The Hockey PDOcast - Episode 69: Passing The Eye Test

Episode Date: March 16, 2016

Gus Katsaros joins the show to discuss how scouting has changed over the years, where it's headed, and the challenge finding a balance between quantitative and qualitative evaluation. Here’s a rundown of the topics covered: 2:30 Finding a balance between data analysis and the eye test 6:00 Entire organization being on same page 9:00 Misconceptions about scouting 13:45 Overvaluing the wrong skills 19:20 Hockey Sense 26:00 Integration of Analysis 29:00 Divide between perceived skills and on-ice results 35:40 The Passing Project *Every episode of this podcast is available on iTunes, Soundcloud, and can also be streamed from our website. Make sure to not only subscribe so that you don’t miss out on any new shows as they’re released, but also take a minute to leave a glowing review. *This episode is brought to you by Freshbooks, an online accounting service designed to save time and help avoid all of the stresses that come with running a small business. They’re currently offering a free 30-day trial to listeners of our show at Freshbooks.com/PDOcast (just remember to enter “Hockey PDOcast” in the ‘How You Heard About Us’ section). *Also sponsoring today’s show is SeatGeek, which is making it easier than ever before to buy and sell sports and concert tickets. They’re giving our listeners a $20 rebate off of their first purchase. All you have to do is download the free SeatGeek app and enter the promo code PDO to get started. Thanks for listening! See acast.com/privacy for privacy and opt-out information. Learn more about your ad choices. Visit podcastchoices.com/adchoices If you'd like to gain access to the two extra shows we're doing each week this season, you can subscribe to our Patreon page here: www.patreon.com/thehockeypdocast/membership If you'd like to participate in the conversation and join the community we're building over on Discord, you can do so by signing up for the Hockey PDOcast's server here: https://discord.gg/a2QGRpJc84 The views and opinions expressed in this podcast are those of the hosts and guests and do not necessarily reflect the position of Rogers Media Inc. or any affiliate.

Are you ready for the most ridiculous internet sports show you have ever seen? Welcome to React, home of the most outrageous and hilarious videos the web has to offer. So join me, Rocky Theos, and my co-host, Raiders Pro Bowl defensive end, Max Crosby, as we invite your favorite athletes, celebrities, influencers, entertainers in for an episode of games, laughs, and of course the funniest reactions to the wildest web clips out there. Catch React on YouTube, and that is React, R-E-A-X-X. Don't miss it. This podcast episode is brought to you by Coors Light.

These days, everything is go, go, go. It's non-stop hustle all the time. Work, friends, family. Expect you to be on 24-7? Well, sometimes you just need to reach for a Coors Light because it's made to chill. Coors Light is cold-loggered, cold-filtered, and cold-packaged. It's as crisp and refreshing as the Colorado Rockies. It is literally made to chill.

Coors Light is the one I choose when I need to unwind. So when you want to hit reset, reach for the beer that's made to chill. Get Coors Light and the new look delivered straight to your door with Drizzly or Instacart. Celebrate responsibly. Coors Brewing Company, Golden Colorado. Sponsoring today's show is FreshBooks, an online service that can help make everyday invoicing and accounting easier for you. You can try your hand at it today with a free 30-day trial that's being offered to all of our listeners.

Just go to freshbooks.com slash PDOCast and make sure you enter PDOCAST in the how you heard about us box to get started. Also sponsoring today's show is Seek. Seekek takes all the grunt work out of buying and selling tickets to sporting events and concerts by pulling all the tickets listed on various different websites and alerting you have the best deals available. It can come in particularly handy with the NHL playoffs just around the corner because now is pretty much as good of a time as any to see if you can get into the building for a reasonable price. If you haven't been to a playoff game yet in your life, I'd highly recommend doing so. The vibe is like night and day when you compare it to the 82 game grind.

That's the regular season marathon. And it can get especially good if you're there to witness an overtime game or a game that's really close down to the wire because you're pretty much just sitting on the edge of your seat the entire time kind of at the whim of the balance of a puck. And seat geeks providing a $20 rebate to all of our listeners for using their platforms. So just to claim that reward, all you've really got to do is download the free Seat Geek app. and then you just go to the settings tab and click add a promo code and enter PDO. After that, Seekkeek's going to send you those $20 once you've gone ahead and made your first ticket purchase with them. And it's really that simple. So all you've got to do is download the free Seek app, enter the promo code PDO.

