Completely outside the topic of the Eastern Conference but how about what is going on in Vegas right now? I mean in all actuality if you try to wrap your head around making The Cup Final in year one it’s almost impossible. There will be a movie made about this one – that’s for sure.

*

This summer for me has been all about trying to find the little idiosyncrasies that you can use to gain an advantage one year leagues next year. I’m excited about some of the modelling I’ve done around shooting percentage and expected goals (ixGF). I’m really looking forward to seeing how stripping out majors and misconducts from penalty minutes can create a more predictable stat line (PIMs). I’ve mentioned before valuing certain positions more than others tends to lend me to draft centers much later.

The data for all of this work came from Corsica.

For that reason it seems unlikely that goal scoring from a center is going to be the big stat you’ll be trying to get from them – instead assists are usually the wheel house for this position. Assists are great except that a secondary assist is usually the luck of the draw and not ratable from year to year. Looking at the last four seasons here is the primary assist rate for the Top 100 scorers each season:

Primary Assist Rate for Top 100 Point Producers by Season

Season	Rate
2014-2015	58.4%
2015-2016	58.5%
2016-2017	58.6%
2017-2018	57.8%

If you tighten this list up the average percentage for the Top 50 scorers in each of these seasons become 60%. This is valuable information! We know that primary assists are a great indicator of point production. I ran the primary assist rate for the top 500 point producers for the last 4 seasons and predicted the point-per-game of those players. The model has a significance level that says it is a predictive value. Check out this graph – basically we see that the higher the Y axis gets the more likely it the primary assist rate will be equal to or greater than 58%.

Similarly if we look at point totals and primary assist totals we can see a nice correlation forming:

Ok so excuse all the graphs but the point is primary assists are a great predictor of point totals. So we know the top 50 scorers tend to have around 60% of their assists being primary assists. We also tend to see that players can sustain in the same ballpark the rate of their primary assists. So using this as a measuring stick we can look at players that off their 3 year average can reasonably increase their primary assist total and as correlated their point totals as a whole.

Since it’s the summer I’ll dump the top 20 players from each division in a table and then over the next couple weeks look a little closer at specific players. The requirements for this table are that the player scored 40 points.

Primary Assist Rate 2017-2018 Metro

Player	A	A1	A1 Pct% 2017-18	A1 % 3 year average	Delta
JEFF.SKINNER	25	13	52.0	71.4	19.4%
KEVIN.HAYES	19	10	52.6	65.9	13.3%
JAKE.GUENTZEL	26	15	57.7	70.0	12.3%
TAYLOR.HALL	52	30	57.7	67.9	10.2%
VLADISLAV.NAMESTNIKOV	26	11	42.3	50.5	8.2%
KYLE.PALMIERI	19	8	42.1	49.5	7.4%
TEUVO.TERAVAINEN	39	23	59.0	64.5	5.5%
EVGENI.MALKIN	56	31	55.4	60.2	4.9%
JOHN.TAVARES	46	28	60.9	65.6	4.7%
J.T..MILLER	35	19	54.3	58.7	4.4%
SHAYNE.GOSTISBEHERE	50	23	46.0	49.1	3.1%
JOSH.BAILEY	51	29	56.9	59.4	2.5%
SIDNEY.CROSBY	59	33	55.9	58.3	2.4%
JORDAN.STAAL	27	14	51.9	54.2	2.3%
TRAVIS.KONECNY	24	13	54.2	56.5	2.3%
ELIAS.LINDHOLM	28	18	64.3	66.6	2.3%
KRIS.LETANG	43	17	39.5	41.6	2.1%
CLAUDE.GIROUX	67	32	47.8	49.8	2.0%
PATRICK.MAROON	25	15	60.0 📢 advertisement:	61.5	1.5%
THOMAS.VANEK	31	22	71.0	71.7	0.7%

Primary Assist Rate 2017-2018 Atlantic


Player	A	A1	A1 Pct% 2017-18	A1 % 3 year average	Delta
RILEY.NASH	27	13	48.1	61.7	13.6%
TYLER.JOHNSON	28	14	50.0	61.7	11.7%
VINCENT.TROCHECK	44	22	50.0	60.6	10.6%
MIKE.HOFFMAN	34	19	55.9	64.2	8.3%
VLADISLAV.NAMESTNIKOV	26	11	42.3	50.5	8.2%
STEVEN.STAMKOS	57	36	63.2	70.4	7.2%
ALEX.KILLORN	31	17	54.8	61.3	6.5%
NIKITA.KUCHEROV	61	34	55.7	61.9	6.2%
ALEX.GALCHENYUK	31	16	51.6	57.5	5.9%
JONATHAN.DROUIN	33	17	51.5	57.1	5.6%
NAZEM.KADRI	23	12	52.2	56.9	4.8%
J.T..MILLER	35	19	54.3	58.7	4.4%
DYLAN.LARKIN	45	25	55.6	59.4	3.9%
BRENDAN.GALLAGHER	23	13	56.5	60.3	3.8%
JAMES.VAN RIEMSDYK	18	11	61.1	64.7	3.5%
BRAD.MARCHAND	51	30	58.8	62.3	3.5%
AUSTON.MATTHEWS	29	18	62.1	65.5	3.4%
JACK.EICHEL	39	23	59.0	61.8	2.8%
VICTOR.HEDMAN	46	23	50.0	52.6	2.6%
ALEKSANDER.BARKOV	48	26	54.2	56.7	2.6%

It is crazy to imagine that Taylor Hall has compared to his three-year career average more room to increase his primary assist rate. Same goes for Vincent Trocheck, after his remarkable season he also has the opportunity to improve on his career average. These percentages give a great baseline for trying to understand where we may see corrections next year; the flipside could be said for the rate of secondary assists produced by a player. I’ll dig a little deeper into that rate next week and look at the specific players on these lists as well.

So to conclude – so far I have written about shooting percentage differential, stripping out penalty minutes from majors/misconducts, expected goal production, and now primary assist rates.

Any questions?

Follow me on twitter at @DH_jcameronmetz

UPCOMING GAMES

Dec 22 - 12:12	NYR	vs	CAR
Dec 22 - 17:12	T.B	vs	FLA
Dec 22 - 17:12	UTA	vs	ANA
Dec 22 - 17:12	WSH	vs	L.A
Dec 22 - 20:12	EDM	vs	OTT
Dec 22 - 20:12	COL	vs	SEA