rakish Posted May 14, 2016 Author Report Share Posted May 14, 2016 So finishing up this section, I will talk about the 3 teams I think are using analytics to draft. Now some may be a false positive, after 2014 I thought St. Louis, Nashville, Detroit, and Tampa were using analytics, but all 4 had terrible 2015 drafts. And there will be more teams drafting this way in the future, as the 2 kids from nhl numbers, who have a tool similar to mine, have disappeared. The Toronto Maple Leafs might have drafted 5 NHL players out the 2015 draft. A couple were on my draft list from last year, Marner, of course, and Timashov. But if you look at the Toronto Chart below, they got 2000 rating points outside of Marner. How did they each play in 2015? 9(th best 2015 season for forwards) MITCHELL MARNER OHL 1091(raw rating)17 DMYTRO TIMASHOV QMJHL 764 Defensemen3 ANDREW NIELSEN WHL 8427 TRAVIS DERMOTT OHL 5189 in 2014 STEPHEN DESROCHER OHL 440 - Overager 15 in 2014 STEPHEN DESROCHER OHL 381 This is a new skill that Toronto has developed, as the chart indicates there was nothing like this before. The Islanders use a tool similar to mine. And like Toronto, this is a new skill, as their 2014 combined rating shows no indication of drafting ability. The Jets also drafted 3000 rating points last season. They are coming off two solid drafts, then a great one. Quote Link to comment Share on other sites More sharing options...
rakish Posted May 15, 2016 Author Report Share Posted May 15, 2016 I know what your're thinking. You're thinking that you'll take a look at Central Scounting draft ranking, because they draft correctly, and see who they like at #8 or #9, I forget which Buffalo has. While I can't grade McKenzie over a long term because TSN lacks the server size to keep McKenzie's old draft lists up, thedraftanalyst.com saves old CSS draft lists, so we'll see the results of drafting correctly. This is the list Buffalo would have taken had they used the Central Scouting and chosen the top ranked North American propect available at their draft pick. I'll also list who Buffalo actually chose, 2005-2010, so six years. The numbers are, of course, years played in the NHL, games, goals, assists, points. The Central Scounting team makes Regier's drafting look visionary. In these six years, had they use the CSS NA rankings, Buffalo would have drafted NHL players Tyler Myers, Mark Pysyk, and I think Jordan Schroeder might still be in the NHL somewhere. 2005:#13 Buffalo Sabres Marek Zagrapan © #48 Buffalo Sabres Philip Gogulla (LW) #87 Buffalo Sabres Marc-André Gragnani (D) 6 78 3 15 18 #96 Buffalo Sabres Chris Butler (D) 8 387 13 71 84 #142 Buffalo Sabres Nathan Gerbe (W/C) 8 394 58 80 138 #182 Buffalo Sabres Adam Dennis (G) 1 0 - - - #191 Buffalo Sabres Vyacheslav Buravchikov (D) #208 Buffalo Sabres Matt Generous (D) #227 Buffalo Sabres Andrew Orpik (F)#13 CSS (NA) Kenndal McArdle (LW) 4 42 1 2 3 #48 CSS (NA) Dan Collins (RW) #87 CSS (NA) Patrick Davis (RW/C) 2 9 1 0 1 #96 CSS (NA) Vyacheslav Trukhno (C/LW) #142 CSS (NA) Nick Drazenovic (C/LW) 3 12 0 0 0 #182 CSS (NA) PAUL KURCEBA not drafted#191 CSS (NA) Jean-Philippe Paquet (D) #208 CSS (NA) CHRISTIAN HANSON not drafted 42 3 6 9 #227 CSS (NA) DEVEREAUX HESHMATPOUR not drafted2006:#24 Buffalo Sabres Dennis Persson (D) #46 Buffalo Sabres Jhonas Enroth (G) 7 147 - - - #57 Buffalo Sabres Mike Weber (D) 8 351 9 44 53 #117 Buffalo Sabres Felix Schütz (C/LW) #147 Buffalo Sabres Alex Biega (D) 2 58 1 7 8 #207 Buffalo Sabres Benjamin Breault ©#24 CSS (NA) Cory Emmerton © 4 139 12 9 21 #46 CSS (NA) Nigel Williams (D)#57 CSS (NA) Ben Shutron (D)#117 CSS (NA) Bobby Hughes ©#147 CSS (NA) GRAHAM POTUER not drafted#207 CSS (NA) Benjamin Breault ©2007:#31 Buffalo Sabres T.J. Brennan (D) 4 53 5 8 13 #59 Buffalo Sabres Drew Schiestel (D) #89 Buffalo Sabres Corey Tropp (RW/LW) 3 148 6 21 27 #139 Buffalo Sabres Brad Eidsness (G) #147 Buffalo Sabres Jean-Simon Allard © #179 Buffalo Sabres Paul Byron (W/C) 6 200 28 38 66 #187 Buffalo Sabres Nick Eno (G) #209 Buffalo Sabres Drew MacKenzie (D)#31 CSS (NA) Tommy Cross (D) 1 3 0 1 1 #59 CSS (NA) Ruslan Bashkirov (LW)#89 CSS (NA) Ben Ryan ©#139 CSS (NA) Dávid Skokan (C/RW)#147 CSS (NA) Matt Fillier (LW)#179 CSS (NA) Paul Postma (D) 6 125 8 11 19 #187 CSS (NA) ERIC DOYLE not drafted#209 CSS (NA) Drew MacKenzie (D)2008:#12 Buffalo Sabres Tyler Myers (D) 7 462 57 136 193 #26 Buffalo Sabres Tyler Ennis (C/W) 7 368 92 131 223 #44 Buffalo Sabres Luke Adam (C/LW) 5 90 15 11 26 #81 Buffalo Sabres Corey Fienhage (D) #101 Buffalo Sabres Justin Jokinen (RW) #104 Buffalo Sabres Jordon Southorn (D) #134 Buffalo Sabres Jacob Lagacé © #164 Buffalo Sabres Nick Crawford (D)#12 CSS (NA) Tyler Myers (D) 7 462 57 136 193#26 CSS (NA) Colby Robak (D) 4 47 0 4 4 #44 CSS (NA) Mikhail Stefanovich (W/C) #81 CSS (NA) A.J. Jenks (C/W)#101 CSS (NA) Philippe Cornet (LW) 1 2 0 1 1#104 CSS (NA) Kelsey Tessier ©#134 CSS (NA) Jacob Lagacé ©#164 CSS (NA) KRUISE REDDICK not