And you can find yourself enjoying the playoff season like never before. Regressing to the mean since 2015, it's the Hockey PDOCast with your host, Dmitri Filippovich. Welcome to the Hockey PEOCast. My name is Dmitri Filipovich. and joining me is my good friend Gus Katsaros. Gus, what's going on, man? Thank you for having me on. I'm Demichu.

I was going to introduce you as your official title is McKean's director of analytics and pro-scouting, Gus Katseros. But that seemed like a very fancy way of just saying that you watch a lot of hockey. So I'm glad to have you on because there's a few directions I want to take this discussion with you. And I start off with this one. It's sort of a broad one.

and then we'll sort of unpack it and get into specific layers afterwards. But I guess the first question is considering, you know, pro scouting is in your title. What's, um, how far are we from finding the right balance between sort of visual analysis in terms of actually either watching from home on Game Center Live or on TV, for example, or even actually being at the rink as opposed, while also merging it with data analysis where you're using a bunch of spread. sheets and combing through data and then putting it together to form the ultimate picture. You know, I often thought that there was this false battle between what you see and what's being recorded that can actually be analyzed through another. I don't think it's that, like, an adversarial relationship that seems to be prevalent on whether you're talking to somebody or online.

I think in the grand scheme of things, what we're seeing is a greater dependence on not only just analytics and the numbers that are better being protected because you can actually have verifiable and measurable data, but there are things that are recorded while you're watching to create different kind of models. And like, I mean, we can kind of go on and on about all of this. I think in the end, what you're going to see

as far as pro-scouting is concerned is you're going to see a melding of data with some kind of a visual and the visual has to be some kind of a hockey idea and that hockey idea is going to have to have to have some kind of a like a current merit for instance something like a face puncher you don't need face punches anymore

right you don't need big slow defense these ideas have already been they've already been talked about and verified and these are things that analytics have provided that will help scouts in order to make some of their decisions a lot easier as well. But in the end, I have to think that data will be the driving force, and there will be more of a coaching presence. I think that pro-scouting takes less of the player evaluation. It's a little bit different.

It's not as pure as it used to be in the past. Now what you're seeing is pro scouts that are looking more for fit, they're looking for players that are fitting specifically into their system or may not necessarily fit into the system of a team that they're scouting against. And what I think they'll end up seeing is a lot more good data being passed on to allow for these pro scouts to be making better decisions based on this fit. I'm not explaining it as clearly as I kind of watch. It makes sense because I think you see with the best, like the model organizations,

and I guess you could say the Red Wings are like this sort of in hockey, although in the past few years maybe they've either people have kind of caught up to the pack or they've trailed back a little bit. But like I look at the San Antonio Spurs or the New England Patriots, for example, right? And there's this sort of stepwise food chain in there, going from the ownership to the general manager to whatever VPs to the coaches, to the assistant coaches, down to the players. And everyone seems to be sort of in line with the same thought process.

And that doesn't mean that I'm sure they have plenty of spirited debates, whether it's strategic things or personnel things. But at the end of the day, you kind of have these meetings and you iron all that out. And as soon as you leave the room, you're all kind of towing that company line where you're all pulling in the same direction, right? Because what's the point of a scout bringing the attention to a player to his GM and he takes that guy or he picks him up in free agency and then the coach doesn't use him properly? And it doesn't really make sense. So you really need everyone to sort of be working in unison for all this stuff to come together. And that's why I think it's going to be

actually more of a coaching element rather than a pure scouting element. For instance, I'll use the least as a prime example. They ended up hiring Shock Lamar. And I'm not sure Shock Lamar is probably the best in terms of, you know, quote-unquote scouting a player, but I mean he's got enough coaching experience and valid coaching experience at a high level, that he's able to look at the game from a different perspective. And he's able to bring different ideas that maybe, you know,

he wanted traditional scout never really brought in the past. Or if they did, it was something in passing that that didn't really have a lot of values. value. So what I think it's going to end up happening is you're probably going to see two major things that come out from protein pro-stouting. First, the whole thing is about information. You send out a scouts essentially to bring you back information, whether it's through a player evaluation or trying to fit a player into your system or even trying to find some kind of reconnaissance. You know, you don't find a player just by, you know, picking on a stretch or even just watching it. You have to do some kind of reconnaissance to understand who's available,

may not be, who may be happy, who may be not, you know what I mean? So in the essence, and again, using the lease in that regard, they hired Dave Morrison to be the director of pro scouting. Now, that guy's got a roolodex. There's a word that we don't really use very much more. He's in a rolloadix that's like pretty fixed. So what you do want is to have as much of a net cast out in order to be able to bring in talent, and then you need to have somebody that's able to verifyably say that, yes,

this talent can actually help us because of X, Y, and C. In the end, I think the essence is make sure that you're getting the proper information, and you're getting that from good, reliable sources. And pro-scouts are essentially going to be that front-line information source. And then you're going to have a whole bunch of coaching elements, and I think that that's going to be the key to moving forward. You need to have, at the pro-level, you need to be able to distinguish who can fit better in a different type of system. You're less involved with, you know, the four-ess and scouting.

