Fantasy Sports: Flyers Board Mock Draft 2017

[Ruck Over]

With the #212 selection in the draft, the Sharks are proud to select a guy, whom we've scrambled to fall in love with because multiple prior targets were taken in previous rounds, there aren't as many fallers we've identified as expected or desired.

And fall in love we did, head over heels, for this smooth skating, Finnish D Man. If we have a type, this is it, Suomi Strong! Our third Finnish D man, arguably one of the best in the QMJHL, playing a solid game of moving the puck and defending the net (LOOK AT THAT PLUS MINUS RATING!), 5'11'', 165 lbs, playing for the Charlottetown Islanders, Saku Vesterinen!

next gm pm'd

 

[BernieParent]

I believe with a limited number of picks, the Senators did pretty well for themselves:

• Conor Timmins (D) – 1st round (20)
• Nikita Popugaev (LW) – 2nd round (48)
• Jacob Paquette (D) – 4th round (113)
• Michael Karow (D) – 6th round (184)
• 2 metric tons of office supplies
• 1 JumboTron
• Bill Daly’s personal cell number (for prank calls)
• 1 Blackhawks luxury box furniture set, with 48†plasma TV

 

[denominator]

In solidarity of the voting, the Rangers ended up with this WHL-heavy draft:

1 (023) Kristian Vesalainen
4 (102) Stuart Skinner
5 (145) Adam Thilander
6 (157) Sami Moilanen
7 (207) Andrei Grishakov

Not bad.... but there are definitely better draft sets out there. The big gaps in between some of the picks certainly doesn't help, and that's an issue I can't solve.

As for judging our draft against the real draft... I can probably set that up fairly quickly in my master excel sheet after the real draft. It might take some small reworking and legwork to make it work but I could pretty easily map out our draft against the real draft.

As for "scoring", I can immediately see three things that would throw kinks in the system:

1) Trades - we all know there WILL be trades on draft day, and trades to Vegas between now and then. How do we judge trades for the team that originally had the pick and the team that now has the pick?

2) Are we judging a "good" pick as one made closest to the actual pick point? Or where our mock pick was lower (a "steal")? Do we give negative points for reaching too early? (and who's to say that the real GMs know better)

3) How do we adjust for different numbers of picks? A team with few picks will be more likely to score better than a team with many picks as more mock picks leads to more chances for misses when trying to match for the real draft.

I think this is a great idea, just needs a few guidelines laid out before we jump into the data.

 

[Rebels57]

With the 213th pick, the Washington Capitals are proud to select, from the Kitchener Rangers of the OHL, Center Greg Meireles

http://www.eliteprospects.com/player.php?player=286948

 

[Chuck Downie]

3) How do we adjust for different numbers of picks? A team with few picks will be more likely to score better than a team with many picks as more mock picks leads to more chances for misses when trying to match for the real draft.

I think this would be the biggest obstacle. Maybe some way to weigh all teams as if they had 7 picks? I like the idea of judging player vs. their actual draft position rather than against the consensus rankings. Going to need some ideas for this one.


We can start a poll for the top 4 drafts right after pick 217 is in. We can let that poll last for a couple days and then the same amount of time to determine the top 2 out of those 4, and the same for the final.

 

[JojoTheWhale]

Nashville selects D Jakub Galvas.

In a rush, sorry. PMing next.

 

[Chuck Downie]

1) Trades - we all know there WILL be trades on draft day, and trades to Vegas between now and then. How do we judge trades for the team that originally had the pick and the team that now has the pick?

2) Are we judging a "good" pick as one made closest to the actual pick point? Or where our mock pick was lower (a "steal")? Do we give negative points for reaching too early? (and who's to say that the real GMs know better)

3) How do we adjust for different numbers of picks? A team with few picks will be more likely to score better than a team with many picks as more mock picks leads to more chances for misses when trying to match for the real draft.

I think this is a great idea, just needs a few guidelines laid out before we jump into the data.

After putting a little more thought into this, I think a scoring system that ignores picks being traded and just basing the score on the players selected vs. where they ended up actually being drafted might work. The score can then be divided or multiplied depending on whether or not the total picks are equal to 7 in this mock. As for whether a pick is ``good`` or ``bad`` it will be based solely on how close the player selected was to their actual draft position. Example: player x chosen at 10 in this mock that was drafted at 8 would get the same score if they were drafted at 12. Players not drafted at all that were selected in this mock would get a score of 217. The lowest score would be the most accurate.