drafted2009:#13 Buffalo Sabres Zack Kassian (RW/LW) 5 234 38 36 74 #66 Buffalo Sabres Brayden McNabb (D) 4 189 5 41 46 #104 Buffalo Sabres Marcus Foligno (LW/RW) 5 267 36 57 93 #134 Buffalo Sabres Mark Adams (D) #164 Buffalo Sabres Connor Knapp (G) 1 2 - - - #194 Buffalo Sabres Maxime Legault (RW)#13 CSS (NA) Jordan Schroeder (C/RW) 4 107 11 16 27 #66 CSS (NA) Ryan Button (D)#104 CSS (NA) Seth Helgeson (D) 2 41 0 3 3 #134 CSS (NA) GABRIEL LEMIEUX not drafted#164 CSS (NA) DANNY MATTSON not drafted#194 CSS (NA) BRENNAN YADLOWSKI not drafted2010:#23 Buffalo Sabres Mark Pysyk (D) 4 125 5 21 26 #68 Buffalo Sabres Jérôme Leduc (D) #75 Buffalo Sabres Kevin Sundher © #83 Buffalo Sabres Matt MacKenzie (D) #98 Buffalo Sabres Steven Shipley © #143 Buffalo Sabres Gregg Sutch (RW) #173 Buffalo Sabres Cédrick Henley (LW) #203 Buffalo Sabres Christian Isackson (RW/C) #208 Buffalo Sabres Riley Boychuk (LW) #23 CSS (NA) Mark Pysyk (D) 4 125 5 21 26#68 CSS (NA) Stanislav Galiev (RW/LW) 2 26 1 3 4 #75 CSS (NA) Ivan Telegin (W/C)#83 CSS (NA) Bill Arnold © 1 1 0 0 0#98 CSS (NA) Patrick McNally (D)#143 CSS (NA) Brandon Davidson (D) 2 63 5 7 12#173 CSS (NA) JAKE FALLON not selected#203 CSS (NA) JOSH SHALLA drafted the next season#208 CSS (NA) CASEY THRUSH not drafted Quote Link to comment Share on other sites More sharing options...
LGR4GM Posted May 16, 2016 Report Share Posted May 16, 2016 I know we have 4 third round picks this year but looking at that list and seeing the CSS rankings, we should trade all 4 of them to get back into the 2nd round. Quote Link to comment Share on other sites More sharing options...
rakish Posted May 16, 2016 Author Report Share Posted May 16, 2016 I don't have a chart handy, but I think there's a real gentle slope in what you get from draft picks, after the first ten or fifteen. So I'm in the camp of wanting 4 thirds rather than one second, but I think you're right, if you're valuing like CSS, you're lost after the first 30 picks or so. Quote Link to comment Share on other sites More sharing options...
LGR4GM Posted May 16, 2016 Report Share Posted May 16, 2016 I don't have a chart handy, but I think there's a real gentle slope in what you get from draft picks, after the first ten or fifteen. So I'm in the camp of wanting 4 thirds rather than one second, but I think you're right, if you're valuing like CSS, you're lost after the first 30 picks or so. I think after the first 30 picks teams try to find value and they overvalue things like character or leadership as opposed to being like "well that kid scored at a ppg pace in his last 30games so we should probably just take him in the 3rd round and hope it works." I am specifically thinking of the 2014 draft when we did not take Justin Kirkland right now. Just an example but I think in your model you value you goals much more than assists and from what I can tell that is the way to go. This kinda brings me back to the discussion on defenders. Who would we label, loosely, as the top 30 defenders in the NHL? I ask because I have a stat project in mind. Quote Link to comment Share on other sites More sharing options...
rakish Posted May 16, 2016 Author Report Share Posted May 16, 2016 My theory is they over-value height after the first 30 picks. It's really hard to skate if you're tall, and only the top athletes can do it. They also don't draft enough overagers. It's rare to find a 17 year-old after the first 100 picks, 18 and 19 year olds are much better odds. They draft too many high school players. LA has only drafted one high school player since 2005, and that was a goaltender. Buffalo has drafted 14 if I remember right. On goals, no, if you look at Eichel's college year, it isn't the number of goals that stands out. If you only looked at goals, you wouldn't think Eichel was the best player on BU, or BC, whichever. Some people pay attention to primary vs secondary assist, but can I get that for everyone, scraping? On defenders, I no longer look at any NHL data. Wildcard seems to want to learn scraping, so I'll be updating that code if you want to get in on that. My favorite metric from when I was doing that was the power play time of your opponent forwards, which is old, but still around here - http://www.limedata.us/1.php. I stopped doing it because I believe to get good data would involve 5 interns taking notes on every game, and I'm not willing to hire 5 interns to take notes on every game. I believe Chicago and Anaheim know how often, say, Mark Giordano, pinches when he can, how often that leads to a chance, and how often that leads to a chance against. I guessing Pittsburgh and Minnesota talk about WOWY Corsi numbers scraped off the NHL's play-by-play, and Calgary thinks he's good. But it's hard to tell, my view of Toronto's drafting totally changed this week. So I'm wrong person to ask about defenders, one of the kids, or Blue, pay far more attention to that stuff. Quote Link to comment Share on other sites More sharing options...
