Claimed:VHLM By the Numbers [Final 6/6]

[Cowboy 20]

VHLM By the Numbers

 

 

*denotes eventual Founder's Cup Champions

Strikeout denotes skew data

 

What exactly am I presenting to you here you might be wondering? In a very basic way, this is what it takes to be a "legitimate" playoff team through the past three VHLM seasons. Due to extreme roster turnovers from season to season, the use of the conference system in the VHLM, and the "tank it for later" mentality of most VHLM general managers, there is a lot of skew data represented in the VHLM.  Inclusion of skew data would cause a general bias and cause the standard deviation to skyrocket based on the huge increase in variance. By neglecting the use of the data taken from illegitimate playoff teams (i.e. the entire European Conference in S39) the data becomes less bias, with less error in consideration of our goal. The goal or aim of this data is to help determine and thereby project what it takes to be a playoff team in the VHLM; in a static world.

 

Drastic changes to the entire league can have a profound effect on the data however. Consider in example if there was a sudden influx of good goaltenders to the VHLM with only a handful of scoring forwards. You would expect that across the league all of the values displayed above would decrease. In consideration of a phenomenon such as that, you have to adhere to the fact that this is HIGHLY theoretical. As we all know, anything can happen when riding the roller coaster known as the VHLM.

 

 

By these numbers a legitimate playoff team in the VHLM will have a Goals For value ranging from 307-393, a Goals Allowed value of 164-260, a Differential value of 95-181, and a Total Points For value of 852-1076.  

 

Now, if I'm going to be brash as hell and use this information determine who is going to make the playoffs this year in the VHLM (which I'm going to do eventually), I need to do some more figuring. To aid in calculations a correlation coefficient (r) between Goals For and Total Points For has been calculated at 0.9906 with a coefficient of determination or r-squared = 0.9906 or 99% (P-Value of <0.00001 at significance level of 0.05). While the data isn't exactly linear, this gives us a good idea that the Goals For variable and Total Points For variable are very well correlated with one another (I can draw this conclusion since r ranges from -1 to 1 with a value of 0 meaning no correlation and values tending closer to 1 indicating a high correlation). That only makes sense since in hockey for every goal scored there is the possibility that three points will be awarded. However, in some instances someone might score a lone goal or a secondary assist may not be awarded. In theory, in a static world, the quantitative relation between Goals For and Total Points For should be 1:3, though I know that isn't the case in the real world or within the VHLM. Since I know there is a high correlation between the two variables I can go ahead and calculate the quantitative or ratio relation between them (should be somewhere near 3). The result is a mean value of 2.75 with a standard deviation of +/-0.0385. So for every goal scored by a legitimate playoff team over the past three seasons in the VHLM 2.75 +/- 0.038 points were awarded. Now that I know this, it will make the backwards calculations much easier once I start doing some predictions.

 

 

Probably the hardest statistic to predict will be Goals Against, since it is so erratic. The skill of the goaltender has a lot to do with it though the collective defensive ability of the other five players on the ice can be just as important. The only way I can figure would be to determine how many shots a goalie would face in a season. There are equally as many variables in determining Total Points For since hockey isn't an individual sport. Since I'm only taking into account the performance of legitimate playoff teams, I may have to go back to the stats and determine the average amount of shots a goaltender on a legitimate playoff team would face, thereby neglecting blocked shots and all that fancy defensive stuff. In consideration there's probably the same margin of error in determining the future Goals Against of a team as there is determining the Total Points For. Therefore, there's never going to be a way to concretely determine any of what follows, though I can still make an educated guess. Consider that statement my disclaimer.

