HFBoards Statistical Analysis, Need Your Help!
 Register FAQ Members List Search Today's Posts Mark Forums Read
 Mobile Hockey's Future Become a Sponsor Site Rules Support Forum vBookie Page 2
 Notices New York Rangers Prospects:    Brady Skjei, D   » Pavel Buchnevich, LW   » Anthony Duclair, LW   » J.T. Miller, C/RW   » Brandon Halverson, G   » Kevin Hayes, C/RW   » Jesper Fast, RW   » Danny Kristo, RW   » Dylan McIlrath, D   » Conor Allen, D   »

# Statistical Analysis, Need Your Help!

 02-13-2010, 09:58 PM #1 CaptainCallahan*   Join Date: Nov 2008 Posts: 239 vCash: 500 Statistical Analysis, Need Your Help! I am working on something HUGE, and I need your help. You guys will be extremely pleased with the finished product, but I won't reveal everything until I am done. Can anyone help me gather data on the following: 3rd Period Goals - Who has them and how many? Successful Penalty Kills - Total Penalty Killing Shifts Per Player minus Goals Allowed While Shorthanded (Example: Lets say Ryan Kesler has 117 penalty kill shifts, and the power play scored 10 times with him out there, he would have 107 successful penalty kills). Help with this stat would be incredible!
 02-13-2010, 10:33 PM #2 ogie Registered User     Join Date: Mar 2007 Location: Staten Island Country: Posts: 1,271 vCash: 500 behold math geek is here to save the day. the math for all this is rather simple gathering the data well that is another story. most box scores will have players ice times including PP and PK time. for example here is the box score from last nights game. http://scores.espn.go.com/nhl/boxscore?gameId=300212016 so if you go through every game you can add up how much PK ice time a player has BUT the box score will not tell you if the player was on ice when a PP goal was scored. I guess you could reference +/- but that is not accurate. So good luck with that ...if you need any help with the math i will gladly assist ...if you want help with gathering data...sorry.
 02-13-2010, 10:36 PM #3 allstar3970 Registered User     Join Date: Sep 2008 Posts: 1,227 vCash: 500 I'd try searching around NHL.com. they have a lot of advanced stats in there. interested to see why you're going after here. not sure what knowing total succesful kills tells us, I would think kill % or GA per min on the kill would tell more. look forward to see what you come up with when you can gather some data, keep us posted on your progress.
 02-13-2010, 11:14 PM #4 CaptainCallahan*   Join Date: Nov 2008 Posts: 239 vCash: 500 Thanks guys, yea I don't need help with the math, just data gathering. I will update on my progress as time goes by.
 02-13-2010, 11:27 PM #5 Inferno HFB Partner     Join Date: Nov 2005 Location: Atlanta, GA Country: Posts: 20,259 vCash: 500 if only i had the time....this kinda stuff is right up my alley.
02-14-2010, 08:49 AM
#6
abev

Join Date: Sep 2005
Location: NY
Country:
Posts: 3,591
vCash: 500
Quote:
 Originally Posted by CaptainCallahan Thanks guys, yea I don't need help with the math, just data gathering. I will update on my progress as time goes by.
Learn REGEX and all data on the internet will belong to you. I can get you started:

3rd\s+Period.*?img.*?">(.*?)\s+\W+.*?

02-14-2010, 09:39 AM
#7
CaptainCallahan*

Join Date: Nov 2008
Posts: 239
vCash: 500
Quote:
 Originally Posted by abev Learn REGEX and all data on the internet will belong to you. I can get you started: 3rd\s+Period.*?img.*?">(.*?)\s+\W+.*?
Sounds interesting, can you explain in detail?

Also, I was able to find the 3rd period goals, albeit somewhat inefficiently. Now if I can just find my PK stats, that would be revolutionary.

And for the record, I am working on the DEFINITIVE statistical/salary breakdown for the Rangers, eventually branching out to the rest of the league. I LIVE for statistics, and I found my calling in life after showing my research to date to my economics professor. My dream is to work for the NHL's statistical data department, and who knows maybe even the Elias Sports Bureau.

My fellow Rangers devotees are gonna be very impressed with what I'm doing and I can't wait to show everyone the finished goods.

02-14-2010, 09:57 AM
#8
abev

Join Date: Sep 2005
Location: NY
Country:
Posts: 3,591
vCash: 500
Quote:
 Originally Posted by CaptainCallahan Sounds interesting, can you explain in detail?
Well, not really. It would take quite a long time time to explain the nuances of regular expressions. I suggest you start here.

Basically what I do is create web pages/services in asp.net (visual studio), parse a web page with regular expression (regex), then dump the data into a sql database where I can do what I want with the data.

It's quite a process but once you have it down you could, for example, have every possible stat from an NHL season in under an hour. Keep in mind there might be 20 hours of prep plus the learning curve.

The reward is that you have great statistical power and if you are truly committed to gathering and analyzing data, it will be well worth it.

