Expected Win Percentages Rest Of The Way
So in a quick discussion around Home and Road goal differentials around the NHL last night on Twitter with Julian (PPP) and Neil Greenberg, the Washington Capitals blogger from the Washington Post, he linked to a story he had posted recently on the Caps likely probability of winning their games for the remainder of the season using the Log5 method originally devised by Bill James of Baseball Abstract and now Moneyball fame.
In an attempt to refine the results, I determined the team win percentages using the Pythagenpuck formula derived by Alan Ryder in his Win Probabilities paper from 2004, which is a modified Pythagorean win percentages model based on goals for and against.
If you care to look into the math in more detail follow the links above, if you don't? Read on and trust my numbers.
Basically the premise here was to look at all of the Leafs future opponents this season, and their home and away goal differentials, in an effort to use their predicted win percentages along with the Leafs' - and thus arrive at the probability the Leafs would walk away with a W.
As you can see from the chart, I allocated any probability above .475 or below .525 as an OTL because I'm trying to be realistic with the point results. Assuming the Leafs pick up some OT or SO wins likely bumps the total a tad... probably 3-4 points at least.
| GP | DATE | H/A | OPP | Prob | W-L-OTL |
|---|---|---|---|---|---|
| 63 | Tue, Feb 28, 2012 | Florida Panthers | 0.555 | 1-0-0 | |
| 64 | Wed, Feb 29, 2012 | @ | Chicago Blackhawks | 0.360 | 1-1-0 |
| 65 | Sat, Mar 3, 2012 | @ | Montreal Canadiens | 0.475 | 1-1-1 |
| 66 | Tue, Mar 6, 2012 | Boston Bruins | 0.365 | 1-2-1 | |
| 67 | Wed, Mar 7, 2012 | @ | Pittsburgh Penguins | 0.348 | 1-3-1 |
| 68 | Sat, Mar 10, 2012 | Philadelphia Flyers | 0.446 | 1-4-1 | |
| 69 | Sun, Mar 11, 2012 | @ | Washington Capitals | 0.379 | 1-5-1 |
| 70 | Tue, Mar 13, 2012 | @ | Florida Panthers | 0.485 | 1-5-2 |
| 71 | Thu, Mar 15, 2012 | @ | Tampa Bay Lightning | 0.433 | 1-6-2 |
| 72 | Sat, Mar 17, 2012 | @ | Ottawa Senators | 0.523 | 1-6-3 |
| 73 | Mon, Mar 19, 2012 | @ | Boston Bruins | 0.314 | 1-7-3 |
| 74 | Tue, Mar 20, 2012 | New York Islanders | 0.642 | 2-7-3 | |
| 75 | Fri, Mar 23, 2012 | @ | New Jersey Devils | 0.501 | 2-7-4 |
| 76 | Sat, Mar 24, 2012 | New York Rangers | 0.391 | 2-8-4 | |
| 77 | Tue, Mar 27, 2012 | Carolina Hurricanes | 0.607 | 3-8-4 | |
| 78 | Thu, Mar 29, 2012 | Philadelphia Flyers | 0.446 | 3-9-4 | |
| 79 | Sat, Mar 31, 2012 | Buffalo Sabres | 0.705 | 4-9-4 | |
| 80 | Tue, Apr 3, 2012 | @ | Buffalo Sabres | 0.432 | 4-10-4 |
| 81 | Thu, Apr 5, 2012 | Tampa Bay Lightning | 0.697 | 5-10-4 | |
| 82 | Sat, Apr 7, 2012 | @ | Montreal Canadiens | 0.475 | 5-10-5 |
So if a couple of those OTLs end up as Ws, the Leafs can maybe shift that to 7-10-3 or so. That would give the Leafs a final record on the season of 36-36-10 and 82 points which likely has them finishing in 10th or 11th in the East.
It should be pointed out that these probabilities are based on the current goal differentials of all of the teams on this list and every single one will see their records and differentials change over the course of the remainder of the season and thus these probabilities will be moving targets going forward.
If you find this depressing - don't feel bad, you aren't alone. I'm also guessing this makes it likely Ron Wilson doesn't come back next year - whether or not we hold him responsible for the team's goaltending, defence, special teams, or decent offensive production.
