45
Photo Credit: Christopher Hanewinckel / USA TODAY Sports

Michael Stone and the difference between looking better and being better

During the intermissions of last night’s 3-1 victory over the Stars, the Sportsnet panel discussed the reasons for the Flames success over the past month or so. Their conclusion was more or less that the defence had improved, which in turn, allowed Brian Elliott to become a better goalie. They also made sure to mention that the Flames were a perfect 9-0-0 with Michael Stone in the lineup. It was quite ridiculous.

As we saw with Garnet Hathaway earlier in the season, there is a near-superstitious drive to attribute the Flames’ wins or losses to one player regardless of their actual contribution on the ice. Hathaway was being praised for merely existing on ice while the team was winning, disregarding his five points in 26 games and average 5v5 icetime of just under nine minutes. This narrative also overshadowed the fact that it was really the 3M line being put together and Chad Johnson playing out of his mind that drove the Flames’ wins. Since being sent down to Stockton, the Flames are 11-1-0, perhaps because his performances and usage were so minor that they really never had an impact on the team’s ability to win or lose a game.

Now, the same narrative is being used with Michael Stone. His presence in the lineup during a winning streak (mostly fueled by Brian Elliott, 13-23-79 being assembled and being a functioning line, and 3M+27-5 being the best five man unit in the league) was the reason for the win streak and not just a correlation.

The argument, on the surface, appears to hold water. Stone is actually being used as a key player, slotting beside TJ Brodie in the top four. If the team placed him in a key spot in the lineup, and the team started winning, there must be some connection, right?

The blind test

Our perceptions of players are, no doubt, influenced by cognitive biases. Some of these are absolutely rational. For example, whipping boy Dennis Wideman gets paid way too much and has been frustrating to watch for the better part of two consecutive seasons. Along with the perceived Wideman Effect and his generally unspectacular play, it is easy and understandable as to why people don’t want him in the lineup.

Conversely, it is easy to understand why people like Stone. He’s new, young-ish, and a cheap acquisition price and cap hit. By these categories, he is the anti-Wideman, which is the thing he has going for him the most. Not being Dennis Wideman is an easy way to get a Flames fan to like a player. You can probably put a bag of hair or a stick of deodorant in the 4D spot and people will be happy.

But what really matters is the results. Say what you want about the bag of hair, if it doesn’t actually improve the team on-ice, the praise is unearned. Our task is to remove said biases from the discussion, and look at this from a neutral, results-focused perspective.

Here’s a table with the Flames’ bottom six defenders and their stats. Your job is to try and guess who is who.

All data from corsica.hockey

Player CF% CF rel% PDO OZS%
A 50.68% -0.05% 97.59 33.44%
B 46.55% -6.07% 101.31 27.93%
C 50.58% 1.24% 97.90 28.28%
D 45.77% -7.84% 99.82 30.19%
E 44.11% -8.34% 104.88 33.53%
F 47.64% -4.44% 107.06 29.67%

There’s a few general conclusions we can grab from this. Players A and C aren’t great, but they’re good, which is what you need from the bottom six. Players B, D, E, and F are all varying shades of the same crap. B and F are slightly better than D and E, but still not as good as A and C.

Let’s complicate this a bit more. Of those six unnamed players above, three of them played with TJ Brodie on the second pairing. Here’s a blind test for their stats.

Player CF% CF rel% PDO OZS%
A 47.45% -3.97% 94.66 24.78%
B 50.99% 0.23% 94.60 33.84%
C 48.83% -2.97% 108.00 27.95%

Again, general observations. Pairing A is not great, neither is pairing C, and pairing B is just okay. Neither option is preferable, but one is better than the others.

Results! For quiz #1, the answers were: Wideman for A, Deryk Engelland for B, Brett Kulak for C, Jyrki Jokkipakka for D, Matt Bartkowski for E, and Michael Stone for F. For quiz #2, A is Brodie-Engelland, B is Brodie-Wideman, and C is Brodie-Stone.

Based on possession results, Dennis Wideman was actually better than Michael Stone, which was a surprise to me too. It very much appears that Wideman was, at the very least, a manageable player sunk by PDO while Stone is the opposite. In the long term, Stone is less likely to be this final piece for the Flames and more likely to be a liability.

