rakish

I've been getting the small gas engine out of my life. I bought a corded electric lawn mower, which is great, except, of course, for the cord. The weed wacker has also gone electric, the neighbor yells 'Judas' from beyond his fence. I put a pellet stove into this pile my girlfriend bought in Virginia.

I'm on Blue's side of this argument. I put a lot of effort into trying to predict future success based upon commonly available data. If you give plus minus any value in assessing a draft pick, you draft a lot of Cory Emmerton and Eric O'Dell.

These are the 45 players, drafted between 2005 and 2015, that exceeded his category teammates by more that 15 in plus/minus (forwards led all other forwards on team, defensemen led all other defensemen). SIDNEY CROSBY 2004 MATT NISKANEN 2013 MARC-ÉDOUARD VLASIC 2004 PATRICK MCNEILL 2005 BRENNAN TURNER 2009 MARK MITERA 2011 CORY EMMERTON 2005 VLADIMIR ZHARKOV 2008 DAVID MCINTYRE 2008 PAVEL VALENTENKO 2010 LOGAN PYETT 2007 THOMAS HICKEY 2008 MAX PACIORETTY 2014 P.K. SUBBAN 2009 CADE FAIRCHILD 2011 PAUL POSTMA 2008 ERIC O'DELL 2007 CALVIN DE HAAN 2009 STEFAN ELLIOTT 2010 BRIAN DUMOULIN 2009 CODY SOL 2011 STEPHEN JOHNS 2014 BRENDAN GALLAGHER 2010 BRANDON DAVIDSON 2009 BEN MARSHALL 2009 SEAN COUTURIER 2010 SEAN COUTURIER 2009 DOUGIE HAMILTON 2012 JAMIE OLEKSIAK 2014 BRANDON SAAD 2011 TYLER WOTHERSPOON 2012 BLAKE COLEMAN 2010 ZACH YUEN 2011 FREDRIK CLAESSON 2013 MATTHEW PECA 2012 SLATER KOEKKOEK 2013 RYAN PULOCK 2011 OLIVER BJORKSTRAND 2014 JAN KOSTALEK 2014 TYLER GANLY 2013 VACLAV KARABACEK 2011 ROLAND MCKEOWN 2013 RINAT VALIYEV 2013 RICHARD NEJEZCHLEB 2012 IVAN PROVOROV 2015

Yes. If he's not on the roster the final two games, he doesn't get the accrued year toward his UFA, but that doesn't affect when he becomes a RFA (This year his ELC burns one year no matter). We disagree on how that affects his value, I think they don't give him the 2 games on the roster, they must trade him, further lessening his value to Tampa, despite also decreasing his value to the receiving team. I would give him the two games, but I'm not a war-time consigliere.

nhlnumbers says Nash's cap hit is 7.8 for 2 years and the rest of this season. Age 31. The fourth period says he has a no trade clause

Yeah, if I was running it the AHL and ECHL would try new rules regularly to see what happens. People like seeing goals, and for as many posts are hit, adding a few inches each way may lead to a lot more scoring. I took from the chart that 18 and under leagues score more than adult leagues. Scoring decreased everywhere 2005-2014, but not as much as the NHL's decrease, so NHL's decrease can be attributed to both skill and equipment, which happens everywhere, and how the game is called, which is unique to the NHL. Hey MODO, do you read this? The SHL looks like scoring has increased the past 3 years, intentional? random?

well, I didn't go back to before 2005, so there wouldn't be a spike, no?

The thick blue line is NHL average team scoring per game. The thick black line is average of each league (not every league is represented in each time period). For me, the change in the black line demonstrates skill level and equipment. The difference between the change in the blue line, and the change in the black line, represents how the NHL is calling the game. The data for the NHL starts in 2005, then, the black line is very close to the blue line, and the NHL is fairly close to the median. By 2015, the NHL is near the bottom of leagues as far as scoring goes. Oh, each league's peak year is noted on the chart. Note that there really isn't a relationship shown between ice rink size and team average goals per game. SHL, KHL, Liiga, and Czech leagues all peak below the NHL peak.

Whenever the 18 to 19 year change happens, that is going to be one sad draft year.

