Sports club stats uses goal-differential to determine team strength, which is why it (over)rates the Bruins so highly, and likes the Habs' chances more than most of the teams around them (note how they're given better odds than Winnipeg despite trailing them in the standings)
Whatever the chances are, I think it's important we move some of our expiring contracts before the Habs get to close to a playoff spot. Guys like Campoli, Gill and Moen. I wouldn't even be against moving Kostitsyn for the right return.
Even if Montreal did move these 4 guys, I don't think it significantly lowers their chances for a miracle run to the playoffs.
If Montreal waits to long, they may end up close enough to a playoff spot that that moving anyone initiates a media frenzy.
Sports club stats uses goal-differential to determine team strength, which is why it (over)rates the Bruins so highly, and likes the Habs' chances more than most of the teams around them (note how they're given better odds than Winnipeg despite trailing them in the standings)
Those numbers are starting to be very telling, especially for Fla. Ottawa too, i expect the number to drop more into the negative
Quote:
Originally Posted by sweat
Whatever the chances are, I think it's important we move some of our expiring contracts before the Habs get to close to a playoff spot. Guys like Campoli, Gill and Moen. I wouldn't even be against moving Kostitsyn for the right return.
Even if Montreal did move these 4 guys, I don't think it significantly lowers their chances for a miracle run to the playoffs.
If Montreal waits to long, they may end up close enough to a playoff spot that that moving anyone initiates a media frenzy.
campoli wont change the numbers, gill maybe, Ak will do. Keep AK and keep moen too. Gill still good, not sold on trading him yet.
Sports club stats uses goal-differential to determine team strength, which is why it (over)rates the Bruins so highly, and likes the Habs' chances more than most of the teams around them (note how they're given better odds than Winnipeg despite trailing them in the standings)
That would explain the 9% jump from the previous day. Habs' increased their goal differential by 5 against the Wings.
Goal differential is better than previous record as a yardstick of team strength, but it's not the be-all end-all either.
It is, however, a good clue that the Habs are a lot stronger than their record suggests. They lose too many one-goal games.
I find goal differential to be rather meaningless tbh. It's much like the pythagorean record in baseball. Blowing out a team one game 7-2 has no predictive value whatsoever as to how that team will fair in subsequent games where injuries and rosters are ever changing.
Not only is goal differential flawed in the sense it doesn't tell us much, but the goal differential itself can be misleading. For example, early in the beginning of the year we had a negative goal differential despite outplaying nearly everyone, so it's bad that goals scored one game are not indicative of the next game, but how the differential came about in the first place is also misleading. However, it's not completely without merit, you obviously can't be hugely in the negative and still make the playoffs, you do have to outscore your opponents to win games, but it's not nearly as valuable as some would like to believe. I agree that it's a good clue that the habs are better than their record indicates, how much better is where the problem lies.
Every year the Jays had an outstanding Pythagorean record and finished anywhere from 3-4 in the division under JP Ricciardi. Interesting to discuss, but not very meaningful in the end.
Sportsclubstats operates by simulating six million seasons then counting which proportion has team X making the playoffs.
While it's mathematically possible that Boston might miss, it only occured in 540 out of 6 000 000 -- far less than 0.05%, so it was rounded up.
It also assumes that teams will remain as good as they were from the start of the season to now, which is why Boston's most likely record until the end is 25-8-2 and Montreal is 18-12-3.
That would explain the 9% jump from the previous day. Habs' increased their goal differential by 5 against the Wings.
That's the flaw in this metric and many other statistics. The 5 extra goals against Detroit doesn't really increase our chances at all, but if you use something as arbitrary as goal differential it would imply that our chances got that much better, when really, they barely improved at all.
The simulation simply weighted the Habs a bit more with this new data and gave an average of 1 or 2 wins more for each of the 6 million runs.
