Ideas for Future Studies

[stlblues5]

Okay so this might be obvious and a useless statistic, but it's just an idea. A players shooting percentage on specific goalies? Like baseball players have a batting average against specific pictures. Again, just a thought, so let me know if this would be useless.

 

[morehockeystats]

Done already!
http://morehockeystats.com/players/face2face/goalie_skater/goals_ratio/filter=

 

[BigRangy]

With Dan Bylsma as the coach of the Sabres, this question is more relevant then ever for me.

I've always wondered if the concept of "over ripening" is a thing about AHL level prospects.

This is something I'd like to see as well. It is (to my knowledge) very, very rare for a player to play 200+ AHL games before becoming an impact NHLer. Just like playing your 100th game in the ECHL means you're probably never going to make it, a players 200th AHL game probably has a pretty similar (though not as drastic) meaning.

 

[stlblues5]

Let me know if this has been done already. I think it would be interesting to see the chance of success based on what pick you get drafted in the first round. So, let's say you get drafted 1st overall, and the success rate from that pick is 75%. Another person gets drafted 16th and their success rate from that pick is 55%. Logically, it would decrease the later you get drafted but I wonder if some interesting data could come out of this. Another thing would be what defines success? Being a full-time NHL starter or backup goalie? Having a certain PPG or GAA?

 

[morehockeystats]

Let me know if this has been done already. I think it would be interesting to see the chance of success based on what pick you get drafted in the first round. So, let's say you get drafted 1st overall, and the success rate from that pick is 75%. Another person gets drafted 16th and their success rate from that pick is 55%. Logically, it would decrease the later you get drafted but I wonder if some interesting data could come out of this. Another thing would be what defines success? Being a full-time NHL starter or backup goalie? Having a certain PPG or GAA?

I sort of considered doing that, but defining a "success" is tough.

 

[stlblues5]

I sort of considered doing that, but defining a "success" is tough.

I think it would be a cool study. I think an easy way to define success would be 100 NHL games played but that doesn't necessarily mean they were a contributor to the team. Another way would be to come up with a "success formula" of sorts. I don't know what numbers or statistics that would entail, but if the formula came out with a number above a certain amount, they would be considered a success.

 

[morehockeystats]

I think it would be a cool study. I think an easy way to define success would be 100 NHL games played but that doesn't necessarily mean they were a contributor to the team. Another way would be to come up with a "success formula" of sorts. I don't know what numbers or statistics that would entail, but if the formula came out with a number above a certain amount, they would be considered a success.

I am working on it right now, first on goaltenders, because their contribution - and I am using exactly this term - is easier to define.

 

[trentmccleary]

Theory: Hockey writers are fans who watch 82 of their home team's games and a maximum of 82 'other' miscellaneous regular season games. That is until tournaments (Olympics, playoffs, etc) dominate the sport and their focus is directed squarely at a handful of star players. The spotlight of these tournaments gives players an advantage when trophy ballots are cast over players who didn't have that spotlight.

Test: Assess whether star players have higher trophy finishes immediately after a 3rd round+ postseason or being a key player on an Olympic medal winning team.

I've compiled a list of the selke votes by year since the lockout in excel (minus 04-05, only found the top 20 vote getters) and have been slowly adding stats per year to see if their might be any sort of statistical correlation between Selke success over the years.

Was wondering if anyone might have some ideas for how I can analyze the data better once it's done (probably will take awhile longer, getting burnt out with excel atm).

I know this post is super old, but if anybody wants to pick up the mantle...
pre-1998... the Selke was a +/- award
post-1998... the Selke was a SH TOI award
post-2008... the Selke was a takeaway award (for a few obvious years anyways)

I sort of considered doing that, but defining a "success" is tough.

I did the top-10 drafted forwards within a decade and chose 3rd highest scoring season to compare. It isn't likely to be an outlier and is essentially the median of their 5 best years. The data collection is very time consuming though.

 

[morehockeystats]

Let me know if this has been done already. I think it would be interesting to see the chance of success based on what pick you get drafted in the first round. So, let's say you get drafted 1st overall, and the success rate from that pick is 75%. Another person gets drafted 16th and their success rate from that pick is 55%. Logically, it would decrease the later you get drafted but I wonder if some interesting data could come out of this. Another thing would be what defines success? Being a full-time NHL starter or backup goalie? Having a certain PPG or GAA?

