Jump to content
Sign in to follow this  
Steve

Advanced stats help

Recommended Posts

Hey everyone, I'm a bit old school and think I understand the gist of most of the advanced stats, but wonder what others thoughts are. How will you use this data and are 20 games in even enough of a data sample to make accurate projections.

 

Could someone also do a quick description of each stat and it's relevance? Thanks.


Share this post


Link to post
Share on other sites

After a quick look through at the 3rd line you played in our last match, i'd say (imo) that all players have the same role, and you had 2 centers on the same line. I wouldn't do that personnaly.

Share this post


Link to post
Share on other sites

There’s some good descriptions in the new helps files. There’s one that lists out all the stats.

 

I think they are a great addition to the game, but not sure how much I’ll use them in making decisions. Sample size is way too small so far as they only factor in the last 5 games or so.


*BISCUIT - SEASON 16 - TARNISHED SILVER BISCUIT PLATE CHAMPION*

 

Share this post


Link to post
Share on other sites

Corsi and PDO are great to determine if player(s) are performing badly over an extended period of time, but at the same time it can be deceptive. They're not something I'm looking first to see when something is wrong. Other indicators will tell me that beforehand.


Share this post


Link to post
Share on other sites
TEAM POS GP S% SV% PDO CF CA CF%
1Red Dragons
 
12thW 50 8,3% 91,4% 99,7% 724 551 56,8%  
2Red Tide
 
2ndE 50 9,3% 94,1% 103,4% 801 637 55,7%  
3Prairie Stars
 
1stW 50 8,0% 94,2% 102,1% 812 646 55,7%  
4Stealheads
 
3rdE 50 9,1% 92,4% 101,4% 705 574 55,1%  
5Rage
 
3rdW 50 9,5% 91,5% 101,1% 776 636 55,0%  
6Dragons
 
1stE 50 9,9% 93,3% 103,2% 779 671 53,7%

 

How? How? How? :D


Share this post


Link to post
Share on other sites
On 12/2/2019 at 7:05 AM, Peekaboo said:
TEAM POS GP S% SV% PDO CF CA CF%
1Red Dragons
 
12thW 50 8,3% 91,4% 99,7% 724 551 56,8%  
2Red Tide
 
2ndE 50 9,3% 94,1% 103,4% 801 637 55,7%  
3Prairie Stars
 
1stW 50 8,0% 94,2% 102,1% 812 646 55,7%  
4Stealheads
 
3rdE 50 9,1% 92,4% 101,4% 705 574 55,1%  
5Rage
 
3rdW 50 9,5% 91,5% 101,1% 776 636 55,0%  
6Dragons
 
1stE 50 9,9% 93,3% 103,2% 779 671 53,7%

 

How? How? How? :D

 

Easy. Remember how good the Carolina Hurricanes have been the past ~7 years in the NHL? No? Well that is odd since over that time they were easily one of the best corsi teams over that time frame almost every year. As usually were the LA Kings. The problem for them was one of the worst team shooting percentages combined with save percentages. Proof that controlling the puck does nothing for you if you can't score and/or stop the puck. They are a talented team. They just aren't doing enough of both to win with regularity.

 

By the way my team is at risk of going the route of the LA Kings.


Share this post


Link to post
Share on other sites
11 hours ago, rainsilent said:

 

Easy. Remember how good the Carolina Hurricanes have been the past ~7 years in the NHL? No? Well that is odd since over that time they were easily one of the best corsi teams over that time frame almost every year. As usually were the LA Kings. The problem for them was one of the worst team shooting percentages combined with save percentages. Proof that controlling the puck does nothing for you if you can't score and/or stop the puck. They are a talented team. They just aren't doing enough of both to win with regularity.

 

By the way my team is at risk of going the route of the LA Kings.

Yeah, except that team that is 1st in Corsi shouldn't be good at controlling a puck, and its no near others on that list in quality of players. Wich is a main reason I posted this.


Share this post


Link to post
Share on other sites
13 hours ago, Peekaboo said:

Yeah, except that team that is 1st in Corsi shouldn't be good at controlling a puck, and its no near others on that list in quality of players. Wich is a main reason I posted this.

Not necessarily true. Remember that corsi is strictly counting shots for and shots against. It actually has nothing to do with puck control directly. It is a statistical assumption that the team taking more shots has the puck more.

 

As for how this applies here? The same as it applies to my team. My team isn't near as talented as yours or the many others that sit at the top of the standings. Yet I have good corsi numbers too. That team is another example, just like mine is, of getting the right players that they need for their system but with two flaws sinking that team. Lack of goaltending and lack of scoring. They have a talented enough team to fight for a playoff spot if they had a good goalie. I have yet to look at the games that team has played so I can't say if they are true puck possession monsters but they have a good enough group talent wise that they can be a solid puck possession team.