I think the best way to use the site is to see where the simulations projet the points threshold to make the playoffs in the east. Right now it says it is around 90 points. The Habs therefore need a 20-10-3 or similar record until the end.
I find goal differential to be rather meaningless tbh. It's much like the pythagorean record in baseball. Blowing out a team one game 7-2 has no predictive value whatsoever as to how that team will fair in subsequent games where injuries and rosters are ever changing.
Not only is goal differential flawed in the sense it doesn't tell us much, but the goal differential itself can be misleading. For example, early in the beginning of the year we had a negative goal differential despite outplaying nearly everyone, so it's bad that goals scored one game are not indicative of the next game, but how the differential came about in the first place is also misleading. However, it's not completely without merit, you obviously can't be hugely in the negative and still make the playoffs, you do have to outscore your opponents to win games, but it's not nearly as valuable as some would like to believe. I agree that it's a good clue that the habs are better than their record indicates, how much better is where the problem lies.
Every year the Jays had an outstanding Pythagorean record and finished anywhere from 3-4 in the division under JP Ricciardi. Interesting to discuss, but not very meaningful in the end.
Its still better than predicting based off team record, which should tell you how reliable using that is.
And the Jays were always a good team in a ridiculous division. That's going to heavily distort the result.
I find goal differential to be rather meaningless tbh. It's much like the pythagorean record in baseball. Blowing out a team one game 7-2 has no predictive value whatsoever as to how that team will fair in subsequent games where injuries and rosters are ever changing.
One game sample ? no **** the predictive value is basically useless, it's not meant for 1 game samples.
Quote:
Not only is goal differential flawed in the sense it doesn't tell us much, but the goal differential itself can be misleading. For example, early in the beginning of the year we had a negative goal differential despite outplaying nearly everyone, so it's bad that goals scored one game are not indicative of the next game, but how the differential came about in the first place is also misleading. However, it's not completely without merit, you obviously can't be hugely in the negative and still make the playoffs, you do have to outscore your opponents to win games, but it's not nearly as valuable as some would like to believe. I agree that it's a good clue that the habs are better than their record indicates, how much better is where the problem lies.
Every year the Jays had an outstanding Pythagorean record and finished anywhere from 3-4 in the division under JP Ricciardi. Interesting to discuss, but not very meaningful in the end.
We maintained a close to nil GD, the jump was due to +5 and a win, and a win against a team that was, statistically, supposed to win that game, by alot i'm assuming, that the 16% chance of making prior to the win was in large part due to the fact it factored the low probabilities of winning that game.
Meaningfulness, well, that's just like your opinion man.
Whatever the chances are, I think it's important we move some of our expiring contracts before the Habs get to close to a playoff spot. Guys like Campoli, Gill and Moen. I wouldn't even be against moving Kostitsyn for the right return.
Even if Montreal did move these 4 guys, I don't think it significantly lowers their chances for a miracle run to the playoffs.
If Montreal waits to long, they may end up close enough to a playoff spot that that moving anyone initiates a media frenzy.
The only 'fear' I have regarding to letting loose some or all of these players (and maybe others) is that Mr. Gauthier will pull a 'Bob Gainey' and just let them go without trying to get something in return (ie. Mark Streit & to a lesser extent, Sheldon Souray).
I don't think he'll do that since he's already signed a player during the season which was, in Mr. Gainey' way, a thing that wasn't done.
One game sample ? no **** the predictive value is basically useless, it's not meant for 1 game samples.
We maintained a close to nil GD, the jump was due to +5 and a win, and a win against a team that was, statistically, supposed to win that game, by alot i'm assuming, that the 16% chance of making prior to the win was in large part due to the fact it factored the low probabilities of winning that game.
Meaningfulness, well, that's just like your opinion man.
No, it's not an opinion. Obviously I used an extreme case of one game to highlight the fact that it means very little, it doesn't change the point. These give you a little insight to what's going on in the bigger picture, nothing more.
PLAYOFFS, HERE WE COME.