Brawndo Posted May 17, 2016 Report Share Posted May 17, 2016 Rakish what do you make of this chart? https://puckalytics.com/blog/2016/05/15/draft-pick-value/ Quote Link to comment Share on other sites More sharing options...
rakish Posted May 17, 2016 Author Report Share Posted May 17, 2016 I think the pick value has some value as to trading up or trading down, though I almost always want to trade down. The question though, indicates I haven't expressed the point I'm trying to make. A third round pick doesn't have a set value, one taken by Central Scouting or Colorado has very little value because the eye test doesn't work that well at the top of the draft, and is pretty much useless past the first round. Over the past decade, Tampa's assessment of talent has been much better than, say, New Jersey. The got Johnson for nothing, Kutcherov for a third, Bishop for almost nothing, Palat for a 7th. The value of Tampa's later round picks are much higher than the value of New Jersey's, or Central Scouting. It's hard to know how much value teams put in the eye-test vs data analysis, but my armchair analysis leads me to believe this is a major reason why Tampa is still playing, and New Jersey isn't. Inkman and Blue(!) have argued this is all luck, yet the hardest-core teams, Calgary, Colorado, New Jersey, Boston, Montreal, Toronto, Buffalo, have been playing golf for weeks. As I said earlier this week, I think Toronto has changed, and are pretending to be hard-core. It's going to get much worse for the eye test people, since someone like Nathan Gerbe will no longer get to the fifth round, since Toronto, Winnipeg, the Islanders, or the next team that figures out how to do this, will grab Gerbe where he should have been taken, near the end of the first round to mid second round. And Inkman, or DFork, will argue 'eff Nathan Gerbe,' but look at my top fifteen or so of 17 year olds from 2012 (the model really nails 2012, so don't take it as a random choice). Do notice the 12th best 17 year old didn't get drafted. Also note that St. Louis took two of these players, and Tampa took two, though Connolly didn't help them, and Gusev hasn't arrived yet, and Anaheim took two. 1 TAYLOR HALL OHL 1210 (15913)2 MIKAEL GRANLUND Liiga 1143 (7651)3 TYLER SEGUIN OHL 966 (14653)4 TEEMU PULKKINEN Jr. A SM-liiga 894 (7456)5 VLADIMIR TARASENKO KHL 774 (13822)6 CAM FOWLER OHL 689 (11939)7 DEVANTE SMITH-PELLY OHL 653 (7537)8 MAXIM KITSYN MHL 636 (2283)9 JADEN SCHWARTZ USHL 631 (8123)10 JEFF SKINNER OHL 614 (8813)11 BRETT CONNOLLY WHL 549 (4583)12 ARTEMI PANARIN KHL 547 (8040)13 ALEXANDER BURMISTROV OHL 511 (4977)14 TYLER TOFFOLI OHL 508 (8737)15 KIRILL KABANOV QMJHL 483 (3151)16 NIKITA GUSEV MHL 470 (6648) Drafting well isn't the only way to build a good team, Washington is terrible, the Rangers trade their picks for talent, Chicago hasn't drafted a valuable player in years. Quote Link to comment Share on other sites More sharing options...
LGR4GM Posted May 17, 2016 Report Share Posted May 17, 2016 I think the eye test will be shifting more to stat analysis and less of, well I think he looked good in those 5 games I saw. Taking a look again at CJH defenders, let's test some theories this draft. It will take awhile for it to matter but Let's look at the top 12 rated CJH defenders and list them by ppg. The rankings correspond to their CSS NA skater ranking. Samuel Girard: 1.104ppg (Q) 38th, 5'9" 165lbs Jake Bean: 0.941ppg (WHL) 15th, 6'1" 170lbs Frederic Allard: 0.922ppg (Q) 32nd, 6'1" 178lbs Cam Dineen: 0.868ppg (OHL) 39th, 5'11" 182lbs Mikhail Sergachev: 0.851ppg (OHL) 8th, 6'3" 206lbs Jakob Chychrun: 0.790ppg (OHL) 4th, 6'2" 214lbs Olli Juolevi: 0.737ppg (OHL) 5th, 6'2" 180lbs Lucas Johansen: 0.710ppg (WHL) 26th, 6'1" 176lbs Kale Clague: 0.606ppg (WHL) 27th, 6'0" 178lbs Luke Green: 0.574ppg (Q) 40th, 6'1" 186lbs Libor Hajek: 0.377ppg (WHL) 31st, 6'1" 198lbs Logan Stanley: 0.266ppg (OHL) 19th, 6'7" 220lbs So what does all this mean in my eyes? First is that Samuel Girard has the best offense but sadly he is woefully short for a defender. If Girard and Dineen are both available at 38, I take Dineen just because that extra 2" and the more important 17Lbs of weight. Defenders do have to worry about size more than forwards. You can take guys with tons of offense though who are under 6' but my issue with Girard is he doesn't weigh anything. While some will say he can add weight and that is true, we are talking he could add a max of like 20lbs. Logan Stanley and Libor Hajek are not even on my draft board in the first 2 rounds. It's great that Stanley is massive but he isn't generating enough offense for a 2nd year OHL player. Hajek has the same problem, he isn't producing enough for a potential 2nd round pick and I sure as hell am not taking either where CSS has them ranked. Just like Girard's size should be a factor, Stanley's size is not because he doesn't produce enough. Jake Bean, without looking at any of the scouting reports and just going off of this has the most interesting potential. His weight is low but which is a concern but he produced at excellent levels. He's a step above production wise and I think is ranked a little lower than he should be at 15th. Cam Dineen likewise is ranked lower than he should just using production. Dineen is not the proper size you want in a defender at only 5'11" but he is producing at a really good rate. The other interesting thing about him is that he is in only his 1st OHL year so he adjusted quick... Just like Sergachev who in his first NA year sits at 5th on this list. Now there is some risk with guys who only have 1 year, what if they just got lucky? The reverse of that is Frederic Allard who is in his 3rd year. While he was a pleasant surprise in the top 3 considering he is ranked 32nd, he is one of the older players on the list. That said, he has a steady increase in points every year going into this one. Seeing a trend is a good sign. The issue with Allard is that he comes out of the Q. The list of top rated defenders come out of the QMJHL is low. The OHL and WHL are better at producing NHL defenders if you look at the past. Would I not take him in the 2nd? No, I would consider it. His size is right about what is should be. I still favor Dineen here because of the Q v OHL difference. Again we are not looking at scouting reports just numbers. Finally the big 3. Numbers wise the pick should be Sergachev. 1st year player, outperformed both other guys and he has the size you want. Chychrun is the most stable of the 3. His ppg from year 1 to year 2 just bumped up slightly which is one reason scouts have raised an eyebrow. That said, he did produce again at a level that slots him 6th on the list by PPG. Finally we have Juolevi who I think is a step below the other two. His ppg is lower and his height/weight is lower. Now it was his first year but if you look at the guys above him with similar height/weights they all produced better. Now only Dineen and Bean are not in the Q so that is a consideration but Juolevi on this chart would be my least likely option at #8 if all defenders were available. So feel free to have at it. I am well aware that there are other factors than height/weight and PPG to consider in the draft. That being said, if it the 2nd round and it comes down to Logan Stanley or Frederic Allard/Cam Dineen/Sam Girard, I am rolling the dice on the smaller higher point producer because that difference is most assuredly worth it. PPG is not the be all end all but as you move into the 2nd round if there is a significant differences between guy A and guy B in regards to PPG, I lean towards trusting the numbers. Quote Link to comment Share on other sites More sharing options...