You're not really looking at, you know, what a player can or can't do. You're trying to fit exactly where he is within your own system. Right. Well, okay, we'll get into the coaching aspect of it a bit more in a little bit. But I first want, I wanted to finish off this topic of sort of melding, using kind of that reconnaissance. As you said, actually being at the rink or talking to people and seeing stuff and then kind of parsing the data and putting it all together to paint one big picture. and I don't know, like, I still think there is obviously a certain, something to be said for being at the rink and seeing certain things because it just provides perspective, right? And it gives you extra layers of stuff to work with.

And it might alert your attention to something that you wouldn't have seen otherwise. But I don't know, like, maybe I just don't have enough experience. Like, maybe if you've been doing it for 20, 30 years, you get a lot better and the action sort of slows down. But like whenever I've been at a game and I've been trying to focus on one thing, I often find that there's just so much going on and hockey is such a hectic sport that it's really impossible to capture everything fully, right? Like you can key in on one certain thing, but then you're going to miss so many other things. And the reason why I personally enjoy looking at the data a lot is because it sort of removes that bias out of the way, right? Like even if you're a very objective person that has no pre-existing biases coming into it, you probably actually do, but it's more so subconscious than you're not really thinking about it.

So your eyes are going to be drawn to certain attributes or certain plays, and you're going to be missing out on a whole lot of other things. I actually don't think that that's correct. And I'll explain to you why. I feel that it's a misconception to think that scouts go and watch every single detail. absolutely cannot, simply for the reasons that you just explained. It's just not, you're not able to do that. And at the same time, if you think that you're able to do that, then you are totally missing.

You're totally missing everything. So what you're doing is focused, and what you're also doing is there's a lot of back references that we don't have privy to, for instance, scouting reports that may go X number of years in the past, right? So it's not like you're going in blind, and it's not like you're going in without any clue about a specific platter. you're going in with a bunch of historical references that you can focus on. Now, if somebody told me to go and record all this kind of stuff for, you know,

grand scheme of the game, then why, well, what's the point? You have data to do that. At the same time, you know, I think that Scouting is a very misconceived craft. I don't think that we're looking at certain things, for instance. I find it wasteful to go and look for things that are already available in the data. if you have to go and verify things with data, sorry, if you have to go and verify your data by watching the game, then your data is wrong.

You know, it's plain and simple like that. Or you're misusing your resource. The resource can be used in a very different way, shape, or form. So I think scouting in its purest form, from a player evaluation standpoint, was breaking down the player into the model that you feel he fits it. Now, it's an inexact science, but you gain that memory by doing all the work,

all those years and watching all these different kind of players to the degree that you're able to kind of put them into a simple little model. And then every other view essentially just has just out greater details about that particular player. At the same time, I don't necessarily need to know how good his own exits are or his own entries are, which seems to be the preconceived notion. Like, it seems like everything that is based out of coming out of analytics seems to be what the basis of scouting is, and it's not. I think they're true very distinct practices.

Breaking down a player and putting him into a specific model that you're able to build upon is very different from seeing how good he is at his own engine and how his own exit and how good he recovers, but he owns in his own zone, or how often he's on the ice or a shot on goal or et cetera. I mean, those things that you can actually look at

and verify through data, why are you bothering even trying to look at that? What you're trying to do is base a player on the four-s skill, speed, smarts, and skating. And once you break them down into that particular model, then you can take a look at where they fit within your own system. Do they fit within the system that they're playing in? And do they do things that are outside of the systems.

