+- Simulation Hockey League (https://simulationhockey.com)
+-- Forum: League Media (https://simulationhockey.com/forumdisplay.php?fid=610)
+--- Forum: SHL Media (https://simulationhockey.com/forumdisplay.php?fid=46)
+---- Forum: Graded Articles (https://simulationhockey.com/forumdisplay.php?fid=545)
+---- Thread: Someone read this, its actually not a presser (/showthread.php?tid=73752)
Pages:
1
2
- Slowpoke - 03-22-2017
Code:
Ready for grading. Word count for write up and spreadsheet 2790 wordsSo, I did a thing again. Time for my semi-seasonal media shitass. Last time I was here I took a look at the fantasy side of the league and made some minor observations. This time I will be looking at the league makeup, try to create my own statistic, and for the most part be wrong about everything.
The Stuff <---Click There, Youre going to want to have this and the link open.
In the linked spread sheet, I have done a few things. I have broken each team up, looked at the number of players that have chosen each type of player (playmaker, sniper, enforcer, etc) and listed the teams total TPE and average TPE per player. There will be some minor issues with that since some teams carry weird composition (Texas carrying 3 goalies). Teams on average had 20 players on their roster with some fluctuations +/- 1 or 2. At a quick glance it is easy to see why some teams have been successful. For example, New England having almost 10000 TPE more in their system than Buffalo. It is pretty obvious on paper that New England and West Kendall should be running the East, which they are for the most part, being #1 and 2 respectively. The West is a little tighter with all teams being within a smaller spread of TPE.
This is where I started to think, and ended up making up a random stat (which I probably fucked up on). I called it Efficiency Score (ES) and it’s a measure of TPE to goals scored. In other words, how much TPE did a team need to spend on each goal that they scored this season. I started to view TPE as a sort of currency that teams could use to purchase players (more on this later) and goals. In looking at the sheet, the lower the ES, the better the team performed in this area, doing more with less. In the West, once ES was determined, it is very obvious at how efficient Edmonton is. Edmonton scored way more goals (165) then their teams TPE (14258) would predict. In the East, West Kendall is on the other end of the spectrum. Just by looking at their team TPE (18218) you would expect the team to be producing at a higher rate than their 124 goals.
As a whole, and without looking at seasons worth of data, I think a perfect team would be performing at somewhere around 85-100 ES. As a Renegade, I know the struggle this season was trying to figure out why the team was underperforming. This can be seen in the team’s ES (128.16) and it’s easy to see that for some reason the team was not scoring enough in comparison to where their TPE levels would indicate. In contrast, Winnipeg scored more, with less TPE, yet they were not performing at an optimal level either. This is where team composition came into play.
I came into the research with certain expectations and assumptions based on my time in the league. I expected a low number of enforcers and power forwards. With injuries off, the need for physical play does not do anything except provide another means of defense. A case study on my player could be used as his defense is his weakness yet he has not had a negative +/- season, although he has been on a good Kelowna squad for the entire time (yes, plug for Kelowna because they kick ass, fuck you Montreal). To go back to looking at Texas, what is the difference in their team comp compared to a similar team, Los Angeles for example. Los Angeles’ forward core is a pretty good split of player types, which one could assume would mean a diverse spread of weaknesses and strengths. Looking at Texas, they are very heavy in snipers as well as playmakers. From a GM perspective that would make sense. Snipers shoot, playmakers get them the puck. However, it isn’t the forwards that caught my eye, it is the lack of an Offensive Defenseman (OD).
The league composition shows that roughly 29% of all defenseman are OD, Texas does not have a single one on their roster. Now, I am no expert but I believe that this might be the weakness in Texas. Now this is where things get all fuckered up. Since goals against (GA) was not figured into ES it’s hard to determine the role that Defensive Defenseman take. Looking at Hamilton, a team stacked with OD, without applying their GA to the ES formula it’s hard to predict what the impact of DDs are.