rakish Posted May 17, 2016 Author Report Share Posted May 17, 2016 (edited) Good liger, you are most of the way there, but your numbers aren't contextual. For instance if Stamkos shows up, all the Sabres' defensemen will score more next year. And context is the hard part. What part of, let's pick one at random, Sergachev, who I've never heard of because I'm still working on the process, and look at his numbers. Mikhail Sergachev has wonderful size, he's really young (born in June), his ppg is excellent, the tough part is figuring out how much of that is due to passing the puck to Christian Fisher or Lemieux, and Fisher or Lemieux putting the puck in the net. If I had enough resources, which I don't, I could get a better idea on the contextuality of Sergachev's scoring, so I need to rely on the relationship in scoring between Sergachev and Chatfield. Sergachev is much better than Chatfield on a scoring basis, so I would be surprised if the model didn't love Sergachev. Now looking back, how did Sergachev do as a 16 year old? He's near the top of his team for defensemen on his MHL team. The Russian international teams love him, what's not to like? In a week or two I'll be able to compare him to past draft picks, and current 17 year olds, by taking that data and resolving it to a number, but my guess is he'll be around Provorov, Werenski, or Hanifan. Bean, on the other hand, plays with a top prospect, likely as a pairing, and is outscored by a lot. So I wouldn't put as much stock in Bean's numbers than Sergachev, even though as raw numbers, Bean scores a lot. Edited May 17, 2016 by rakish Quote Link to comment Share on other sites More sharing options...
rakish Posted May 17, 2016 Author Report Share Posted May 17, 2016 I think, in the end, what the question requires is a long hard look at Cam Fowler, PJ Subban, and Mark Pysyk's numbers at age 16 and 17 to try to pull apart what matters, and what doesn't. Quote Link to comment Share on other sites More sharing options...
LGR4GM Posted May 17, 2016 Report Share Posted May 17, 2016 Good liger, you are most of the way there, but your numbers aren't contextual. For instance if Stamkos shows up, all the Sabres' defensemen will score more next year. And context is the hard part. What part of, let's pick one at random, Sergachev, who I've never heard of because I'm still working on the process, and look at his numbers. Mikhail Sergachev has wonderful size, he's really young (born in June), his ppg is excellent, the tough part is figuring out how much of that is due to passing the puck to Christian Fisher or Lemieux, and Fisher or Lemieux putting the puck in the net. If I had enough resources, which I don't, I could get a better idea on the contextuality of Sergachev's scoring, so I need to rely on the relationship in scoring between Sergachev and Chatfield. Sergachev is much better than Chatfield on a scoring basis, so I would be surprised if the model didn't love Sergachev. Now looking back, how did Sergachev do as a 16 year old? He's near the top of his team for defensemen on his MHL team. The Russian international teams love him, what's not to like? In a week or two I'll be able to compare him to past draft picks, and current 17 year olds, by taking that data and resolving it to a number, but my guess is he'll be around Provorov, Werenski, or Hanifan. Bean, on the other hand, plays with a top prospect, likely as a pairing, and is outscored by a lot. So I wouldn't put as much stock in Bean's numbers than Sergachev, even though as raw numbers, Bean scores a lot. I agree and I want to create a model where we value ppg, heigh/weight and then figure out how to add context to that. Sean Monahan will always be my best example of this. In his draft year he played on a god awful team and had something like 40% more points than the next closer player on his team. I am just starting to delve deeper into the numbers on defenders and it will probably be an ongoing thing. I am scheduled to take a statistics course in the fall and I also think that will help. My thought is calculating team total points scored (assists and goals) and then seeing what percentage that particular player has in their draft year. Also as we have discussed before, goals are more valuable then assists. The issue with defenders is they are naturally set up to be assist machines. Hopefully by the 2017 draft I will have figured out how to value defenders better. Thanks for the input and keep it coming. Quote Link to comment Share on other sites More sharing options...
rakish Posted May 17, 2016 Author Report Share Posted May 17, 2016 If you want to do this research, I can provide some scraping code Quote Link to comment Share on other sites More sharing options...
Derrico Posted May 17, 2016 Report Share Posted May 17, 2016 Good liger, you are most of the way there, but your numbers aren't contextual. For instance if Stamkos shows up, all the Sabres' defensemen will score more next year. And context is the hard part. What part of, let's pick one at random, Sergachev, who I've never heard of because I'm still working on the process, and look at his numbers. Mikhail Sergachev has wonderful size, he's really young (born in June), his ppg is excellent, the tough part is figuring out how much of that is due to passing the puck to Christian Fisher or Lemieux, and Fisher or Lemieux putting the puck in the net. If I had enough resources, which I don't, I could get a better idea on the contextuality of Sergachev's scoring, so I need to rely on the relationship in scoring between Sergachev and Chatfield. Sergachev is much better than Chatfield on a scoring basis, so I would be surprised if the model didn't love Sergachev. Now looking back, how did Sergachev do as a 16 year old? He's near the top of his team for defensemen on his MHL team. The Russian international teams love him, what's not to like? In a week or two I'll be able to compare him to past draft picks, and current 17 year olds, by taking that data and resolving it to a number, but my guess is he'll be around Provorov, Werenski, or Hanifan. Bean, on the other hand, plays with a top prospect, likely as a pairing, and is outscored by a lot. So I wouldn't put as much stock in Bean's numbers than Sergachev, even though as raw numbers, Bean scores a lot. I can't wait to see this analysis. Good work rakish. Quote Link to comment Share on other sites More sharing options...
