Earlier today Winnipeg Jets head coach Claude Noel was asked about Ondrej Pavelec's absence from ESPN's top 25 goaltenders. He stated that "I think that's probably fair" based on the netminder's numbers. Not to be one to throw his goaltender under the bus, Noel added in that he believes they have the goaltending to make the playoffs...
...well... I agree one hundred and ten percent, on both accounts.
We'll ignore the fact that the article wasn't based of statistics, as otherwise the order would be much different. Otherwise, Noel is spot on. Thus far, Pavelec has not deserved or earned the privilege to be named one of the better goaltenders in the league when looking at the on-ice results. Over the last three years combined, Pavelec ranks 35th for stoping pucks from going into the net. The order of that list kind of matches with the eye-test, doesn't it? This is a very significant difference, where shot quality could only change so much (we'll get back to this later).
However, there is a difference between what occurred in the past and what is going to happen in the future. The two are related as talent causes a correlation between past and future performance, but weighted probabilities are not destinies. Ondrej Pavelec could do better, and I for one hope he does. I think it's unlikely but far, far from impossible.
In fact, I have already shown that a difference in player management could have tipped last year's team into the playoffs. There's also the "Chris Osgood effect" in how the Detroit Red Wings were a dominant team for many seasons in-spite of middling goaltending. Pittsburgh Penguins have done well in recent years with Marc-Andre Fleury, a goaltender who on average is very similar in Pavelec when it comes to stopping goals.
Is this new news or has Pavelec always been below average?
Pavelec's history should be haunting and telling, but because of the move, very few Winnipeg Jets fans know of it.
In the summer of 2005, the Atlanta Thrashers drafted a very promising goaltender in the second round who had very minimal professional goaltender coaching. Ondrej Pavelec had just came off of two back-to-back 30+ game season's with impressive save percentages. In his last season in the Czech junior leagues Pavelec came fourth in a league of 20 teams.
The next season Ondrej moved to North America to play for the Cape Breton Screaming Eagles in the QMJHL. He posted a very impressive 0.929 sv% in his rookie year, the highest in the QMJHL of regular goaltenders. At this moment the Pavelec Hype Train started rocketing to the sky. Then something new happened.
Pavelec reported to the Screaming Eagles training camp out of proper condition. Things entered his scouting report like "added extra weight", "inconsistent", "poor positioning", etc. Pavelec's save percentage fell to a pedestrian level on a playoff calibre junior team, and he has yet to post above average since.
The next two season's Pavelec played in the AHL for the Chicago Wolves. In his first season Ondrej Pavelec posted a 0.911, which was the 32nd best Sv% for goalies with 10+ games. The Chicago Wolves were a strong team though, and were able to pull themselves into the playoffs and eventually the Calder Cup.
The next season Pavelec improved to a 0.914 save percentage for 30th in the AHL, but Pavelec had enough of the AHL and wanted to push away one of Kari Lehtonen or Johan Hedberg. Allan Walsh -- Pavelec's agent -- went on record saying that Pavelec would not report to the AHL any further. He gave an ultimatum stating it was either trade, NHL or KHL (while Alexander Burmistrov is being crucified for the same thing although Alex was at least not under contract).
So, the Atlanta Thrashers bent over, traded Lehtonen and graduated the young Pavelec to the NHL. This is one of the few situations where I can find evidence in the Thrashers "rushing" a prospect and likely having long-term negative affects. Pavelec finished his season in the NHL posting a 0.906 SV% that season, right where his career average sits for the remainder of his career.
Pavelec's off-ice conditioning and dietary habits continued to haunt him. This all eventually peeked into a horrific event where Pavelec collapsed early in the first period and had to be taken to the hospital.
Since the move to Winnipeg, Pavelec has only had one month where his Sv% has been above league average. In December 2011, Ondrej posted an 0.939 Sv% and the Jets came out of that month with a 7-3-1 record. Other than that the results have been predominantly mediocre or worse, which as you can see isn't exclusive to his time in Winnipeg or Atlanta.
So... Can Pavelec improve?
Not good enough? Want to know why I think he could?
Pavelec has long been touted as a very athletic goalie who has good size to fill out the net; however, his poor positioning and conditioning have also been a long areas of concern for Pavelec. Here are the old Hockey's Future scouting notes in regards to Pavelec back to his amateur days:
On the downside, Pavelec needs to keep his conditioning in check and remain consistent. He allows a good number of goals through his five-hole and needs to play better in position when making the saves.
His stickhandling ability is rather average and some of his moves can lead to risky plays.
It's remarkable to how much this still remains to be true.
As I mentioned above, it's long been my opinion that Pavelec was rushed into the NHL. It's already been shown by Tyler that there is a strong link between AHL and NHL success, both being something Pavelec has yet to obtain. Pavelec held about a 0.912 save percentage in the AHL, with his final year being at 0.914. In a study of 37 goalies who jumped from the AHL to the NHL, 60% of them had more than 0.915 for their final AHL season. Keep in mind some of these goaltenders ended up career backups or regressed back to the AHL and also league average save percentage was lower then.
So, we have a player who was never taught positioning well and was rushed to the NHL. These type of players have been fixed, but the window to rectify the issues is closing fast. Pavelec is at the age where most goalies peek for save percentage, meaning he has a strong chance in trending downward over the next few seasons.
