With another year complete, Flames Nation is in the midst of the post-mortems. Over the next couple of weeks, we’ll look at some interesting EV data from timeonice that might shed a bit of light on why the season ended the way it did. First up, overall team performance and a look at Calgary’s number one goalie.
I had a sense for most of the year that the Flames were a decent team, and likely a playoff team on balance, even if they were clearly outside the elite of the Western Conference. They ended the season in a slightly worse spot than that, so it seems reasonable to investigate potential reasons for things going sour.
One of the best ways to discern team strength is to look at EV shooting data when the game is tied. That eliminates any score effect, since teams that are trailing invariably attempt more shots while leading teams sit back a bit to defend that advantage. When the game is even, though, we get more light and less heat, if you will.
With that in mind, I collected the EV tied data for the top ten WC teams for this year from timeonice, so that covers the Canucks down to the Flames. The data collection scrapes away all empty net situations, but does include 4v4 play.
First, let’s look at the general team numbers for all the squads, and I’ve arranged them by their Fenwick ranking, since we don’t have scoring chance data for all the teams and Fenwick % matches to chance % slightly better than Shots% or Corsi%:
The Flames were, at least by this metric, clearly ahead of Dallas, Phoenix and Anaheim, and slightly ahead of Nashville and the Kings, with a clear gap between Calgary and the elite of the conference. JLikens covered off in some detail how these metrics correlate with successful team performance here and here, so based on that criteria, the Flames should have expected a finish in the 5th to 7th place range.
Of course, as we all know, stuff happens. Percentages aren’t always equally fair, so a team that isn’t very good on the face of it, like the Ducks, can ride hot shooting and a superb PP to home ice in the first round. The Ducks’ PP numbers have some backing, but at EV that just isn’t the case, and EV play is about 75% of the game. The other obvious imposter here is Dallas, and the fact that they missed the post-season is nothing more than justice.
The percentages don’t just inflate crap teams, either. They can accentuate differences between good teams as well. Here are the EVSH%, EVSV% and PDO for all ten clubs, with the ordering by SH%:
As I said, the Ducks really rode the bounces, and the Hawks were openly unlucky. We often talk about PDO numbers ( the combination of EVSH% and EVSH%, expressed as a number based from 100), and if you add these totals, you might also note that Vancouver, beside being good, rode fate a touch as well. .939 and 8.2 gives a PDO number of 102.1. That’s not ridiculously high, but it’s high, and the Canucks took advantage of good play and good fate to amass a goal differential of +23 when the game was tied at EV. Calgary’s full season differential in all situations was +11. Must be nice.
Calgary’s PDO number was 98.1, and the Hawks were the only other team below 100 at 98.4. Even with the low PDO number, I suspect that a Hawk team that had been able to dress Dave Bolland before last night wouldn’t be down 3-1 to the Canucks at the moment. Of course, if Quennville would have clued in as to how bad Marty Turco was a bit sooner, they wouldn’t be facing the Canucks, but that’s another story.
As for the Flames, the numbers paint an interesting picture. The team was a decent outshooting club, and weren’t the worst SH% club when the game was in the balance, which does seem to point at one particular area for further examination.
The numbers don’t often lie, and that EVSV% was a killer. Miikka Kiprusoff seems to have a certain teflon quality amongst the fan base and the local press, but his personal performance is likely what separated the Flames from the post-season.
Now, before any of you head for the hills yelling “shot quality”, I’d like to mention two things. First, if you can prove it exists in any meaningful way, head over to Behind the Net and relieve Gabe Desjardins of the 150 bucks. I’m certain he’ll be happy to hear from you.
Second, even if there is the potential for high shot quality against to have affected Kipper, the one place that collects that sort of data, Hockey Numbers, doesn’t help your cause. Chris Boersma’s info suggests the Flames were the third best team in the league at limiting shot quality at EV. The team also allowed the 4th fewest shots/60 at EV, so again, he wasn’t exactly peppered with an unstoppable amount of rubber.
The data is really grim when you examine how Kiprusoff stacks up against the other regular starters in the West. Here are the EVSV% tied numbers for the regular goalies on the ten teams in question:
Ouch. That sort of performance just won’t do, and it marks the third time in four years that Miikka Kiprusoff has been substandard. A careful reader might also note that his cap hit is higher than every last one of those other gents, and he has three years remaining on his contract. Sigh.
To put these percentages in some context, the Flames allowed 676 shots on goal when the game was tied at EV. If Kiprusoff was as good as the goalie that finished next to last, Jimmy Howard, he would have permitted nine fewer goals in that game state. Normally, a nine goal cut in differential would result in three points in the standings, but I have a sense that nine fewer goals allowed when the game was tied might have a slightly greater impact. Oh, and on the list of things I never planned to write, the suggestion that the team would have fared better if Kipper had only been as good as Jimmy Howard would be right up there.
One final set of numbers seems worth looking at. I wanted to break down the year into three parts, largely because the season seemed to have three distinct sections. The first was the opening 35 games that had the world screaming “rebuild”, the second was the 30 game segment that calmed the waters, and final section was the scuffling 17 game finish from the March 2nd loss in Chicago onward. Here’s Miikka Kiprusoff’s full set of EV tied numbers in that order, and note that all figures are for when he was in net, so anything where Karlsson was playing has been excluded.
That first set looks a lot like last year, doesn’t it? Kipper was playing well, and nothing went in at the other end. Those outshooting numbers were quite good, so the suggestion by a few of us that the team wasn’t a 14th place club was on fairly solid ground.
The middle section was where the club took off, and even with the juicy SH%, they weren’t that bad of a value for their record. On a personal level, Kipper really struggled in mid-January before having a good finish to that stretch.
In the final set of games, losing Morrison and Moss certainly didn’t help the outshooting numbers, but it wasn’t like the Flames fell right off the cliff. When your goalie bats .878, though, you aren’t winning anything, even if you’re the 77/78 Canadiens. I don’t necessarily buy the premise that fatigue was a factor, but that drop is surely troubling, no matter the cause.
As I mentioned earlier, Kiprusoff gets a lot of rope from the fans and media because of his past history, and he did have a fine season in 09/10. At some point, though, reality has to set in. You just can’t pay 5.833M a year for this type of work, no matter how you want to sugar coat things.
I have no idea why Kipper’s play has hit the ditch, but my guess is that the removal of Jamie McLennan isn’t the solution, just like replacing Marcoux wasn’t the solution, so the team is in a bit of a pickle. There can’t be much of a market for a goalie that has played poorly and has a large contract, and hoping he just snaps out of this funk doesn’t seem like any sort of acceptable plan, either.
Of course, the Flames have other players as well, and it wasn’t like every skater covered themselves in glory, only to be foiled by their hapless stopper. In the next installment, we’ll look at the position players of note on the roster using the same collection of data.