Certified Legendary Thread Race for the flag, in squiggly lines

[Jtree]

But how would you calculate this? Would you divide the world into provinces ala Europa Universalis and then count the number of provinces each controls?

Yeah what constitutes good defence? Least amount of soldiers conceded? As a percentage of total army.... Or the Russians would have a pretty leaky backline!

 

[BF Tiger]

I dont get how it predicts adelaide winning every match from here except against geelong but somehow has them losing two games. Went through and checked, it has adelaide beating Carlton by 8 goals and yet the ladder adds a loss to the crows and hence pushes them down the ladder. Kind of important.

It takes into account that probable wins aren't certainties, so it will take into account the 25% chance that we drop this game, the 10% chance that we drop that game.

Just to expand a little... look at the tips and you will see a tipped winner/loser but under each team is a number that is the probability of a win according to Squiggle. It might be Adelaide v Carlton and under Adelaide it will say 0.81 wins and under Carlton it will say 0.19 wins. It is these figures that are added and then rounded to give total wins.

So if Adelaide before Carlton are already on 10.53 wins for the year it is rounded up to 11. They then get the added 0.81 wins which takes them too 11.34 wins, which is rounded down to 11.

Similarly if Carlton were on 8.37 wins before Adelaide, rounded down to 8, the extra 0.19 would take them to 8.56 and rounding up would give them 9 wins, even though they lost.

Hope this helps. And sorry for not using the actual Squiggle probabilities and totals, too hard to do on my phone.

 

[Deleted member 170709]

You can get round 13 tips by going into the forecast and clicking the + on home and away

Can't seem to find the forecast feature....could you point me to it?

 

[BF Tiger]

Can't seem to find the forecast feature....could you point me to it?

http://maxbarry.com/squiggle/ and hit the forecast button up the top

 

[Deleted member 170709]

http://maxbarry.com/squiggle/ and hit the forecast button up the top

Thanks, thought this might be a sure fire way of getting my weekly tips right, but doesn't seem to be unfortunately

 

[Suma Magic]

Geelong now tipped to win the flag

I think they are 4th on the flagpole.

 

[WhiskeyTangoFoxtrot]

Thanks, thought this might be a sure fire way of getting my weekly tips right, but doesn't seem to be unfortunately

It manages to get a decent number on average. I think Final Siren has published graphs showing the Squiggle's tipping accuracy previously in this thread.

 

[Deleted member 170709]

It manages to get a decent number on average. I think Final Siren has published graphs showing the Squiggle's tipping accuracy previously in this thread.

Ok i must look into further, thanks. I like to have a small multibet every week, looking for any advantage i can get

 

[WhiskeyTangoFoxtrot]

Ok i must look into further, thanks. I like to have a small multibet every week, looking for any advantage i can get

It has been getting the scores for both teams spot on for quite a few games. It doesn't adjust for weather though, so something like the Sydney vs Melbourne game was much lower scoring than expected.

 

[Deleted member 170709]

It has been getting the scores for both teams spot on for quite a few games. It doesn't adjust for weather though, so something like the Sydney vs Melbourne game was much lower scoring than expected.

Sounds sensational. I notice it averages about 6 winners every week tho which is a 66% strike rate - unless I'm missing something. Although good, looking for something a bit surer, that is, without fully looking into it in depth yet. I will certainly keep an eye on it in the next few rounds

 

[Final Siren]

Sounds sensational. I notice it averages about 6 winners every week tho which is a 66% strike rate - unless I'm missing something. Although good, looking for something a bit surer, that is, without fully looking into it in depth yet. I will certainly keep an eye on it in the next few rounds

It's 70.2% so far, which is a touch below its 10-year average, and a little above the 20-year average.

Squiggle definitely isn't a bookie-buster, or even a very sophisticated algorithm of the kind you can find on sites like Footy Maths. But it is solid for how simple it is, and, I hope, a tool to increase your knowledge.

Here is a good graphic from Figuring Footy:

Squiggle got 139 last year according to this criteria, which is no finals and draws being counted as wrong tips.

So it's generally better than your average punter, but unlikely to win your tipping comp. For that, you either need to watch a crapload of football, or else use a model as a tool to cover gaps in your knowledge, while still backing your own insight. Smart human + model should always beat model alone.

Or you could enter hundreds of times and tip randomly. That would work, too.

P.S. Flagpole got 145 last year by this criteria (152 overall), which looks like top 1% stuff. So that's impressive. But I think it just had a lucky year.

 

[Bob Neil]

Here is a good graphic from Figuring Footy:

What is the vertical axis in this graph?

 

[cryptor]

It's 70.2% so far, which is a touch below its 10-year average, and a little above the 20-year average.

Squiggle definitely isn't a bookie-buster, or even a very sophisticated algorithm of the kind you can find on sites like Footy Maths. But it is solid for how simple it is, and, I hope, a tool to increase your knowledge.

Here is a good graphic from Figuring Footy:

Squiggle got 139 last year according to this criteria, which is no finals and draws being counted as wrong tips.

So it's generally better than your average punter, but unlikely to win your tipping comp. For that, you either need to watch a crapload of football, or else use a model as a tool to cover gaps in your knowledge, while still backing your own insight. Smart human + model should always beat model alone.

Or you could enter hundreds of times and tip randomly. That would work, too.

P.S. Flagpole got 145 last year by this criteria (152 overall), which looks like top 1% stuff. So that's impressive. But I think it just had a lucky year.

Obviously games that are expected to be close are the hardest ones to pick. Have you ever looked to see at what (if any) margin Squiggle tips have a significant jump in reliability?

So it's 70.2% right now for 2016 considering all tips. If you were to only consider tips where the margin was tipped to be 1 goal or better either way does the Squiggle get significantly more of these correct?