Need some opinions/input on this. It is early in the morning here this seems to make sense. Maybe I am missing something?

 

[Ghosts Beer]

Nashville selects D Jakub Galvas.

In a rush, sorry. PMing next.

I was about to ask if I could sign him as an undrafted free agent. Can't believe he almost fell out of the draft.

 

[Ruck Over]

The Sharks are proud to select with the #215 selection (the last of this draft) in this exercise of futility to which the Sharks brain trust really cares about being proven correct about and will cirlce back in 4 years to either lament their first NHL mock draft or ungraciously brag about for no good reason if these picks prove fruitful to anyone who would listen, a young goalie (the first in our draft), American born with a great chin, bountiful raw skills with a sizable frame (large goalies are so in now), who will be playing at the prestigious University of Denver. Strong glove hand that will lead to left handed larceny, the 6'2", 192 lbs. Dayton Rasmussen.

Dayton is not irrelevant, that's reserved for the poor chap 2 picks later.

The Sharks final selections for the 2017 NHL Draft - all sure to be HHoF inductees

1-21 Sharks - Nick Suzuki, C, Owen Sound (OHL)
3-83 Sharks - Jonas Rondberg, RW, Vaxjo J20 (Superelit)
5-143 Sharks - Kasper Kotkansalo, D, Sioux Falls (USHL)
6-159 Sharks (from Coyotes) - Adam Kalaj, Kladno U20 (Czech U20)
6-174 Sharks - Aarne Talvitie, F, Espoo U20 (Jr. A SM-liiga)
7-205 Sharks - Ivan Kosorenkov, RW, Victoriaville (QMJHL)
7-212 Sharks (from Blackhawks) - Saku Vesterinen, D, Charlottetown (QMJHL)
7-215 Sharks (from Ottawa) - Dayton Rasmussen, G, Chicago (USHL)

Next GM pm'd.

 

[Blackhawkswincup]

With the 216th overall pick in 2017 HF Flyers Mock Draft the Chicago Blackhawks select

RW Jordan Timmons - Cedar Rapids (USHL)

My complete draft
1st round = C Alexei Lipanov - Balashikha (VHL)
2nd rouund = G Maksim Zhukov - Green Bay (USHL)
3rd round = D David Farrance - US NTDP (USHL)
4th round = LW Mick Messner - Madison (USHL)
5th round = C Nikita Anokhovsky - Yaroslavl (MHL)
5th round = D Leon Gawanke - Cape Breton (QMJHL)
5th round = C/W Noah Cates - Omaha (USHL)
6th round = C Renars Krastenbergs - Oshawa (OHL)
6th round = D Jesse Bjugstad - Stillwater High (USHS)
7th round = RW Jordan Timmons - Cedar Rapids (USHL)

 

[SanBlom]

With the final pick of the fake 2017 NHL Draft the friggin Penguins select Goalie, Adam Ahman. We are only selecting him because he is still on the board and we are quite surprised (we have him in our top 5). We know that Murray really sucks and has gotten awfully lucky so we decided to snag some of our favorite Goalies in hopes that one is good enough to choke against the Flyers in the Eastern Finals in 2019, 2020, 2021, and 2022.

Crosby for president!

 

[SanBlom]

Job well done Mr. Downie. Lets do it again!

 

[Chuck Downie]

WOW! The draft is complete!!!!!

A big thank you to everyone who participated. I am quite pleased that we were able to complete this ahead of the draft. Now, I will set up a poll for those who are interested to vote on who you think had the best draft. I will try to get it up today.

Well done everyone!

 

[denominator]

Okay, I gave a bunch of thought to judging our draft versus the real draft, and scoring it. I think Chuck is mostly right and here is my suggestion (should be simple enough to run the numbers 1/2/3 days after the draft):

1) For the first round, most of the picks are going to be "hits" or "near misses" because it's fairly well established. Big advantage to NJ and PHI as they're going to either hit or miss by 2 at most, whereas the bottom half is likely to hit within 8. Outside the first there is a much higher miss chance. Nothing can be done about this, but it means any bonus we attach to hits is going to unfairly advantage the teams picking top-10.