I did something like that.
First, I calculated the averages and the weighted averages for each draft pick:
http://morehockeystats.com/drafts/pickstats
Then, I compared whether the player picked at that position exceeded, met or failed the expectations:
http://morehockeystats.com/drafts/picksuccess

I know this is a bit crude, but something to work with...
This will be the last update on my site until the new season starts. Gotta rework data collection, optimize the page publishing and improve the performance of the website.

Always open for new suggestions!

 

[morehockeystats]

As I understand, the board does not have an API to post from scripts, does it?

 

[stlblues5]

I once did a study on teams percentage of drafted prospects turning out to be quality players(100 or more nhl games played) and they all turned out to be working 0.3 percent of each other.

 

[Bear of Bad News]

As I understand, the board does not have an API to post from scripts, does it?

Nothing I'm aware of, no.

 

[oilerbear]

I thought someone had looked at scoring by defensemen?

I can post a table of scoring tiers for all players (including d-men), proportional to the number of teams in the league (e.g. 1-30, 31-60, ... 331-360). I would have posted this quite a while ago, but it looks better as a graph (and can't figure out how to post a graph). It uses simple adjusted points, so any fluctuations indicate that a certain tier is doing scoring a larger proportion of points.

That would be me, it is one of the 14+ base hockey theories I created in the last 12 years.
From observations that have stayed with me from the age of 4 (50yrs ago) till now.

1.Offensive Dmen are OFFENSIVE!:
A. almost all offensive Dmen generate even offence at a 9,10,11, 12, 13, 14, NHL Fwd and 1st line AHL (15, 16, 17 NHL) fwd level.
B. By system structure they have to abandon the defence of free path to their own HD area. (what was called ricki's for the first 10-6 years ago.).
It is one of my 4 most critical theories to define cap value of Def play.
Because they occupy forward spce down low they need to be compared to the players space they are occupying!
C. Offensive Dmen yielf high HD Shot rates. Often have very high EVGA rates.

2. Failed offensive play (fwds/Off D) establishes the CA rates.
it is only a measure of the players active in the Offensive Structure of the Team.

3. Dpairs Establish the average save % a goalie performs around.
A. there is a standard rate for HD shots in a 30 shot game.
B. HD Theory values.
Avg HD shot .8250 Save% 17.5% success.
Avg LD shot .9650 Save% 3.5% success
17.5:3.5 = 5:1
C. AVG HD Dpair
10.5 HDSH/60 X .8250 = 8.6625 saves
19.5 HDSH/60 x .9650 = 18.8175 saves
(8.6625sv + 18.8175sv)/30 Sh = .9160 Save%
D. Elite HD Dpair
7.5 HDSH/60 X .825 = 6.1875 saves
22.5 LDSH/60 x .965 = 21.7125 saves
(6.1875sv + 21.7125sv)/30 SH = .9300 Save%
E. Poor HD Dpair
14.0 HDSH/60 X .825 = 11.55 saves
16.0 LDSH/60 X .965 = 15.44 saves
(11.55sv + 15.44sv)/30 Sh = .8997 Save%

 

[oilerbear]

Isn't this what GVT (Goals Versus Threshold) is, or at least attempts to be?



International Sports Management

This is a utter waste of time without the correct reference.

My first 3 critical Base theories:
1. Hockey is like playing chess:
A. it is a series of dynamic moves, that must be tracked individually.
B. 2-10 groups of measure result in multiple levels of permutations
C. Hockey must be analized with an empirical approach.
D. Trying to identify the players best and worst at critical performance skills.

2. Seasonal expected mean and performance variance range +/- based on team, comp, ZS.
A. the first critical thing is defining true ZS. not based on just FO,
but also the affect of coaches decision to send a player on when their team is with or without procession.
B. I generated
8 groups upper & lower 1st, 2nd 3rd, 4th line Team
x 8 groups upper & lower 1st, 2nd 3rd, 4th line comp
x 8 groups of ZS based on Standard deviations
= 512 groups of 3 axis blocks in a 3D performance matrix.
C. By holding 2 Axis constant you can see an area graph of Variable affect on any important stat over that given axis range.
D. It allows you to allows you compare (stack) a series of area graphs over a 2 axis range.
E. It is apparent you need all 3 variables to establish a stats mean to compare a player against.
Not much long after I presented this on Lowetide, Rob Vollman created a player usage chart based on 2 variables.
2 variables makes on performance reference false.
F. My 2 favourite Stats:
-Goal diff (+/-) avg range fro all 512 groups. -29 to +30.
- EVP/60 .25 to 3.05

 

[morehockeystats]

This is a utter waste of time without the correct reference.

