Nov
22

NHL Expectancy of New York Rangers Prospects – A Cohorts View

November 22, 2017, by

lias andersson

As you all know, I absolutely love looking at prospects and the upcoming draft. With more information readily becoming available, I am enjoying incorporating everything from individual scouting to the underlying numbers that have been unearthed due to research projects in the analytics community. Over the weekend, I came across a tweet by Byron Bader of the Athletic here that shows how likely a prospect can become an NHL player by using NHLe as a key. This piqued my interest.

I then incorporated Christian Roatis of NHLNumbers NHLe calculator to get an accurate NHLe for some of the Rangers forwards prospects. Byron’s tweet came at just the right time for me, I’ve noticed people have been complaining about picking Lias Andersson over Casey Mittlestadt due to point production, and while we can make the argument in regards to potential, the point argument in my opinion is invalid. When comparing NHLes both players have a nearly identical number which would also put them in the same cohort for Byron Bader’s NHL likeliness/superstar key.

NHLe and Potential Setbacks:

NHLe is an interesting tool we have for prospects when it was originally pioneered by Gabriel Desjardins and improved upon by Rob Vollman. Essentially, NHLe compares how difficult it is to score a goal and point in the NHL vs all of the other leagues in the world. You can compare a player in one league’s point production to his point production in the NHL the year after, assuming that the talent within the 1 year period would remain consistent. This allows us to potentially predict a player’s production in the NHL the year after. By having this, we can standardize a prospect’s production for each league and compare players who would otherwise have vastly different competition.

There are certainly some possible setbacks though. First, forwards provide more points than defensemen. Including both in calculation for a single league NHLe can be misleading. This is something I intend to research in the near future and see how NHLe for forwards and defensemen can differ for each league.

Another thing that I am concerned about is that NHLe uses points per game. This is fine in the sense that we can calculate it for every league, but if we can somehow incorporate ice time and maybe a 5v5 NHLe, we may be able to also get a finer reading. Unfortunately not all leagues provide such stats so this can only be done for a few.

I feel the P/Gm may also dis-proportionally hurt KHL players. A majority of players under the age of 20 get very little ice time, that can lead to having 0 points in the KHL and an NHLe of 0. A prime example is German Rubtsov, the Flyers 22nd overall pick in 2016. Last year he had an NHLe of 0 in the KHL playing mostly 4th line minutes. He left halfway to the QMJHL and had an NHLe of 29 there.

In any case, all information is useful to provide a deeper report on a prospect. With a level head and some caution, we should leave this post with maybe more information than we all had about a week ago. I have separated my visual into two parts. The first will be following Byron Bader’s NHLe key based on draft year production. The second part is comparing the percentage of being an NHL player from Byron’s key and comparing it to the percent likeliness of a draft pick in a certain round being an NHLer from two sources. This allows us to maybe see if the Rangers would be operating with a prospect that is currently exceeding expectations based on former picks or not.

NOTE: Byron’s key provides little opportunity for some players to have an NHL likeliness of over 80. Travis Yost noted that 80% of first rounders make the NHL, so  I removed Filip Chytil and Lias Andersson from that analysis.  If one were to be interested in where they stand in NHL Likeliness,  I can gladly provide it in the comments section. 

Part 1 – NHL Likeliness

Click to enlarge

First thing we can immediately see is how difficult it is to produce points when you don’t get much ice time. Filip Chytil didn’t have a high NHLe during his draft year, but the model put value in him producing as a 17 year old in a pro league. We see that now as the Wolf Pack’s first line center, Chytil’s NHLe flies up to 47.1. So honestly welcome to the Filip Chytil show. He looks in line to not only be an NHLer but likely a first liner that can be a major part of the Rangers core.

Lias Andersson has been the target of a little bit of vitriol by some fans commonly in relation to Casey Mittlestadt. Interestingly enough, Lias Andersson had a higher NHLe than Mittlestadt, who played in high school last year, and they are at an identical NHLe this season as well while Casey is in college. The case study between them seems a bit overblown and perhaps it is with people valuing the NCAA a bit too much relative to the SHL.

We are also seeing something pretty incredible regarding Tim Gettinger and Ty Ronning. While they are late round picks, their production is actually near a 50/50 chance to become an NHLer, with Ronning currently exceeding that this season. They like to say a prospect either becomes an NHLer or not, but the reality is a 5th and a 7th rounder likely don’t have that chance.

