What's new
By:
USCHO Fan Forum

This is a sample guest message. Register a free account today to become a member! Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members through your own private inbox!

  • The USCHO Fan Forum has migrated to a new plaform, xenForo. Most of the function of the forum should work in familiar ways. Please note that you can switch between light and dark modes by clicking on the gear icon in the upper right of the main menu bar. We are hoping that this new platform will prove to be faster and more reliable. Please feel free to explore its features.

RPI 2013 Part V: Who Wants Screech?

Re: RPI 2013 Part V: Who Wants Screech?

With the week long PWR postulating on here, I was expecting real time updates on RPI's changing odds...:D How far off the board are we slipping ?? Can I make other plans for next weekend or what ?
 
Re: RPI 2013 Part V: Who Wants Screech?

Wicked Slappaahs said:
With the week long PWR postulating on here, I was expecting real time updates on RPI's changing odds...:D How far off the board are we slipping ?? Can I make other plans for next weekend or what ?
Click to expand...

Well, I already laid out all of today's scenarios :).

RHamilton said:
WCHA 'Quarterfinal' 1:
Minnesota State wins: http://pwr.reillyhamilton.com/minnst.html (25.63%)
Wisconsin wins: http://pwr.reillyhamilton.com/wis.html (14.58%)​
WCHA 'Quaterfinal' 2:
if Minnesota State won:
Colorado College wins: http://pwr.reillyhamilton.com/minnst-cc.html (26.06%)
North Dakota wins: http://pwr.reillyhamilton.com/minnst-nd.html (25.38%)​
​
If Wisconsin won:
Colorado College wins: http://pwr.reillyhamilton.com/wis-cc.html (15.22%)
North Dakota wins: http://pwr.reillyhamilton.com/wis-nd.html (14.20%)​
Click to expand...

Other than regenerating KRACHes between each game, not much else to do :p.
 
Re: RPI 2013 Part V: Who Wants Screech?

Wicked Slappaahs said:
I certainly didn't like Burger's salute, but it was a guttural reaction to Bailen having been cheap-shotted and injured due to a flagrant leg check that went uncalled...
Click to expand...

I seem to recall that a five-minute major and game misconduct was assessed on the play.

How much more of a call did you want?
 
Re: RPI 2013 Part V: Who Wants Screech?

Additionally, here's something else for people who like numbers to chew on:

http://pwr.reillyhamilton.com/games.html

The above page says... said:
This page lists the number of outcomes for each result of each game when RPI is in the tournament. For each game, the number represents the number of in-the-tournament outcomes that RPI would have remaining if that result occurred. The result with the greatest number of outcomes is the preferred; the disparity between the numbers of outcomes illustrates how important that game is to RPI. No KRACH weighting has been performed. Games that have gone final will have only one 100% result. Any game that is missing a result represents a case where RPI cannot make the tournament if that (missing) result occurs. There are currently no missing results -- no single result can spell doom for Rensselaer.
Click to expand...

It's effectively a quantitative cheering guide. Cheer for the team with the greater number/percent of outcomes.
 
Re: RPI 2013 Part V: Who Wants Screech?

RHamilton said:
If there are any other breakdowns you'd like to see, let me know!
Click to expand...

Just waiting on that web app so we can go through different situations without bugging you, but I understand it probably takes work. If anything, what would our chances be if BU, Providence, and Wisconsin all lose tomorrow?
 
Re: RPI 2013 Part V: Who Wants Screech?

lugnut92 said:
Just waiting on that web app so we can go through different situations without bugging you, but I understand it probably takes work. If anything, what would our chances be if BU, Providence, and Wisconsin all lose tomorrow?
Click to expand...

41.6815%, if I entered everything correctly. My queries are win based, so that's BC, Lowell, and SCSU.

I think that still leaves the possibility of 3 auto-bids, which substantially pushes down on our odds.

Will hopefully have it out tonight, but I'm on West Coast time :).
 
Re: RPI 2013 Part V: Who Wants Screech?

lugnut92 said:
Just waiting on that web app so we can go through different situations without bugging you, but I understand it probably takes work. If anything, what would our chances be if BU, Providence, and Wisconsin all lose tomorrow?
Click to expand...

So, here's 1/2 of the tool I was talking about:

http://pwr.reillyhamilton.com/query.php

It may take up to a minute or so for some queries. The more teams you switch from "Don't Care" to something else, the quick it will run. Results are cached, though, so repeated visits to the same query should be instantaneous.

Let me know if you have any issues running it. We'll see how long it stays up, and if my webhost complains!
 
Re: RPI 2013 Part V: Who Wants Screech?

RHamilton,

You mentioned earlier that your code runs significantly faster than it did two years ago despite the fact that you are using an old computer. How much of the speed increase is due to there being one fewer consolation games and thus only one third as many cases to run? (Or am I wrong about that? I don't remember when the WCHA started having two play-in games. It may be the same number of games, but one fewer that can end in a tie, and thus only a 33% decrease in the number of cases.)
 
Re: RPI 2013 Part V: Who Wants Screech?

Today is 22 March 2013. There are either 7, 8, or 197 days until RPI's next game. ;)


This is based upon the round of 16 of the NCAA tournament being on 29 and 30 March 2013 and 5 October 2013 being the start of next season.
 
Re: RPI 2013 Part V: Who Wants Screech?

RHamilton said:
So, here's 1/2 of the tool I was talking about:

http://pwr.reillyhamilton.com/query.php

It may take up to a minute or so for some queries. The more teams you switch from "Don't Care" to something else, the quick it will run. Results are cached, though, so repeated visits to the same query should be instantaneous.

Let me know if you have any issues running it. We'll see how long it stays up, and if my webhost complains!
Click to expand...
Nice. :)
 
Re: RPI 2013 Part V: Who Wants Screech?

Ralph Baer said:
RHamilton,

You mentioned earlier that your code runs significantly faster than it did two years ago despite the fact that you are using an old computer. How much of the speed increase is due to there being one fewer consolation games and thus only one third as many cases to run? (Or am I wrong about that? I don't remember when the WCHA started having two play-in games. It may be the same number of games, but one fewer that can end in a tie, and thus only a 33% decrease in the number of cases.)
Click to expand...

Right, there are 1/3rd as many outcomes as there were two years ago (~1.2 million outcomes down to ~400,000). So, there's a third amount of the time gone right there. Additionally, I re-wrote everything from scratch much more efficiently. On the same hardware (this same laptop), I was running at 4 outcomes per second last time. This year, it was running at closer to 30 outcomes per second. I additionally designed the database side of things a bit better, which helped to dramatically speed up the queries when generating reports.

For those interested, I uploaded the generating part of the source code here:
https://github.com/hamilr2/NCAA-Hockey-PWR-predictor/

It's not exactly as I ran it, in fact it apparently break;s right in the beginning of the PWR generating loop, as I as probably testing something when I committed the file.

The PWR generating function is here (it calls a bunch of other functions that can be found above it in that file):
https://github.com/hamilr2/NCAA-Hockey-PWR-predictor/blob/master/include.php#L295

This code is VERY sloppy, and a ton of stuff is hardcoded. My rewrite this year only took probably 4-6 hours of programming, and was designed to be quick, not necessarily extendable. It's not runnable without a LAMP/WAMP stack, and you'd need a DB schema (which I haven't yet provided).
 
You must log in or register to reply here.
Back
Top