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 2009-10 Part IV: The Boys are Back in Town

Status
Not open for further replies.
Re: RPI 2009-10 Part IV: The Boys are Back in Town

jmhusker said:
Not so fast there RC. The model also states as follows:

All other future unplayed games are assumed won/lost with a probability based upon relative team strengths.

In other words, it is saying all teams play as well as they have thus far no better and no worse. This leads to little movement in the standings and doesn't calculate all possible permutations. If we were to win the last 4 and SLU did the same, and then we ran the table until the final we would finish 25-14-2 which would put us in the 16-18ish RPI. The model as stated does not take this scenario into account because it assumes that SLU only takes about 5 of the remaining 8 points. How our PWR would fare from there is the question but I'll bet there are scenarios where we could beat out 3 or 4 teams above us and sneak in.

Keep hope alive!
Click to expand...
I was looking a while back as to what exactly they do, but I couldn't find it. I assume that they run a Monte Carlo simulation. That is they run thousands (maybe even millions) of sets of game outcomes assigning for each game an outcome each time via those probabilities.
 
Re: RPI 2009-10 Part IV: The Boys are Back in Town

Another season reset is up at Without a Peer, featuring very detailed descriptions of potential results for the women's team.
 
Re: RPI 2009-10 Part IV: The Boys are Back in Town

Red Cloud said:
Well, they also use an X in each table for when achieving a position is statistically impossible. And, well, we've got an X in the "Win Out Minus CF" column, where we didn't before losing to Dartmouth. Other teams have at least "<1%" where there's at least a statistical possibility of it happening.

Personally, I'm in complete agreement. Based on what I know about PWR, there's not much opportunity for the Engineers to get into the Top 15.

Regardless, if you're counting on an at-large bid, you're fooling yourself.
Click to expand...
All I'm saying is that is it not statistically impossible using a number of scenarios besides "past performance guarantees future outcome". I would only be fooling myself if I thought it was impossible. I agree that it is quite unlikely.
 
Re: RPI 2009-10 Part IV: The Boys are Back in Town

Ralph Baer said:
I was looking a while back as to what exactly they do, but I couldn't find it. I assume that they run a Monte Carlo simulation. That is they run thousands (maybe even millions) of sets of game outcomes assigning for each game an outcome each time via those probabilities.
Click to expand...
When you say "those probabilities" do you mean via the winning percentages so far?
 
Re: RPI 2009-10 Part IV: The Boys are Back in Town

jmhusker said:
When you say "those probabilities" do you mean via the winning percentages so far?
Click to expand...

I would hope that it is via the likelihood of each team winning in each game based upon KRACH to date, but as I said, they don't see to have divulged what they actually did.
 
Re: RPI 2009-10 Part IV: The Boys are Back in Town

Ralph Baer said:
I would hope that it is via the likelihood of each team winning in each game based upon KRACH to date, but as I said, they don't see to have divulged what they actually did.
Click to expand...

It looks like KRACH at least plays some role in their calculations.
 
Re: RPI 2009-10 Part IV: The Boys are Back in Town

jmhusker said:
All I'm saying is that is it not statistically impossible using a number of scenarios besides "past performance guarantees future outcome". I would only be fooling myself if I thought it was impossible. I agree that it is quite unlikely.
Click to expand...

Calculate it out, then. Under what scenario does RPI not win the ECAC championship, yet still makes it to the big dance?

Honestly, if there's an X there, I trust it to be mathematically impossible. Unfortunately we are not dealing with very large quantities of 2, so 2 + 2 will never equal 5.
 
Re: RPI 2009-10 Part IV: The Boys are Back in Town

Ralph Baer said:
I would hope that it is via the likelihood of each team winning in each game based upon KRACH to date, but as I said, they don't see to have divulged what they actually did.
Click to expand...
Well whatever they did the model doesn't reflect the reality of the RPI and PWR. Cornell is ranked 9th in PWR today. They are 23rd in this model with only a 29% chance of making the tournament by going 7-0 and then potentially losing to Yale (who is 7th in PWR) in the final. I should think that under that scenario Cornell would be a lock to make the tournament.
 
Re: RPI 2009-10 Part IV: The Boys are Back in Town

jmhusker said:
Well whatever they did the model doesn't reflect the reality of the RPI and PWR. Cornell is ranked 9th in PWR today. They are 23rd in this model with only a 29% chance of making the tournament by going 7-0 and then potentially losing to Yale (who is 7th in PWR) in the final. I should think that under that scenario Cornell would be a lock to make the tournament.
Click to expand...

The eagle eyed among us noted that their numbers haven't been updated since Sunday morning.

There's still five weeks left in the season. There are an extreme few teams who are locks for the tournament. Like, maybe two.
 
Re: RPI 2009-10 Part IV: The Boys are Back in Town

FlagDUDE08 said:
Calculate it out, then. Under what scenario does RPI not win the ECAC championship, yet still makes it to the big dance?

Honestly, if there's an X there, I trust it to be mathematically impossible. Unfortunately we are not dealing with very large quantities of 2, so 2 + 2 will never equal 5.
Click to expand...
I agree that it is mathematically impossible under the assumption that they stated. I question whether their assumption is valid. It is probably as good as any other scenario but it is not a comprehensive look at all of them. Nor does it factor in the idiosynchrasies of RPI and PWR, as evidenced by Cornell's situation.

I'm fairly confident that I can put together a scenario that would get us into the top 15 in PWR. Let the top 10 teams (or so) win out. SLU wins their next 4 but RPI wins out until the final. Let the underdogs win every other game including in the playoffs. Mix in some wins for people RPI fares well against in PWR (Yale, BU, UNH) and losses for those we don't (Union!). I'll bet that gets us in. It's a ridiculous enough scenario but it is possible. When I get a chance I'll try it out.
 
Re: RPI 2009-10 Part IV: The Boys are Back in Town

jmhusker said:
Nice try. I'm not going to escalate.
Click to expand...

Why would there be anything to escalate? I was just opining that it's sad to derive pleasure from irritating other people or from seeing someone irritated.

You really don't understand me at all. You think you do, you do not.
 
Re: RPI 2009-10 Part IV: The Boys are Back in Town

Red Cloud said:
Well, they also use an X in each table for when achieving a position is statistically impossible. And, well, we've got an X in the "Win Out Minus CF" column, where we didn't before losing to Dartmouth. Other teams have at least "<1%" where there's at least a statistical possibility of it happening.

Personally, I'm in complete agreement. Based on what I know about PWR, there's not much opportunity for the Engineers to get into the Top 15.

Regardless, if you're counting on an at-large bid, you're fooling yourself.
Click to expand...

Hey RC, are you sure that takes the league consolation game into account? And why does RPI have a 6% chance under current win%s? Does that mean they have a 6% chance of winning the tourney? I like pushing you.
 
Re: RPI 2009-10 Part IV: The Boys are Back in Town

Oh look, my stalker is back.

Too bad for him he's on ignore.
 
Status
Not open for further replies.
Back
Top