As of right now the Rangers are in the presence of two prospects that can provide them great value. Looking at this for forwards I do tend to wonder: At what point do you feel a prospect is worth trading as opposed to hoping he makes the NHL? Is there a certain percent-chance that some people consider a threshold? Does it also depend on their draft position? Would a player drafted in the 2nd or 3rd round with a likeliness at 30% be traded before a 6th rounder who has a 15% likeliness?

Part 2 – Difference in Draft Expectancy

Click to enlarge

For this stage of the analysis. I compared the percentages that were provided in work done by Travis Yost on TSN (linked above) and Adam Gretz on SBNation. The big thing I wanted to look at here was how the Rangers have drafted based on Bader’s key. I also wanted to see how the Rangers prospects are faring as they move years away from their respective drafts.

For this I took the NHL likeliness and subtracted the percent likeliness from the articles mentioned above. Anything above 0 suggests the Rangers may have found some good value and they have a better chance with this prospect than the average picked in his position. For negative it is vice versa.

We can immediately see that regarding the draft, every prospect the Rangers drafted had an NHLe that correlated to an NHL likeliness that exceeded the reports (Lakatos/Virta start at D+2 as they were drafted as overagers). In case you do not want to open the articles here are the percentages for the rounds that apply to the NYR prospects.

Round Ranking (NYR) Expectancy on Round via TSN Likeliness via SBNation TO appear Likeliness via SBNation TO 100 Gms
Round 7 18.4 27.1 11.6
Round 6 19 30 14.3
Round 5 20 29.9 14.2
Round 2 44 65.7 31.1

The Rangers have showed that they are adept at finding value in the late rounds, all players were more likely to make the NHL than the average in their round. Lets start with the bad news first. While Dominik Lakatos and Patrik Virta have looked alright with Virta even improving, given their age you’d like them to take the next step offensively. As mentioned earlier there may be issues with using NHLe as it currently is for predictive purposes but it does seem at the very least like a good key. They still have a chance but you want to see a bit more offense for their age to be more of a sure thing.

Ryan Gropp and Gabriel Fontaine also started off pretty strong. Gropp was right in line with his fellow 2nd rounders, though you can make the argument for the Barzal boost here if you so choose. Fontaine had looked like a pretty nice bet relative to his draft round to at least get a cup of coffee, with a shot to potentially be an NHLer. Unfortunately, both players had their likeliness take a nose dive as soon as they made the AHL. Shoddy production on both ends can be derived from many things, but I’d say Gropp, given his reputation as a point producer, is in a bit more danger than Fontaine. That said, the Rangers took players who were valued above their draft round selection.

But again, how good do Gettinger and Ronning look? I think it is so funny how opposite they are, but they are proving to be significant value for the Rangers. As mentioned in the earlier part, they are at the point in their development that it looks like they’re hovering around a 50/50 shot to make the NHL. That is huge value for a 5th and 7th round respectively.

It is so much higher than the expected rate for their rounds that you have to tip your cap to Clarke and the scouting staff because as of now it looks like we are more likely to get some NHLers than we were on draft day. I should also mention that in Ronning’s short stint in the AHL last year, his NHLe was identical to the same season in the WHL. Not only does this reinforce NHLe as a transfer coefficient but it also shows what Desjardin’s mentioned in his post:

“player’s skill level is approximately constant over this two year period, the ratio of his performance in each league can be used to estimate the relative difficulty of the two leagues.”  – Gabriel Desjardins

Morgan Barron, coming off his surprising 7 game point streak, is also starting to look like a value pick. He produced enough to be considered a good pick for the draft but NHLe for his league isn’t exactly known. Given that he played in Canadian High School I used Roatis’ Highschool coefficient so his draft year number may not be accurate. In any case, his production at the NCAA level at the moment is proving to be quite significant and can be a prospect to watch.

Overall, it is hard to project prospect success, but we can maybe gain some peace of mind knowing that the Rangers seem to be a pretty good team in drafting forwards in the later rounds. We can be getting excited knowing that Filip Chytil is starting to look more and more like  a top line NHLer. Perhaps we can stop also ragging on Lias Andersson, seeing that his NHLe is in line with someone like Casey Mittlestadt at the moment.

On the other end maybe it is also time to start looking at Gropp and ask ourselves, do we believe he can bounce back with more ice time in the AHL, or is he starting to look like a prospect to trade while he still has value? Perhaps we can revisit this all in January or February and hopefully he’d have more opportunity by then.

