Ryan Whitney and PDO

At plus-12, he’s the lone Oiler in double-digits by plus-minus, and has more than double the next best player (Ales Hemsky, plus-5).

Ryan Whitney’s spectacular plus/minus on a spectacularly bad team has garnered a lot of attention, even from people who don’t generally believe in plus/minus. I’ve quoted myself above, but the numbers has shown up in the columns of most of the local papers – from the Sun to the Journal.

Is it important, or is it a mirage?

Advertisement - Continue Reading Below

Before we consider that question, I’d like to talk about a statistic tracked at Behind the Net called PDO. PDO is very simple: it is simply the addition of on-ice save percentage and on-ice shooting percentage (please note the difference here between shooting percentage and on-ice shooting percentage; the former is the shooting percentage of a single player, while the latter is the shooting percentage of that player and all of his teammates while he is on the ice). The idea is simple: since on-ice percentages are affected by teammates and opponents and fluctuate wildly from year-to-year, a player with a high PDO is likely to see that number drop, while a player with a low PDO is likely to see it improve.

We’ve already seen one example of that – the difference between Shawn Horcoff in 2009-10 and 2010-11. In 2009-10, he had awful on-ice percentages; this season the numbers have been much better. The impact on his plus/minus (and the positive/negative press he has received at least partially as a result) has been undeniable.

What does this have to do with Ryan Whitney? It turns out that Whitney has a very high PDO, compared to the rest of the Oilers defence (all numbers from 5-on-5 play, and courtesy of Behind the Net):

Advertisement - Continue Reading Below
Player SF/60 SA/60 Corsi/60 On-ice SH% On-ice SV% PDO
Ryan Whitney 22.1 31.3 -11.86 12.50 93.5 106.0
Theo Peckham 22.0 28.3 -11.53 9.26 91.4 100.7
Kurtis Foster 25.4 27.8 -5.80 7.36 91.4 98.8
Ladislav Smid 24.1 26.4 -5.48 7.57 90.8 98.3
Jim Vandermeer 25.2 28.9 -4.77 7.69 90.6 98.3
Tom Gilbert 24.6 25.8 -3.30 8.61 88.8 97.4
Jason Strudwick 22.4 30.3 -21.16 5.42 89.9 95.3
Jeff Petry 25.6 23.6 0.74 7.59 87.2 94.8

Ryan Whitney is leaps and bounds ahead of the pack in terms of on-ice percentages. The team was not only far more likely to score on any shot they took while Whitney was on the ice, but they were also far more likely to save any shot against while Ryan Whitney was on the ice.

The question then, is this: was Whitney somehow responsible for those incredible percentages, or was he simply getting the bounces? Obviously it would be preferable if he was somehow causing it, although given the number of other players on the ice that seems extremely unlikely.

However, if Whitney were somehow making his goaltenders better through defensive positioning and his shooters better through pinpoint passing, it should be something he’s done previously for his other teams. How has Whitney’s PDO compared to his defensive comrades historically?

Season Team PDO Rank (among defensemen with 25+ games)
2009-10 Anaheim 101.5 3rd of 9
2008-09 Anaheim 99.2 7th of 9
2007-08 Pittsburgh 102.9 3rd of 8

Despite generally playing with offensive players and not facing top opponents, Whitney’s on-ice percentages have never been especially high in comparison to his teammates. In fact, over the past three seasons, his on-ice save percentage has never been better than this season: it has alternated between a very low 0.904 and a decent 0.917. It’s never been close to the 0.935 behind him this season. Whitney’s on-ice shooting percentage has been above the NHL average, ranging from 8.80 to 11.21 (during his time with the Penguins, who led the league in shooting percentage that year). Again, it’s never been as good as the 12.50 number Whitney’s been on the ice for this season.

If historical trends repeat themselves, both Whitney’s on-ice save and shooting percentage numbers will drop off sharply next season. Whitney’s plus/minus, which is a function of the percentages and shots for/against, will also plunge. The alternative is to believe the 28-year old suddenly developed an ability to sharply influence the percentages of others in his seventh professional season.

Advertisement - Continue Reading Below

  • Tha Legion

    Hmm, I know I’m not the biggest supporter of advanced stats but this is selling Whitney a bit short, it was very visible that he has improved greatly this season and logically deserves that +/-

  • Whitney was the only Oiler that consistently made a good pass out of his own zone and most of the Oilers offence was created by that. Thus the high PDO.