Like, for instance, I use this word a lot, freestowning. If a player has actually got such skill that he freestyles out of this system, if things are going wrong because he's doing that, then he's a detriment to that team rather than a plus for the skill that he's bringing. and I see that a lot in the minor leagues, right? You see guys that score tons, but they do it more of a lone ranger style. And, you know, at the same time, when it works, it looks good,

but when it doesn't, that actually creates strain on the team that he's freestyling it. So it's a lot of different factors that I think go on when you're actually looking at a particular player. And I don't think that we need to double up and verify our data by what we're looking at. I just think it's a waste. You're looking at certain things that you're expecting a player to be, and you're expecting that out of your data. And if it's not, is it wrong? Does it necessarily have to be wrong? Well, here's my question, because you mentioned the 4Ss, for example, right?

And I think there's sort of this idea we've all built up in our heads, and obviously some of us have different, different sort of outlines of this, but it's what a good player looks like, right? and the problem I have with that sort of thing is that good players come in so many different shapes and sizes. And I think that let's say a player is really strong or really fast or he does a lot of these physical things really well. But at the end of the day, especially from the, you know, I'm not talking about junior hockey, for example, or even the AHL where the results are less important and you're sort of trying to work on your craft and get there at the end of the day. But in a pro scouting perspective, if a guy, does a lot of individual things really well, but at the end of the day, it's not translating to results.

Then I feel like there's a divide there maybe we're just overvaluing the wrong things, you know what I mean? You know, at the same time, I've actually... Individual skills. The thing is, individual skills are useless if they are looked at in a bubble.

Right. You know, a player that can skate well and a player that can, you know, shoot well, it means nothing if they can't do those other things, right? And I'll use Ovechkin as an example, all right? Well, Vetchian isn't the best because he can shoot all the time or because he has a great shot. It's because he can skate.

He puts himself into positions where he can use that great shot. So it's not just skills or individual skills. It's the integration of skills that makes that player great. So, like above and beyond just skating in a good shot, Oveciman has got this game's beat. That's absolutely crazy. So he'll go into the corner and get the puck for what? To pick the shot?

he'll go and and he'll skate like crazy in the zone and he jumps at a defense means gaps. Why? Because you could take a shot. Because it's the integration of skills rather than the individual skills that we should be assessing. Unfortunately, that's a tough

stuff. It's something that's you need the experience and the ability to have seen certain things over and over and over to really kind of develop that particular model. If you don't look at that one individual skill that's making it successful, you need to look at the peripheral

that that player does in order to overutilize that one particular skill. I always say that every player has one elite skill in them, and it's how they utilize every other skill that makes that one elite skill so important. Would you agree that maybe, I don't know, five, ten years ago, or maybe even longer than that, that one skill for a lot of guys was generally size, and now people have realized that if that's your one elite skill, maybe you need to also have something else to support that? That probably was, I don't think it was a really fantastic rule,

but I think it was very important from a scouting perspective. Everybody looked at size because of big brash, physical game, et cetera, et cetera. Clearly that changed with the first lockout, you see it more fluid, et cetera. And teams that understood that adapted. Right, you know what I mean? They were using less bigger guys on the blue line. They were using more skilled guys. Now, a skilled guy doesn't necessarily mean skill.

at the end of the stick. It could be something like, you know, preventing a zone entry. And they do that by a particular skill. They either have good lateral agility or at the same time, like, I mean, a player that does good on zone entry, like a player that really does good

that doesn't allow zone entries is only as good as the support system around them. Right. So, I mean, you could have all the skill in the world, but if you have players that don't necessarily, they're not able to adapt to whatever you're trying to do, then those skills are pretty much more. wasted. So it's tough to think that it's always been one particular skill that's led as far as

size goes. I think that's done at this point in time. But you need to have other things that are able to offset that one elite skill. And that's something I think a trained eye kind of really doesn't need to know. That's what I think closeouts will end up. I don't know if the word is modifying in terms of how they watch the game, but like I think what they'll do is start to integrate less size, more skill. Where do these skills integrate? How do they integrate and how do they integrate with our system and our particular team? Before we keep going with the show, let me give a quick shout out to the sponsor of today's show of FreshBooks.

FreshBooks is a godsend for freelancers and small business owners specifically, but really it works for anyone who doesn't want to deal with the time-consuming cluttered nature of keeping track of all their financial transactions. Its interface is incredibly easy to use and it'll pretty much do all the work for you. I think it's most handy features that it allows you to fire off invoices a matter of seconds, and it requires just a simple information of who you're sending it to and what you're charging them for. And then after that, it really does all the work. It keeps tracks of whether your clients open the invoice, and it even goes so far as to remind them if they've fallen behind on their payments.