Share this post


Link to post
Share on other sites

Yeah, but to have shots you have to have puck possesion. So basically, Corsi would determine how well you play up until puck goes in the goal.

Its still something off here, looking at roster. Or very big statistical error, or there is much more in the variables then just ratings (traits?), maybe too much. We still need tools to better understand when something like this happens.


Share this post


Link to post
Share on other sites
8 hours ago, Peekaboo said:

Its still something off here, looking at roster. Or very big statistical error, or there is much more in the variables then just ratings (traits?), maybe too much. We still need tools to better understand when something like this happens.

I don't have a strong opinion on it, but here's why I wouldn't necessarily agree with this.

1) High Corsi doesn't necessarily mean that team is dominating games.  It could be as simple as, the team goes down early and plays Offensive when trailing.  This could lead to more shot attempts, but not necessarily quality chances.  Or the team could play Defensive and just not allow many shots, but not get many quality scoring chances.  Either way, in this specific case, the Red Dragons are facing relegation so the high Corsi isn't helping.

2) The game has an element of randomness to it, much like the sport itself.  That's a good thing.  Sometimes certain outcomes just don't make sense, other than the element of randomness.  Overall, the game is directionally accurate and the "better" teams tend to place higher in the standings/have better stats.

3) If we had the tools to determine exactly why stuff like this happened, then every manager would use the same tactics, build the same roster, play the same focus.  That would get pretty boring.  I like the fact that every manager is given a clean slate and able to customize their team to their liking without there being a "right" or "wrong" way to do it.

4) Personally, I think statistical outliers add fun to the game.  It's fun when an inferior team is high in the standings - or a completely stacked team isn't as good as everyone thought.  Although I would admit that this isn't necessarily fair... but sometimes sports aren't fair (or life for that matter).

5) If you take a close look at this specific team, they have a very unique set up.  An Offensive Defenseman playing 1st line wing.  A Two Way D playing 4th line wing.  That would definitely screw with the simulation.  But I think we all agree the game needs more micromanagement, so I see that as a good thing.  Figure out unique ways to set your lineup.

If you compare their defense to mine, they are actually better.

Red Dragons - 90 89 88 88 88 85 (528)

Stealheads - 92 88 88 87 86 81 (522)


*BISCUIT - SEASON 16 - TARNISHED SILVER BISCUIT PLATE CHAMPION*

 

Share this post


Link to post
Share on other sites
13 hours ago, Peekaboo said:

Yeah, but to have shots you have to have puck possesion. So basically, Corsi would determine how well you play up until puck goes in the goal.

 

Yes, you have to have puck possession to get a shot. However that isn't really what corsi is telling you. Going back to Carolina as for why. Carolina, since they became "good" statistically corsi wise, has had the mentality of shoot the puck from anywhere, anytime. They weren't dominating possession of the puck so much as they were throwing a lot of bad shots at the net. The net result of that was averaging ~35 shots a game for with ~27 against. This served to heavily bloat the corsi and offensive zone face-off numbers of the team even though they were actually mid pack, at best, puck possession wise. Thus Carolina was gaming the statistical system to look good when they really weren't that good at all. This is why I think the actual puck possession % is the most important team tracking stat. It will actually tell you if you are getting dominated, if the game is close or if you happened to be playing "Carolina" that game.


Share this post


Link to post
Share on other sites
On 12/5/2019 at 11:40 AM, Paul T said:

4) Personally, I think statistical outliers add fun to the game.  It's fun when an inferior team is high in the standings - or a completely stacked team isn't as good as everyone thought.  Although I would admit that this isn't necessarily fair... but sometimes sports aren't fair (or life for that matter).

As I say this, my team continues to lose to teams at the bottom of the standings and the game winner is scored by an 81 with 67 Shooting rating.  LOL

I'm going to try to tank on purpose and see what happens.


*BISCUIT - SEASON 16 - TARNISHED SILVER BISCUIT PLATE CHAMPION*

 

Share this post


Link to post
Share on other sites

For me, game needs some tweaking. The formulas. There were cases in both buzzer, attack and cage where players with serious lack of certain skills dominated chats and are way out of statistical oddity.

Also, there is something wrong with that underestmation formula, I think it does impact the game too much. Yeah, I get that there is trait part of the game, that has impact on players perfomances, but its too shady.

I am really trying to understand why is my winner instinct in blue, while clearly some other teams that are in green and if you ask me they shouldn't be there. Whatever you say, I really don't find it fun trying to guess formula to get good winner instinct.

Also, I do think that having too many determined and herioc players being contraproductive regarding winner instinct is a serious nonsese. Green traits should be best, while red should be bad. Simple as that. It would give us all room to balance our squads and trades. Like for example, trading 90 guy with bad traits to 86 rated guy with green traits, for the sake of the squad.

