Sunday, December 27, 2009

timeonice.com

I have finally gotten around to updating the playershots scripts at timeonice.com. What follows is a 'how to' for those interested in using these apps. All data is for even strength only, empty net situations are excluded.

http://timeonice.com/playershots0910close.php?team=EDM&first=20001&last=21230&daniel=5&henrik=71

The word close refers to the score being within one goal in the first or second period, or a tied score in the third period or overtime.

Omit the word close to get the overall numbers, or replace it with tied, lead or trail to generate data for these self explanatory game states. ' Lead' refers to leading in the third period, or leading by more than one goal in the first or second period. 'Trail' is the photo negative of that.

The team can be switched by changing the three character team abbreviation. So replacing EDM with CGY in the above URL would generate the Flames' numbers. A full list of NHL.com play by play abbreviations can be found here.

The URL near the top of the page also outputs the combined data of Smid (5) and Visnovsky (71). This result is shown as a fictional player with the jersey number 99. If you add "&kesler=27" to the end of this URL, the results of player 99 will reflect events that occurred when all three of Smid, Visnosky and Penner were on the ice at even strength, with both goalies in the net, and with the score close. The players can be changed by using different jersey numbers in the URL.

21 Comments:

Blogger Hawerchuk said...

This comment has been removed by the author.

12/27/2009 2:16 pm  
Blogger Olivier said...

Yay! Christmas comes early this year!

Err...

12/27/2009 2:58 pm  
Blogger Earl Sleek said...

or replace it with tied, leading or trailing to generate data for these self explanatory game states.

Actually, I'm finding that the words to insert are "lead" and "trail", not "leading" and "trailing".

But yeah, effing awesome, man.

12/27/2009 6:02 pm  
Blogger Sunny Mehta said...

sick

12/27/2009 6:16 pm  
Blogger Sunny Mehta said...

my only suggestion would be to tweak Sh% and Sv% so that corsi is the denominator.

12/27/2009 6:22 pm  
Blogger Scott Reynolds said...

Thank you very much Vic! This is tremendous! That is all.

12/27/2009 6:34 pm  
Blogger Hawerchuk said...

Fuck, sorry for that stupid request. I clearly can't even read. Great job, Vic!

12/27/2009 8:36 pm  
Blogger Kent W. said...

Yay!

12/28/2009 12:28 am  
Blogger Vic Ferrari said...

D'oh! Thanks Earl, I've changed the original post.

12/28/2009 7:39 am  
Blogger Vic Ferrari said...

Sunny,

I agree, but will leave it as is for this season at least. It's simple enough to drop that into a spreadsheet and generate those corsi shootings and save%s yourself.

12/28/2009 7:42 am  
Blogger Sunny Mehta said...

vic,

yup, no prob.

btw, not sure if you saw the discussion at JLikens' recently, but basically the spread of Sh% this season seems to be a bit wider than last. idk if you've looked into it at all, but i'm curious to get your thoughts.

12/28/2009 11:29 am  
Blogger Vic Ferrari said...

I didn't see the thread at objectivenhl.blogspot.com. Link?

12/28/2009 2:21 pm  
Blogger Sunny Mehta said...

http://objectivenhl.blogspot.com/2009/12/even-strength-data-for-games-1-500.html

12/28/2009 2:29 pm  
Blogger Vic Ferrari said...

Thanks Sunny, terrific conversation there. I have Jlikens on my feed reader and he's a first stop if he's posted something new ... somehow I missed that, though.

A lot of the variation may well be coming from the fact that tied scores contains random competition effects as well. Not something that would have occurred to me, except a poster at LT or MC (knighttoen maybe?) posted some stuff last year showing the late-season Oilers to be very good at EV tied corsi and getting absolutely bootstomped when leading. It turned out that over that stretch most of their tied minutes were vs poor teams, and, by chance alone, they managed to get the early lead against almost all the good teams and hang on for dear life while everything seemed to hit Roloson for a couple of periods.

That's why I started using "close score" numbers. Because "while tied" clearly excludes all playing-to-the-score effects, in absorbs a lot of randomness by it's nature.

I agree with your comments re corsi save/shooting% as well.

When time permits, I'll send you and Jlikens a script to get the game by game data. I've got some scoring chance stuff written that I'll clean up and send to the scoring chance guys eventually as well. That said, I think all of us realize that while teachers are paid to educate kids that are dumb as stumps ... we aren't. And you are only ever as wise as others are foolish.

12/28/2009 3:43 pm  
Blogger Jonathan Willis said...

Thanks for doing this, Vic. It's absolutely essential stuff and I appreciate it.

12/31/2009 4:16 pm  
Blogger Hawerchuk said...

Vic,

I think your 'close score' is a good idea. We've convinced ourselves that the trailing team has a higher shot rate, but how much of that is late-game desperation? Have you ever looked at shooting rates vs time remaining in one-goal games? I was going to take a look, but I didn't want to duplicate your efforts.

1/01/2010 11:37 am  
Blogger slipper said...

Vic. you magnificent bastard.

1/03/2010 7:06 pm  
Blogger RiversQ said...

Good. Thanks Vic.

1/04/2010 2:07 am  
Blogger YKOil said...

Not to sound too much like a fanboy but...

Vic...

You Rule.

1/07/2010 5:39 pm  
Blogger Sunny Mehta said...

Vic,

Can I manipulate the URL to see only home or only road games, or is there not anything in the script for that?

3/09/2010 12:11 pm  
Blogger Jim Philips said...

I will take a look because it sounds pretty interesting and it helps me a lot with the price per head free trial

5/30/2013 11:17 pm  

Post a Comment

<< Home