2) In my current consensus rankings, the largest grouping in the first round is 14 spots (25-38). This means that [if my model is realistic], we can justifiably say that a good draft will hit a player within 7 spots either way (ie - mock pick at 30 could be a player that is in reality drafted anywhere from 23 to 37). Thus, my suggestion is to award "points" for near misses up to 7 spots away.

3) Following the above thinking, a hit (mock pick at 30 is real pick at 30) is worth 8 points. A miss by one (29 or 31) is worth 7, a miss by two (28 or 32) is worth 6, a miss by three is 5 (27 or 33), four is 4 (26 or 34), five is 3 (25 or 35), six is 2 (24 or 36), and seven is 1 (23 or 37). A miss by anything greater than seven spots is worth 0 points.

4) With the exception of the last 7 picks (#211-217), mock drafting a player that goes undrafted in the real draft is worth -1 point. This rewards smart value picking later and penalizes fun/homer picking that is unrealistic. The last 7 picks are exempt because they theoretically could still have gained points if the draft extended past #217.

5) To adjust for uneven numbers of picks, once each team's raw score is tallied, it is adjusted to compensate. Each team will be scored using the formula of

(TOTAL SCORE / NUMBER OF PICKS) X STANDARD PICKS (7)

Thus adjusting everyone to the same number of picks. I originally thought we could standardize to the highest number of picks but this would just artificially inflate the numbers and give greater weight to fewer hits.

Thus, a perfect score would be 56. The worst possible score would be -7.

I will build an excel table for our picks between now and the draft, and it will be fairly easy to run the numbers after the draft if everyone is good with the above suggestion (or something similar).

 

[Ghosts Beer]

That would be a way to judge who predicted the outcome of the real NHL draft best.

I don't know how much it would say about who actually drafted the best players, though.

By that system, let's say someone had 7 picks in the mock, and somehow all 7 went 1 thru 7 in the first round. The proposed system would give them a far worse rated draft than someone who picked a 5th rounder in the 5th round, a 6th rounder in the 6th round, etc.

 

[Captain Dave Poulin]

Thanks again for doing all this work for us, Chuck

 

[sobrien]

Great job Chuck and everyone else who took part.

Semi shocked we finished it. I think over 200 picks went faster than a typical WJC fantasy draft

 

[Chuck Downie]

Okay, I gave a bunch of thought to judging our draft versus the real draft, and scoring it. I think Chuck is mostly right and here is my suggestion (should be simple enough to run the numbers 1/2/3 days after the draft):

1) For the first round, most of the picks are going to be "hits" or "near misses" because it's fairly well established. Big advantage to NJ and PHI as they're going to either hit or miss by 2 at most, whereas the bottom half is likely to hit within 8. Outside the first there is a much higher miss chance. Nothing can be done about this, but it means any bonus we attach to hits is going to unfairly advantage the teams picking top-10.

2) In my current consensus rankings, the largest grouping in the first round is 14 spots (25-38). This means that [if my model is realistic], we can justifiably say that a good draft will hit a player within 7 spots either way (ie - mock pick at 30 could be a player that is in reality drafted anywhere from 23 to 37). Thus, my suggestion is to award "points" for near misses up to 7 spots away.

3) Following the above thinking, a hit (mock pick at 30 is real pick at 30) is worth 8 points. A miss by one (29 or 31) is worth 7, a miss by two (28 or 32) is worth 6, a miss by three is 5 (27 or 33), four is 4 (26 or 34), five is 3 (25 or 35), six is 2 (24 or 36), and seven is 1 (23 or 37). A miss by anything greater than seven spots is worth 0 points.

4) With the exception of the last 7 picks (#211-217), mock drafting a player that goes undrafted in the real draft is worth -1 point. This rewards smart value picking later and penalizes fun/homer picking that is unrealistic. The last 7 picks are exempt because they theoretically could still have gained points if the draft extended past #217.

5) To adjust for uneven numbers of picks, once each team's raw score is tallied, it is adjusted to compensate. Each team will be scored using the formula of

(TOTAL SCORE / NUMBER OF PICKS) X STANDARD PICKS (7)

Thus adjusting everyone to the same number of picks. I originally thought we could standardize to the highest number of picks but this would just artificially inflate the numbers and give greater weight to fewer hits.

