There’s been discussion at On The Forecheck about how much of a difference it would have made if Pekka Rinne had been healthy and playing at an elite level for this season. Some people say that a healthy Rinne would have put the Predators in the playoff hunt. Others believe that his work wouldn’t have been enough. I decided to test the issue. I’ve done this in the past for the Lightning and it’s fairly straightforward, if a bit detailed.
What I wanted to know was how many more points the Predators could have gained if they had goaltending at the level that Pekka Rinne was providing before he went on injured reserve. He had a .926 even-strength save percentage in 9 games this season. That’s a terribly small sample size, and it’s an analytical mistake to equate having Pekka Rinne with having a .926 even-strength save percentage. There’s no telling what kind of performance he would actually have put up if he were playing. It could be higher or lower than .926; we don’t know. Still, it works as a thought experiment because (a) that’s the data that people are using in their discussions, (b) it’s pretty much what he’s done in the past, and (c) it’s all we have to work with.
The Method
I got all the game data from extraskater.com–all 5v5 goals and shots, as well as actual game scores. For most of this analysis, I’m using even strength because power plays have such a disproportionate effect on results. You don’t spend very much time on them and every goaltender sees a different combination of penalty kill and power play minutes, but a lot more goals get scored. So rather than trying to untangle the effect of those minutes, it’s easier to leave them off to see if the other 40-55 minutes of play have anything to tell us. Then I’ll figure power plays, penalty kills, and overtime back in to find out how many standings points Rinne could theoretically have gained (or lost) the Predators so far this season.
First I marked all 18 regulation and 1 overtime win. No way to gain any points there, although some could be lost, in theory, if a .926 would have cost enough goals to lose a game that was won in real life. But we’ll stipulate those as wins regardless. And I removed any game Rinne played in (9 games), leaving 36 games (15-16-5). (Although if anyone’s interested, a .926 at 5v5 would have turned one of Rinne’s regulation wins into a regulation loss and one of his OT losses into a regulation loss, losing the Preds 3 points. It wouldn’t have been enough by itself to turn any game lost into a game won or to send a regulation game to overtime.)
The process here is to compare actual 5v5 results to what the results would have been with Rinne’s .926 5v5 save percentage. His PK save percentage is .861, lower than either Carter Hutton’s .871 or Marek Mazanec‘s .864.
The Data
First off, 14 non-Rinne games had higher save percentages than .926: 10 were regulation wins, 1 was an overtime win, 3 were losses. We can safely say, however, that the three losses were not due to goaltending at even strength. In two, the Predators were shutout (11/23 vs the Rangers and 11/27 vs the Oilers). In the third, (12/17 vs the Blackhawks) the final score was 3-1, but only one goal (for the Blackhawks) was scored at even-strength. In all fourteen cases, the Preds would have been worse off if the goalies had posted a .926 even-strength save percentage.
That leaves 22 games where a .926 could possibly have made a difference. The question is how much of a difference and would that have been enough to actually change the outcome of a game?
In 5 games the improvement was less than one goal. For instance, in the January 5 game against the Hurricanes, Carter Hutton faced 23 shots against and gave up 2 goals, leaving him with a .913. If he had had a .926, he would have given up 1.7 goals. In these 5 games, the improvement would have been insufficient to change the score at all.
Of the remaining 17 games, the Preds won two in real life. Better goaltending wouldn’t result in any extra points.
There were seven losses where a .926 would have resulted in a positive 5v5 goal differential, while in real life there was a 0 or negative goal differential. In other words if Hutton or Mazanec had put up a .926 in those 7 games, the Preds would have outscored their opponents at 5v5. In three of those, the Preds earned a point by going to overtime, so we’re looking at a potential of 11 standings points.
There were 8 games (real-life losses) where even with a .926, the goal differential at 5v5 would have been negative.
Next I factored power play goals, both for and against, back in:
[Actual Preds Overall GF] – [Opponent 5v5GF at .926 save percentage + Opponent actual PP GF]
This process leaves the penalty kill and power play rates alone to isolate the impact of that .926 save percentage. It doesn’t include any overtime or shootout goals, because we’re also trying to see if any games would have been thrown into overtime.
|
Game |
Preds’ actual GF |
Opp 5v5 GF @ .926 |
Opp. PP GF |
Opp Overall GF |
Overall GD |
Outcome |
Pts Gained |
|
2013-10-26: Blues 6 at Predators 1 |
1 |
1.924 |
2 |
3.924 |
-2.924 |
L |
0 |
|
2013-10-31: Predators 4 at Coyotes 5 (SO) |
4 |
1.406 |
1 |
2.406 |
1.594 |
W |
1 |
|
2013-11-08: Predators 0 at Jets 5 |
0 |
1.85 |
0 |
1.85 |
-1.85 |
L |
0 |
|
2013-11-10: Predators 0 at Devils 5 |
0 |
1.184 |
1 |
2.184 |
-2.184 |
L |
0 |
|
2013-11-15: Predators 1 at Penguins 4 |
1 |
1.924 |
1 |
2.924 |
-1.924 |
L |
0 |
|
2013-11-30: Flyers 3 at Predators 2 (SO) |
2 |
0.962 |
0 |
0.962 |
1.038 |
W |
1 |
|
2013-12-05: Hurricanes 5 at Predators 2 |
2 |
1.332 |
2 |
3.332 |
-1.332 |
L |
0 |
|
2013-12-07: Predators 2 at Capitals 5 |
2 |
1.998 |
1 |
2.998 |
-0.998 |
L |
0 |
|
2013-12-19: Predators 2 at Lightning 4 |
2 |
2.146 |
1 |
3.146 |
-1.146 |
L |
0 |
|
2013-12-21: Canadiens 4 at Predators 3 (OT) |
3 |
1.554 |
0 |
1.554 |
1.446 |
W |
1 |
|
2013-12-23: Bruins 6 at Predators 2 |
2 |
2.22 |
3 |
5.22 |
-3.22 |
L |
0 |
|
2013-12-27: Predators 1 at Stars 4 |
1 |
2.146 |
0 |
2.146 |
-1.146 |
L |
0 |
|
2014-01-02: Predators 2 at Bruins 3 (OT) |
2 |
1.406 |
0 |
1.406 |
0.594 |
W |
1 |
|
2014-01-04: Predators 4 at Panthers 5 (SO) |
4 |
1.776 |
0 |
1.776 |
2.224 |
W |
1 |
|
2014-01-09: Ducks 4 at Predators 3 |
3 |
1.702 |
1 |
2.702 |
0.298 |
OT |
1 |
The Results
That’s a total of 6 standings points gained, all by changing the Preds’ overtime record. It would put the Preds at 50 points, placing them ahead of the Dallas Stars but still three points out of a wild card spot and behind the Phoenix Coyotes in that race.
Six standings points is a pretty big contribution to attribute to one player, especially just over half a season. And what we’re really talking about here are the contributions of three players, not just one, if I were to get technical. If the backups had performed somewhere better than they have, we wouldn’t be having this conversation. But it does show something about what the Predators and Predator fans actually expect out of Rinne and gives some objectivity to the claim that Rinne is more crucial to this team than he would be to a team built differently.
Obviously the Predators would be better off with better goaltending results. Who wouldn’t? But having a healthy and elite Pekka Rinne would not turn this season into a winning season and in the majority of cases it wouldn’t overcome the scoring issues the team continues to have.
