Post 4954
Yes, still wrong given the port win in 2001. I'd guess we are not far off 50% in away finals, which is very good given the stronger team should be getting the home final anyway and the inherent advantage that comes from a home crowd.
Follow along with the video below to see how to install our site as a web app on your home screen.
Note: This feature may not be available in some browsers.
AFLW 2024 - Round 6 - Chat, game threads, injury lists, team lineups and more.
Post 4954
The squiggle does not care; the squiggle only processes.Lol at people having a go at a reasonably accurate data set presented without comment.
And the squiggle crashes and burns again. Picked 1 winner out of 4 games. Even my kids could do better than that. Squiggle is getting very close to having the credibility of Roby's Power Rankings.
Games Hawthorn (the club) won or lost many years ago before this current Hawthorn side was formed with these particular players have no statistical significance on how they'll perform.
As pointed out the side that plays away in a final tends to do so because they are the poorer side. Hence the poor record of travelling sides in finals and why Hawthorn (this side) haven't played an interstate final previous to last week since 2010.
The reason this Hawthorn side is being rated and tipped to beat Fremantle in a PF in WA in spite of what a particular set of data suggests is because this Hawthorn side isn't poorer but just inconsistent in being able to apply themselves - and many must think the reality of having no further chances will force them to apply themselves.
Personally, I'm not too sure. I feel like Hawthorn making the GF from here is maybe a 30-40% proposition. A certain OutSpoken Fremantle supporter will likely still argue it's 0%.
I'm convinced that the winner of the Hawthorn v Adelaide final will go into the PF with favouritism
It's more likely than a sydney v fremantle grand final, so no I wouldn't think it is the most unlikely outcome, and definitely not by a long way.Actually, no it doesn't. 10% of qualifying finals losers make the grand final. Hawthorn has not won a final outside Victoria in the history of the AFL. A Hawthorn v WC grand final is statistically the most unlikely outcome by a long way.
Actually, no it doesn't. 10% of qualifying finals losers make the grand final. Hawthorn has not won a final outside Victoria in the history of the AFL. A Hawthorn v WC grand final is statistically the most unlikely outcome by a long way.
Nail on headIf you just blindly follow the almighty Squiggle and don't account for the things it doesn't account for itself, like form (it tries to account for form, but there's a difference between winning, and winning convincingly) and especially injury, then get 3 out of 4 tips wrong, you've brought it on yourself.
Go back to the beginning re whether Squiggle takes account of the 'convincingness' of a team's victory (yes it does if you're talking about margin—and it also factors in whether it's home or away, the number of points given up as a measure of defensive quality, as well as raw margin, etc; see below if you're talking about, 'well, regardless of the final margin, they just didn't look that convincing in victory').If you just blindly follow the almighty Squiggle and don't account for the things it doesn't account for itself, like form (it tries to account for form, but there's a difference between winning, and winning convincingly) and especially injury, then get 3 out of 4 tips wrong, you've brought it on yourself.
This is why I think the squiggle has the potential to be amazing if certain things are implemented. Like player value/team value. Different home ground advantage, e.g. instead of 12 point subi, make it 17.5 or something closer to what it actually is.Go back to the beginning re whether Squiggle takes account of the 'convincingness' of a team's victory (yes it does if you're talking about margin—and it also factors in whether it's home or away, the number of points given up as a measure of defensive quality, as well as raw margin, etc; see below if you're talking about, 'well, regardless of the final margin, they just didn't look that convincing in victory').
When it comes to things like injuries, individual player form, 'resting' players, the vibe of how switched-on a team looks during a match, etc—I think you'll find that people are saying, 'this is an inherent limitation of a statistical model; you can allow for that stuff yourself when placing your tips/being a pundit if you want, but to expect things that are not generally amenable to being expressed purely in numbers, to be catered for by a foolproof statistical model, is crazy talk'. If you're not interested in statistical models, then there are plenty of 'Do you reckon that Richmond are the greatest chokers of all time?' (or whatever the made-up controversy of the day is) threads. Is the statistical model more than interesting, and actually useful?; i.e. does it get better results than the average, 'gee, I thought that Team X were looking a bit tired in the final quarter last weekend so I'll tip against them' impressionistic tipster? Looks like it does. Not perfect, but better than the average.
FIFY.Frustration as a Hawks supporter seeing what we're capable of (the squiggle peak) and then delivering 0.068%. Collingwood 2011 feel about it
I have seen this a lot actually. Can someone give some context? I am out of the loopFIFY.
The reason I said convincingness is because it can be a hard thing to gauge just on a statistical model. I mean, Fremantle just beat a top 4 side by 9 points - and even though Squiggle actually predicted they'd lose (not being able to interpret the Swans' injuries, mind you), the win was still far from convincing. Yet, overall their Squiggle benefited from the game.Go back to the beginning re whether Squiggle takes account of the 'convincingness' of a team's victory (yes it does if you're talking about margin—and it also factors in whether it's home or away, the number of points given up as a measure of defensive quality, as well as raw margin, etc; see below if you're talking about, 'well, regardless of the final margin, they just didn't look that convincing in victory').
When it comes to things like injuries, individual player form, 'resting' players, the vibe of how switched-on a team looks during a match, etc—I think you'll find that people are saying, 'this is an inherent limitation of a statistical model; you can allow for that stuff yourself when placing your tips/being a pundit if you want, but to expect things that are not generally amenable to being expressed purely in numbers, to be catered for by a foolproof statistical model, is crazy talk'. If you're not interested in statistical models, then there are plenty of 'Do you reckon that Richmond are the greatest chokers of all time?' (or whatever the made-up controversy of the day is) threads. Is the statistical model more than interesting, and actually useful?; i.e. does it get better results than the average, 'gee, I thought that Team X were looking a bit tired in the final quarter last weekend so I'll tip against them' impressionistic tipster? Looks like it does. Not perfect, but better than the average.
Not really a confidence interval, but yes, there was a good linear relationship between the predicted margin and the correctness of tips last time I looked at it, where PROBABILITY = 0.52 + (0.01 x MARGIN). That is, for every 1 point of margin, the tip is 1 percentage point more likely to be right, levelling out around 97%/3%.Does the margin forecast translate directly to a confidence interval? (equal = 50% 6 points = 55% etc.)?
Perhaps they just peaked too early. They were cherry ripe in August, and now have injury concerns to a number of players and they look tired and sore.Frustration as a Hawks supporter seeing what we're capable of (the squiggle peak) and then delivering 0.01%. Collingwood 2011 feel about it