On the other end of the spectrum, it can also keep track of your expenses in one tidy location by importing and categorizing everything you do with your visa or your debit card directly from your online banking system. You can do all that from your phone or tablet if you choose. So for those of you that are always running around doing errands and going to meetings, FreshBooks Now, as well as an app that you can download to ensure that you're keeping your business in the palm of your hand. To start the free 30-day trial they're offering to everyone listening to the show,

all you've got to do is go to freshbooks.com slash PDOCast, and remember to enter PDRCast in the how you heard about a section. That's freshbooks.com slash PDOCast. Well, so another thing that is definitely a skill, but it's taken some time for myself personally to kind of, get a better grasp on it and I still think I'm way away so off from actually being confident in throwing it out more frequently but it's like the the kind of catch-all term of hockey sense right where you sort of it's a very loose definition where you can't really kind of actually say what it is until you really see

it and the perfect example for me is a guy like justin shaltz for example where you watch him play and he has a lot of physical tools, but for whatever reason, and he's been a lot better in Pittsburgh, obviously, kind of they've made life easier for him, and he may just well be a guy that was given too much, and he couldn't really handle it now that he's in a better system. He's doing a lot better for himself, and he has that support, as you mentioned, but he was always out of, out of place, and it seemed like he never really knew where he was supposed to be or where his guy was that he was tracking. Like, he'd always just go for these random laps in the defensive zone, and be, and look so silly at times.

And I think, I don't know, a lot of that probably has to do with coaching, but I think it's also an inherent thing, because then you watch a guy like Chris Tanev, and he doesn't have overwhelming physical skills, right? He's not the fastest skater, even laterally. He's not the biggest guy. He doesn't shoot that hard. He's not an, you know, exhilarating passer.

But at the end of the day, he's just always in the right position to make a play, and there's something to be said for that. You know, it's a typical thing to judge hockey sense, because I think that it's actually similar to statistics. I think it's something that you're looking at from the past. You don't read the person's hockey sense prior to make it a play. You see it after it's done, right? So it's a difficult thing to assess.

However, going back to your point, how much of that is actually the player and how much of that is actually coached. And I can tell lots of players that have really good hockey sense that are stifled by a coach that says, don't do that, or do more of this, do less of this. So now you're kind of putting this idea into the player's head that may not necessarily agree with his natural instincts.

So addressing hockey sense is difficult from the perspective of whether or not it's directly from the player or directly from the instructions that he's getting, whether it's on ice or from his poach and staff. So as far as I can see, I try to look for a player that, he should have gone left, he went right. He did it once. Did he do it twice? Did he do it three times?

Does he do this consistently game to game? Does it, and is it a coaching thing? Look at the system that this team is playing. So you can judge hockey sense. It's just a very, it's probably the most inexact of the inexact of the inexact sciences that's involved in coaching. You know what I mean? So it's not, it's not something that you'll ever get 100%.

As long as you have a good general idea of how smart a player is, you can use him in a variety of fashion. So it's not necessarily hockey sense, but smarts. You give me an intelligent player, as long as the skills are going to fade, and you know the skills are going to fade. Those smarts won't, and he'll be able

to compensate for a fading of skills. Jarimir Yeager. I mean, he's clearly not the score that he ever was, but he knows what he has to do in order to still be competitive at this day and age, and he knows where to be based on his lack of skill. At this point,

to his heyday, right? So his hockey sense is high because he understands what he needs to do right now. We could probably go on and on and on and on about this, but as long as the player kind of looks like he's doing what he's doing innately, you can kind of give him the benefit of the doubt. You have to really be careful, though. Hockey sense is not something that can be measured in terms of, you know, oh, he's got a good one because he does this.

Too many variables involved. really good point you made earlier about sort of the post hoc analysis of it where you generally do it after something's already happened as opposed to being like, okay, watch this. This is going to be great hockey sense. And for people like myself who are maybe more analytically driven or more based their analysis on data, it always makes me a little bit uneasy kind of because there's two different components, right? Yeah, it's, we're evaluating what's already happened on the ice, but it doesn't necessarily help us in prognosticating what's going to happen moving forward. and those are two completely different things

that people sometimes kind of blur the lines on. You know, the other thing about being analytically inclined, I mean, I think I've watched enough players to be able to analyze, and that's essentially what you're doing from an analytical perspective, right? You're analyzing whatever your inputs are in order to create some kind of an output right or wrong. Database analysis is very different from the I test, which is a phrase I absolutely hate.