"NHL Expectancy of New York Rangers Prospects - A Cohorts View", 5 out of 5 based on 13 ratings.
Categories : Prospects

47 comments

  1. Richter1994 says:

    Great as always Josh, thank you.

    Gettinger has been very interesting to me for a while. Gem in the rough pick?

    • Reenavipul says:

      I wrote a while back that you hope he can be a Mike Rupp and his scoring rate puts him pretty much there(he’s on a 38g pace, which is NBD in junior for a 19yr old.) He’s not a like to fight guy, but will go, so there’s box number 2 ticked. They shouldn’t send him back as an overager as he’s done all he can in the Soo.

      A very good skating Mike Rupp? Not bad for a 5th.

      • Richter1994 says:

        thanks for the info pal, you seem to know your stuff on these prospects.

        any time a 5th round pick makes the NHL then that’s a good pick.

    • Joshua Khalfin says:

      Depends what you mean by Gem. I don’t think he will be a top 6 guy but his production relative to ice time is nice even if you factor in age as he was one of the better guys in the O last year. As a 5th rounder I think we just all hope for an NHLer and I believe we have a better chance with Gettinger right now than maybe some other guys other teams picked. He is progressing hopefully it works.

      • Richter1994 says:

        Thanks Josh, yeah, if Gettinger can make the NHL in any capacity then the 5th rounder is a very good pick.

  2. Reenavipul says:

    NHLe has too many variables for junior eligible players playing in European men’ leagues due to seniority issues. Lots of noise, need to somehow include prior J20-J18 numbers to have a more accurate projection.

    Watched TPS last night, might be worthwhile selling high on Virta. Does some nice things, but I’m pretty sure his skating isn’t good enough for the NHL. Not even sure it’s good enough for Hartford. Didn’t look strong enough to win puck battles. Never a good sign when being your scoring rate is less than a guy twice your age who had less than 100 NHL games in his career.

    Gropp needs to earn more ice time, hit the gym. Likely a year away from putting it all together.

    • Walt says:

      Reem

      To your comment about hitting the gym, he is a big framed kid who seems to lose the battles along the boards. Is it his strength, or lack of desire? Interested to hear your response……..

      • Reenavipul says:

        Virta not a big kid. Not just losing battles on the wall, any 50/50 puck. Not strong on the puck in the least. He’s a one touch kind of player, for better or worse: his goals especially come that way. They drafted him as a 20 yr old, didn’t even invite him to camp. There’s easier ways to find AHL fodder than a draft pick.

        Gropp needs to fill in, playing the right game in the wrong role if you expect him to score from day 1. Being on the line with Fogarty will be good for his all around game.

        • Walt says:

          Reem

          Thanks for the reply on Gropp, needing to fill out his frame. To your point, he can score, but maybe they are trying to teach him to play defense as well. Could be a nice third line guy, with a scoring touch down the road!!!

          • Reenavipul says:

            He’s got the skill to play a power game and is willing to go into traffic, but without the strength to play that game, you have to have sublime skills like Buchnevich. Gropp does not.

            So it’s bulk up or ship out. He’s got all of this season and the off season to do that, then turn heads at camp.

            Only thing stopping him is him. If he gats to 200lbs this year, his game will pop in the A. 205 by camp and he will be able to play a 4th line role.

  3. SalMerc says:

    Let’s not forget we also draft for our AHL team. It is likely that some people are drafted to augment that team with little focus of taking the next step.

    • Reenavipul says:

      If you’re drafting for an AHL team, you need to be on UI. You’re drafting skill ceiling and trying to fix the flaws that limit reaching that potential you can have em put on weight(easier for some than others) you can teach them to skate better(easier…) You can’t teach height(which itself means nothing without skill), you can’t teach guts, you can’t teach tough.

      You can teach a guy with no skill( who can skate) how to play a wholly defensive role. A guy who has skill who can’t skate is having a tough time in this league right now. If Daryl Sutter ever gets back to coaching, getting him all the Strome brothers(at a discount) might work.

      • SalMerc says:

        All I am saying is that teams draft in the later rounds hoping for a diamond in the rough, but also know they need to put a roster together for their AHL affiliate.

        Do they draft using these tools with high expectations? Of course, but reality sets in for more than 50% of those drafted that a call to the big club is just never going to happen.

        • Reenavipul says:

          Just look at Hartford for how they handle the CBA. They’re putting players on ELC in a place to succeed, then signing UDFAs to take a few more chances, then AHL vets for skill and supporting the kids. Rare is the prospect who gets a 2nd contract who doesn’t get a few NHL games in by year 3.