My first 3 critical Base theories:
1. Hockey is like playing chess:

No, not really.
Said by a US Chess Master and someone who considers Elo function as the best evaluation method.

 

[OvermanKingGainer]

I would like to see if there is any way to acquire or estimate TOI numbers from the 90s and prior. I see the ATOI for guys like Sakic, Lemieux, Bure and these guys played 25-28 minutes a game and it wasn't even the 80s yet.

 

[Doctor No]

I would like to see if there is any way to acquire or estimate TOI numbers from the 90s and prior. I see the ATOI for guys like Sakic, Lemieux, Bure and these guys played 25-28 minutes a game and it wasn't even the 80s yet.

Iain Fyffe's done some work there - if I get a chance, I'll see if I can find it.

 

[Doctor No]

Found a thread on it:
http://hfboards.mandatory.com/threads/principles-of-estimating-ice-time.1731329/

I recall that Iain discussed this in greater depth in a published work as well.

 

[tinyzombies]

I have a silly idea for filling in missing assists (assuming they are missing) from early NHL days. In advanced stats I think there are percentages for how chances are created (off the rush, in-zone, etc.). I'm guessing then there must be stats for how goals are scored. Can we not apply this method to the different sets of rules that were employed in the early and pre-NHL? I assume most goals were scored off the rush or rebounds in those old carry systems, and even when the forward pass was put in. So, if Nighbor led the league in assists with 13, those are all rush assists because accounts say that he would rush the puck and look to make a play, whereas maybe Morenz would have driven to the net instead. If we bring Nighbor into today's league and put him on a counter-attack team, can we not then fill him in with the other types of assists he would have amassed (inzone, etc.) to fill in missing assist numbers?

 

[Canadiens1958]

I have a silly idea for filling in missing assists (assuming they are missing) from early NHL days. In advanced stats I think there are percentages for how chances are created (off the rush, in-zone, etc.). I'm guessing then there must be stats for how goals are scored. Can we not apply this method to the different sets of rules that were employed in the early and pre-NHL? I assume most goals were scored off the rush or rebounds in those old carry systems, and even when the forward pass was put in. So, if Nighbor led the league in assists with 13, those are all rush assists because accounts say that he would rush the puck and look to make a play, whereas maybe Morenz would have driven to the net instead. If we bring Nighbor into today's league and put him on a counter-attack team, can we not then fill him in with the other types of assists he would have amassed (inzone, etc.) to fill in missing assist numbers?

Rebound assists were not always awarded. Would have to check when the NHL started awarding rebound assists.

 

[olli]

A players points per game in 1 goal games vs 2 goal games vs 3 goal games etc.

 

[trentmccleary]

There isn't really a news section here... Hockey Reference appears to have added their era adjusted stats to their searches in Play Index.

 

[Hasbro]

I would love to see a study on if any correlation exists between playersize and durability.ch

 

[Toews2Bickell]

How do teams assess a fair value of a player?

Is it all comparable analysis (age, ufa/rfa, etc) or do teams try to determine how to translate production into wins (something like war in baseball) and then value the wins with a dollar amount ($/war). As if the player has an intrinsic value and they can use that to determine their offer price.

In theory, a player is worth the war over the duration of their contract. Teams could run sensativity tables to forecast production and different $/war figures to assign a fair value, but I’m not sure teams approach it like this in the NHL.

Can anybody shed some light on this?

Something I’d find interesting is looking at how efficiently GMs assign values to players via trade or free agency. The price to acquire an asset has an implied production attatched to it. So if you pay X for a player in cap space or draft picks it implies Y production, then you can compare a few years later if the GM generated value or lost value for his club given price paid. I don’t think hockey is at this point at scale, but maybe it is. It would be very interesting to see.

 

[krutovsdonut]

i am interested in comparing age vs mileage as a predictor of drop off in performance.

intuitively you would think the north american pro game at any level is harder on the body than junior, ncaa or european play, and that this might shift the bell curve drop off for players who debut later in north american pro leagues.