LGR4GM Posted May 17, 2016 Report Share Posted May 17, 2016 If you want to do this research, I can provide some scraping code I am not exactly sure how to do that research. Might take me until June to even get around to working on it. Quote Link to comment Share on other sites More sharing options...
LGR4GM Posted May 17, 2016 Report Share Posted May 17, 2016 How would I control for the fact a player missed time. For instance if you add up all the points a team scored and then use that number to divide the total points the player scored, you get a number but it doesn't factor into games missed. I will toss this out for now. Chychrun played for the worst team of the big 3 defenders. His team averaged 9.18tppg (so they had a total of 9 goals/assists handed out per game) and he factored into 7.8% of those tps (total points scored) Sergachev's team had 10.06tppg and he factored in on 8.3%tps Juolevi's team is the one that makes me not want him at all even though ik he has been great in tourney's too. His team had 12.96tppg and he factored in on 4.8% of tps. Still looking at this and it is important to note that Juolevi missed 11OHL games, Chychrun missed 6 and Sergachev missed 1. Quote Link to comment Share on other sites More sharing options...
rakish Posted May 17, 2016 Author Report Share Posted May 17, 2016 Missed games create some noise in the system, I don't know of a way to get it better without an enormous increase in data acquisition time. Consider a player being traded midseason. Tavares was traded midseason, so his percentage of points in your system doesn't really work, since he would have a very small percentage for each team. I use ppg vs the best (or often) the second best player on the team. That way Tavares is compared to the second best scorer on team 1, then compared to the second best scorer on team 2. Quote Link to comment Share on other sites More sharing options...
rakish Posted May 24, 2016 Author Report Share Posted May 24, 2016 (edited) Today it's defensemen. It's more difficult to value defensemen because they get better later. I charted a handful of players who have been talked about lately, and added two players to show what a late-round success looks like, Parayko (Blues)(grey line) and Lindell (Stars)(royal blue). With less KHL and Czech to NHL interaction, Kempny's(lime green) numbers will have the highest likelihood of error. Kempny's latest two years have been pretty good. To me, it looks like he will score like Jake McCabe(red line) did in 2015, but since he's about 3 years older, I wouldn't expect him to get much better. NHL hockey player? yeah. Might be the perfect BlackHawk third pair. Additionally, Casey Martin(black) Griffin Reinhart(purple) Brycen Martin(baby blue) Edited May 24, 2016 by rakish Quote Link to comment Share on other sites More sharing options...
rakish Posted May 26, 2016 Author Report Share Posted May 26, 2016 Defensemen day 2. For me, this one is kind of interesting. The model was designed to value 17 year olds, so I don't use any data that I can't get for most leagues, but it seems to work OK for NHL players. These are defensemen who were very good at a young age, Bogosian and Myers, once firmly part of this group, are no longer near the rest. Quote Link to comment Share on other sites More sharing options...
rakish Posted May 27, 2016 Author Report Share Posted May 27, 2016 (edited) 2014Since Liger was talking about McCann this morning, it works out well that I getto the 2014 draft. Notes on my draft contest below. The model has evolved a bit, this is what it the current modelwould have thought at the time as the top 17 year olds, I ended up with Reinhart,Holmstrom, Point, Barbashev, and Glover. Had I used the current model, I would havedrafted Martin instead of Glover, but, except for that, things are pretty much thesame. These are the 20th highest rated forwards, 9 highest rated defensemen, at age 17. The number following each name is their total score for ages 16 and 17, so you will notice that the year 18 scores are similar.1 WILLIAM NYLANDER 12762 SAM BENNETT 12213 SAM REINHART 11814 LEON DRAISAITL 9955 AXEL HOLMSTRÖM 8966 NICK RITCHIE 8767 BRAYDEN POINT 8588 ROBBY FABBRI 8429 KEVIN FIALA 80910 MICHAEL DAL COLLE 79911 SONNY MILANO 79412 JAKE VIRTANEN 76713 KASPERI KAPANEN 75514 NIKOLAI GOLDOBIN 74515 IVAN BARBASHYOV 69816 NIKOLAJ EHLERS 65617 CHASE DE LEO 65418 JAYCE HAWRYLUK 62119 JARED MCCANN 61620 NICK SCHMALTZ 6141 