This is nearly exactly what people were saying about Wideman in the 2014-15 playoff run season. All numbers suggested that his suddenly improved performance was not likely going to be sustained, and it wasn’t. He was the same old defensive liability Dennis Wideman.

So why does it matter?

The purpose of this post is not to put a wet blanket on the Flames, nor is it a call to put Wideman back in the lineup (heavens no). There is nothing the Flames could do, or should do, to fix the defence, mostly because no option will actually make a major improvement for the final stretch of the season.

It’s a warning for the future. When management decided that Wideman needed to be replaced, they made a judgement based on PDO. Who’s to say that they won’t make a similar decision about re-signing Michael Stone when better options exist?

In the long run, Stone will not be a solution. Winning should not cloud the real problems with Stone, because he’s much more likely to turn into another Wideman than be a Wideman replacement.

  • redricardo

    I’ve said it before, I’ll say it again. After our top 3, we drop off to AHL/replacemt calibre. The fact that Stone looks decent and narrative exists that he’s helping says more about how awesome Brodie is than anything else.

  • Joeymullen'sghost

    Thanks CHcristiN. Great article to make us really think about our perceptions as Stones numbers are worse then Wideman’s. But I caution it also just shows advanced stats are an incomplete tool not a silver bullet. There’s more to hockey then corsi and ozs% thankfully! Because stats be dammed no way should we see Wideman dress. Just create a new stat for Wideman “ppp#” or press box popcorn pieces consumed per hour

  • RedMan

    Ok, CORSI is the best all end all.
    CORSI is infallible.
    CORSI says Wideman is better than Stone.
    So Wideman is preferable.
    But you say no.
    Please explain why, using CORSI, you choose stone over Wideman??????🤔

    • As I said, I don’t really care which player they use in the 4D spot because it’s probably not going to make a difference at this point in the season. The purpose of the article was to try and dispel the narrative being forwarded by many that Stone is a better defender than Wideman and that his presence has somehow changed things simply because he is not Wideman. He is likely to be a future Wideman, and I wanted to point out the potentially critical error of pumping the tires of a liable player in a contract year.

  • Juan Valdez

    The stat that really matters is high danger scoring chances against, not corsi. Stone’s job is to neutralize the offensive threat and dish the puck up to the forwards. Every championship quality team needs this type of player shut down player to make a deep run in the playoffs. I’d argue that without Robyn Regher, the Flames don’t make it to the Stanley Cup Final in 2004, so to suggest that Stone has real problems is a bit of a stretch.

  • canadian1967

    Corsi is stupid.
    I don’t make money answering your stupid narratives, so I don’t have the time to rebut them, but suffice to say that Stone is the most viable partner for Brodie of who is currently on the team (likely so in Brodie’s mind as well, which is very important, but not quantifiable by numbers) but you go ahead and spout off about your “made up, cherry picked numbers!

  • Arminius

    Not buying it sorry. All one has to do is look at Brodie’s game. He was a league worst -23 if I remember before Stone came along. He looked disjointed and nothing like the Brodie we were used too.
    Enter Stone and almost immediately TJ was flying up the ice, scoring an OT winner, getting scoring chances and generally kicking ass.
    (Last night’s boner move aside of course. )
    I know it’s almost blasphemous to day Brodie is playing badly but you look at him with Wideman and he had no answers

    • T&A4Flames

      I agree. #’s are a tool. Brodie apparently feels comfortable to do his thing playing alongside Stone. To me, that’s what matters most. Besides, we can’t have all $4.5-$5 mil dman in our top 4-5. We have to think bang for buck a bit as we get deeper into our depth. Stone was making Brodie play better. That’s the key.

  • Baalzamon

    If you had done this analysis before last night you would have come to the opposite conclusion (Brodie and Stone had a 51.4%CF together going into last night’s game).

    The sample sizes we’re dealing with here are too small to draw conclusions from.

      • Torchy

        Who exactly are you interacting with that is giving Stone full credit for the win streak?

        The Flames have had great success since Stone was acquired. This is a casual observation and it is not incorrect. The framing of this as a “narrative” is a bit much in my opinion. Fact is that the Flames were 9-0 with Stone in the lineup. It’s a FUN story and you can’t fault the TV crew for mentioning it.