          Guys with AHL ceiling like Keegan Kolesar don’t even get offered. Leedahl Dawson(who had similar counting numbers than Kolesar) got a short ELC as a UDFA and is likely in the ECHL until it expires.

          With the 50 man roster it’s tough to get on the treadmill, let alone stay on it. Doing a Krzysztof Oliwa is pretty much impossible now.

  4. Reenavipul says:

    With Morgan Barron, the level of competition is negligible, you just have to look at prior scoring rates among peers. Barron was ahead of solid NCAA players. Add to that his frame & size, you’ll get a 4th liner out of him at a minimum.

    That school is gonna be a college & NHL pipeline for years to come as they have 1st rate facilities. Now this might limit their ceiling in some respects, but you’ll know what you’re getting.

  5. Uptown Girl says:

    I didn’t see any data on Letteri. What does the data project for him?

    • Joshua Khalfin says:

      Hello. I didn’t do anything for Lettieri since he is already 3 years passed his original draft day and over 22. That said Lettieri especially one the first line with Chytil has been looking like an NHLer. His NHLe is 29.4. He is definitely older than the key, but that is a pretty nice NHLe to at least be strongly considered for the NHL. Hopefully soon eh? His last year in College he also had an NHLe of 27 so it is not like this is a massive flash in the pan. I think we should have an NHLer here.

  6. Reenavipul says:

    For using TOI as a variable, it’s sort of overkill to see that data. Just looking a where they slot in scoring totals by position and discounting accordingly should give you all you need to know. It would be great to get really granular on the data, but P60 isn’t giving you much more than you need.

    The one thing you might want to sift through is the point drop off curve from line to line and see who is outside the curve. Gets you a similar result with a lot less work.

    • Reenavipul says:

      Would love to see a counter argument than a thumbs down, but the internet is filled with gutless wonders.

    • Joshua Khalfin says:

      I think line to line is harder since there is no way to standardize lines among teams. P/60 provides us with ability for projections. Not only would we factor in the time they are playing but we can also likely use it to predict the point production if said player gets to play on a certain line in the NHL. If their P/60 is something like 1.56 then we can in theory project what points they’d get on line 4 duty, line 3 duty etc. I think I understand your line drop idea but I feel in regards to projection which is really the purpose of NHLe it might be unnecessary. Regarding overkill, I’d rather overkill than not have info. If there are major discrepancies we can go even deeper and try to find the source of that.

      • Reenavipul says:

        I understand your wishes, but deploying the resources necessary at junior to make it happen are unlikely without NHL central scouting kicks in some serious money. Even then, what you’re getting is marginal in relationship to counting stats bar D performance.

        In a stats only scouting environment every data point helps, but if you’re in your draft year and you’re not on the PP, you’re getting drafted in the last 2 rounds or not at all. This is where even video scouting helps, just to know who’s on the PP & PK and who’s not. You’ll get 95% of the info needed to adjust your formula on a player with a hell of a lot less work.

        • Joshua Khalfin says:

          Absolutely you get information from video but your other point about PP is why I mentioned ice time. A good number of leagues especially juniors do in fact have the 5v5 ice time. We can and I try to in my visuals allow a section to compare even strength production and all situations.

  7. Joshua Khalfin says:

    Depends what you mean by Gem. I don’t think he will be a top 6 guy but his production relative to ice time is nice even if you factor in age as he was one of the better guys in the O last year. As a 5th rounder I think we just all hope for an NHLer and I believe we have a better chance with Gettinger right now than maybe some other guys other teams picked. He is progressing hopefully it works.

  8. wwpd says:

    Josh, interesting work and always appreciate the perspective. Have a great Thanksgiving!

  9. Ric Erdheim says:

    Very interesting article and another important datapoint.

    But using the points generated by draftees doesn’t necessarily provide a complete picture of a player.

    You made reference to the relatively uninspiring point output so far this year by Ryan Group. Maybe Ryan Gropp is going to be a bust and drafting him was a mistake. But Gropp, if he has any potential, is a player who isn’t going to generate much offense on his own. He needs a linemate who can put him in positions where he can use his speed and shot. Clearly having Matthew Barzal as his center helped Gropp. But at Hartford, for the last few games his center is Stephen Fogarty. Whatever one thinks of Fogarty he is not someone who is generating chances. In 17 games this year Fogarty has exactly 0 assists. And so the question is whether Gropp’s lack of production is because he is not that talented, or he still has some maturing to go through or he is not being put in a position to succeed offensively.