Now, on to the interesting part. As a member of the site who is not all that active I spend most of my time observing. Some of the major things I have noticed are a concern about site activity, SMJHL TPE cap, and money generation and use. After looking at the role of TPE on team, which it is obvious that more TPE means more success (I’m looking at you, underperforming Wolfpack (imagine if they had the same ES as Edmonton, fuck)) I believe that a change in how the league handles the salary cap may be needed. Changing a team’s salary cap to be based on TPE vice money would drastically change the landscape. By utilizing something akin to league average TPE across all teams +/- 10% as the cap it would force teams such as New England to strategically handle drafting, call ups, and free agency. How does this affect those concerns I identified earlier, site activity for instance? Well, imagine if New England needed to break up due to cap constraints. Now you have them trading off players like Dieter Dominique or Mia Landvik. The free agency market would be alive with active members of the site. It would have the potential to also reenergize the site as new rivalries would develop (familiarity breeds contempt after all) as veterans either get traded or dropped all together.
Since it’s double monies week for media, I figured I would expand upon some of the work I had done. After talking to a few people briefly and realizing that my initial attempt at measuring a team's efficiency was not quite the best approach, I decided to go a little deeper. Simply looking at team TPE/GF did not give the entire picture. I revised the name of that stat to be Offensive Efficiency Score (OES). I needed to look at a myriad of things. GA, GF, hits, assists, and a few other things. So I applied a weighted formula to each statistic based on how important I felt that they were to the success of the team. That's where True Efficiency Score (TES) came about. So let's go over that first. If you look at the 3rd sheet, True Efficiency, the higher the number, the better the team.
<div align="center">Goals: 0.75
This is a simple one. Goals win games, and thus are weighted the most.
Assists: 0.65
Assists are a measure of how well the team moves the puck. There will always be outliers as sometimes a goal doesn't have a secondary, or a primary assist to it. I weighted this so heavily since defenseman have a lower chance to score goals, but I did not want to diminish their contributions to the team.
SOG: 0.075
This is a measure of how many shots actually reach the goalie. Since there are low quality and high quality chances this is a middle importance statistic.
Blocked Shots: 0.1
Since the goal was to get a grasp on how effective the team was, this is a way to tell how good the defense is when in their own end. I weighted this more than shots since it is slightly more important to block a shot then it is to take a shot.
Hits: 0.15
I may be slightly bias, but I view this as one of the few ways to directly take the puck from the opposition. Without the ability to accurately measure intercepted passes, hits will bear the brunt of the turnovers, even when not all hits result in a turnover. It doesn't mean I don't favor them over other things.
FOW/FOL: 0.01 each
Measuring a team's ability to win or lose faceoffs is a direct indication of how well they control the game. A good center can take the pressure off of the PK on the PP, as well as set the team up for chances on the offensive side. These are weighted low due to the raw number of faceoffs a team takes each season.
GA: 0.2
Goals against is the opposite of goals for, seems simple. I didn't give them an equal weight to goals since I did not bring goalies into the situation, I need to leave something else to do next time I need money, and a bad goalie could skew efficiency.
Shots Against: 0.075
These are measured the same as shots for pretty much the same reason.</div>
So, with that done I think it helps paint a better picture of the league. If you look at the 3rd sheet I think it is pretty obvious that New England is the team to model in a way, with the Chiefs on the other end of the spectrum. Some of the differences I note with this being done is how off from some of my earlier statements were, Texas may not be one or two moves away and may need to do a little more and take a deeper look at their lines and composition.With everything I have done, I do wholeheartedly believe that Winnipeg will be a very tough team in the next 1- 2 seasons as everything predicts they should be better than they have been. Now with all that said, I can submit this.
- fgh - 03-22-2017
Fucking GOAT, give this man all the money.
- CBear - 03-22-2017
Awesome stuff man
Pretty interesting to see this side of things
Interesting to see how efficient the east teams are with their relative stats but have the 'easier' conference. Also, interesting to see how big of a gap there is in total TPE count with most east vs west teams and how much more developed the west seems to be.
good work dude
and thanks for posting actual interesting things on here.
- Possum Queen - 03-22-2017
I think goal differential might be a better indicator than goals alone. Not sure how it affects the efficiency numbers, but a team can score a lot and still do poorly if they're weak defensively, and a team can dominate while scoring fewer goals if they're shutting their opponents down. In the end, after all, the way you win isn't by scoring goals - it's by scoring more goals than the other team.
Edit: I hate that I always do this. The article was fantastic, and I really like the way you broke things down. Stuff like this is what the league needs, and I didn't mean to immediately jump in with "this could have been done better" and make it seem like I was tearing it down.
