu Bickel against Dallas over Justin Fontaine who has the team’s best usage adjusted Corsi For among forwards, or benching Erik Haula during the roster freeze and playing Bickel instead. (Yes, Haula needed to be sent a message, but a struggling Haula is still more valuable than Bickel and the message can be sent after the roster freeze so you can move someone else into the lineup.)
Nonetheless, this team isn’t at a point where firing Mike Yeo accomplishes anything. Their record is significantly better if they got even league-average goaltending. Their record is probably better if they didn’t have crap luck with illness this year, with the mumps and a Norovirus ravaging the blue line. (Also, injury stripped the team of their third pairing left shot defenseman, likely for the rest of the year2.) But these are the things that happen during a season and a lot of the things that happen during a season are out of a team’s control. Teams have to win anyway.
One way of trying to regain control of the season is getting back to a style of play focused on possession, one the Wild proved they could be successful at early in the year and can help the team start to earn points again. If this team doesn’t want to play the Connor McDavid sweepstakes3, things need to be corrected quickly. It’s never easy to right the ship and it’s going to be especially tough with the team’s next five games being vs. San Jose, vs. Chicago, vs. Nashville, at Chicago, at Pittsburgh.
Here’s one way to think about how this has to go for the Wild to get into the playoffs: The Wild grabbed the top wild card spot last season with 98 points. Dallas grabbed the second wild card spot with 91. (We’ll use that number as a spot where the Wild have a good chance of getting into the playoffs as it’s the only season where playoff berths were gained under the new system.) The Wild have 45 games remaining and would need to get 51 points in those games to get to 91 points. So they could go 23-17-5 and hit that number. That’s a lot better than they’re doing right now and we’re setting the bar pretty low. There’s a fair chance 91 points wouldn’t be enough. They really need to be even better than that. 97 points would probably get them. So let’s adjust that record to 26-14-5. (Or 25-13-7, whatever combination you want to use to bring them up to 97.)
There’s a lot of work to be done if the Wild are going to turn in a record like that. Getting back to a game that focuses on possession would go along ways toward that end.
1 I’m, for convenience going to continue saying possession, but I think there’s probably a need for a better term there. It does measure possession, but it’s imperfect and we actually know what it is measuring more precisely than possession. Anyhow. Discussion for another day.
2 With Michael Russo’s great piece on Zach and JP Parise reminding us that this is just a game, it’s worth noting here that we all still wish the best for Keith Ballard‘s recovery. I get a little sad every time I think of the quote he had in the offseason talking about how he doesn’t want concussions to wind up causing permanent damage to the point where it would effect his relationship with his kids.
3 A game they shouldn’t play because the odds of them winding up with one of the top two picks is nill. This is a deep draft, but if you aren’t tanking for Connor McDavid or Jack Eichel, it’s not worth tanking. Dylan Strome, Noah Hanifin, and others are great players to draft, but not worth throwing a season away over. So before you get too excited about #DishonorForConnor, know that that’s another game the Wild will lose.
Bonus. Here are a couple of other graphs from Micah Blake McCurdy that he tweeted out yesterday that help visualize the Wild’s decline this season. In this one, SVO adjustment is adjustments for score, venue, and opponent.
25-game rolling averages of SVO-adj possession over the past (calendar) month. pic.twitter.com/CbF1Lv8cow
— Micah Blake McCurdy (@IneffectiveMath) January 4, 2015
Scoring rates over past 70 days; 25-game rolling averages. pic.twitter.com/7i86V3lJiO
— Micah Blake McCurdy (@IneffectiveMath) January 4, 2015