        No doubt the Flames made moves on defence because they were unhappy with how their group were playing. I am curious how you narrowed this down to PDO.

        Take us deeper man! Is there any depth to the analysis you performed beyond pulling raw numbers? A lot of people are sensing that the team is better with Wideman out and Stone in, why not delve deeper into the comparison and see if in game events support the surprising statistics? (For example: if Stone surrenders relatively more weak shots than his peers, but relatively fewer dangerous shots…wouldn’t that be more helpful info to have on hand come free agency vs “Stone’s RelCorsi stinks”?)

        • In what ways is this not a PDO move? Wideman has always been an acceptable option for the team, minus the pricetag. If a move was possible, they would make it. Not to move on from Wideman, but to move on from his cap hit. Since there was none, they were fine with playing him in the top four, giving him PP/PK time, and the A. During the really bad 4-0 losing streak, he was the whipping boy for a lot of it even if not everything was solely his fault (the entire team deserves blame, really). That’s when, and why, they made the Stone move.

          The team is incredibly susceptible to PDO based judgement (signing Troy Brouwer, re-signing Karri Ramo, waiving Paul Byron, re-signing Lance Bouma), so it’s best to nip it in the bud immediately.

          As for your depth comment, there’s not a lot more to dive into. Corsi/fenwick/whathaveyou are all game events. Stone is on the ice for a lot more shot attempts than people are willing to acknowledge. People are fine with this as long as the team is winning.

          • Kevin R

            Holy cow Christian! You can interpret stats all you want & debate whose right. They have been doing that with the Bible for Centuries. Here is one little tidbit that you don’t take into account, with Wideman press boxed & Stone in, I have found the officiating playing field levelled to some extent & that is just my eye test. But check out a Blog from Paul Stewart & his reaction to the appeal being ruled in favour of Wideman. The venom of the officiating community is real any way you slice it. Not sure how you want to factor that into PDO & CF percentages.But it certainly impacts the game being played on the ice.

      • OKG

        Hamilton Giordano Backlund Tkachuk Frolik play together most shifts. Every other player is inherently going to struggle in rel stats when these five are throwing up like 60% CF together every night. Stone is not Hamilton… about as big a revelation as “Monahan is not Backlund”.

        • That’s not how corsi rel works though. It’s CF% on ice minus CF% off ice. You would have a point if his corsi rel was something like -0.5% or something, but if it’s -4% or more, it’s likely that he’s the issue. It means that when Stone is on the ice, shot attempt production drops by 4% compared to when anyone else is on the ice.

  • jupiter

    Why do analytic’s get interpreted as FACT.And when the result doesn’t match the narrative, along comes randomness luck and sample size.While I agree that the eye test can be flawed, I don’t believe the analytics are much better.

  • RealMcHockeyReturns

    Based on stats only, seems like Kulak could be another 4D-Brodie partner option by put eye-test suggests Stone is best there. Brodie clearly benefits from having a veteran reliable partner opin Stone.

  • TheoForever

    Funny article, make an assumption that your stats are the truth. Dismiss every other opinion with ‘cognitive biases’, and you have won the argument, all that is left is to pat yourself on the back. Good job.

  • Davebot

    For the love of… He’s not saying Wideman is better than Stone. He’s using the “Wideman better than Stone” statement as a rhetorical device to make the point that Stone may not be responsible for the winning streak and we should be careful about signing him based on what may be coincidence. The winning streak simply coincided with Stone joining the team. Bartkowski also joined the team at the start of the winning streak. Is Bartkowski some sort of 3rd pairing defense messiah? The point the article makes is that we shouldn’t judge Stone based on the current winning streak. We should judge him on measurable performance indicators. It’s not a crazy idea.

    My only criticisms of the article would be sample size (a point that has already been made) and the lack of scoring chance info (also mentioned). If anyone is interested I looked up SC stats on Natural Stat Trick:

    Player GP SCA SCA/GP SCF% HDSCA HDSCA/GP HDSC%
    Wideman 55 380 6.91 48.92% 151 2.75 50.00%
    Stone 9 59 6.56 53.91% 21 2.33 57.14%

    If you want to argue in favour of Stone those stats look promising, BUT, this is a small sample size. Stones HDCF% in Arizona this year? A whopping 35.3%

    I’m not ready to shout “Sign Stone!” yet. I suspect his recent success is a result of a small sample size and a run of great goal tending.