But, let us break this down further. For if we know his faults, we can see where improvements can occur.
Pavelec needs to be more consistent
In normal use, consistency is essentially clipping out the extremes, becoming what you are most often more often. Hopefully this is not what people are wanting of Pavelec. Quality Starts shows that Pavelec actually is the 4th least likely starter to either post above league average save percentage in a game or let in less than three goals. Real consistency would likely make Pavelec a worse goaltender if not just the same on average.
In fact, Eric has shown over at NHL Numbers that the idea that there are two groups of goalies, one being consistent and the other not, is in actuality a false idea. Eric compared how Marc-Andre Fleury, Ilya Bryzgalov, Henrik Lundqvist, Pekka Rinne, Jaroslav Halak and Carey Price's standard distributions in results fared versus what a consistent robot would. There was no difference.
When fans and media use the term consistency, what they are really saying without knowing it is be good more often or be bad less often. In other words: better on average. Coaches likely know a bit better then the average fan, with some rare exceptions. I'd be willing to bet usually it's just a euphemism for being better.
So, if it is not about being more consistent, what is it?
As noted above Pavelec has long been criticized for his poor positioning. This is where the "eye test" tends to fail viewers.
Pavelec often pushes himself into poor positioning, causing extreme plays where he either makes a miraculous save or the opposing team scores. When he does succeed, this is a very low-occurance high-impact play. A psychological bias then forms where the spectator remembers the high-impact result more clearly than the other moments throughout, thus distorting the perception of what is the average outcome. On top of this, the action creates a "wow effect" which causes the individual to think that the play is more indicative of Pavelec's true self; all he needs to do then is be his true self more often, thus the false idea of consistency arises.
But, but shot quality!
Shot quality DOES indeed exist. You can't convince me that there is no difference between Sidney Crosby and Chris Neil in the offensive zone other than just the amount of scoring chances they accumulate.
The problem is not that it doesn't exist; it's that it is far from the dominant factor for a goaltenders save percentage over an extended period of time. Essentially the more dominant factors mask the lesser factors. Shot quality could push it one way, yet the other factors overpower and push it wherever they want. The shot quality factor essentially becomes insignificant (stress) in this particular case (unstress).
The dominant factors tend to be goaltending skill and random variance (i.e. luck), while the magnitude of each is dependent on what sample you are taking from.
How do we know shot quality isn't the cause for Pavelec's save percentage being below average?
From years and hours of studying, making predictions, experimenting and hard work:
1) A defensemen does not have the ability to sustain a save percentage for their goaltenders. Showing how that a defensemen's on-ice Sv% is outside of their control by it being largely unrepeatable and the differences pattern similarly to random chance.
2) A look at the correlation between scoring chances and shot totals. Since scoring chances correlate closely to shot differentials, shot quality differences are negligible and don't add much information. Also, there is also no persistence in shot quality between seasonal data of a team's line.
3) The "Shot Quality" Fantasy. When a goaltender changes teams, the Sv% is just 0.0011 larger than random chance. Also, the direction of change is not always in the direction that one would expect or hope with moving from worse-to-better teams and vice versa.
4) Better Late than Never. The best shot-quality-influencing system era reduced Fenwick shooting percentages by about 0.0015 (or 1-2 goals per season).
5) Does Shot Quality Exist. Tom Awad has written some articles in regards to shot quality that are often falsely quoted for the existence of shot quality as a repeatable skill. He in actually shows here that there is very little sustainability in the shot quality metrics he brought up. This therefor shows that it is not a team talent and really more is a demonstration in the natural variance that occurs with sampling effects.
6) Goaltender Talent. "The takeaway here is if you want to estimate a goaltender's true talent, you may as well just look at his even-strength road save percentage"
7) Fooled by Randomness: Goaltender save percentage. Showing the confidence intervals you can have of certain goalies save percentage. (side note: this means you can be over 90% confident this order of goaltenders is correct in being a proxy for talent levels)
8) The Dave Tippett Effect and The Ken Hitchcock Effect. This showed that the two most defensively mind coaches did not improve their goaltenders as much as the common fan and media seem to believe. In fact this effect would be even less significant now, since the St. Louis Blues and Phoenix Coyotes goalies regressed this last season to more normal numbers. I also did a statistical test showing how the results were not statistically significant, which also predicted Mike Smith and Brian Elliott's eventual collapse to earthly numbers.
The take home to all of this is that there is some possibility that team effects could have lowered Pavelec's save percentage; however, it would be impossible for team effects to have pushed Pavelec to this level if he was indeed an above average goaltender.
In fact, it is even possible that team effects did the opposite, inflating Pavelec's save percentage, as the eye-test of most fans (including the Jets') seems to be poor in evaluating how much shot quality the Jets' allowed relative to other teams. I compared Pavelec to goaltenders on other teams who are well known both qualitatively and quantitatively for being poor defensive teams.
The truth is: Pavelec had it easier than many. The Jets' are not a top 10 team for minimizing quality or quantity of scoring chances, nor are they a bottom 10. They are quite average.
Claude Noel got it right. Pavelec has yet to earn being on the top 25 list, but the Jets still have the chance to be a playoff caliber team with their current goaltending.
We know Pavelec could get better, but we don't know if he will. Odds are he won't, but I still pray for the less likely result... because otherwise it's an even steeper hill to climb.