rakish

Sure there's a strong correlation, how did Gagner do? I bet his even strength points were extremely high, but like Ehlers, it matters more who you play with. At 17, Gagner played with Patrick Kane at 18. This year, Ehlers played with Drouin at 18. If you don't account for this, you're drafting Gagners all day long.

I'm not against the use of the stat, I'm against the value he puts on it. Points are important, points are consistent over time, you cannot draft purely on points, because other things matter, how much? you can guess, sure, but he's constructing a model out of his butt. Even strength points are exactly the same, how much does it matter? You are guessing until you can model stats at 17 to NHL success. They don't bother.

A few thoughts on the Edmonton link which is here There are three main problems with their analysis. First, they don't account for size. Virtanen is prototypical NHL size, and should go off the board well before half of the prospects on his list. Ekblad isn't loved for his scoring, he's loved for his scoring by a big man. Is he a better pick than DeAngelo(who the writer didn't bother to note, using this method, would be the best player in this draft)? Of course Ekblad is a better pick, Ekblad should go 4th, not 8th or 9th, whatever he has Ekblad at. Second, they are modelling based upon thin air. They believe that even strength scoring is the key to finding talent, or scoring per 60 minutes, or whatever. This is based upon? A hunch? If they spent the effort (and actually had the stats) for Gilbert Brule, and were making the argument, this is why you don't pick xxxx, because he's like Brule. Instead, they have a new stat than has no basis as to how it works. They don't even take the time to argue something like "This works because the LA Kings beat the hell out of people in even strength minutes" Thirdly, age isn't a straight line. Being born in November (like Reinhart) seems to be worse than being born in January, it isn't. It actually isn't very far from being born in June (like Bennett). This is actually just the second problem again, because they haven't bothered to correlate age to success, they just have a hunch it's important.

For me the interesting thing about this draft is the strength of this draft, at least in the second round, is small players that can score. They're all smaller than what Murray wants. At 31, Scherbak or Ho Sang or Point or Gunnarsson will be there, and I think all of them are a higher caliber player than is normally available at 31, but I think Murray takes Lemieux.

If Bennett does fall to 6, 6 and Horvat would be a steal

Thanks, Liger, A lot of it is the same as the work that you've been posting, trying to separate out the second assist, or talking about the even strength points is trying to get the stats to be fair regardless of the players around you. I think the biggest question is how much does Drouin affect Ehlers, the model thinks a lot, certainly more than the people looking at it by eye.

I finally finished my model. It ended up being a study in size, I no longer try to get a rating for Bennett and a rating for Ekblad to compare them, I try to get a rating of Bennett as compared to Seguin and one for Ekblad as compared to Pietrangelo, but for everyone. http://www.limedata.us/blog/b40.php I tried to write it so you can form your own conclusions. The model loves most everyone everyone here loves, with only a couple exceptions. For those without the time to read, my conclusions are: Buffalo's board at 2: Reinhart or Bennett Second round: Holmstrom, Glover, Engvall, Sanheim, Kempe, Point Third round: Kirkland, Iverson, Mayo, Amadio, Bristedt, Kontos, Tkachev, Gunnarsson, Wood

Yeah, Nylander's a possibility, but I think Nylander is close to consensus number 5, so I think he would have said Nylander. Instead, he kind of phrased it as 'someone you are not thinking about.' So I've been thinking who he meant, I have a hard time thinking either Murray or Nolan think Nylander might be the right answer, or Ehlers, maybe Milano, Nick Ritchie?. Where as I thought I remember Murray say he liked mobile defenseman like Karlsson, which leads me to DeAngelo. As far as top 10, risky yes, but my model calls this 'The Tony DeAngelo Draft'. He's a tool.

Regarding DeAngelo, Devine was talking about the draft last week, said the Sabres were looking at 5 players, Sams, Ekblad, and Draisaitl, and the 5th was a surprise, and not Dal Colle. He also said he wouldn't be surprised by a surprise pick in the top 5, I'm thinking he was talking about DeAngelo. He's certainly not big, but his scoring numbers are incredible. I study 2002-2011, and 1.39ppg is way above everyone else. He scored about the same ppg as Dal Colle, as a defenseman. We agree that he'd be a perfect fit for the Sabres, there's no one like him in the system. While a perfect fit, still not needed as much as either Sam, so I want to stay at 2 rather than trading back for Horvat and 6, and expect to get DeAngelo there. A lot of people think he might make the second round. If I was Toronto, I'd take him at 8, though as a data guy, I know nothing of his attitude.