Right now I have no clue what I am to do about my winner instinct. Its just a guess game and its stupid.


Share this post


Link to post
Share on other sites

3) If we had the tools to determine exactly why stuff like this happened, then every manager would use the same tactics, build the same roster, play the same focus.  That would get pretty boring.  I like the fact that every manager is given a clean slate and able to customize their team to their liking without there being a "right" or "wrong" way to do it

 

I have to totally disagree with this and its really not the point of what I was talking about. You would use tactics depending on your preference, your squad, ratings, traits, etc. You missed my point completely - like I am asking for a "win" button.

I want to know how to calculate winner instinct correctly.

I want to know more precise formulas so I can choose my stlye/focus/roles better. Actually, I don't need formulas, but even more detailed explanation of roles and line setups. For example, in this new guide, where can I see how different twoway or scoring winger play in twoway line? For example. The fact that I can't read from the player stats (unless I am tracking my own excel sheet), I have no clue which role would produce more shots, would it have less its, etc etc etc. You get what I am saying? :)

etc.

It has nothing to do with everyone playing the same. If I lose to 82 rated squad, I would like better tools to determine if there is just "luck" part of the game, and I am fine with that, or there was something else? Captains? Underestimation? Tactics? So for, there is assistant, but from his report, you can conclude anything and is not reliable.


Share this post


Link to post
Share on other sites
4 hours ago, Peekaboo said:

I have to totally disagree with this and its really not the point of what I was talking about. You would use tactics depending on your preference, your squad, ratings, traits, etc. You missed my point completely - like I am asking for a "win" button.

I wasn't accusing you of anything.  I was simply providing my own opinion.  By making comments like " I am really trying to understand why is my winner instinct in blue, while clearly some other teams that are in green and if you ask me they shouldn't be there. Whatever you say, I really don't find it fun trying to guess formula to get good winner instinct." leads people to believe you aren't really enjoying the aspects of the game that are meant to be fun and that winning is your most important priority.  And if winning is the most important part of the game, that's fine too.  But most others don't feel that way.  Many people enjoy the trial and error aspect of the game.

 

4 hours ago, Peekaboo said:

For example, in this new guide, where can I see how different twoway or scoring winger play in twoway line?

 

4 hours ago, Peekaboo said:

 I have no clue which role would produce more shots, would it have less its, etc etc etc. You get what I am saying? :)

No, because I enjoy the trial and error aspect of the game.  I think that's what makes good managers.  If I wanted to know which roles produced more shots, I would either look at the existing data/stats or set certain lineups to test it out.  Look at player stats and see which players are taking more shots.  See which role combinations are getting more shots in comparison to their ice time.  See if certain line focuses are getting more shots.  See which position C vs. F are getting more shots.  Even go beyond my own team to see how the teams that are getting the highest number of shots are set up and which players are getting those shots.  It's all stuff I've enjoyed figuring out over the past 10 slow seasons.

4 hours ago, Peekaboo said:

It has nothing to do with everyone playing the same. If I lose to 82 rated squad, I would like better tools to determine if there is just "luck" part of the game, and I am fine with that, or there was something else?

   \/

On 12/5/2019 at 11:40 AM, Paul T said:

2) The game has an element of randomness to it, much like the sport itself.  That's a good thing.  Sometimes certain outcomes just don't make sense, other than the element of randomness.  Overall, the game is directionally accurate and the "better" teams tend to place higher in the standings/have better stats.

 

 


*BISCUIT - SEASON 16 - TARNISHED SILVER BISCUIT PLATE CHAMPION*

 

Share this post


Link to post
Share on other sites

I just try to understand, trial and error is fine if you can draw conclusions from it. Right now we dont have tools for that really.

Eh, where did I say you accused me? Plz don't add drama where its not. I disagreed with that part of your post.

Furthermore, you are implying that I am not enjoying the game, not sure why would you do that, besides trying to create argument.

I do stay by my opinion, having more resources, better analysis would help alot.


Share this post


Link to post
Share on other sites
13 hours ago, Peekaboo said:

Also, I do think that having too many determined and herioc players being contraproductive regarding winner instinct is a serious nonsese. Green traits should be best, while red should be bad. Simple as that.

 

It isn't that simple though. Having too many determined and heroic players can be counterproductive in real life. Look at Johnny Gaudreau. He is definitely a heroic type. However if things go bad for him or the team he has a very bad habit of trying too hard to do things on his own which ultimately ends up being counter productive in helping him and the team have success.


Share this post


Link to post
Share on other sites
17 hours ago, Peekaboo said:

Furthermore, you are implying that I am not enjoying the game, not sure why would you do that, besides trying to create argument.