I think you are pointing out how terrible the Democratic Primary is run. The D's don't need to win Mississippi to take the presidency, but they must win Michigan. Therefore the election in Mississippi is scored way too high, and the election in Michigan much too low. The Republicans, though likely inadvertently (more below), run a much better primary because their candidate must be strong in Ohio and Florida to win their primary, since as winner take all, they are much more important than non-battleground states. This, in addition to D's god-awful super delegates, makes, for me, the R primary much better organized than the D. The inadvertent part is, I might be wrong here, that R's made winner take all states not by their battleground nature, but to favor establishment favorites in Florida (Bush), Ohio (Kasich), and SC (the guy that dropped out... Graham ... thank you google)

Where are the remaining candidates on this?

For me, this has been the most interesting question in the election. 538 won't tell me, my Republican inlaws don't know. The best argument I've heard is that the party worries Trump takes his supporters and runs independent, which would be much worse for the party than Trump winning. They don't fear Rubio or Cruz running independent, so they make anti-Rubio and anti-Cruz ads. Great argument? maybe not, but it's all I got.

I think that's feasible, though for me, much less likely because of the risk/reward he was taking on by bluffing. Either scenario, (A:Babcock was bluffing or B:Babcock changed his mind after talking to Buffalo for a couple days), Buffalo would think it was scenario A

I think, Hoss, that's the crux of my issue. From how I see it, if Toronto was still negotiating, they wouldn't have a scheduled hiring. Therefore it makes sense to me that they has been told they were out. Then Dudacek, why had Toronto have a hiring scheduled unless they were told by Babcock he was coming? Of course, Tom, the press conference was a bunch of lies. He can't go on TV and say Tim Murray is effin crazy, so they can't win, or Kim Pegula is effin crazy, so they can't win. I don't care much what people say in a press conference. If I have the story right, Toronto's behavior makes it look like they were told they were out, and who else would tell them they were out but Babcock.

I have a question about the phrase 'pulling a Babcock.' My understanding of the Babcock contract battle was that when Babcock announced he was going to Toronto, Toronto was set to announce their new coach. If this is true, then Toronto must have been told at some point that they were out of the running. Now the current SabreSpace narrative seems to be that Babcock always intended to go to Toronto, but if someone intends on going to Toronto, they would never tell Toronto they were out of the running. If this is all true, then the only reasonable narrative is that Babcock changed his mind. Whether this was from talking to Tim Murray or from talking to someone in the ownership side, I don't know. So my question is, what do I remember wrong?

You guys are confusing correlation (team scores less when Lehner plays) with causation (Lehner causes Sabres to score less). Without spending the time to figure it out or actually watch the Sabres play, I'm certain it's a trend of the team, not a difference in goaltending producing offense. Most of Lehner's games are recent, so a recent trend by the team of decreased scoring will cause correlation between Lehner and goal scoring. If you want my guess as to the cause of the trend, the team has some injuries, and Blysma, in roundyball lingo, has a really short bench. O'Reilly plays almost a minute more per game than than the second highest forward in the league. Kane is third in the league in forward minutes per game. My theory is that O'Reilly and Kane are tired, and stopped scoring. I wrote about it more a few months ago here (it's in the Bylsma post) when 5 players were in the top 50 league wise in minutes per game for forwards. Since then Blysma has cut back minutes for some, but not for O'Reilly and Kane.

While the top 50 list isn't great, I'm taken by how many weren't taken from the top of the draft. Someone on this sight was opining about how only the top 15 picks matter, and other pundits put percentages on late first, and later round picks, but it seems to me that if about half the good talent comes from outside the top 15 picks, it's something worth worrying about.

At least for O'Reilly, you wouldn't draft him based on his offensive expectation. I try to take defensive play into consideration when I did my draft board last year. For instance, my first pick (using Winnipeg's draft position) I got Harkins at 17 partially because of his reputation as a defensive player, therefore I placed him fairly high on my draft board. It's interesting for me that Harkins didn't get selected until 47 because the NHL didn't value this defensive reputation, or maybe the reputation was wrong. This is why it's hard to backdate players because there is a certain level of buzz that has to be listened to (for instance, Vainio doesn't play much defense, because only looking at his offense, Vainio would have been a top 10 pick). Getzlaf is interesting because most players on that level don't spend that much time in development, about 2 1/3 years. Many of the bigtime players played in the NHL immediately, either as a top pick, or the team immediately realizing what they had, like O'Reilly or Bergeron. Benn spent a little less time in development than Getzlaf. I think, Wildcard, it's a different process that ends up finding different types of players. Pysyk, for example, didn't score much in junior, but people love watching him skate. McCabe scores much better, always has. So this process will end up with players like McCabe instead of players like Pysyk because I'm trying to predict how they will score. That might a good thing, or it might not.