How does Halifax use their lines? Ehlers with Drouin? Change things a lot?

Much of the article is excellent. Where he/she? fails, I think, is worrying too much about the price being paid. For instance, Ekblad could go first, my model thinks he should go 5th, that's not going to change how Ekblad plays. Bo Horvat is either going to be a difference maker, or not, and because he's a 9th overall pick, doesn't change that. As my model evolves, my model and the writer agree on one thing, Reinhart is going to be really good.

Myers, Ehrhoff, Stafford, Hodgson, Grigo

We think about this very differently. The question isn't how many pounds Bennett has to put on, the question is do the players in Bennett's weight class succeed. If you look at my chart for Bennett, he sits right in the middle of 7 players, Evander Kane, Huberdeau, Nugent-Hopkins, Bennett, Phillips, Seguin, Brassard, and Stajan. It's only RNH who is very much lighter than the rest, well, Huberdeau is lighter as well. The Bennett group size are slightly different players than the group the Reinhart is in, well below Hall, just above Voracek and Loui Eriksson. You don't find many bargains in either group in later rounds (there is lots of correlation between their play as 17 year olds, and their NHL play) either they score enough, or they don't, so it isn't where you take a 4th rounder. I'm fine with Bennett, I'm OK with Reinhart, I think Murray is picking Draisaitl. What interests me with these charts is that it says that Barbashev's a hockey player, and at least one of those guys is getting to the second round. As I've said before, what bothers me about thinking about pullups as data points, or the dozen of other things, is that there's no connection between previous players and the success they have, so the metrics come from thin air.

It'd be nice to have pull-up numbers to see how they correlate.

Regarding Milano, I think it was Derrico who recently said there're about 30 really good players here. I think Derrico is right, and Murray will sit at 31 to see who he gets. I'm hoping he gets DeAngelo. The intern in Vancouver looks at his stats and says 1.39 ppg, for a defenseman? My model thinks he's the best player in this draft, but the model looks at the world like the Vancouver intern, ignoring the fact that DeAngelo is small. A couple of the mocks that Liger posted had DeAngelo available at 31, I can't believe it will happen

Actually I find the defensemen are easier to find than the forwards, therefore the drafting a forward in each first round is probably the best strategy. And I agree with LPF that good forwards correlate to winning more than good defensemen do, therefore early picks should be used on forwards. No shred of information that wasn't already available at the time was used. They limit their picks to the players taken by everyone else before the next Canuck pick, which they would not know at the time. They would do better by not valuing Q points so high, would save them Zagrapan, but would cost them Giroux.

Ok Ghost, What I get is Playoffs: 53g .915 Last 10: 90g .913

Who was the defenseman beaten?

Looking at that website liger, it looks to me like they have: Winnipeg pick in the Moulson trade Minnesota pick in the Pommer trade, which they trade up to a first if Miller signs their own pick The LA pick acquired in the Regier trade seems to have gone back to LA in the Fasching trade, it doesn't necessarily have to be LA's pick, it could be Buffalo's pick, the site didn't know which second went back to LA so 3, extremely potentially 2 Now I understand the chart a bit better, definitely the Kings pick was traded back to the Kings, which might go to Columbus

I invite you to take a closer look at the 2006 NHL draft

Last year I made some charts comparing player size to wins. The three biggest teams were Florida, Winnipeg, and Carolina. To become one of the biggest teams, you must believe that big wins, therefore I believe that Florida will take either Ekblad or Draisaitl. If it was a team that uses a model similar to my model, like Anaheim or Montreal, I would expect them to take Bennett. As to the Barkov is to Jones as a Sam is to Ekblad, yeah, my confidence is probably unfounded, I now have that feeling where I need to change my mind, feels like a screwdriver in the ear, I love that feeling

Draisaitl is slightly bigger than Lucic at the same age

Sabres draft both Sams Major differences on defense Tallinder, McBane, out, Zadorov Risto, Ruweedal in Ehrhoff bought out. If you don't, in 2019 you may need to choose between your Norris winning Zadorov, Smythe winning Bennett, and Vesna winning Ullmark, because you won't be able to afford all 3. For forwards, Hodgson out, two of Moulson, Calahan, and Steve Ott in As for this prize, I will need to wait for the 2019 Conn Smythe award

marcellus, what about tomatoes into soil, what do you recommend?

No way, I said that a month ago