__________________
Fantasy sports betting at FakePuppy - Pick moneylines, over/unders and spreads. Totally free.

02-14-2010, 10:40 AM
#9
CaptainCallahan*

Join Date: Nov 2008
Posts: 239
vCash: 500
Quote:
 Originally Posted by abev Well, not really. It would take quite a long time time to explain the nuances of regular expressions. I suggest you start here. Basically what I do is create web pages/services in asp.net (visual studio), parse a web page with regular expression (regex), then dump the data into a sql database where I can do what I want with the data. It's quite a process but once you have it down you could, for example, have every possible stat from an NHL season in under an hour. Keep in mind there might be 20 hours of prep plus the learning curve. The reward is that you have great statistical power and if you are truly committed to gathering and analyzing data, it will be well worth it.
Thank you so much!

02-14-2010, 12:58 PM
#10
DutchShamrock
Registered User

Join Date: Nov 2005
Location: New Jersey
Country:
Posts: 5,001
vCash: 500
Quote:
 Originally Posted by Inferno272 if only i had the time....this kinda stuff is right up my alley.
Me too. I wanted to do my senior thesis on the correlation between spending money and winning games in the NHL. Then I wanted to do it for fun when I went another route.

02-17-2010, 08:34 AM
#11
CaptainCallahan*

Join Date: Nov 2008
Posts: 239
vCash: 500
Quote:
 Originally Posted by abev Well, not really. It would take quite a long time time to explain the nuances of regular expressions. I suggest you start here. Basically what I do is create web pages/services in asp.net (visual studio), parse a web page with regular expression (regex), then dump the data into a sql database where I can do what I want with the data. It's quite a process but once you have it down you could, for example, have every possible stat from an NHL season in under an hour. Keep in mind there might be 20 hours of prep plus the learning curve. The reward is that you have great statistical power and if you are truly committed to gathering and analyzing data, it will be well worth it.
Just an update, this is extremely painstaking but I am going through every single play by play chart for all 62 games we have played, and tallying up all the players on ice for a successful penalty kill, or power play goals against. You guys will be shocked by my final analysis.

Are you saying learning REGEX would save me this incredibly tedious process?

02-17-2010, 09:15 AM
#12
abev

Join Date: Sep 2005
Location: NY
Country:
Posts: 3,591
vCash: 500
Quote:
 Originally Posted by CaptainCallahan Just an update, this is extremely painstaking but I am going through every single play by play chart for all 62 games we have played, and tallying up all the players on ice for a successful penalty kill, or power play goals against. You guys will be shocked by my final analysis. Are you saying learning REGEX would save me this incredibly tedious process?
The short answer is yes. I sent more info in a PM. I have to warn you, it's not something you are going to learn overnight

 02-17-2010, 11:27 AM #13 Fletch Registered User   Join Date: Feb 2002 Location: Brooklyn Posts: 21,469 vCash: 500 nhl.com used to have advanced box scored that detailed who was on the ice for each goal. Now it's under game summary.
 02-17-2010, 12:07 PM #14 Blueshirt Special HFBoards Sponsor     Join Date: Feb 2007 Location: NJ Country: Posts: 3,208 vCash: 500 Here's one for you CaptainCallahan (Ogie too?) What would the current standings look like with a 3-point regulation win in effect? Trying to think how to go about this....is there anywhere you can find regulation wins broken out? Then it's just a matter of adding a point for each to the total? Anyone feeling ambitious?
 02-17-2010, 12:13 PM #15 morffin Registered User     Join Date: Sep 2009 Location: CT Country: Posts: 807 vCash: 500 Great thread. Any way to determine PPG / PP minutes? I have always felt that the PP% is not a good way to gauge how good a PP is since some PP only last a few seconds. Some sort of Goals per Minutes on the PP...like a pitchers ERA would be a good way to track PP proficiency. The same could be done for PK.
02-17-2010, 02:23 PM
#16
CaptainCallahan*

Join Date: Nov 2008
Posts: 239
vCash: 500
Quote:
 Originally Posted by Blueshirt Special Here's one for you CaptainCallahan (Ogie too?) What would the current standings look like with a 3-point regulation win in effect? Trying to think how to go about this....is there anywhere you can find regulation wins broken out? Then it's just a matter of adding a point for each to the total? Anyone feeling ambitious?
What would the points breakdown be? 3 Points for a regulation win, 0 for a loss, and than what else?

Quote:
 Originally Posted by morffin Great thread. Any way to determine PPG / PP minutes? I have always felt that the PP% is not a good way to gauge how good a PP is since some PP only last a few seconds. Some sort of Goals per Minutes on the PP...like a pitchers ERA would be a good way to track PP proficiency. The same could be done for PK.
Very interesting, and I'll look into that, but what I'm working on is individual based, not team based. So if someone wants to tackle that project.....

02-17-2010, 10:15 PM
#17
Blueshirt Special

Join Date: Feb 2007
Location: NJ
Country:
Posts: 3,208
vCash: 500
Quote:
 Originally Posted by CaptainCallahan What would the points breakdown be? 3 Points for a regulation win, 0 for a loss, and than what else?.....
3 points for a regulation win
0 for a regulation loss
2 points for OT/SO win
1 point for OT/SO loss

Forum Jump