61 comments
|
0 recs |
Do you like this story?
Comments
Would 10/11th in the East really be a top 5 pick? Is the Eastern Conference that bad? (Excluding CoLOLmbus.) Why am I asking you all these questions?
"I myself am made up entirely of flaws, stitched together with good intentions."
11th in the East is the 7th pick right now.
Pension Plan Puppets
I hope YouTube comes down to film this.
You think the Hockey Gods would shine on Leafs Nation B1rky? haha I’m sure we’re eternally screwed!
Follow me on Twitter: https://twitter.com/#!/cyf8
by Cory Fawcett on Feb 28, 2012 10:44 AM EST up reply actions
Isn’t Galchenyuk not available to be drafted until next year?
Riding the Leafs bandwagon since 1991.
Check out my art!
by CanadianMaple09 on Feb 28, 2012 10:12 AM EST up reply actions
no. he’s definitely available in this year’s draft.
Pension Plan Puppets
I hope YouTube comes down to film this.
I don’t know what I was thinking. I was home in Sarnia last week and saw a couple of Sting games. Yakupov lit it up both games. Too bad that Galchenyuk is out for the season.
Riding the Leafs bandwagon since 1991.
Check out my art!
by CanadianMaple09 on Feb 28, 2012 10:15 AM EST up reply actions
We will be good enough to not draft a good player, and they probably won’t draft a Russian anyway. Wouldn’t fit in Toronto you know…smh
by Cory Fawcett on Feb 28, 2012 10:40 AM EST up reply actions
Lucky for us, Galchenyuk was born in Milwaukee!
The Roar of the Masses Could be Farts.
by Self Destructive Zones on Feb 28, 2012 10:57 AM EST up reply actions
Crazy
It’s predicting a 10 game losing streak. But if you look at the schedule and how the Leafs are currently planning, it’s not unreasonable.
Agreed
I was being roasted for even suggesting it on twitter last night. They can conceivably go 13-7 to finish the season with 91 points, and that’s with 2 losses to both the Bruins and Flyers. The question is if they can somehow find it in them to beat Florida, Montreal and Buffalo twice. The way they’ve been playing, it’s hard to imagine they have it in them. js
by Cory Fawcett on Feb 28, 2012 10:38 AM EST up reply actions
Is there a good summary of Log5 method anywhere? Or can you show me your work/formula?
I will stand beside him with an axe! @theninjagreg
I ask because Mr. Greenberg’s article talks about Log5 but not how to calculate it, and the pdf by Mr Ryder has some logarithms, but nothing explicitly named “Log5”.
I will stand beside him with an axe! @theninjagreg
by theninjagreg on Feb 28, 2012 12:20 PM EST up reply actions
It should be noted that this doesn’t have us winning a single road game. I understand (ish) where the numbers come from, but teams win road games. Even the Leafs.
In the court of public opinion, Steve Simmons is Lionel Hutz.
Right
This is why the Leafs still have to play the games.
Pension Plan Puppets: A Toronto Maple Leafs blog and a group therapy session.
Like reading thoughts confined to 140 characters? I'm on Twitter too.
by PPP on Feb 28, 2012 11:25 AM EST up reply actions 4 recs
Was that a “obviously this is just a projection” comment or a “this projection has serious limitations” comment?
In the court of public opinion, Steve Simmons is Lionel Hutz.
55% of the time the projection would be right 100% of the time.
by Leafswinthecup on Feb 28, 2012 12:17 PM EST up reply actions
Both
Pension Plan Puppets: A Toronto Maple Leafs blog and a group therapy session.
Like reading thoughts confined to 140 characters? I'm on Twitter too.
Leafs are basically fighting for one spot with two of Jets, Caps & Panthers. I say two because one of them will win their division and the third seed. But they won’t make it because these teams will still pick up a OT/SO point even if they keep losing. Down the stretch about 25% of games end as 3 pt games.
Having said that I can see the Leafs winning their last 5 of 6 games as they are against teams that will miss the playoffs and those teams will basically focus on player development than meaningless wins. This means Leafs will get nowhere close to a lottery pick.
Yup. The Leafs fucked themselves in February. They saw a month full of winnable games followed by a month of extremely hard games (at least compared to first 1/2+ season Leafs) and decided that “yup were going to just give up now”.
