Login

Coops · 10-16-2018, 05:00 PM

Quote:1422 words and lotta research.

Alright, so if you didn't see my last piece, here it is.

Basically, I'm using a well known and well traveled single stat called "point shares" to reflect on how players in the SHL and, next, the SMJHL did in S43. Albeit, point shares with a slightly different take and approach.

If you missed the last article or are too lazy to click on the link, here's a brief primer: point shares are determined by attempting to map out the offensive and defensive contributions ("shares") that an individual skater has on the team's success. Offensive shares are mapped by such incredibly rare stats as "goals" and "assists." Whatever those are. Defensive point shares are calculated with the craziness of trying to map out time on ice, shots against, positional comparisons, goals against, and of course, plus minus.

There's been a couple changes to make the stat more "to the norm." To that effect, I've done the math and magically transformed point shares to win shares. I'll let you guess the secret formula that was used. Definitely not just dividing by 2. I also, for the SHL, incorporated weighted historical win shares. So skaters who played in S41 and S42 will have their win share nominally weighted by those scores. I didn't use that in this season in review article, but I'll be using it in future season preview articles. And finally, I conceded to the current popular hockey statistical community and weighted each score against "replacement level." AKA rookies in the SHL. The thought process there was pretty simple: rookies in the SHL generally represent the best of the SMJHL. I also added goalies, albeit without historical weighted stats.

Boom! You now have another meaningless statistic that I am going to write almost exclusively about. So I hope you *somewhat* like it.

Speaking of which, presenting the league leaders from S43. Since this is a recap article, I'll stick to only looking at S43 and ignoring the weighted historical impact from S42 and S41.

So first of all, what a season for the first three guys. To top 3 win shares over a replacement SMJHL player is a crazy good achievement. Only 3 players did that in all of S42 and all of S41 combined, so this is a very strong season by three S43 players, especially Jason Vissers who not only beat the 3 wSAR mark but provided another 1/3 of a win.

To underscore just how good a season those guys had, there's a comparison on the right of the list of top 10 skaters that breaks down skaters over threshold in S43, S42 and S41. You can see it was a historic season for those guys. Only one player in S42 managed to break 3 win shares over replacement (that being Oisin Fletcher, who just missed the top 10 this season with 2.31 wSAR).

So round of applause for the veterans. Let's take a look at how the S43 rookies did - both in terms of the cohort and historically.

Well it was quite the year for our two top wSAR rookie who almost managed to squeak into some historic (recent historic mind you, I don't have long term season data), and the over 1.0 win shares above replacement for rookie club. In the past three seasons, only two rookies have managed the feat. Any guess on those two S41 rookies?

Dekens had a great season, putting up a 0.96 wSAR. His offensive production was off the charts - 4th on the Manhattan Rage. It was his defensive point share that kept him from breaking the 1.0 mark and joining the exclusive club (mostly a nasty look -17, second worst on the Rage). That just speaks to his offensive point shares which were spectacular.

Flacko Lagerfield had a much more balanced approach to getting to 0.90 wSAR. With almost double the defensive point shares of Dekens, Lagerfield almost provided a whole team point based solely on his defensive play. Lagerfield also only had 24 minutes on the powerplay, compared to Dekens' 124 minutes at advantage - as well as Deken getting 2 minutes more per game. With some changes to time on ice - while keeping the same production - Lagerfield might've hit the 1.0 rookie club.

Here's a break down of the S43 Forwards, Defenseman and (new!) Goalies:

So it's pretty clear that either the formula or the sim favours forwards, then defensemen, and far less goalies. STHS veterans know that this is a reality of the sim - save percentages do not closely resemble that of real hockey. And mixed with this imperfect formula, boom, you have the making for some weird wacky goalie statistics.

Since we don't have many rookie goalies usually, the replacement is considered to be 80% of the average. And since the goalie share formula greatly favors saves and shots against, you'll notice losing team goaltenders tend to get highly rated. There's an argument to be made that facing 1600 shots and letting in 158 is as, or more in these terms, impressive than facing 1082 shots and letting in 113. That's the disparity between Chris Patridge of the Pride, rated by wSAR as the second best goalie this season, and Geezus Kryyst, not even in the top 10 Goalie wSAR (he has -0.28).

Here's the rundown of this phenomena: win shares have to split by the team in the formula. If the skaters get extra win shares for say, not allowing shots, the goaltender has to get less share of the credit. For a goaltender on a team that allowed 1600 shots, obviously any wins that happen can be more attributed to them. It makes a certain amount of sense, but it does screw with our thinking of "good" and "bad."

But enough about goalies - we can also look at how the forwards and defenseman fared in the SHL. I was interested in how these forwards and defensemen fair against the last three seasons, so I compared thresholds for forwards and defensemen to see how they stacked up.

It was a good season for forwards and defensemen in the SHL. Both continued to see more and more players get higher wSAR scores, indicating that either players are producing more or that the rookie average ("replacement value") is getting lower and less impacting scores. It doesn't appear to be the latter though, since you can clearly see that the number of rookies in either thresholds of win shares has actually decreased season over season.

In either case, there is a marginal increase in forwards making the jump to the next threshold of win shares. Where there were 15 forwards over 2.0 wSAR in S41, there were 17 last season and 19 this season. If the trend continues, more and more forwards might be contributing 2 wins (of 50) to their teams. In context, that means more forwards might help move their team from golf in April (Renegades 25 wins S43) to playoffs (Dragons 27 wins S43).

The increase season on season is even more pronounced on defense. While only 3 defensemen in S42 and in S41 managed to get above 2.0 win shares, a whopping 8 managed it in S43. The gains spread down the table, with more and more defensemen providing benefit to their teams leading to a record low number under 1.0 win shares above replacement. Take a look at how the defensemen over 1.5 win shares doubled from S41 to S43.

Alright, that's gonna wrap this S43 Season in Review article. Since I'm a stats nuts, I'm going to stick to writing about this stuff with a SMJHL in review coming out later this week / early next. Keep eyes out for other wSAR related articles - since I like to find an angle and grind it out.

***luke*** · 10-16-2018, 05:06 PM

This is what I like to see from people. Good job coops. great article

ArGarBarGar · 10-16-2018, 07:37 PM

Looks like I'm getting another Stevens nomination, good work!

Alonzo Garbanzo Final Tallies (Among Defensemen):
2nd in Goals (208), All-Time Assists Leader (765)*, All-Time Points Leader (973), 3rd in Hits (2587), All-Time Blocked Shots Leader (1882)*
*All-Time Leader Among All Skaters
Player Profile | Update Thread
[Image: IeEV7Iv.png]

monkey10pts · 10-18-2018, 11:32 AM

The more I see the stats between Lagerfield @JSS and myself, I just can't help but think the guy gets the Jesster. I mean his entire site username is in the award name for goodness sake!. For real he had an awesome season and he truly battled through the playoffs as well. He is as deserving as anyone.

Navigation

Extra Menu

About us