AARON EKBLAD 10432 BRYCEN MARTIN 8213 HAYDN FLEURY 6064 ROLAND MCKEOWN 6045 ANTHONY DEANGELO 5936 ALEXIS VANIER 4877 JACK GLOVER 4448 MARCUS PETTERSSON 4249 JULIUS HONKA 389As 18 year olds in 2014-2015, this is how they are rated.Garland was an overager drafted in the 2015 draft. Korshkov, undrafted, was selected in the 2015 draft contest by me.1 DAVID PASTRNAK NHL 15502 SAM BENNETT OHL 13163 MICHAEL DAL COLLE OHL 12374 LEON DRAISAITL WHL 11935 WILLIAM NYLANDER SHL 11006 DYLAN LARKIN NCAA 10247 KEVIN FIALA SHL 10118 JOSHUA HO-SANG OHL 8669 AXEL HOLMSTRÖM SHL 80310 NIKITA SCHERBAK WHL 80211 ROBBY FABBRI OHL 78112 JAKUB VRÁNA SHL 77713 NIKOLAJ EHLERS QMJHL 77614 CONOR GARLAND QMJHL 76215 JARED MCCANN OHL 75416 NIKOLAI GOLDOBIN Liiga 71817 KASPERI KAPANEN Liiga 71218 SAM REINHART WHL 71019 NICK RITCHIE OHL 69920 BRAYDEN POINT WHL 68921 NICK RITCHIE OHL 68022 SONNY MILANO OHL 675 23 KEVIN LABANC OHL 66824 OSKAR LINDBLOM SHL 65125 BRENDAN PERLINI OHL 64726 YEGOR KORSHKOV MHL 64527 ROURKE CHARTIER WHL 64128 ALEX TUCH NCAA 60729 JAYCE HAWRYLUK WHL 57630 KEVIN FIALA AHL 57431 MICHAEL BUNTING OHL 56932 JAKE VIRTANEN WHL 56633 IVAN BARBASHYOV QMJHL 56234 JUSTIN KIRKLAND WHL 55035 CHRISTIAN DVORAK OHL 53636 PIERRE ENGVALL SuperElit 5171 AARON EKBLAD NHL 19072 ANTHONY DEANGELO OHL 7853 TRAVIS SANHEIM WHL 7744 JULIUS BERGMAN OHL 4785 JONI TUULOLA Liiga 4546 JULIUS HONKA AHL 4287 ANTHONY DEANGELO OHL 4108 ALEX LINTUNIEMI OHL 3969 JOHN NYBERG SuperElit 35410 BRETT LERNOUT WHL 34711 CARL NEILL QMJHL 33612 DYSIN MAYO WHL 31813 HAYDN FLEURY WHL 31414 BRANDON HICKEY NCAA 31215 RYAN MANTHA OHL 29516 BLAKE SIEBENALER OHL 29117 PETTER HANSSON SuperElit 28818 ALEXIS VANIER QMJHL 27619 KYLE WOOD OHL 27020 DOMINIK MASIN OHL 27021 JAKE WALMAN NCAA 262As 19 year olds, in 2015-2016, this is how they rated.Desroacher and Mangiapane were drafted as overagers the following year.1 LEON DRAISAITL NHL 15872 WILLIAM NYLANDER AHL 15503 JAYCE HAWRYLUK WHL 13754 CHRISTIAN DVORAK OHL 12855 DYLAN LARKIN NHL 12336 SAM REINHART NHL 12277 KEVIN LABANC OHL 12208 MICHAEL DAL COLLE OHL 12059 DAVID PASTRNAK NHL 117610 SAM BENNETT NHL 107911 NIKOLAJ EHLERS NHL 106412 ROBBY FABBRI NHL 103613 NICK SCHMALTZ NCAA 102614 KASPERI KAPANEN AHL 90415 JAKUB VRÁNA AHL 87916 CONOR GARLAND QMJHL 85917 FRANCIS PERRON QMJHL 83618 MAXIM LETUNOV NCAA 83319 AXEL HOLMSTRÖM SHL 81420 OSKAR LINDBLOM SHL 80521 NICK RITCHIE AHL 79922 BRAYDEN POINT WHL 79223 ANDERS BJORK NCAA 77324 ANDREW MANGIAPANE OHL 72225 KEVIN FIALA AHL 70126 JAKE EVANS NCAA 68927 VLADISLAV KAMENEV AHL 66928 NIKOLAI GOLDOBIN AHL 66129 ALEX TUCH NCAA 65730 JOHN QUENNEVILLE WHL 64531 RYAN MACINNIS OHL 64532 JARED MCCANN NHL 60233 JAKE VIRTANEN NHL 60034 BRENDAN LEMIEUX OHL 60035 ONDREJ KASE AHL 59036 SONNY MILANO AHL 58137 JOSHUA HO-SANG OHL 57138 NICOLAS AUBÉ-KUBEL QMJHL 56639 ANTHONY ANGELLO NCAA 52940 BRETT POLLOCK WHL 52041 CHASE DE LEO AHL 51942 SPENCER WATSON OHL 50843 NIKITA SCHERBAK AHL 49544 RYAN DONATO NCAA 49245 LUKAS VEJDEMO SHL 49046 BRETT SENEY NCAA 48847 AUSTIN POGANSKI NCAA 46948 DAKOTA JOSHUA NCAA 45349 ERIC CORNEL OHL 45350 YEGOR KORSHKOV KHL 4521 AARON EKBLAD NHL 19702 TRAVIS SANHEIM WHL 10823 JAKE WALMAN NCAA 8174 JULIUS HONKA AHL 5525 ANTHONY DEANGELO AHL 5076 HAYDN FLEURY WHL 4987 KYLE WOOD OHL 4718 ROLAND MCKEOWN OHL 4279 RYAN REHILL WHL 42110 STEPHEN DESROCHER OHL 44011 DOMINIK MASIN OHL 42012 JONI TUULOLA Liiga 420 13 CARL NEILL QMJHL 39815 ALEXIS VANIER QMJHL 35716 JACK DOUGHERTY WHL 35117 GUSTAV FORSLING SHL 33818 MICHAEL PRAPAVESSIS NCAA 335Last time we looked at the 2014 draft contest,Hockey News was a bit ahead. After 2015, to meit looks close between me, the Hockey News, andCraig Button.For 2015, I have:6 - (among forwards) Reinhart55 - Barbashev19 - Holmstrom34 - (among defensemen) Glover84 - Engvall22 - Point36 - Mayo33 - Bristedt (overager)78 - Kontos (overager 2 years)The Hockey News:10 - Bennett55 Barbashev8 McKeown16 Dougherty22 Brycen Martin36 Dyson Mayo27 Adam Ollas Mattsson19 Axel Holmstrom23 Jacob MiddletonButton:6 - Sam Reinhart27 Vlad Kamenev22 Braydon Point95 Anton Karlsson35 Ondrej Kase22 Brycen Martin44 Spencer Watson26 Vladislav Gavrikov37 Emil JohanssonMurray:6 - Sam Reinhart34 Brendan Lemieux49 Eric Cornel100 Vaclav Karabacek(goaltender) Jonas Johansson22 Brycen Martin96 Maxwell Willman (overager)92 Chris Brown37 Victor Olofsson (overager) Edited May 27, 2016 by rakish Quote Link to comment Share on other sites More sharing options...
