Tuesday October 1 NHL Tor-Mon

[ProlinePlayer]

props 10-01.jpg (61.63 KiB) Viewed 12838 times

Anyone have any thoughts or insights into this?

PLP

[Scotty43]

Looking at some correlations here (sort of like Rosie play of Hky/HKS)

1. Kessel Under, Mtl Penalties Under, Mtl saves Under 6.1
2. Briere Under, Tor Penalties Under, Tor saves Under 5.5
3. Kessel Over, Mtl Penalties Over, Mtl saves Over 4.1
4. Briere Over, Tor Penalties Over, Tor saves Over 4.4

Tickets:
1: 1 & 2 - 33.6-1
2: 1 & 4 - 26.84-1
3: 2 & 3 - 22.6-1
4: 3 & 4 - 18.0-1

Will tread lightly to see if these correlations are true. A couple of fights throws things out of whack and Leafs led the league in fights last year.

[MattyKGB]

Numbers for TOR PIM and hits are high. I think they were at or near the top of the league last year and they still weren't that high on a per game basis.

I think we have to factor in some "regression to the mean" for teams with extremely high or low values last year in some of these categories.

PLP it would be great if you can get us the NHL version of that NFL data set you posted.

Strongest correlation I can see would be both teams PIM. Either there's a fight or there's not.

The million dollar question is, what's the correlation between TOR faceoffs and MTL faceoffs? Assuming it's negative, is it negative enough to justify playing both sides and crisscrossing O/U? I will attempt to do the math later using something called a mixed Poisson-Binomial distribution.

I like 1u 3u 4u 7u 8u 9u so far...

[MattyKGB]

Anyone calculating per-game averages based on last year's stats - remember that only 48 games were played last year, not 82. I almost made a huge mistake.

[kappa]

I like 1u 3u 4u 7u 8u 9u so far...

Nice to see we are on the same page. I keyed 3U/7U and and rotated 1U/4U/8U/9U. No correlation with 3 & 7 but the numbers seem too high.

[ProlinePlayer]

PLP it would be great if you can get us the NHL version of that NFL data set you posted.

.

will get to work on collating some data for this stuff

[ProlinePlayer]

The million dollar question is, what's the correlation between TOR faceoffs and MTL faceoffs? Assuming it's negative, is it negative enough to justify playing both sides and crisscrossing O/U? I will attempt to do the math later using something called a mixed Poisson-Binomial distribution.

an interesting idea. Only problem i see is that a team going over may just be a case of a game having a higher than average number of faceoffs rather than winning a higher percentage. In those cases both teams could go over or under

Definitely worth exploring.

PLP

[frostillicus]

just curious plp but are the odds for the hockey correlations reflect the new tie rules alc now has?

[ChinMusic]

I like 1u 3u 4u 7u 8u 9u so far...

I'm also playing all of those, plus 2o. Part of me is worried about the game turning into a PIM war, but the numbers don't lie. 3U and 4U are the way to play it.

[MattyKGB]

The million dollar question is, what's the correlation between TOR faceoffs and MTL faceoffs? Assuming it's negative, is it negative enough to justify playing both sides and crisscrossing O/U? I will attempt to do the math later using something called a mixed Poisson-Binomial distribution.

an interesting idea. Only problem i see is that a team going over may just be a case of a game having a higher than average number of faceoffs rather than winning a higher percentage. In those cases both teams could go over or under

Definitely worth exploring.

PLP

Yep...using a mixed Poisson-Binomial model, the correlation actually shows slightly positive. Go figure.

Next question, is there a home ice advantage in the faceoff circle? There's some kind of rule about the visiting player having to put his stick down first, right?