    • cberg

      Nice deeper data, clearly showing Stone much greater value, albeit in a small sample size. The other thing that Stone provides no one else provides (exc. Gio, occasionally) is the ability to pin guys on the boards which allows Brodie to take control. CT can throw all the simplistic stats at Stone that he wants, but anyone with eyes can see the Flames overall D is much better since he arrived.

    • That’s an interesting point. Stone was injured, in fact, his knee was reconstructed. Unless the Flames are way ahead of the medicinal science world, that knee will always remain a problem. Why exactly are you convinced that this is something he can quickly overcome? And he was definitely on a bad team, but his possession stats on Arizona and his possession stats in Calgary are pretty much identical. At what point is it the player and not the situation?

      • TheoForever

        Possession stats are not the start and end of it. Assumption that he is a bad player because of his possession stats is the beginning of the difference of opinions here. He is the type of a player Flames need, strong in front of the net and in the corners, able to stop the cycle, defensively responsible, etc. I believe he can be an effective #4 for a while, until Rasmus matures. Sure, we can try to get somebody else to pair with Brodie, not sure whom, just please don’t say Fransen. Anyway, for the rest of the season Stone is the best choice, after that we will see.

  • Newbietwo

    Stone wasn’t acquired to be Brodie’s pivot long term.. so I would caution that notion because he was acquired to be on our 3rd pairing next year..

    We will go to market and get another defender but I do want to throw this out there.. the same was said about Engelland and overall he has been a positive for the team.. one can also make the argument that Brodie has impacted the metrics with whoever he has played this year because he hasn’t been great like usually.. metrics are fine but not a complete story..

    I also heard there is a very good chance Anderson slots in with Brodie next year

    • BeerCorsi

      Engelland has not been good. I don’t know when the narrative changed to him somehow being this reliable defensmen. His analytics across the board are truly terrible.

  • cjc

    Looking at score adjusted numbers, Stone has a 10%lead over Wideman in xGF%. 57.36% to 47.44%. Same story with scoring chances, 62.16% to 47.69%. Stone has been better than Wideman, hands down.

  • dontcryWOLF88

    Yeah, always good to check eye test and compare with analytics.

    As for stones numbers in Arizona. ..well can’t forget that a players stats are highly effected by their teammates, as well. Is there anybody on the yotes with good underlying numbers?

  • Jobu

    Disagree. You still sign him to a nice 2 million contract and play him as a fourth fifth or sixth over then couple years. The guy JUST moved to a team and a new D pairing whose system is difficult to understand. He’s also had a tough year of injury and being on a losing team. Right now he’s playing it safe. Once he gets comfortable his numbers will improve. Once again, I plead patience people!!

  • Just.Visiting

    I think this story is a great example of why it is dangerous to look at stats in isolation. My eyes tell me that Stone plays a disciplined, physical game in his own end by rubbing guys out along the boards on attempted entries, by being a presence in front of his net, by having some semblance of a plan about what he’ll do with the puck when in his own end and by allowing his goalie and Brodie to play with greater confidence. I’m not sure if he’s the ultimate number 4 guy, but he looks a lot better than anyone else we’ve put out there in a while. Of course, that’s just my eyes telling me that.

  • MWflames

    Wow – talk about a sensitive topic. Christian, all in all, a good article.

    Here’s my qualitative response, something we got used to sayng under Hartley:

    Quality of chances, shots, CA, etc. Lately wideman has been good for at least one goal against per game as a result of blown play, turnover, or bad decision. Is it crazy to suggest Stone isnt being a liability like this? Won’t explain stones high profile, but it would be interesting to see wideman on ice sv%, and see if that is the major contributor to low pdo as a result of high danger scoring chances conceded…

    Anyways, while stones transition game needs work, I’d bet he’s doing a better job in his own zone keeping the play towards the outside. And when your goaltender is own, and tracking the puck, he’ll make a lot of the saves from the outside.

    I believe there was an article on here about a week ago that showed stone was given up much less high danger scoring chances than wideman. Would be nice to revisit that