- bilbo - 03-22-2017
I'll have to take a better look at this later, but proud of u. Looks very fancy and colorful. Six french toasts for u.
- Slowpoke - 03-22-2017
Quote:Originally posted by Tanner@Mar 22 2017, 08:24 PM
I think goal differential might be a better indicator than goals alone. Not sure how it affects the efficiency numbers, but a team can score a lot and still do poorly if they're weak defensively, and a team can dominate while scoring fewer goals if they're shutting their opponents down. In the end, after all, the way you win isn't by scoring goals - it's by scoring more goals than the other team.
Edit: I hate that I always do this. The article was fantastic, and I really like the way you broke things down. Stuff like this is what the league needs, and I didn't mean to immediately jump in with "this could have been done better" and make it seem like I was tearing it down.
I thought of that as well. My concern was if the team is a negative I would need to do some higher level math and may require some stuff I dont have. I would love to have shots for and against, corsi stuff, but without digging thru everything Im not sure how to do what I want.
I also attempted using the league average TPE to put everyone on the same level to easily see which teams are better but the math got a little wonky and it wasnt giving me new info. The rankings were all the same and it just gave me what I already knew.
- Slowpoke - 03-22-2017
Quote:Originally posted by bilbo@Mar 22 2017, 08:32 PM
I'll have to take a better look at this later, but proud of u. Looks very fancy and colorful. Six french toasts for u.
I mentioned trading you, hold off on that french toast until you read it.
- EuroCap - 03-22-2017
TLDR
- Jenny - 03-22-2017
sick analysis and cool idea re: making the cap based on TPE, that would definitely be interesting if nothing else
- Slowpoke - 03-22-2017
Quote:Originally posted by EuroCap@Mar 22 2017, 08:35 PM
TLDR
MY MAN!!! Just what I like to hear.
- Possum Queen - 03-22-2017
Quote:Originally posted by Slowpoke@Mar 22 2017, 08:33 PMThis actually reminds me. Give me a moment, I'll post this season's GVT spreadsheet.
I thought of that as well. My concern was if the team is a negative I would need to do some higher level math and may require some stuff I dont have. I would love to have shots for and against, corsi stuff, but without digging thru everything Im not sure how to do what I want.
I also attempted using the league average TPE to put everyone on the same level to easily see which teams are better but the math got a little wonky and it wasnt giving me new info. The rankings were all the same and it just gave me what I already knew.
- Slowpoke - 03-22-2017
Quote:Originally posted by CBear@Mar 22 2017, 08:10 PM
Awesome stuff man
Pretty interesting to see this side of things
Interesting to see how efficient the east teams are with their relative stats but have the 'easier' conference. Also, interesting to see how big of a gap there is in total TPE count with most east vs west teams and how much more developed the west seems to be.
good work dude
and thanks for posting actual interesting things on here.
I think the east is "easier" just because of the gap between top and bottom. NE and WKP have that shit on lockdown, where the West is literally the wild west. I mean Texas and Winnipeg as 1 or 2 moves from competing, if not being on top.
- CBear - 03-22-2017
Quote:Originally posted by Slowpoke@Mar 22 2017, 06:43 PMabsolutely. After the the wheeling and dealing the Jets did at the trade deadline ( i miss you, come back!!!!!!!
I think the east is "easier" just because of the gap between top and bottom. NE and WKP have that shit on lockdown, where the West is literally the wild west. I mean Texas and Winnipeg as 1 or 2 moves from competing, if not being on top.
they were one of the hottest teams except for those pesky dragons and blizzard. And in the end, possibly 2 games short of contending.Makes sense about the east. With those 2 teams dominating..the rest of them were pretty close in the standings and playoff race.
- bilbo - 03-22-2017
Quote:Originally posted by Slowpoke@Mar 22 2017, 08:34 PM
I mentioned trading you, hold off on that french toast until you read it.
OHHHHH BOYYYYYY. HOW COULD YOU PROPOSE SUCH A THING, SP?
You know what? You still get your french toast because I'm going to say THANK YOU for mentioning Deets and Mia. Any team would be lucky ducks to get one of us. Makes a team 100% better, can confirm. But I must note that Deets had a killer reg season and just forgot to care in the playoffs rip
- Slowpoke - 03-26-2017
This is not ready, because apparently math is hard.