rakish Posted May 30, 2016 Author Report Share Posted May 30, 2016 (edited) 2015My results for 2015 weren't good. My biggest fault was,I believe, that I decided I should should look forimprovement, meaning that I took the 17-year-old rating andsubtracted the 16-year-old rating. With ratings withso much noise, it ends up that looking for improvementends up selecting the most likely to be anoutlier as a 17 year old. Live and learn. The 2015 draft contest (see below) usedWinnipeg's draft, somehow I missed Kyle Connor at #17,which was a no-brainer, since he should have beenpicked as 4th the 5th forward, but, instead of Harkins,I also could have easily chosen Merkley, who also had a bad 2015.This is what last year looks like withthis years model. I may confuse some forwardswith defenseman here, we're on a budget. Againthe number after is 16 year ranking plus their17 year ranking.1 CONNOR MCDAVID 29942 JACK EICHEL 19333 MITCHELL MARNER 12974 KYLE CONNOR 11535 DYLAN STROME 11506 NICHOLAS MERKLEY 11047 MATHEW BARZAL 9348 DANIEL SPRONG 9249 ROOPE HINTZ 89510 SEBASTIAN AHO 85711 MARTINS DZIERKALS 84812 MIKKO RANTANEN 83113 YEVGENI SVECHNIKOV 81014 JEREMY BRACCO 77415 ANTHONY RICHARD 73616 JANSEN HARKINS 73617 ANTHONY BEAUVILLIER 71218 TRAVIS KONECNY 70319 JÉRÉMY ROY 70020 DMYTRO TIMASHOV 66021 PAUL BITTNER 64222 TIMO MEIER 62423 PAVEL ZACHA 62024 JAKE DEBRUSK 61625 FILIP CHLAPÍK 61526 YAKOV TRENIN 61027 COLIN WHITE 60028 JULIUS NÄTTINEN 58929 FILIP AHL 58430 CAMERON HUGHES 55531 COOPER MARODY 55532 JAKOB FORSBACKA KARLSSON 53933 BLAKE SPEERS 53734 CHRISTIAN FISCHER 53435 JACK ROSLOVIC 53436 RYAN GROPP 53037 ERIK FOLEY 52538 KIRILL KAPRIZOV 52439 ALEKSI SAARELA 52340 ROBIN KOVACS 51841 DENIS GURYANOV 51342 BROCK BOESER 51243 THOMAS NOVAK 50744 CONOR GARLAND 50045 DENIS MALGIN 48746 MICHAEL SPACEK 48647 PAVEL KARNAUKHOV 47248 ETHAN BEAR 45949 JOEL ERIKSSON EK 458 50 DENNIS YAN 44751 JORDAN GREENWAY 44552 NIKITA KOROSTELEV 4421 ZACH WERENSKI 10572 NOAH HANIFIN 9563 IVAN PROVOROV 9514 RASMUS ANDERSSON 9215 RYAN PILON 7576 NICOLAS MELOCHE 7437 VEETI VAINIO 7048 NOAH JUULSEN 6999 TRAVIS DERMOTT 60210 THOMAS SCHEMITSCH 54811 VINCE DUNN 50412 JAKUB ZBORIL 45113 OLIVER KYLINGTON 447This is their 2015 rankings, theliiga numbers are probably a bit high.1 CONNOR MCDAVID NHL 34132 JACK EICHEL NHL 14403 DYLAN STROME OHL 13914 KYLE CONNOR NCAA 13035 SEBASTIAN AHO Liiga 12526 BROCK BOESER NCAA 11797 MIKKO RANTANEN AHL 11349 MITCHELL MARNER OHL 109110 TRAVIS KONECNY OHL 90011 COLIN WHITE NCAA 87212 MATHEW BARZAL WHL 86813 ANTHONY RICHARD QMJHL 86714 ROOPE HINTZ Liiga 84315 ANTHONY BEAUVILLIER QMJHL 80916 ALEKSI SAARELA Liiga 799 17 TIMO MEIER QMJHL 92918 KIRILL KAPRIZOV KHL 726 19 TRAVIS KONECNY OHL 71320 FILIP AHL SuperElit 66321 CHRISTIAN FISCHER OHL 62622 JULIUS NÄTTINEN OHL 61023 COOPER MARODY NCAA 60224 DMYTRO TIMASHOV QMJHL 60025 JEAN-CHRISTOPHE BEAUDIN QMJHL 59426 CAMERON HUGHES NCAA 58627 GIORGIO ESTEPHAN WHL 57928 JAKOB FORSBACKA KARLSSON NCAA 57729 NICOLAS ROY QMJHL 57130 YEVGENI SVECHNIKOV QMJHL 56931 DANIEL SPRONG QMJHL 55832 LAWSON CROUSE OHL 55033 ADAM GAUDETTE NCAA 54934 JOEL ERIKSSON EK SHL 54935 THOMAS NOVAK NCAA 54736 JACK ROSLOVIC NCAA 54237 PAVEL ZACHA OHL 50938 LUKAS JASEK Czech U20 50839 MATHIEU JOSEPH QMJHL 50340 JORDAN GREENWAY NCAA 50041 ROBIN KOVACS Allsvenskan 4951 NOAH HANIFIN NHL 10262 ZACH WERENSKI NCAA 9713 ANDREW NIELSEN WHL 8424 IVAN PROVOROV WHL 7955 RASMUS ANDERSSON OHL 6586 ETHAN BEAR WHL 5447 TRAVIS DERMOTT OHL 518The 2015 draft contest looks like this. Even thoughI did a terrible job with the first pick, the rest ofthe picks were OK, so I'm still hoping to win this thing.The World Juniors part of this competion wasn't asexciting this year, as only one pick of mine and twoof Pronman's played in the WJC. I had the most pointsas Korshkov had a big tournement. It was a far cryfrom last year, when Button had 6 WJC players and Ihad 5.This year the 4 participants who chose Connor at pick 17look like they each only got one NHL player out of thisdraft, mostly because Merkley had a bad year.Winnipeg's #198 pick probably puts him out front.Viewed singly, each may be a bit different, since anotherparticipants choice previous to yours will preclude yourchoice.Who Winnipeg took:17 Connor 4th among forwards24 Roslovic 39th47 Harkins 71st78 Foley 51st108 Spacek 68th168 Appleton 54th (2016)198 Niku 14th among defensemen203 Gennaro 103rdrakish:17 Harkins 71st25 Andersson 5th among defensemen47 Vainio 38th among defensemen78 Ahl 20th in superelite, though lower in higher leagues108 Pilon Where did you play last year, and why did you play there?168 Levelle 13th among defensemen (undrafted in real world)198 Abols 82nd (2014) (undrafted)203 Korshkov 50th (2014) (undrafted)Hockey News17 Connor 4th25 Merkley 49th47 Roy 34th among defensemen78 Foley 51st108 Korostelev 80th168 Fitzgerald 98th (undrafted)198 McKenzie 78th (undrafted)203 Noel 97th (undrafted)Button17 Connor 4th25 Merkley 49th47 Vande Sompel 39th among defensemen78 Foley 51st108 Gavrikov 26th among 2014 defensemen168 Bouramman 41st among defensemen198 Marsh 76th203 Noel 97th (undrafted)McKenzie17 Connor 4th25 Merkely 49th47 Siegenthaler 61st among defensemen78 Booth (goaltender)108 Pilon Where did you play last year, and why did you play there?Pronman - taken from a reddit copy and paste, so may be inaccurate.17 Konecny 10th25 Kylington 43rd among defensemen47 Looke 92nd78 Pilipenko no stats on 2015108 Wotherspoon 27th among defensemen168 Aho the Swede 24th among defensemen in 2014 draft (undrafted)198 Tkachev 102 in 2014 draft (undrafted)203 S Olsson the forward? the goaltender? I'm going with the forward 111th (undrafted) Edited May 30, 2016 by rakish Quote Link to comment Share on other sites More sharing options...