I didn't mean to imply that.  Sometimes when I start writing my posts, I continue to expand on them based on where I am with my own thoughts.  I try to give others a glimpse of the logic behind my thinking as opposed to trying to attack what the other person is saying.  I've been doing that since the beginning of time... check some of my posts from 2017 😁

So, sorry if it came across that way.  I enjoy having analytical discussions about the game, but don't enjoy when it becomes argumentative.  There's plenty of room for that in other aspects of my life... like work... and marriage 😝  I think at this point we are just thinking about the game differently.


*BISCUIT - SEASON 16 - TARNISHED SILVER BISCUIT PLATE CHAMPION*

 

Share this post


Link to post
Share on other sites

/\

I will add to the above, which I have stated before, that I STRONGLY feel there is an element of randomness to this game.  As there should be!  I think some of the information that you are looking to obtain just isn't as simple as it seems (like rainsilent said), and that you're going to end up not enjoying the game as much as you should.  So I don't mean to imply that you aren't enjoying the game, but I think that your current mindset will lead you to not enjoy the game.

The randomness that I talk about could be anything from:

  - puck luck (that crazy biscuit takes some weird bounces sometimes).

  - players just having off games (they do in real life)... Ovechkin had a game a few nights ago where he scored 3G + 1A, then followed that up with 3 games of 0+0 and a -1.

  - the fact that the player ratings aren't as far apart as most people view them.  A sub-90 rating doesn't mean the player sucks at that specific skill.  Even an 85 rating in certain skills is considered a natural for certain roles.

  - there's probably 5-10 other things I could list here, but I'm losing my train of thought.

Sometimes inferior teams beat more dominant teams.  Sometimes teams with low winner instinct win close games.  By the way, Winner Instict is something that I have NEVER been able to figure out.  I've had a Heroic team, at most, for a couple of games, before it slid back down the bar.  I'm going to be pumped when I finally figure it out, but I hope it's by chance... I don't want someone to give me the magical formula.

And after seeing your second comment about you implying that you aren't enjoying the game, here's some reasons why you SHOULD enjoy the game

  - you are current GHL champ

  - you are currently in 1st place

  - you have a great team now and into the future (this doesn't mean that sometimes you won't get beat by an SHL team here and there... it happens).

  - despite it's minor flaws, the game is still REALLY F*CKING COOL 😜

You are a good manager.  I like the league being filled with good managers.  Simple as that.  I was bummed out when Eric left (Fairbanks Union)... I was bummed out when Matt H left (Brass Castle Berzerkers).  It was clear to me that they both left because their teams weren't doing as good as they wanted them to.  But in the end, nobody really cares.    To me that is just insane.  The object of the game is to win.  Sure.  I want to win.  But that's not why I enjoy the game.  It's the escape from real life that I enjoy.  Just as much as winning, I also enjoy scouting the draft, developing players, trying players in new roles and positions, chatting with Wendel and Naekfor about various things.  There are times I get BS after a loss to an inferior team, but then I step back and have a "who gives a sh*t" moment.  Failure isn't losing a game or having a losing record or getting relegated.  Failure is getting upset in real life about fake game failure and not being able to just enjoy it for what it is (that isn't directed at anyone in particular... it's directed at EVERYONE!).

See - I went off on another tangent again.  Take from it what you will. 

 

 

 


*BISCUIT - SEASON 16 - TARNISHED SILVER BISCUIT PLATE CHAMPION*

 

Share this post


Link to post
Share on other sites

I agree on that we need more tools and more transparency to determine what works well and what does not. The advanced stats are great, but they only go so far. I understand that RNG is always a part of the game, and it should be, a random chance of luck is a large portion of a real hockey game as well. Sometimes, and quite often actually, it feels like stuff is happening without clear understanding of what exactly is wrong, and that is not something I like in a manager game. It leaves me in the dark concerning pretty much everything. Great example is the current season. I started strong and won the first five games, then I dove deep and borderlined a crisis team, and then, out of nowhere, everything clicked and I began to perform well.

 

Why? I didn't do anything drastic.

 

I had a balanced team to begin with. My winner instinct sucked in the beginning and it keeps on sucking to this day, but it doesn't seem to have any effect to my team's performance. Suddenly everything just seems to work like a dream. My 1st and 2nd goalies are both doing well now, but they both had a streak of really bad games... why? A single game is easy to swallow. Okay.. shit happens in a real hockey game too, but over a longer period of time good players can turn it around and excel. In GPHM it seems to be quite random.... why?

 

I just want to know more about why things happen the way they happen,  so I can adjust accordingly.. you know.. like a hockey team manager would. :)


Share this post


Link to post
Share on other sites

Create a GM profile or sign in to comment

You need to be a member in order to leave a comment

Create a GM profile

Sign up for a GM profile in our community. It's free & easy!

Create a GM profile

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this  

  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...