Measuring Teams’ Draft Success Using Analytics Pt. 2

Let’s keep this gravy train going, shall we? Only this time, let’s break the draft up by player position.

GravyTrain

If you haven’t already, today’s post won’t make much sense without first reading last day’s post. For a quick refresher for those who need it though, and because a picture is worth a thousand words, I’m going to let this graph do most of the talking:

GVT7_Trendline

GVT = “Goals Versus Threshold,” a measure of how many goals a player is worth to his team.

What you’re looking at is the average GVT/year of all players taken in the 1992-2008 drafts, by draft slot, using only their first seven years’ worth of GVT – and a nice, pretty red line. If we break the graph down by player position though, things start getting pretty interesting. Let’s take a look at just forwards:

GVT7_Forwards_All Years

Ok, ok, this isn’t the most interesting plot in the world, I’ll admit, but one point of interest that some of you may have already picked up on is that dip after pick 29. That’s pretty significant, and almost makes it seem like the difference between the 1st round and 2nd round for forwards is literally night and day. That discussion might be had another day though. Moving on to defencemen:

GVT7_Defencemen_All Years

Now that’s a little more crazy, ain’t it? The forwards plot was more or less a bumpy slope. This- is a little more bumpy and a little less of a slope. What does this tell us? Probably what we’ve been told all along, actually. It’s almost common knowledge to hockey fans that follow the draft that forwards are easier to predict and project than defencemen. Well, the numbers look like they whole-heartedly agree.

It’s also important to remember though that because there are always less defencemen drafted than forwards, the forwards plot would naturally be expected to be smoother, if only somewhat. Less defencemen drafted = more weight given to each, individual defencemen’s numbers = higher variablity = bumpier graph. Speaking of higher variability:

GVT7_Goalies_All Years

Footage of my reaction to seeing this for the first time can be found here. Holy jumpin’, is that nasty or what? I can sum up what to take away from this graph in one sentence: You have a half-decent chance of picking an actual NHL goaltender if he’s talented enough to be taken in the first round of the draft- or bad enough that he’s picked with the 200th pick- THERE IS NO MIDDLE GROUND.

In all seriousness though, if you weren’t adverse to drafting goalies before, hopefully this helps you see why much of the “fancy stats” universe shies away from picking goalies, at least with higher draft picks anyways. Yes, you probably have a higher chance of selecting an NHL starter if he’s talented enough at the time to warrant consideration for using a 1st round pick, BUT the variability in goalie selections is so huge that you will most likely be safer taking a skater in place of the ‘tender. That’s a conversation for another day though, as there’s A LOT that can be said about that topic.

Moving on to the meat and potatoes of today’s post, let’s see who’s good at drafting what, starting with the forwards:

GVT7_Results_Forwards_00-08

For another refresher, what this chart tells us, looking at Edmonton as an example, is that Edmonton’s forwards drafted from 2000-2008, produced on average, a GVT/year that was 0.03 below what you would have expected them to produce, given their spot in the draft.

Quick Thoughts

1) Hey, Edmonton isn’t a dumpster fire at something, we’re only below average!

2) Poor Florida, both the team and the rest of the state.

3) How does Buffalo go from drafting like that, to finishing like they did this year? Talk about mismanaging your assets.

4) San Jose’s management team still doesn’t get enough credit.

5) Chicago, by this measure was only slightly better at drafting forwards than Calgary, averaging a -0.28. There has to be quite a few busts to go along with Toews and Kane there.

Moving on to defencemen now:

GVT7_Results_defence_00-08

Quick Thoughts

1) San Jose’s management team doesn’t get enough credit.

2) I don’t know about you all, but I’ve always heard it said that Nashville was a defenceman-producing factory. This makes Nashville look like that and then some. Props to them.

3) Poor Florida.

4) BOSTON, WHAT THE HELL ARE YOU DOING DOWN THERE?! I’d be lying if I said I wasn’t surprised to see them so low.  What a dumpster fire.

5) Speaking of dumpster fires- you know where I’m going with this. Quick, name the 24 defencemen that Edmonton drafted in that timespan! Petry, Greene- ugh.

6) Go ahead and gloat, Flames fans, you have something to be proud of.

And now onto the netminders:

GVT7_Results_goalies_00-08

Thoughts:

1) There’s not a whole lot to say about this list. Just looking at the amount of goalies picked by each team, for many teams you really only need one sure-fire pick to pan out for your team’s goalie-drafting record to look good. Take this all with a couple grains of salt.

2) Except for Calgary… let us all just assume, with no bias whatsoever, that they are terrible at drafting goalies.

3) Is it not incredibly ironic that the Flyers, the team that always seems to have goaltending problems and controversies, picked the most goalies in this nine year window, and has very little to show for it?

4) At least Edmonton didn’t waste too much time drafting goalies. That’s a plus, right?

5) I wonder how much of the “Brodeur effect” had to do with New Jersey drafting the fewest goalies out of any team, a mere four, in this timespan. I’ve heard that he had some say in the team’s handling of other goalies, including drafting, but those could just be rumours. Still makes you think, though.

Want to See More?

Obviously you can dissect these charts as you wish, and there’s much and more you could take away from them all. Being the hip and cool guy that I am, I spent a lot of time staring these charts down until my eyes near fell out. Now I’ll allow you fine people to do the same. Go crazy. It makes for some good hockey talk.

It’s also easy for me to adjust the search criteria for these GVT charts, so if any of you have any curiosity at all, feel free to leave a comment below or tweet at me and I can whip up a batch of whatever charts you’d like to see. I can alter the draft year, player positions and draft round. Enjoy.

This entry was posted in Uncategorized. Bookmark the permalink.



Leave a Reply