It's quite basic. How you score at 17 correlates strongly to how you score later in your career, but there are a lot of thing that affect your scoring. DeBrincat's a good example. He is scoring a lot because he's a good player, and also because Erie as a team scores a lot. He's too small to be considered a 3rd or 4th line player by current NHL thinking, so he will need to score as a top six player. What the model tries to do is separate how good DeBrincat is from how his stats are affected by playing for Erie. We also have to adjust for how old the player is, since their scoring increases about 50% per year, an August born player (who is young) gets bonus points, a January born player gets some deduction. It's also very important to be getting better, which is why I did an early season valuation to compare to the rest of the season. So in the end, will DeBrincat score as P Kane, or Ennis, or Nylander, or Marner? This, of course, misses big things on whether someone is a good hockey player or not. but so does going to games and watching kids play. The model loves William Bitten, check his team stats, he's outpointing his teammates by a half a point a game. That's a huge number. If he was bigger, it would be better for him as a prospect, but as of now, most top prospect lists don't mention him. McKeen's has him at 22. As Liger will ask when people talk about drafting best player available, how exactly are you comparing the value of forwards to the value of defensemen? Even for this board, there's a spectrum of opinions on the value of Weber, Ennis, Pysyk, even Risto. There are guys out in Vancouver doing the same thing, they write for nhlnumbers.com

In my system 1000 pretty much means you are ready, so I would say no. Even then, Nykvist (sp?) on Detroit spent a couple years in the minors while over 1000.

I talked a little about Puljajarvi upthread. It's hard to tell with people playing with men. He's playing Liiga. Laine has right about the same numbers, also in Liiga. In the end I'll compare big forwards to big forwards, and defensemen to defensemen of the similar size. But I'm not there yet. It ends up that the 6'1 210 player gets underdrafted, and the 6'4" player gets overdrafted, especially outside of the first 15 picks.

It's early. He certainly won't be drafted that high. It will help a lot if he stays for the World Juniors break, to see how he does without Strome. I haven't even looked yet as to who he plays with. The point of this isn't to find Debrincat's drafting position, it's to get a baseline on his season. If he rates a 1000 the last third, I know he's getting better, which is a good thing. It's also really hard to tell with players playing with men, so Pul-etc isn't ranked very high. I'll have to rely a bit on his age 16 year, which I haven't looked at yet.

SabreSpace: So rakish, why your own thread? rakish: Well, there's a few reasons, but mostly, trying to data porn the draft makes people angry, so I'm trying to stay out of the way. SabreSpace: Why post at all? rakish: In the end, I need stuff dated. This process might work, or it might be a disaster, but if it works, I need a date on it. SabreSpace: So what nonsense do we have today? rakish: Total nonsense. What you are looking for in the draft is increase in productivity. So I'm trying to see who is good next Summer by how much better they are in the Spring related to their Fall. SabreSpace: So these are the 10 players you don't want? rakish: Absolutely, not yet anyway. SabreSpace: Any qualifications? rakish: Well, I don't have many NLA players in the database, and Matthews has been injured, so his score is more meaningless. Chychrun doesn't score enough, so he'll never get the model's love. I'm thinking the smalls are overvalued, because there's so many of them. SabreSpace: How does this compare to last year? rakish: Well Eichel and McDavid were over a 1000, Marner was at 850, Strome was about 625. So as an early look, expecting the Matthews rating to be way off, then replacing Matthews with Eichel or McDavid, pretty much the same, except for one star instead of two. SabreSpace: So one less McDavid, and two more Marners? rakish: yep Keller: 891 Matthews: 853 Bitten: 710 Tkachuk: 676 DeBrincat: 662 Nylander: 553 Dube: 537 Burke: 527 Mascherin: 497 Rubtsov: 473 Pethrus: 471

As Swamp says, the lost goal when losing a challenged goal, and the goal added when winning a no-goal challenge, throws the odds of the result. using my fingers, I get games where a challenged goal was lost, teams were 11-21, which isn't too far from my guess on how much a goal is worth, but it's totally a guess games were a challenged goal was won, teams were 10-9, almost exactly even games were a no-goal challenge was lost, teams were 5-6, almost exactly even the only surprising number is that the only two teams that won their no-goal challenge, lost the game thanks Hoss for finding the data, I was thinking about looking into this, it would have taken some time

I was looking at DeBrincat's stats yesterday, thinking the same thing. I also think the same of Bitten, but I'm sure if Bitten continues to score, his ranking will become more reasonable.