 

I'm looking forward to watching Thomas Duddy's career evolve this season in Ottawa

 

Just by scanning the Shots Against stats over the past three seasons for what I have judged to be legitimate playoff teams, the observed average is somewhere under 2000 Shots Against. However, the actual mean is 1979 +/- 313 Shots Against. That's pretty hard to determine what number we're going to use to do the calculation. So, by removing the outliers that cause significant variance we come up with a much more manageable number. By removing the S39 Wild (2362 SHA), S40 Gladiators (2243 SHA), and S41 Outlaws (2582 SHA) we can trim the Shots Against variable down to 1823 +/- 172. Not by coincidence, these teams finished the lowest in the regular season for their respective years when considering legitimate playoff teams. With that in mind I don't feel bad about discounting those values since eight samples is still fairly significant for this near-impossible task.

 

 

The formula for determining Total Points For and thereby Goals For is far more pedestrian; no median and no standard deviation will be involved here. I will go through the "finalized" rosters of the VHLM teams and figure out this value by analyzing the players on the team and their respective past performances. The hardest part will be compensating for activity, change in team/line skill, line combinations, and for any of the VHLM rookies with no previous data. One thing I've learned from STHS in the past is that although hockey itself isn't an independent game, STHS can make it out to be at times. I've witnessed good players stranded on a poor team relatively keep up with their career pace in scoring time and time again. There are obviously always exceptions, but I personally believe this to be true. I will account a deviation of 10% both for and against based on skill changes within line combinations. Defense has effectively been taken care of with the Shots Against variable, though I have a standard deviation of +/- 172 to play with. With that in mind I will also try to factor in that deviation in the same manor as mentioned above for defensive pairings. CPU players will also have to be taken into account. This portion of the task will be the most monumental but may yield some interesting results.

 

In order to do this I've gone through the rosters of each team as they would appear at this very moment, barring future trades, player improvement, and waiver acquisitions - the following are my predictions for S42.

 

 

NORTH AMERICAN CONFERENCE

 

BRAMPTON BLADES

Season 42 should be the year for the Blades to come into their own in a very weakened North American Conference. With Mikko Koskinen (69G, 70A, 139P), Yannik Zug (25G, 80A, 105P), Valentin Taneyev (36G, 37A, 73P), Jack Harkness (21G, 75A, 96P), and Ivars Klajums (39W, 0.890%, 4.15GAA) all returning one could say that not a lot has changed in Brampton. That would be a very naive statement to make however. While they haven't stepped back in any one particular area, they have managed to improve mightily on both sides of the puck. With the acquisitions of Vincenzo Valentini (21G,12A, 33P), Travis Willcox (43G, 59A, 102P), E'Twaun Delicious (23G, 32A, 55P), and Bronson Faux up front it is rumored that Mikko Koskinen may be on his way out. The only place the Blades didn't manage to improve is between the pipes, though with Gunzerker Salvador patrolling the blueline it should take some of the pressure off of Klajums. I went with a reserved value of 930 points for the Blades this season since I didn't want to look to foolish and biased if they don't achieve what I think they can. That said, I think the true number should be somewhere around 1020 points. The Blades should take the North American Conference easily and contend for the Founder's Cup unless Fong goes crazy.

 

Prediction:
930P//338GF//220GA//118DIFF

OTTAWA LYNX

 

If I had said that the Ottawa Lynx were going to make the playoffs in Season 42 during their 3-win season last year people probably would have laughed at me...maybe they are right now? I don't know. Returning only Carl Sledgehammer (40G, 38A, 78P) and adding a whopping 12 new faces through various means, the Lynx are literally a whole new team. Newcomers like Thomas Duddy, Kewl Running, Pablo Escabar, and enforcer Doug Glatt give the Lynx a lot to look forward to. Duddy looks like a complete stud and should perform admirably on a roster with free agent acquisition Hunter Backenbauer (77G, 62A, 139P) and Legion prospect Benjamin Zeptenbergs (50G, 57A, 107P). Morgan and Escabar give them a defensive pairing, though Holy the Goalie is a big question mark. Whether he can supplant OTT G before the beginning of the season remains to be seen. I figured on a factor of 646 base points for this team but I'm expecting big things out of some of these young players by the end of the season.