Thus, a perfect score would be 56. The worst possible score would be -7.

I will build an excel table for our picks between now and the draft, and it will be fairly easy to run the numbers after the draft if everyone is good with the above suggestion (or something similar).

This system looks fine to me. I'd be grateful if you could set it up.

That would be a way to judge who predicted the outcome of the real NHL draft best.

I don't know how much it would say about who actually drafted the best players, though.

By that system, let's say someone had 7 picks in the mock, and somehow all 7 went 1 thru 7 in the first round. The proposed system would give them a far worse rated draft than someone who picked a 5th rounder in the 5th round, a 6th rounder in the 6th round, etc.

I think the poll will do that, at least by the opinion of others on the board. We could evaluate this mock again next June (and seasons after) to get a better look at who merited where the GM selected them in reality.

 

[Chuck Downie]

Job well done Mr. Downie. Lets do it again!

Thanks again for doing all this work for us, Chuck

Thanks!

Great job Chuck and everyone else who took part.

Semi shocked we finished it. I think over 200 picks went faster than a typical WJC fantasy draft

Me too. It took nearly 2 months from when I proposed to do this to the last pick. The original 4 round mock idea seemed doable to me, but I definitely had my doubts about a 7 round mock.

 

[Captain Dave Poulin]

Me too. It took nearly 2 months from when I proposed to do this to the last pick. The original 4 round mock idea seemed doable to me, but I definitely had my doubts about a 7 round mock.

There was one page with rankings that I used to keep track of the names, and I had that tab open the entire time we were doing this

 

[Chuck Downie]

There was one page with rankings that I used to keep track of the names, and I had that tab open the entire time we were doing this

You probably weren`t alone with that approach. It`s a good way to not pick a player who was already selected

 

[denominator]

This system looks fine to me. I'd be grateful if you could set it up.

Will do. I'll try to get it into googlesheets so everyone can access.

I think the poll will do that, at least by the opinion of others on the board. We could evaluate this mock again next June (and seasons after) to get a better look at who merited where the GM selected them in reality.

I agree. The polls, and later revisits will be the best judge of this.

I have been trying to figure out a way to calculate which mock draft team got the best value (based on actual draft results), but I can't come up with anything that is fair. The problem is that any valuation I can come up with will automatically favour later picks (it's far more likely to have mock picked a player in the 7th round that was real drafted in the 4th than it is for a 3rd to 1st).

 

[denominator]

I have my excel sheet all set up, now just need the draft to actually happen so I can calculate everyone's scores.

In doing so, I was also giving some thought to tracking this long-term. I'm obviously going to keep doing my pre-draft rankings for a while yet (I actually still have my list for 2015 and 2016, just wasn't posting them here), so it's rather easy for me to keep our mock list and potentially track long-term winners.

My thinking would be that once per year, at the end of the season, I'd add in all the games played for this draft class. The GM that has the most GP is the winner (games played is, I think, the best measurement of a good draft - balances for goalies/defenders/2-way forwards). I've included the mock GMs - even if you don't take the same team in next year's mock, I could potentially start tracking users here over multiple drafts. Of course, this is all projecting years out but it is feasible and would end up as the best measurement of a successful draft. Plus, bragging rights.

The only hang-up on that system is that a couple of people were drafting for multiple teams. I'm okay with uneven picks across the teams as that's simply the way the NHL is, but it seems unfair that one mock GM would have twice as many picks just by drafting two teams. Thus, I would ask that prior to the real draft: chinatown81592 (Boston and Detroit), morinisbear (Colorado and Pittsburgh), RebelBully (Philadelphia and Washington), and sobrien (Columbus and Minnesota) pick one of their teams as their long-term set of players. If I don't hear back, I'll just assign them whichever has more picks.

 

[Chuck Downie]

I have my excel sheet all set up, now just need the draft to actually happen so I can calculate everyone's scores.

In doing so, I was also giving some thought to tracking this long-term. I'm obviously going to keep doing my pre-draft rankings for a while yet (I actually still have my list for 2015 and 2016, just wasn't posting them here), so it's rather easy for me to keep our mock list and potentially track long-term winners.

