ATD 2017 Draft Thread IV

[jarek]

All I see is a bunch of random points scattered around in a general downward trend.

Where on earth did you even get all those super high point scores? Nobody on the list I showed you scored more than 121.

This doesn't pass the smell test for me.

Quite a bit- a flat number or sum of numbers gives you very little insight what a player actually did in-season. VsX doesn't tell you that, and neither does a lump ESP total. Finishes do. That's the magic of context.

You do realize that VsX is computed based on percentages on a season-by-season calculation, right? The scores seventies computed for Ramsay and Luce ES scoring are percentages based on the exact same criteria. You'll have to ask him for the exact benchmark he came up with but it would have been applied consistently to both players.

There's a really good reason why the vast majority of us have largely moved away from treating scoring finishes like they're all that meaningful.

Then why did one have significantly better ES scoring finishes? Could it be one was simply a better scorer at ES, or is this another "coincidence"?

I have another explanation: one spread out their ES scoring more evenly, thus having fewer chances to place high on the ES scoring list. seventies already showed, based on the percentages, that they were fairly equivalent at ES production. Percentages paint a more accurate picture because two different 14th place finishes, for example, are not necessarily equal.

 

[seventieslord]

Quite a bit- a flat number or sum of numbers gives you very little insight what a player actually did in-season. VsX doesn't tell you that, and neither does a lump ESP total. Finishes do. That's the magic of context.

We used to care about "finishes" like 7 years ago. Now we care much more about the quantities associated with them. Quantity matters much more. If you care about context as much as you say, then you should greatly prefer a quantitative measure like VsX over "ES points finishes", because the latter really doesn't mean even close to the same thing from one decade to the next or even from one season to the next. With the tightly packed leaderboards, there could be virtually no difference between 15th and 30th, or a strangely large difference between 5th and 6th. Finishes aren't good context, not for us.

 

[seventieslord]

Okay, let's do that:

Anddd this would appear to be pretty straight-forward- less points, more Selke votes.

I appreciate the effort. A few things:

1. This was about ESP, so you should be using ESP. Padding your raw point totals with PP points says nothing about how much you are concentrating on defense in even strength situations, right?

2. There are 24 data points here. You have a solid top-10 that's usable for every season in this range (no crap 1-3 point totals), giving you 80 data points. To establish a trend, you should use more data than you have, instead of just finalists.

3. (edit) - actually, it looks like you took totals for 24 players over this period, right? I think that's horribly flawed. If you want to prove that selke votes in a season are negatively correlated with ESP in a season, then you should look at every relevant player in every season separately.

 

[Voight]

Indiana Ice will select C/RW David Backes!

Did some research and it seems he's played RW fairly often.

 

[jarek]

I appreciate the effort. A few things:

1. This was about ESP, so you should be using ESP.

2. There are 24 data points here. You have a solid top-10 that's usable for every season in this range (no crap 1-3 point totals), giving you 80 data points. To establish a trend, you should use more data than you have, instead of just finalists.

3. (edit) - actually, it looks like you took totals for 24 players over this period, right? I think that's horribly flawed. If you want to prove that selke votes in a season are negatively correlated with ESP in a season, then you should look at every relevant player in every season separately.

The data I gave him was total points for each player in the top-10. I did not use ESP.

I can generate a new list with just ESP, but I'm going for lunch now so it won't be for a little while.

3. (edit) - actually, it looks like you took totals for 24 players over this period, right? I think that's horribly flawed. If you want to prove that selke votes in a season are negatively correlated with ESP in a season, then you should look at every relevant player in every season separately.

.. Really? That makes no sense.

 

[Sturminator]

Indiana Ice will select C/RW David Backes!

Did some research and it seems he's played RW fairly often.

If you meet BB's team in the playoffs, how many shifts before Backes and Benn throw down?

 

[seventieslord]

Indiana Ice will select C/RW David Backes!

Did some research and it seems he's played RW fairly often.

The only seasons Backes hasn't taken scads of faceoffs are his rookie season and this year:

http://www.hockey-reference.com/players/b/backeda01.html

I wouldn't personally put him on the wing based on that little evidence, but I guess shakier claims have been accepted before.

 

[jarek]

If you meet BB's team in the playoffs, how many shifts before Backes and Benn throw down?

 

[Voight]

The only seasons Backes hasn't taken scads of faceoffs are his rookie season and this year:

http://www.hockey-reference.com/players/b/backeda01.html

I wouldn't personally put him on the wing based on that little evidence, but I guess shakier claims have been accepted before.

He's likely going to be a full time RW going forward given Boston has Bergeron-xxx-xxx down the middle so that's where I was going with it. But we'll see if I can nab someone else to take his spot next pick.

 

[tony d]

While I do need a 4th line right winger I'm going to draft my 1st extra here. A guy that I thought about picking had I kept 408 way back when, he's still available at 493. He did go to early last year but right now he's going to be a bench defenseman who can come up and fill in if one of my defensemen has to sit out due to injury.

Welcome to Toledo with Pick 493 defenseman PK Subban

Subban won the 2013 Norris Trophy.

next has been pmed.

 

[seventieslord]

The trendline on EE's first chart would have you believe that a 50% drop in points leads to approximately 4X as many selke votes. The truth is much less staggering.

As the trendline moves from 5 voting points to over 180, points decrease from about 46 to about 43.

Excel calculates a correllation of -0.07.

Pointless exercise, really.

