Login

***hhh81*** · 01-06-2022, 10:50 PM

Welcome back to my fool’s errand of applying my Bojo Score to the S62 Alexis Metzler race (best defensive defenseman). If you missed Part 1, go back and read it right now. This won’t make sense until you do.

Part 2: Digging Deeper in the shortlist

Once I identify a short list, I then look into figuring out the context of that player/team. I want to first say that this is going to be a tough analysis. These are 12 very good players, who all have reasons to be considered elite defensively. Naturally, some are going to have to get cut if I’m ever going to get to a finalist group.

[Image: unknown.png?width=911&height=568]

[Image: unknown.png?width=911&height=568]

One of my first steps is to compare to peers on the same team. A lot of the advanced stats have some degree of reliance on team performance; a player on a good team will naturally have higher differentials, CF%, etc and a lower GA/60. How does the player sit within their own team? This is where I think the Corsi and Fenwick rel stats can be useful.

In this case, I think looking at Newfoundland’s four in the top 12 is a good starting place. The team was very good, obviously, based on the regular and postseason results. All 6 of their defensemen played between 18:30 and 21 minutes per night (so ice time was very even). As a team, their GA/60 is incredibly low, ranging from 1.7 to 2.1, and every player except Big Boss had over 60 CF%. When you dig into the relative possession stats, you noticed pretty quickly Jannick Nordstern has a team worst -5.9 Corsi rel, and the weakest shot/goal differentials among these four Berserkers. Part of that could be due to playing with a less defensively sound rookie (Rand al’Thor) based on the data here. He also played the least SHTOI among the top 12 comparison group, just barely qualifying for this group at 1:00 SHTOI. He did get points in 6 of the 7 Bojo Score categories and had all around solid numbers in several spots—he tied for the third most takeaways in the league (48). Despite that, I think the negative relative possession stats, combined with the lack of PK time makes for an easier cut.

Moving onto the other three Berserkers, all scored well in the Bojo Score. The stats make it seem like O'Neil-Metzler was the top PK pair, with Williams II joining Nordstern on the 2nd PK unit. Metzler is the crown jewel of this group, having the highest Bojo Score at 6, only missing points in SB/CA. When looking at the top 12 reference group, he was top 3 in 5/7 categories, landing 3rd in TA-GA, SHTOI, 2nd in both goal and shot differential. He and teammate Jay O'Neil tied with a 1.7 GA/60 that led all defenseman in the SMJHL (and this group). Metzler had a superb season, and I see no reason to cut him at this point.

Jay O'Neil finished second on Newfoundland in takeaways (39), hits (115), and blocks (90), while having a team worst 67 giveaways. Only three SMJHLers had more turnovers last season. Despite this, O’Neil led the comparison group with the best GA/60 (tied with Metzler), goal differential, and shot differential. While Newfoundland as a whole didn’t spend a lot of time blocking shots (why block a shot when you always have the puck?), O’Neil had the fourth highest SB/CA among the reference group and the best on Newfoundland, but the 2nd lowest TA/GA among the reference group.

Williams II had goals against, goal, and shot differential comparable to his teammates, though a little lower. This difference likely is negligible when adjusted for Metzler and O’Neil’s slightly elevated PDO. In the reference group, Williams had the worst TA/GA at 0.462, and was also 2nd to last in SB/CA at just 5.7%, along with some average or below average results within the comparison group. He certainly wasn't a bad defenseman, but I have a hard time including him in the elite defensive defenseman this season without a single standout positive statistic. O’Neil certainly has merit in the discussion, though I will cut him here because I think that low of a TA/GA ratio really diminishes some of his strong points. Metzler is as good or better in most categories and was a leader across the top reference group in a way O’Neil really wasn’t.

Looking into the two nominees from Vancouver’s top pairing, Conner Tanner and Landon Fischerman have very similar numbers. It’s clear they carried the blue line for the Whalers, with the only positive relative Corsi/Fenwick among their six defenders. Both had incredible seasons. Head-to-head, Tanner has the edge in SHTOI, TA-GA, hits, blocks, goal differential, SB/CA, and HIT/PIM, albeit many of those being very slight. Fischerman only had the (slight) edge in rel CF/FF and shot differential. When compared to the rest of this “top” reference group, Fisch did not land in the top 3 in any of the categories, while Tanner had the third highest goal differential at a 2.0. In a more subjective sense, looking at the two players’ seasons side by side, Tanner’s stat line “looks more” like a defensive standout, versus the all-around elite season from Fisch, but Tanner having the stronger GA-TA and SHTOI figures. I drop Fisch from consideration here.

ANC's Vejonis got points in the Bojo Score from every category except SB/CA and finished in the middle among the comparison group in other categories. He wasn’t bottom 3 in any category among these 12 and tied Devin Williams III for third in GA/60 (1.9). His 38 TkAs tied for ANC's lead and was top 20 league wide, and he led Armada defenders in SHTOI. He and Cup Noodle had the lowest SA/60 for ANC as well as the highest possession stats on a team that tied for the best record leaguewide. His stats are very impressive, especially his 1.267 TA/GA, which was the only one above 1 among the four Armada defenders who met the SHTOI minimum. His 6.86% SA/CA was the highest among qualifying ANC players, though pretty low outside the team. There’s only one issue with Vejonis (more on that later).