Dat rebuild
totally gave up, didn’t try at all, scored no goals, made no saves, hell, I don’t even think they practiced
Y'all heard about me, you just didn't know it was me
by JaredFromLondon on Feb 28, 2012 11:39 AM EST up reply actions
That’s why teams shouldn’t look at their expected win percentages.
In the court of public opinion, Steve Simmons is Lionel Hutz.
I think expected win percentage is very useful if interpreted correctly. It’s not a perfect predictor of W/L but it is useful in the sense that it gives you an unbiased approximation of how well the team performs down the stretch based on how they performed historically. Can the Leafs magically become better and go 12-5-3 down the stretch? Absolutely. Is it likely based on how they’ve played so far this season? I don’t think so.
by Leafswinthecup on Feb 28, 2012 12:20 PM EST up reply actions
hyperbole, eh? fightin’ dirty, huh?
imma comatchu like a alliteration alligator.
I shot a moose once, in upstate New York.
by not norm ullman on Feb 28, 2012 12:11 PM EST up reply actions
Hey Burtch
I’ll bet you $19 that this isn’t the exact set of results the Leafs end up with.
Come on, tough guy.
Let’s see some green.
I shot a moose once, in upstate New York.
by not norm ullman on Feb 28, 2012 11:45 AM EST reply actions 3 recs
Ok, ok.
$18.50
I shot a moose once, in upstate New York.
by not norm ullman on Feb 28, 2012 2:38 PM EST up reply actions
What I take away from this is the Leafs simply need to survive the next three weeks.
Get past Game 73 in Boston, and the Leafs close with a manageable schedule. 6 at home, 3 on the road, only 3 games against teams currently in the playoffs.
Right now, the Leafs are 3 points out and have 2 teams to hop over. Let’s set that as the benchmark. If three weeks from now the Leafs are in the same position or better, I think they make the playoffs.
Cynically Sarcastic
Сертыфікаваны Grabbo Палюбоўнік
“Survive” was going to be their mantra if they hadn’t taken a month off from getting any points. Their mantra now might as well be “Wait til next year”.
Pension Plan Puppets*
* Blog contains less than 2% puppet content by weight.
by Chemmy on Feb 28, 2012 12:29 PM EST up reply actions 1 recs
Who the hell knows?
The team’s played from behind down the stretch the last four years. Maybe they’re just a team that breaks late
Cynically Sarcastic
Сертыфікаваны Grabbo Палюбоўнік
Nothing can be done to substantially improve the team until June now. Might as well root for some wins.
Cynically Sarcastic
Сертыфікаваны Grabbo Палюбоўнік
Oh you’re cute
Fairweather fans can go to hell
Equal oppurtunity asshole and a proud member of the PPPPP
Follow me on twitter: leafer1984
Cheering for your team to win is something that fans of other teams do . . .
The plural of "anecdote" is not "data".
Internet Champ. I'm on Twitter too! - draglikepull
by Draglikepull on Feb 28, 2012 3:02 PM EST up reply actions
This information makes me wish I worked in a high rise with openable windows.
Offizielles Mitglied des MĂĽller / Holzer Fan-Clubs. Ich fordere mehr Deutsche in diesem Team ....... tweet me @Alex_Scotian
Why do you hate the Maple Leafs?
Why do you insist on telling me my favorite team, the best in the world, may not be very good at ice hockey? You are a mean person and secretly hate the Toronto Maple Leafs! I’m going back to a place where the Maple Leafs are the 20 time defending Stanley Cup Champions.
The First Certified Grabbo Lover
I’m going back to a place where the Maple Leafs are the 20 time defending Stanley Cup Champions.
Ah NHL 12 – the place that dreams are made of.
My rambling tweets
Winning is a habit. Unfortunately, so is losing - Vince Lombardi
I smell a 9-game win streak to end the season
Lombardi and Armstrong finally round out into mid-season form and their line becomes our third first line. Jake Gardiner becomes the next Karlsson. Leafs make the playoffs on a Kadri shootout goal. Carey Price pulls his groin on the play.
