(Via Pawel Dwulit at the Canadian Press)

Toronto goaltending is a weird thing. It's pretty well been established that Toronto haven't had a reliable #1 starter since Ed Belfour and attempts to fix this via trades for Andrew Raycroft and Vesa Toskala horribly backfired.

This season, Toronto are the 22nd best team in the NHL in overall save percentage but are actually the 13th best team with even strength goaltending according to Behind The Net. Unfortunately, the team's overall goals against total (3.10) reflects more the overall team save percentage than that at even strength, due to a well-publicized brutal penalty kill.

On my own, I don't have the information necessary to properly evaluate the Leafs PK, but I do have the means to compare the two Leafs goaltenders this year. After a strong second half of last season, James Reimer came into the season as the opening day starter, but after taking a hit from Brian Gionta at a game in Montreal earlier this season and subsequently knocked-out with upper body injury-like symptoms, he's come back and played less than what he had been previous.

But out of nowhere, the Leafs have got help from an unfamiliar friend: Jonas Gustavsson, who came into this season with an underwhelming .909 even strength save percentage and a quality start** record of just 35.6%—only 21 of the 59 times he'd taken to the net this season and been blown-up 15.3% of times.

**(A quality start is a Hockey Prospectus statistic for goaltenders that is more reliable than goaltender wins for predicting future success. A goalie gets a quality start if he stops 91.3% of shots in a game or more, or allows 2 or fewer goals and stops 88.5% of shots. The cousin to the quality start, the blown-up rate, created by Thomas Drance at our sister website Canucks Army, measures starts where a goalie fails to stop 85% of shots, or allows 5 or more goals despite facing fewer than 40 shots.)

This season, though, the trend has reverse. After earning the first three starts in the new year, Gustavsson and the Leafs have won all three, which is good enough for Ron Wilson. But check out how the numbers have shaken up this year:

(Numbers come from NHL.com stat pages and game logs)

Not only has Gustavsson earned more quality starts than Reimer this season, but he also hasn't put up a start so bad that he doesn't give the Leafs a chance to win, which has been Reimer's downfall. A lot of this is penalty kill luck, after all, Reimer is still performing much better at 5-on-5. So what's the overall difference between Gustavsson's success and Reimer's failings?

I am absolutely not convinced that penalty killing save percentage is a reflective quality upon the goaltender, but with the Leafs fortunes on any given night teetering with how well the penalty kill is playing, it shouldn't shock us that Gustavsson's play a man down has really contributed to his quality start record.

Not necessarily agreeing with the "win and you're in" mentality that Ron Wilson is preaching to his goaltenders, it at least re-assures me that Gustavsson has got it done with a pair of quality starts in a row. The healthy competition isn't hurting the Leafs too much right now, but I have to think that Reimer's record improves through the season.

But the fact that Gustavsson has yet to allow five goals in a game this season is indicative of something. Maybe he's turned a corner, maybe it's just fluke over a small sample size.

UPDATE: The fact that Gustavsson has yet to allow five goals in a game this season is actually indicative of the fact I can't count. I updated the chart to include the games I overlooked. I will add, though, that a 12% rate is a little less than league average when it comes to BU%.