 

Prediction:
726P//264GF//253 GA//11DIFF
 

Aside from the Lynx and Blades, the remainder of the North American Conference is looking pretty poor. Minot will probably snatch the final playoff spot on the merits of Jeltz and Wingate alone. Saskatoon and the Rush are going to have rough seasons.


EUROPEAN CONFERENCE

 

BRATISLAVA WATCHMEN

 

With their two biggest names in AIM-11 and Mario De Rossi gone, along with their combined 360 points, the landscape looks a little different in Bratislava heading into Season 42. Lord Karnage has also graduated to the big club and Jax Barnstormer appears to have moved on as well. Returning players to the Watchmen roster include Sam Teibert (59G, 79A, 138P), Kyle Kingma (66G, 55A, 121P), Superbman (21G, 57A, 78P), Sepp Reiter (2G, 22A, 24P), and Raphael Pettyfer (24G, 65A, 89P). I'm not going to lie, that's a far cry from the team they put on the ice last year. With a big hole in net, the loss of their top-two forwards, and best defenseman, Bratislava is going to have to bring in more than Paul Ready and Soren Douffet. Kingma, Pettyfer, and Teibert are more than serviceable as a top offensive unit in the VHLM and should continue to serve the Watchmen well into Season 42. Perhaps the biggest hole remains in net with Soren Douffet, who didn't play at all last season. When he did play in Season 40 he was extremely porous. A shadow of their former selves, though still dangerous, the Watchmen take a stumble but should emerge victorious in the European Conference.

 

Prediction:
934P//339GF//220GA//119DIFF

 

OSLO STORM

 

The Storm are returning Til Baumann (31G, 83A, 114P), Emidas Blair (59G, 55A, 114P), and Aleksander Petrov (30g, 58A, 88P) from their 22-win team in S41. With the loss of some of their top players like Splorg Hansburger and Alex Vaklav come some new faces who are destined to have some very positive effects. Defenseman Patrick Demko, the 4th overall selection in the S42 VHL Entry Draft, along with starting goaltender Coca Cola should shake things up in Oslo following a couple off season trades. While their offensive herd has been thinned, it should allow guys like Mairis Ziedins, Til Baumann, and Emidas Blair to take more of the load and step up. I personally don't see Heffley repeating the 111 point season he just had, but that might turn out to bite me. Demko will be a most-welcomed addition to an almost non-existent back end, while Cola doesn't have very big shoes to fill. OSL G cost the Storm a lot of games in S41 and Cola was brought in to give the team a chance.

 

Prediction:
760P//276GF//225GA//51DIFF

 

Jack Merica should be fun to watch in S42 as well

 

BERN ROYALS

 

The second worst team in the VHLM last season with only 7 wins, the Royals are looking to start turning things around in Season 42. With Atticus Finch (16G, 27A, 43P) and Severin Von Karma (17G, 36A, 53P) returning amongst a sea of fresh faces, things are looking brighter in Bern perhaps. With Ken Anderson, and Golden Jedtsson now added to the back end things should tighten up a little. Without a legitimate goaltender however, I doubt the Royals can make much of a splash. Newly-acquired forwards Josh Merica and Matej Nemecek are expected to takeover the offense moving forward. Merica seems like he has what it takes to make things happen at this level and I'm expecting Bern to win some games this season off of his stick. Bern and Moscow are likely to battle for the final playoff spot in the European Conference but I give Bern the upper hand.

 

Prediction:
630 points//229GF//287GA//-58 DIFF

 

The Outlaws are going to be near the bottom and Moscow could very well beat out Bern to make the playoffs and "surprise" me.

 

 

 

Edited February 17, 2015 by Cowboy

[Baozi 1,638]

So much math hahaha

[DollarAndADream 3,356]

Content: 3/3 - Ridiculous read. Not in a bad way, as it was very interesting. I should give you a 12.

Grammar: 2/2 - Only a couple of things I noticed, which is tremendous for such a big article.

 

"this information determine" - to determine

"in the same manor" - manner?

Appearance: 1/1 - Good!

Overall: 6/6

[Frank 4,819]

Claimed