Because I think, see, the reason, is because I think for the most part, what the statistics have done, and not the animal of the statistics, it's elevated the general perception of what we look at about hot squares. At the same time, I think what it's also done is

it's inadvertently given permutations to different types of aspects that we have online or in conversation, you know, hashtag scouting. And that's not really the case. I think that it's a lot more intricate than that. And I can go and watch a player 10 times,

and I use an analytical mindset in order to develop my output. Right. But one is either data-driven and one is not. The future is definitely data-driven, because why would you be stupid enough not to use that? Like, that's just... Right. If it's there, and that's all fine and good,

but as long as we understand what that data-driven analysis is for, and what other analytics, clinically inclined analysis is for. So, to me, I think that the eye test, especially the way that it's being portrayed most of the time online, is based on something that, oh, yeah, okay,

well, analytics says this, this player doesn't do this, he failed the eye test. Right. And it kind of goes back to my original point. If you need to watch the game in order to verify your data, there's something wrong with the data. There's nothing wrong with the eyes. Right. Yeah, and I definitely think

it can kind of go overboard sometimes, and there can be a straw man built up on both ends of the spectrum where, and I'm personally responsible myself sometimes, where I just, a few minutes ago, called myself a more analytically inclined individual, and I think that there doesn't really need to be, you know, like sort of definitions and stipulations placed on it. Like, I watch 40 hours of hockey a week almost, and I also spend a lot of time digging into the numbers,

and both those things work together. It's not, you're not picking one or the other. If you are, you're missing a large component of the whole basketball. And what you're actually doing is putting the best of both at test, right? Like, I mean, you develop your staff's-driven analytical skills more so than your scouting-based analytical skills, and that's all fine and good. I mean, there's absolutely nothing wrong with that, you know? And I would bet, I would bet even though you probably aren't really actively doing this, it's probably sinking into your head about certain things. I bet you got one idea that I planted about every player having one elite skill.

you probably see it. You don't recognize it or haven't to recognize it because you're not really looking for it, right? Right, of course. But everybody can distinguish something like that. Everybody knows that one guy

has great so nominal shot, right? And it kind of goes back to my simple point of it's not about the individual skills, it's about the integration of all those skills that makes player better. I love the fact that I can look at data right now and find things that

I just don't see on the edge. It's not necessarily from the skills perspective. It's stuff that makes him want to go, hey, you know, why is this guy used in this particular situation? Why are the coaching staff not utilizing the skills that he really uses in order to really be successful? Yeah. You know what I mean? So it's the integration of both types of analysis that I think will be, it's probably essentially the model that I would try to. If I had an organization, I would try to get a little bit of both.

And I'm not talking about one over the other or one is better than me. I'm just thinking that they need to be working in conjunction with each other. They need to hide all the holes that are possible are available through both types of analysis. You need to be able to cover every single aspect, as much as you positively can. Nothing is perfect. But we try to be as good as that. And things like, for instance, like salary and Claireworth and all of that,

like, I mean, skills are one thing, but I think that that is all really statistically driven. And as long as you understand that, and you don't answer crazy questions about scouting with numbers. And you don't ask stupid questions about numbers from the scouts. The scouts are they probably know, but they won't be able to give you the same kind of answer that you're looking for based on what we're most likely expecting. Let's say an online writing or something like that. So it's tough. Both things have to work in conjunction with each other. If they don't, you're wasting your time.

For sure. All right. I think the most interesting case studies on an individual player to player basis are the ones where there is that sort of divide where you can't really, for whatever reason, kind of fill those holes, those blind spots, right? So, like, obviously there's guys like Patrice Bergeron, for example, where everything comes together beautifully. And as you said, it's kind of a waste of time, like, to be, you don't need the numbers to support what you see or you don't need to see what, like, there's a marriage there where. Patrice Bergeron is a remarkable player and pretty much everything he does translates into any sort of number

you're going to look at that tells you that he's a really good hockey player. But then it's not as simple for a lot of other guys and a guy that I threw out to you when we were discussing kind of what we wanted to talk about in this show was John Carlson where there appears to be a massive divide between how he looks physically

like if you look at those individual skills where he's a big guy but he skates well he has puck skills, but then I don't know if it's sort of a system thing or how he's being used or what the divide is there, but it doesn't really translate into the numbers. Like the capitals aren't any better when he's on the ice as opposed to when he's not on it, and I'm kind of wondering what's going on there. Yeah, well, that's the thing, right? If you're looking at specifically from the perspective, okay, well, he produces X, but he looks like Y doing so. the Y is probably a lot more difficult to really explain. And at the same time, it's not like I can technically tell you,

okay, he does this because of one, two, three, four, five. I think it goes back to the combination of skill. For instance, John Carlson has a fantastic shot. And we see it at the point. But he doesn't just stick to the point. He uses his space up there. And at the same time, teams probably know that he has such a good shot. So it's actually funny.