 

[Cyborg LeClair]

The Borg Collective finish our main roster by adding a defenseman as "steady as the Empire State Building". A big, strong, defensive defenseman who was always consistent defensively, and despite his low point totals, showed up to score in some clutch situations. He was an Olympic Gold medalist for Team USA and 4 time Stanley Cup Champion on the Islanders Dynasty. He'll bring balance to our bottom pairing and complete our penalty killing units. The Borg Collective proudly assimilate "The Wolfman": Ken Morrow

A.K.A Ken Morrborg ​

 

[jarek]

The trendline on EE's first chart would have you believe that a 50% drop in points leads to approximately 4X as many selke votes. The truth is much less staggering.

As the trendline moves from 5 voting points to over 180, points decrease from about 46 to about 43.

Excel calculates a correllation of -0.07.

Pointless exercise, really.

Did you use total points or ES points? The values I provided were total points.

 

[jarek]

Read the title of the graph

I realize that. I'm not sure if seventies used the totals I provided (total points), or if he got the ESP by himself.

Anyways, that data suggests exactly what I had thought the whole time - there is very little correlation between offensive production and Selke votes.

 

[Dreakmur]

Consistently good? Sure. Recognized for his elite play in the form of Selke votes? Only when he sacrificed enough offense. That's not a knock, just a truth- if you are going to focus more on D, you spend less time attacking. I don't get why that's controversial.

It's controversial because it's just about the exact opposite of how Selke voters generally vote. As a general rule, offensive production increases likelihood of Selke recognition. Go through some of the defensive forwards - you'll see their Selke placements follow their point totals almost exactly.

An accomplished historian like you obviously knows the purpose of the Selke has shifted in its time. We now look at it more as the award for the scorer who plays the best defense, with the odd John Madden sneaking in. Back then? Not so much.

Go through some of the defensive forwards - you'll see their Selke placements follow their point totals almost exactly. Carbonneau, Poulin, Kasper... even Ramsay....

I'll say it again: play more defense, thus making yourself more likely to be recognized for your defensive play, score less points. It happened to Clarke as well. His Selke came in his fifth-highest scoring season, FYI.

Well, Clarke's Selke win came in his 7th highest scoring season. 5 of those higher scoring seasons came before the Selke existed. His 6th highest scoring season was the 1st season of the Selke, and Clarke did well in voting.

Since the Selke was created, Clarke's 3 best offensive seasons... he received his 3 best Selke voting results.

 

[Sturminator]

Bure = The Bure not with Candace

 

[chaosrevolver]

The Cobalt Silver Kings will select a power forward who was a fix on the infamous Goal A Game Line. He was a tough player and I think he will compliment my other two fourth liners quite well. He is a 2nd team all star and was named to Canadas 1972 Summit Series Team (and left in Russia)..

The Cobalt Silver Kings select LW - Vic Hadfield

 

[seventieslord]

The Cobalt Silver Kings will select a power forward who was a fix on the infamous Goal A Game Line. He was a tough player and I think he will compliment my other two fourth liners quite well. He is a 2nd team all star and was named to Canadas 1972 Summit Series Team (and left in Russia)..

The Cobalt Silver Kings select LW - Vic Hadfield

As overrated as he often is, I'd take him over Wendel Clark.

 

[Hawkey Town 18]

Oh good we're still doing these...

Stamkos: Injured Needs Doctor
Tikkanen: Bothering Rigorously Often
Goring: A Meaningful Aquisition

 

[chaosrevolver]

As overrated as he often is, I'd take him over Wendel Clark.

Yeah Im not looking for a ton out of him either. Just looking to someone that can get to the net, be physical and finish some passes off from Oatman or clean some rebounds up from McGee rushes. I have a very defensive third line so didnt really want to go in that direction with my fourth line.

 

[Hawkey Town 18]

As the table I posted when drafting him shows, Tony Esposito's playoff woes are overblown, but we're still going to take a backup who's a great big game performer just in case. The Chicago Shamrocks select Hap Holmes, G

 

[BenchBrawl]

While I do need a 4th line right winger I'm going to draft my 1st extra here. A guy that I thought about picking had I kept 408 way back when, he's still available at 493. He did go to early last year but right now he's going to be a bench defenseman who can come up and fill in if one of my defensemen has to sit out due to injury.

Welcome to Toledo with Pick 493 defenseman PK Subban

Subban won the 2013 Norris Trophy.

next has been pmed.

I think Subban is a good pick here tony.Subban's peak is very high for this range, he brings it in the playoffs, is physically strong and can be used as a PP triggerman.

 

[jarek]

As the table I posted when drafting him shows, Tony Esposito's playoff woes are overblown, but we're still going to take a backup who's a great big game performer just in case. The Chicago Shamrocks select Hap Holmes, G

It was between him and Gump for me. I still have no idea if Worsley is actually better or not.

 

[chaosrevolver]

I think Subban is a good pick here tony.Subban's peak is very high for this range, he brings it in the playoffs, is physically strong and can be used as a PP triggerman.

Yeah and while Karlsson is definitely ahead on career..I am not sure that there is this much of a gap. Though..one far better season can make a big difference for sure.

Note: I should add that im not saying PK is really close to Karlsson but rather that defensively and physically he has had the edge for much of his career but EK has been an offensive dynamo since 2013 and I know that sticks out more and it should given his defensive improvements that have been made.

 

[Stoneberg]

West Lafayette will round out it's forward group with Ed Sandford, LW.

Sandford ticks off pretty much every box you want for a fourth liner.

TDMM bio