Shameless optimism
9 game winning streak? Let’s go 20-0-0! Destroy all competition! Autobots, Transform! Up and away! To infinity and beyond!
and so on and so forth
by Goosemonster on Feb 28, 2012 4:56 PM EST up reply actions
Can he become the next Karlsson without being balls in his own end?
The Roar of the Masses Could be Farts.
by Self Destructive Zones on Feb 28, 2012 5:12 PM EST up reply actions
And I’m the negative one!?
All hockey all the time at The Globe and Mail
by James Mirtle on Feb 28, 2012 4:55 PM EST reply actions 1 recs
who called you negative?
I mean aside from all those times I did earlier this season.
"Success is the ability to go from one failure to another with no loss of enthusiasm."
- Sir Winston Churchill
I'm pretty sure he's talking about the Leafs.
by Steve Burtch on Feb 28, 2012 5:11 PM EST up reply actions
Losing 8 of the past 9 and not acquiring magical unicorns for beans will do this to fragile fan base.
"They build a statue, they knock it down and piss on it, and now they will be out there building it again."
by ThickSkinnedAlive on Feb 28, 2012 5:25 PM EST up reply actions
Another way of looking at this is
the Leafs have a better than 50% chance of winning 7 of their last 20 games… plus another 7 games with a 40% chance of coming away the victors. So it’s not impossible that they could win 14 of the last 20. The math just says it’s unlikely.
Should also be noted that the Leafs are playing backsides of b2b’s with travel @ CHI, @PIT, @ WAS, vs. NYI, vs. NYR… so 5 times out of those 20 games. Only 1 of those was an expected win, but that makes it less likely they win them.
"Success is the ability to go from one failure to another with no loss of enthusiasm."
- Sir Winston Churchill
I'm pretty sure he's talking about the Leafs.
Amazingly we can’t beat teams on the second half of their back to backs and can’t win on ours.
"They build a statue, they knock it down and piss on it, and now they will be out there building it again."
by ThickSkinnedAlive on Feb 28, 2012 5:33 PM EST up reply actions
I am confused
Tampa has a -38 goal differential. We have a -6 goal differential. Why would they be the favourites? Does being hte visitor make that much of a difference?
(Same with Fla -17, and Wash -7)
About time that people finally realized how awesome Gunnar is...
Certified Gunnar & Kule lover!
My new goal: To get the nickname Hebrew Hammer for Mike Brown to take off.
Their name starts with a T.
I shot a moose once, in upstate New York.
by not norm ullman on Feb 28, 2012 7:09 PM EST up reply actions
You're using
overall goal differential, I used home and road differentials – since there are huge swings for many teams around the NHL between home and road goals for and against.
For instance TB’s goal differential at home is +10, but on the road is -46. Florida is -1 at home, but -10 on the road. Washington is +20 at home, and -24 on the road.
So yes it does make that big of a difference.
"Success is the ability to go from one failure to another with no loss of enthusiasm."
- Sir Winston Churchill
I'm pretty sure he's talking about the Leafs.
by Steve Burtch on Feb 29, 2012 7:19 PM EST up reply actions
The expected number of points you estimate based on these probabilities is a bit low, I think. You are awarding only 1 point to games that the Leafs have a greater than 50% chance of winning. Based on these probabilities, the expected number of wins is 9.442, and if we assume that approximately 20% of the Leafs losses are overtime losses (based on their record this season) then their expected number of points is 21.12 + 65 = 86.12.
It's an expected win percentage
In a game where more than 2 points are available this is a significant problem, and there’s no accurate way to account for OTL’s.
The original method was devised in baseball where there are no ties or points for OT.
Don’t use it as a perfect approximation.
"Success is the ability to go from one failure to another with no loss of enthusiasm."
- Sir Winston Churchill
I'm pretty sure he's talking about the Leafs.
by Steve Burtch on Feb 29, 2012 7:17 PM EST up reply actions
Haha
You didn’t even predict the first game right. Case closed, math sucks.
I will stand beside him with an axe! @theninjagreg
This is the formula
according to Mr. Burtch:
log5 formula is just: (p – pq) / (p + q – 2pq) where p and q are the expected win% of the two teams in competition.
Saving for future reference.
I will stand beside him with an axe! @theninjagreg

by 

