when I was looking at some of the passing data from the Capitals power play prior to Christmas, the major C that we all expect that's going to come is from the other side of the boards, and it's not John Carlson at the top of his own. So, I mean, you're looking at a guy like that and seeing his individual skills, but when you start putting things into conjunction, for instance, he rushes the park much better than probably any defenseman on the caps right now. So he replaces Mike Green in terms of what he does there. that may not necessarily show up on the score sheet

in terms of the numbers that we're expecting, but that's an important factor. If you don't have anybody that can rush that puck, you're defaulting to players that can only do outlet passes. That's a Pittsburgh Penguins problem, especially with Christopher Tang out of the lineup. And you're looking at trying to adapt to what Carlson brings from the opposition's perspective,

too. You're trying to limit his rushes. You're trying to limit his own time at the top of the zone on the parplay. It may not necessarily look like he's contributing in terms of the way that we see things online, especially from a statistical perspective. But it's those little subtle things that eventually will show up somewhere, somehow. RFID will eventually take care of a lot of that. But because it doesn't show up in our current ability to measure items, we look at it as more of a detriment.

And I don't think so. I think that that might be one of the perfect occasions where you're actually looking for somebody to say, well, why is this happening? And you start looking at the player and you put him into that model that I was talking about earlier. And if he doesn't fit that model, then you have issues. Then you need to start making decisions about, okay, is this really the player that we were expecting? Or is this really the player that he's becoming, et cetera, et cetera. So it's really, really difficult to say, well, he looks bad, but his numbers are good.

Or his numbers are fantastic, but he looks bad, you know? I think you could say that about a million other defense and right? I think that based on sort of the information we currently have presently available to us, there is a massive kind of shortcoming there. There's pitfalls we can get into an evaluating defenseman because you can't really point to, you know, big point totals or you can't really, you can get into trouble when you try to overvalue glaring mistakes they make. Like there's certain guys that give the puck away, for example, on occasion,

and it might lead to a bad, goal against and then fans latch on to that and be like, oh, look at this guy. He's such a liability. But then you miss out on 10 other sort of subtle plays he made where, as you said, he rushes the puck out and sets his team up for a few possessions to come. Or, you know, he does a certain positional things or he sets up his teammates to do things themselves. And we just don't really have a lot of information right now readily available to kind of put distinct qualities on that. But it is something that it will eventually come better over time. It's difficult really to do that kind of an assessment with defensemen in particular.

I mean, you know, going back to my, actually this is a pretty good example. John Carlson, I said it was a good puck rusher, and I think that it's a distinct difference between a rusher and a mover. A rusher is able to do, take a Thomas Cabrillet or Drew Dowdy or somebody that's able to go into end and not limit or at least put more defensive risk into the gunnoblin. limiting the defensive risk of moving the puck up the ice. An outlet guy, a puck mover is someone that might stay to the blue line or the top of the circle and move the puck out with a pass.

Now you're putting more risk in there because there's the chance of interference, et cetera, et cetera. The evaluation methods are just not conducive to what we want measured at this particular point in time. You know, I'll use Cody Franzen. It's actually a pretty decent example. Here's a guy that's, I mean, to me, I can't understand why any team would want to sign him long term, but yet his statistics are incredible. Like, I mean, you can see that he's able to do things in the offensive zone that a lot of other defensemen may not be able to do. But at the same time, he can't rush.

He's not a very good puck over. I mean, he makes decent out that passes, but you can find other players that are able to do that kind of stuff. at the same time, he has physical attributes that are detrimental. He can't do lateral turns, his pivots aren't very good, et cetera, et cetera. So the positive stuff that he does from an offensive perspective are really balanced by a detrimental. Damn, I'm going to have to use this word, the way that he looks. You know, like his quote-unquote eye test, it looks awful.