    Obviously his defensive numbers will be affected by his ability to move the puck out of his own zone.

    He was THAT much better than his teammates and thus that stat.

    His defensive stats are a product of his offensive abilities.

    If you are moving the puck out of your own zone, no one is scoring on you!

    • Steve Smith

      You’re going to have to explain how “his ability to move the puck out of his own zone” has the first damned thing to do with his PDO. Otherwise, your comments are going to fit in awfully well here.

      • Steve Smith

        If you don’t know how moving the puck in your own zone can help scoring chances for, then I don’t know how to help you.

        Do you really think scoring chances for are only created in the offensive zone?

        • Steve Smith

          Of course they affect scoring chances. Now explain how they have the first damned thing to do with PDO.

          Edit: Just noticed that you’re the same guy who said that Whitney’s PDO will come down as a consequence of the rest of the team improving. I’m officially satisfied that “PDO” does not mean what you think it means.

    • Kodiak

      If he is such a great puck moving Dman, how is it when he is on the ice there are 142 shots against for every 100 for? By contrast, when Gilbert is on the ice we allow 105 shots against for every 100 for. It makes no sense to think Whitney can keep allowing that many shots and not eventually have his +/- take a serious hit.

  • Bar Qu

    Outstanding article. You have really brought into focus how a string of luck can markedly affect a players stats and you have done so with substantial sample size. It speaks volumes about how little we should read into boxcars taken from very small sample sizes. A contemporary example would be 2011 draft eligible junior hockey playoff statistics.

  • Bucknuck

    I think anything plus minus on a team with as inconsistent a goaltender as Khabibulin is suspect. Particularly when looking at less than half a season.

    Everyone’s plus minus should go up next year when he isn’t starting so much. Anyone who stays above 0 on this crappy team should get a bonus.

    • Quicksilver ballet

      Anderson had a .893 in Colorado. How do you explain his .944 in Ottawa…..same goaltender.

      Said suckage is due to the lack of depth and injuries, Khabibulin will be just fine with a decent team in front of him, much like Andy’s doing in Ottawa.

  • Shapeman

    You have to take into account that he only played 35 games or so as well. If he had played the whole year his humbers might have dropped off when our offensive studs went down.

  • Whatchu talkin’ bout Willis?

    Damn you man! Stop ruining my inner dream life which has become the only thing keeping me going lately! An almost undeniable argument however. If I were a betting man I bet on Whitney having a more realistic season next year. I reiterate: damn you!

  • “Hmm, I know I’m not the biggest supporter of advanced stats but this is selling Whitney a bit short, it was very visible that he has improved greatly this season and logically deserves that +/-”

    Great. Let’s bet on it. You take the side that says Whitney will have a 106 PDO next year and I’ll take the side that says “lower”.

    I’m sure we can find someone here (JW) with a reputation willing to escrow our bet, how much would you like to wager?

    • Tha Legion

      My argument is more along the lines of I don’t care for that particular advanced stat, if you would line to friendly wager that Whitney wont be a plus player I’ll gladly take the will be side, hit me up on twitter if you want to

  • Feynman

    Interesting statistic. One thing I found very interesting, although not a surprise to me, is that Omark has the worst PDO on the team at 942. This also ranks 25th worst in the entire league. The guy seems to have a lot of bad luck when he’s on the ice and I think it is almost inevitable that it should change soon. On Whitney, he’s definitely an above average defenseman, but I feel he has been extremely overrated by Edmonton fans this year. However, as a die-hard Oiler fan I’m relying on hope and I pray that Whitney is a key piece on defense for the Oil going forward.

  • @ Quicksilver:

    The problem with your analogy is that it places the Ottawa Senators as a “decent” team.

    Players go up and down for a variety of reasons. Anderson’s numbers in Ottawa are much closer to his numbers in Colorado last season and his numbers in Florida – i.e. his actual level of ability.

  • beezkneez

    The reason he was so heads and tails ahead of everyone on the ice on the Oilers, is he was heads and tails above everyone on the Oilers

    The stats are not wrong, the rest of the Oilers defence was awful especially in their own zone.

    His PDO will naturaly get worse as the rest of the teammates will improve.