rakish Posted May 31, 2016 Author Report Share Posted May 31, 2016 (edited) This is a list of prospects drafted in the past 12 drafts who were within one inch of Matthews 74 and ten pounds of his 194. Reading from the left, we have player, his total score, where he played at 17, that score (averaged if more than one), where he played at 16, his 16 score, and his age, which you can ignore because it's already figured in to his score. Note that Matthews gets very little score for his 17th year. Not many NHL players come from NLA, and, Swede's at least, don't let you score your first year in that league (SEL), so I have little faith in this 17th year number. I would need to put a lot of effort into getting better numbers, but it doesn't matter. He's the best player in the draft by a wide margin. 1 CONNOR MCDAVID 2994 -OHL 2132 -OHL 861 17.72 JACK EICHEL 1933 -NCAA 1245 -USDP-USHL 689 17.93 AUSTON MATTHEWS 1681 -NLA 760 -USDP-USHL 921 18.04 ANZE KOPITAR 1410 -SuperElit 794 -Slovenia-Slovenia U20 616 17.15 MATTHEW TKACHUK 1342 -OHL 869 -USDP-USHL 473 17.86 JAKUB VORACEK 1326 -QMJHL 840 -Czech U20 487 17.17 BRAYDEN SCHENN 1259 -WHL 724 -WHL 534 17.18 SAM REINHART 1181 -WHL 785 -WHL 395 17.99 DYLAN STROME 1150 -OHL 924 -OHL 226 17.510 JESSE PULJUJÄRVI 1117 -Champions HL-Liiga 813 -Mestis-Liiga-Jr. A SM-liiga 710 17.411 JAMES VAN RIEMSDYK 1105 -USDP-NAHL 1823 -NAHL-USDP-USDP 194 17.412 LEON DRAISAITL 995 -WHL 736 -WHL 259 17.913 JOEL ARMIA 973 -Liiga 785 -Jr. A SM-liiga 188 17.314 TOMAS HERTL 951 -Czech 711 -Czech U20 240 17.815 MIKHAIL GRIGORENKO 903 -QMJHL 713 -MHL 190 17.316 SEAN MONAHAN 902 -OHL 546 -OHL 356 17.917 PIERRE-LUC DUBOIS 901 -QMJHL 632 -QMJHL 268 17.218 ROOPE HINTZ 895 -Liiga 579 -Jr. A SM-liiga 316 17.819 JONATHAN TOEWS 868 -NCAA 696 -USHS 172 17.420 YEVGENI SVECHNIKOV 810 -QMJHL 511 -MHL 299 17.921 KYLE BEACH 809 -WHL 336 -WHL 473 17.722 ZEMGUS GIRGENSONS 765 -USHL 474 -USHL 291 17.7 Edited May 31, 2016 by rakish Quote Link to comment Share on other sites More sharing options...
rakish Posted May 31, 2016 Author Report Share Posted May 31, 2016 (edited) These are the top USDP forwards at age 16 1 AUSTON MATTHEWS USDP 11872 JACK EICHEL USDP 7633 MATTHEW TKACHUK USDP 6334 J.T. MILLER USDP 4835 DREW SHORE USDP 3246 JAMES VAN RIEMSDYK USDP 3037 MAX JONES USDP 2488 JIM O'BRIEN USDP 2419 NICOLAS KERDILES USDP 22710 AUSTIN POGANSKI USDP 21411 BLAKE GEOFFRION USDP 19112 SHANE EISERMAN USDP 173 These are the top USHL forwards at age 16 (USDP players play half a season in USDP, and half in the USHL. It used to be the NAHL, but you don't care about that) 1 AUSTON MATTHEWS USHL 6552 JACK EICHEL USHL 6143 MATTHEW TKACHUK USHL 3144 ZEMGUS GIRGENSONS USHL 2915 KIEFFER BELLOWS USHL 2776 NICOLAS KERDILES USHL 1687 JAKOB FORSBACKA KARLSSON USHL 1328 DANIIL ZHARKOV USHL 829 MAX JONES USHL 8010 J.T. MILLER USHL 79 These are all the players that played in NLA at age 17 drafted by the NHL in the past 10 years: 1 AUSTON MATTHEWS NLA 760 Edited May 31, 2016 by rakish Quote Link to comment Share on other sites More sharing options...
LGR4GM Posted May 31, 2016 Report Share Posted May 31, 2016 I love having a solid number to compare players too. Man if I could learn all about the stats and do player evals for a job I would be there in a heartbeat. Quote Link to comment Share on other sites More sharing options...
Derrico Posted May 31, 2016 Report Share Posted May 31, 2016 Rakish, your numbers always have Mathews on a significantly higher level than Eichel. You think he's a much better prospect? Or am I reading your numbers wrong? Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.