Right. You know, and how do you base an assessment at this point? based it on the fact that he does all these good things in the offensive zone, or are you looking at the liability that he could possibly be if he's on the ice and some team really keys in on those liabilities? So defensemen, at this point in time, are a very, very difficult thing to assess, specifically just from a stats-driven analysis. I think that once you put both together, you've got a fantastic picture

of what this player's possibility is and what type of liability. abilities you may become in your own mind. Yeah, no, for sure. All right, one final thing before we get out of here. You've been doing a lot of work in your recent post, kind of diving into the passing project, as you mentioned earlier, and digging into it and applying it, which I think is the important thing,

because sometimes it's easy just to regurgitate numbers, but the difficulty is getting practical use out of it, and I think that you do a really good job in that regard. And I'm a big believer in these sort of micro-tracking components, of the game because they can. You can piece them together and get a bigger picture, as you said. So kind of let's discuss that a little bit. Where do you think that's going to go in the next few years?

And what have you been doing with it recently? Well, I absolutely love the passive project. Personally, going back to that thing about pro-scouting being more of a coaching thing, that's essentially what I think is going to end up becoming the future. RFID data is good. But unless you start putting players into that PlayStation kind of, kind of tracking suits that tracks everything from head to toe.

I don't think that you're going to get a very good skills evaluation, but you can do a hell of a lot of good stuff with strategy. And that's why I think the PASC project gives us a nice little glimpse as to what teams could possibly be doing with some RFID or even proprietary data. So what you're trying to do here is

try to find the strategy in the game. What are you doing? What does your team do? Does this data conform to what your team is doing? If not, where is the problem? So what you're doing is isolating issues using a massive set of data filtering down to what you feel might necessarily be the problem. Whether that problem is identified directly through the data or a scout or somebody else, a coach, for instance, decides, you know, maybe we should key in on this particular area. That's where the future of the passive project is going to be fantastic in the public forum. I use an example of stretch passes

and we go back to that puck rushing versus puck mover and I looked at stretch passes for some teams and see, do they actually do move the puck very well? An example right off of that, Patrick Kane, Patrick Kane, if he receives a pass in the defensive zone, he can take it in all on his own and take a shot. If he takes a pass in the neutral zone, the chances are he won't end up with that particular shot.

Or if it does, it's not necessarily from a good shooting position. So now you have this. different variables that may end up coming out of the data that you're actually analyzing, which could be mind-blowing and really, really overwhelming at sometimes, especially if you're really looking at a big data set. So it's key to focus in on the things that you're trying to improve or problem areas that you're trying to avoid. The one detriment of the Passen project is there is no data if there isn't a shot of that. And I think that that's a big detriment. There is a lot, a lot of information that you could really, really clear up as to why you

didn't this play end up as a shot? Like, what happened? Did it get... Like, did it die at the blue line? Did the defensive team turn it over? Did the offensive team turn it over? Did the defensive team stifle, etc., etc. There was so much information that is kind of missing, which unfortunately is part and parcel

with this project in itself. I mean, we're doing it on a manual basis with a bunch of volunteers on their own free time. It's tough to get everything right now. But if you have RFID data that is able to kind of reproduce what we're looking at here, you can have some fantastic insight that was never available in the past.

So the passive project on the public forum is a fantastic tool in the future to start gauging what teams and players are doing. In a private sphere, I feel that they probably use that data with things like sports logic or RIFID data, you know, GPS within the Jersey or whatever the case. So I love the passing project of what it's able to give us in the public sphere. I would absolutely love to dig into some of that private data as well. Yeah. Well, keep fighting a good fight, man.

And thanks for coming on the show. I think we honestly could have talked about all of this stuff for another couple hours. There's so many nuances that we didn't really have time to get to. So how about we just put a bookmark in it here and agree to do this again sometime soon? Dude, the pleasure is all mine. Thank you very much for the invitation. And I would end-time be it. So people can find you on Twitter at Katz Hockey.

and I don't know, how often are you writing? I feel like you write pretty often. I'm doing one column a week on Rotoworld, specifically for an analytics perspective, and my other stuff is on mixing hockey. So that's a little bit of a sports sport. It could be analytics. It could be statistical stuff.

It could be, I really love systems. To me, I think that systems are a big, big thing that we don't at this point assess very well. So I'd like to delve a little bit more into that, whether it's through data or specifically through video. Cool. I wholeheartedly checking out, recommend checking out Gus's work.

And Gus, we'll talk soon, okay? Thanks very much, dude. The Hockey PDOCast with Dmitri Filipovich. Follow on Twitter at Dim Philipovic and on SoundCloud at soundcloud.com slash hockeypedocast.