  • Shapeman

    I think the best way to rate a valuable a player is to his team is to do the following..

    -Take the Time On Ice SH% ex. 10.0%
    -Add to that the Time On Ice SV% ex. 91.0%

    And your total is 101.0

    That should be that players On Ice Rating..
    Only problem is someone like A.Burrows might have a rating of 110 while Jarome Iginla might have a rating of 102..but Jarome Iginla is clearly the better player on a lesser team…

    • Wax Man Riley

      For the record, I actually believe that trade was a win-win for all parties.

      Anaheim got Vis…. we know how that has turned out.

      Vis gets on a playoff team in his prime.

      Oil get Whitney, a younger, bigger, cheaper olympian that fits in better age-wise and has the potential to be a very good #1 as we saw in the first 35 games

      Whitney got the opportunity to be a #1


      • Kodiak

        Vis leads the league in D scoring with the Ducks and oft injured Whitney misses 47 games for us.
        The Ducks got faster and more offensive and we got Strudwick in the lineup.

        How is that trade a win/win?

        • Wax Man Riley

          So you think Visnovsky would lead the league in D scoring if he played on this Oilers team?

          I agree he missed some games, but that can happen to anyone. Crosby missed some games this year too. Still a good trade.

  • There’s no question his +/- was exaggerated because of his high PDO but if you are going to say that his PDO is out of line because of his past performance then would you agree that his past performance sets a reasonable expectation? His career average would still give him the highest PDO on the team by a decent margin.

    I’d be curious to see how a PDO of around 102 would affect his +/- number, because it seems as though his +/- would still be unexpectedly high compared to his teammates. I don’t have time to do the math right now (I am writing an opinion on legal access to micro-hydro power – good times let me tell you) so I’m hoping somebody else will…

  • Kodiak

    Great article JW, I totally agree. When you are continually giving up more shots than you are generating, in the long haul it should balance out that you are being scored on more than you are scoring. Whitney has the worst for/against shots totals among regulars. It is bound to catch up with him.

    Whitney was also playing with our “A”team, prior to all our injuries. Personally I think Gilbert being even over his last 22 games with a depleted lineup and still generating almost as many shots for as against is a lot more impressive than Whitney’s fortuitous numbers or actual play.

  • Kodiak

    I wish I knew what your predication success rate is. Do you keep track of that? I find your statistical analysis to be boring and tedious. I’m just waiting on you to start bashing Hall and Eberle. I guess that will start soon enough.

    People improve with time and sometimes they actually improve performance. With such a poor team surrounding him I’m amazed Whitney has shown what he has.

  • New flash the world is not in fact flat and there are no such thing as witches.

    I’m sorry but it’s way too funny to hear someone say they hate statistics then try to argue how wrong you are.


    Willis does a damn find job breathing some level of interest in a train being driven by a drunk conductor. It’s crazy that people take plus minus, assists and goals as gospel metrics but refuse to bother understanding that leaps and bounds of improvement are being made by people like Willis (compliments primarily to Dejardins) by pulling apart the underlying numbers and sussing out a much clearer picture. Wake up. Already you have analyst just starting to scratch the surface of what these guys have been doggedly pursuing for free for years. Terms like even strength scoring per 60 minutes and such are starting to leech into commentators insight for good reason. They actually do mean something.

    When you wake up in 10 years and more and more of these analytics are discussed and utilized by teams and on tv, pray you aren’t ignorant enough to pretend you weren’t previously a troglodyte.

  • Wax Man Riley


    Is it easily doable to pull apart our scoring rates for the 1st 35 games compared to the remainder?

    It was mentioned but the fact that Gilbert has been pulling his numbers up amongst a mounting body count is a lot more impressive than a lot of people are willing to really credit.

    I see Whitney and Gilbert as decent 2/3 defensemen. A healthy Souray would be a solid 4.

  • Wax Man Riley

    It’s funny to think 5.5 mil in souray with another 1.8 mil for Foster (souray lite w/ aspritane) is on the books.

    That’s 7.3 million and we still don’t have a number 4 defenseman on the roster.

  • Of note is I have not once ever tried to counter any arguement by trying to diminish one’s use of the english language.

    I’m smart enough to recognize my intellectual weaknesses.

    Never really been that concerned about it as I can read my T4 just fine.