To discuss Nevada’s Melvin Majestik-Moose, I would be remiss to omit fellow Battalion member Adelie de Pengu from the discussion. She earned a 4/7 on the Bojo Score, getting partial points in 4 of the 6 categories she qualified for. Even then, her 50 takeaways (2nd in the league of any player) and 1.923 TA-GA ratio are incredible. Her 1:28 SHTOI led Nevada (and was 22 seconds more on average than the Moose). Majestik-Moose was also 2nd on Nevada with 46 takeaways. De Pengu had more hits (154 to 144), a better TA-GA, a lower GA/60 (2.2 to 2.7), and a +0.4 goal differential to Majestik-Moose’ -0.3. How then did the Bojo Score give Moose the edge? While his shot differential was slightly better (+4.7 versus +3.1) and his SB/CA edged her out (11.82% to 10.48%), none of those gave a points edge in the Bojo Score. Majestik-Moose’s 10.286 HIT/PIM was drastically better than de Pengu’s 4.4 and earned Majestik-Moose a full point lead on de Pengu. This is a great example of how the Bojo Score is a tool, but not 100% of how I decide my nominees. I think de Pengu’s edge in TA/GA, SHTOI, goals against, and goal differential weighs more to me than Majestik-Moose’s edge in other categories. Here, I think I would have to bring de Pengu into my top group for the S62 Metzler. When added into the reference group, de Pengu has the 3rd most blocks, 2nd most hits, a SB/CA in the top 5, and the best TA/GA at 1.92.

Turning our attention to the pair from Great Falls, Arturs Turmovs and Jonathon Hagan are intriguing players. Both played over a minute and a half short handed for the Grizzlies, and Turmovs was top 20 in takeways (42, tied for 13th), and 2nd in the league with 194 blocks. He had the highest SB/CA in the SMJHL at a whopping 15.76%. Unfortunately, he also was middle of the pack among the Grizzlies on stopping shots and goals (2.9 GA/60, 31.4 SA/60), and some of the weakest figures among the reference group. He was bottom two in GA/60, goal differential, and shot differential, and also had negative relative possession stats. This could reflect being used as a truly defensive defenseman by Great Falls, leading to less opportunities for positive possession. I wouldn’t judge someone for nominating Turmovs; with how the SMJHL is set up, truly elite shot blocking is incredibly rare. Turmovs was the best in blocking quality attempts, and with Regina’s Chris Chelios Jr, was one of only 2 players to meet the 15% threshold for my Bojo Score.

Then we come to Hagan, who the initial Bojo Score really liked. His 6.5 was the highest this season, as he only fell short of full points in SB/CA. While Turmovs had higher results in the raw hits, blocks, and takeaways, Hagan thoroughly looks better in this model with a HIT/PIM over 2 higher, nearly a goal against less per 60, drastically stronger goal/shot differentials (+1.5 goals, +11.3 shots), better relative and raw possession stats, and a TA/GA right at 1.0. It’s everything you want to see out of a defensive defenseman in my mind. While a little lower than Turmov’s SB/CA, Hagan still finished 4th in the league with a hair over 13%, good enough for second among the reference group. He also had the third highest HIT/PIM in the top 12, which reflects a clean physical game to me. In the end, I think Turmovs had a great season and deserves some recognition. With the amount of great defensive performances here, I cut him from consideration based on the areas where he was the weakest.

Finally, we have our two contenders from Maine. Robertson and Mr. the Walrus were the clear primary penalty killers on the blueline for the Timber, averaging 1:41 and 1:59 SHTOI, respectively, good for 1st and 2nd in the reference group. Unlike pairs from other teams, they each excelled in different categories in the Bojo Score. Duncan the Walrus had the second best TA/GA in the reference group at a 1.45 (he also tied for the third most takeaways on the season, with 48). Robertson led the group with his 13.25 HIT/PIM (only Moose was also in double digits). While others in the reference group were well-rounded in their defensive game, the Timber representatives were very specialized. Walrus had the lowest HIT/PIM, exacerbated by only taking 59 hits on the season. Physicality wasn't a major part of his game. On the flipside, Robertson had a shockingly low number of blocks (31), leading to an unsurprisingly low SB/CA. Neither had particularly strong goal or shot differentials. Within the Timber lineup, Walrus' possession numbers look much stronger especially when you factor in Robertson's 104 PDO. No full-time defender in the SMJHL had a better puck luck last season as reflected in PDO. I could understand an argument for including either in nominations; I think I would lean more toward Walrus if I had to pick one. Because of the boom-bust aspect of the relevant stats, I feel I would probably include both in the 2nd tier of “valid, but not my top group” with O’Neil, Turmovs, and Majestik Moose, rather than my finalist group.

Who are the finalists?

After this analysis, I’m left with Merlin Metzler (NL), Adelie de Pengu (NBB), Connor Tanner (VAN), Valdis Vejonis (ANC), and Jonathon Hagan (GFG). An impressive group, to say the least! This is where I regret my life choices in having to narrow this field to a top 3.
I start with the easiest choice—Vejonis was inactive for the entirety of S62. Philosophically I believe awards should be given to active players, so despite Vejonis recently updating after 3+ months away, I will leave him out.

Four remain, and I honestly don’t even want to try to break it down further from here.

[Image: unknown.png?width=911&height=180]

[Image: unknown.png?width=911&height=180]

I’ve been working on this media since Monday, and probably put at least 15 hours into trying to narrow this whole process down. Each of these four stands out as a great defensive defender, and I believe deserve to be a Metzler finalist. If you forced me to pick a winner, I would probably put Hagen or Metzler first on my ballot with de Pengu third. All of these players deserve some recognition for the seasons they put together, though.

And there you go, my analysis of the S62 Metzler race using my brains and my Bojo Score. What do you think? How do you think my analysis compares to the actual results?

Here is a spreadsheet of my work for anyone who wants to look into the stats.