My thinking would be that once per year, at the end of the season, I'd add in all the games played for this draft class. The GM that has the most GP is the winner (games played is, I think, the best measurement of a good draft - balances for goalies/defenders/2-way forwards). I've included the mock GMs - even if you don't take the same team in next year's mock, I could potentially start tracking users here over multiple drafts. Of course, this is all projecting years out but it is feasible and would end up as the best measurement of a successful draft. Plus, bragging rights.

The only hang-up on that system is that a couple of people were drafting for multiple teams. I'm okay with uneven picks across the teams as that's simply the way the NHL is, but it seems unfair that one mock GM would have twice as many picks just by drafting two teams. Thus, I would ask that prior to the real draft: chinatown81592 (Boston and Detroit), morinisbear (Colorado and Pittsburgh), RebelBully (Philadelphia and Washington), and sobrien (Columbus and Minnesota) pick one of their teams as their long-term set of players. If I don't hear back, I'll just assign them whichever has more picks.

This is good. I'm curious to see how this turns out for sure, I watched a lot of non-NHL hockey this past season (actually for the last 30 years!!) and would like to see how my opinion stacks up.

After some more thought, the points system still looks good, but might need a little tweaking (if not now, in the future). Example - picking player X in the mock at 50 who went 60 in the actual draft should have some sort of value as opposed to the same player picked in the actual draft at 160. What I'm thinking is maybe a pick that is no less than 31 spots away from where they were drafted should get a point and you could increase the value of your proposed system by 1 (or more points to make it make more sense). The first couple rounds would be where the points are most likely to be earned, naturally, just as it is the easiest to predict. Just an idea.



A little off-topic, I was thinking of starting a thread later in the summer once the early starting leagues commence so that posters on this board can post what they are watching. I watched two thirds of Kunlun Red Star's games this past season and would be a good person to ask about the players on that team. I wanted to watch more Victoria Royals games but they are just too late for where I live but would have liked to ask some questions on this board about what NA West coasters saw (or whomever stays up that late). Getting familiar with what posters on this board tend to watch could be useful for all and could help with forming opinions for future mock drafts as well.

 

[denominator]

This is good. I'm curious to see how this turns out for sure, I watched a lot of non-NHL hockey this past season (actually for the last 30 years!!) and would like to see how my opinion stacks up.

After some more thought, the points system still looks good, but might need a little tweaking (if not now, in the future). Example - picking player X in the mock at 50 who went 60 in the actual draft should have some sort of value as opposed to the same player picked in the actual draft at 160. What I'm thinking is maybe a pick that is no less than 31 spots away from where they were drafted should get a point and you could increase the value of your proposed system by 1 (or more points to make it make more sense). The first couple rounds would be where the points are most likely to be earned, naturally, just as it is the easiest to predict. Just an idea.

I read your post this morning and got to think about it all day while wandering around in the forest.

You're right, ultimately. The problem is trying to score a first-round pick in the same way as a seventh-round pick. They're different beasts - anything you do to reward a good late pick has huge rewards for an early pick.

I am thinking of a similar points system as I initially proposed but taking into account your thoughts:

1 point for mock picking a player in the same round as he is actually drafted.
1 point for mock picking a player within 30 positions as he is actually drafted.
1 point for mock picking a player within 20 positions as he is actually drafted.
1 point for mock picking a player within 10 positions as he is actually drafted.
1 point for mock picking a player within 5 positions as he is actually drafted.
1 point for mock picking a player within 3 positions as he is actually drafted.
1 point for mock picking a player exactly where he is drafted.
1 point for mock picking a player 60+ positions later than he is actually drafted (value bonus).
-2 points for mock drafting a player that goes undrafted.
-1 point for mock drafting a player 60+ positions earlier than he is actually drafted (reach penalty).

The points stack, so a perfect pick is worth 7 points, a near miss is worth 6, and a wide miss is worth 1/2 points. Then I'd adjust the score to 7 pick average for all teams again, so a perfect score is 49 while a complete fail is -14.

I'm all for adjusting the points, but keep in mind if you're trying to reward hitting a late pick, it rewards hitting an earlier pick (NJ and Philly are essentially guaranteed 6 points in the first 2 picks using this system).