    I had the opportunity to see one Hartford game a few weeks ago in Hershey and I thought Gropp showed good speed and had a few good chances. If this one game is representative — and it’s not much of a sample size — then there still may be hope for Gropp.

    My point is that while the data that you have in the article is extremely interesting and useful, I think we need to see if the eye test confirms the data. I agree with your conclusion that it is hard to project prospect success.

    • Joshua Khalfin says:

      For sure man. Stats of course aren’t everything but I’d rather be operating with a little more information with this than maybe having to rely solely on other people’s reports. None of us really have the time to watch every prospect game so I think the stats and the projections like this allow the common fan to be working from a stronger position. There is always a chance of failure but in Gordie we trust.

    • Reenavipul says:

      1st month back w/Barzal, Gropp did nothing, then caught fire.

      Fogarty not only has zero assists, his goals are all unassisted. Line so focused on being shutdown, generating zero possession.

  10. joe from newburgh says:

    NHL drafting is, to some extent, always a crap shoot, even among first rounders. Remember Pavel Brendl (Rangers)? How about Alexandre Daigle (Ottawa)? Or Patrick Stefan (Atlanta)? Nikita Filatov (Columbus)? Alexandre Volchkov (Washington)? Despite all the scouting and statistics, some prospects just can’t take the next step, and nobody can say why. Getting a couple of solid NHL players in a draft is a good thing. Getting a star is a real long shot, unless the team has the first or second overall pick. Below that, toss a coin.

    • Joshua Khalfin says:

      Absolutely agree Joe. This is also why I enjoy looking at the statistics. Of course it is less likely for a 7th rounder to make the NHL as opposed to a 1st rounder and doesn’t mean a 1st rounder can’t fail. I did this to see if perhaps the Rangers are operating from a position of strength and finding guys who are more likely to be in the NHL in their draft position than the rest of the league

    • Reenavipul says:

      Stefan lost the concussion lottery, Daigle had never seen adversity, Russians gonna Russia. Brendl was the one out of nowhere unless you had your ear to the ground.

      The thing with North Americans are normally self sorting, unless you’re DeAngelo and a handful of others. With Euros, it’s largely a function of lack of information, which google alerts should fix.

  11. Eddie!Eddie!Eddie! says:

    Josh-

    This is absoultey phenomenal. It’s articles like this that make this site the best Rangers blog on the planet. Thanks for all the time and effort you put into this!

    • Joshua Khalfin says:

      Appreciate it Eddie. If I do recall, you were a Lettieri fan. somewhere in the comments I did talk about him. He looks good too.

      • Eddie!Eddie!Eddie! says:

        I don’t think that was me. I tend to be very neutral on all prospects. I basically assume all of them, to quote Niedermyer from Animal House, are “worthless and weak” until they prove to me otherwise. :). Just years of being teased by the next supposedly “great” player leaves me perpetually skeptical when it comes to prospects.

        That being said, it appears Lettieri is making good strides. Perhaps we shall see him soon.

        Again, great work!

  12. Chris F says:

    Josh,

    Do you think it’s fair to say that NHLe actually undervalues player potential?

    You mentioned that the NHLe is based on talent remaining the same from current year to next. But given that this is a model for prospects, the majority of the subjects are young players developing their game. With that in mind, shouldn’t we expect that for most players in this model, their skill level will actually increase from current year to next?

    • Reenavipul says:

      If you know role, then you can contextualize. If you know the context, you can project. It’s not as granular as P/60, but you work with what you can get.

      You show me a team scoring list and break it down by position, I can get within 95% of TOI. 95% of the time it doesn’t matter unless you’re on London or Windsor or Portland.

    • Joshua Khalfin says:

      As Reenavipul mentioned, it’s all about context. It gives us quality info but we gotta use it within reason.

  13. Pete says:

    I am confident that both Chytil and Andersson will be centering lines for the Rangers in the not too distant future. Chytil has possibly elite skills and Andersson does everything well with a high hockey IQ and enthusiasm. Those who rag on Andersson haven’t seen him play much, if at all. They are why I do not favor spending valuable assets on a quick fix Center this season.

    • Walt says:

      Pete

      Well said, that was also the case with Oscar Lindberg, who took a few years to develop into what he is today……I remember seeing him at Hershey, when the Pack came in, and was excited about the kid, and posted so….

    • lv says:

      But Andersson was a stretch pick too, like Malhotra. If he tops out as a 3rd or 4th line center, then it was not a great 1st round pick.