Finding this sweet spot/s if it exists would be a good step in the direction of making the Squiggle a viable betting help tool.

 

[Sirlemsip]

What is the vertical axis in this graph?

No data for the vertical.
It is just the data points for the horizontal axis being spread out so you can see the relative number of tipsters at each value.. i guess that is somewhat of a data value.

Basically it just shows a relatively standard 'normal' distribution of data for tipping.

 

[Billy ray]

I want to make my child's middle name
Squiggle.

Deal. So, er, can I buy you a coffee sometime?

 

[sataris]

Deal. So, er, can I buy you a coffee sometime?

Billy Ray Squiggle has a nice ring to it.

 

[Theprofessional]

Great thread and very interesting- funny to see Swans so close to 2005 although weather obviously has a huge part in that, probably closer to 2012 form in reality which is still a bit of the normal premierships.

Hawthorn still very dangerous imo if they get players back, still a tough team but Geelong look poised for a top two finish. Looks like percentage could be more important than most other years so pretty important for Hawthorn to notch up a big win this weekend.

Good stuff

 

[GreyCrow]

Billy Ray Squiggle has a nice ring to it.

No. Stop this nonsense now. This is how the world gets Miley Squiggle

 

[Deleted member 170709]

It's 70.2% so far, which is a touch below its 10-year average, and a little above the 20-year average.

Squiggle definitely isn't a bookie-buster, or even a very sophisticated algorithm of the kind you can find on sites like Footy Maths. But it is solid for how simple it is, and, I hope, a tool to increase your knowledge.

Here is a good graphic from Figuring Footy:

Squiggle got 139 last year according to this criteria, which is no finals and draws being counted as wrong tips.

So it's generally better than your average punter, but unlikely to win your tipping comp. For that, you either need to watch a crapload of football, or else use a model as a tool to cover gaps in your knowledge, while still backing your own insight. Smart human + model should always beat model alone.

Or you could enter hundreds of times and tip randomly. That would work, too.

P.S. Flagpole got 145 last year by this criteria (152 overall), which looks like top 1% stuff. So that's impressive. But I think it just had a lucky year.

70% success is a very impressive strike rate indeed! It's finding the other up to 30% that would be nirvana.

As I mainly watch only the Hawks these days (I'm sure that's a major part of being behind the eightball) and haven't come across these models you speak of until now i guess, I'll have to stick to raw numbers analyses and a bit of hunch in randomly selecting 4-5 winners + margins together with my calculated 5-4 tips. Funny thing is i have won a few multibets on tipping on fewer games but always bomb out when going the whole hog in 9 games + margins. Guess ive been greedy, but all it will take is to win it once for an almighty return considering i play small amounts for fun but also for astronomical returns.

Love your work mate keep it up. Btw love to know who/what Flagpole is. Top 1% he should be rolling in it!

 

[WhiskeyTangoFoxtrot]

70% success is a very impressive strike rate indeed! It's finding the other up to 30% that would be nirvana.

As I mainly watch only the Hawks these days (I'm sure that's a major part of being behind the eightball) and haven't come across these models you speak of until now i guess, I'll have to stick to raw numbers analyses and a bit of hunch in randomly selecting 4-5 winners + margins together with my calculated 5-4 tips. Funny thing is i have won a few multibets on tipping on fewer games but always bomb out when going the whole hog in 9 games + margins. Guess ive been greedy, but all it will take is to win it once for an almighty return considering i play small amounts for fun but also for astronomical returns.

Love your work mate keep it up. Btw love to know who/what Flagpole is. Top 1% he should be rolling in it!

Getting 9/9 in a round by itself is pretty difficult, let alone trying to get the margin right as well.

 

[Hobbes]

Getting 9/9 in a round by itself is pretty difficult, let alone trying to get the margin right as well.

Some rounds are a lot easier than others - big differences between teams, predictable outcomes. Other rounds are really hard, either because of closely-matched teams or because somebody set the form guide to "randomise".

 

[WhiskeyTangoFoxtrot]

Some rounds are a lot easier than others - big differences between teams, predictable outcomes. Other rounds are really hard, either because of closely-matched teams or because somebody set the form guide to "randomise".

You are correct. It's still fairly rare for even an educated tipster to get 9/9 correct in a week lots of times per season as there's often an unexpected upset or two each round.

 

[Deleted member 170709]

Getting 9/9 in a round by itself is pretty difficult, let alone trying to get the margin right as well.

I know mate, almost impossible but I'm a deluded romantic always dreaming of a $5 investment for a $20k return. Punting $5 to win $300 bores me, and I know I should follow this line for an overall positive bank balance at the end of the season. Having said that, I only missed on 9 + margins when Blues rolled the Cats a few weeks back argh

 

[WhiskeyTangoFoxtrot]

I know mate, almost impossible but I'm a deluded romantic always dreaming of a $5 investment for a $20k return. Punting $5 to win $300 bores me, and I know I should follow this line for an overall positive bank balance at the end of the season. Having said that, I only missed on 9 + margins when Blues rolled the Cats a few weeks back argh

That just proves how hard 9+ margins is to get though!

 

[Here2tellyouwhy]

I know mate, almost impossible but I'm a deluded romantic always dreaming of a $5 investment for a $20k return. Punting $5 to win $300 bores me, and I know I should follow this line for an overall positive bank balance at the end of the season. Having said that, I only missed on 9 + margins when Blues rolled the Cats a few weeks back argh

That just proves how hard 9+ margins is to get though!

https://www.bigfooty.com/forum/threads/2016-afl-impossible-multi-aka-money-for-jam.1126561/

There's a thread every year. Still can't remember anyone winning it though.