OT Football prediction sites question

-

jleves

Awesomeness
Joined
Aug 27, 2011
Messages
3,275
Likes
4,695
I participate in a football pool every year where you have to pick two teams to win each week, no point spread, just pick the winner. Once you select a team, you can't pick them for the rest of the season. Week one correct picks are worth 1 point each. Week 2 are worth 2 points, etc through week 16 worth 16 points each and that exhausts all the potential teams. Obviously you have to get rid of bad teams early and count on big points for the final 8 or 10 weeks to have any chance of winning. This involves having to really plan for when your best matchups are.

I've been using the game predictor on ESPN for the last several years which gives win% for each team based on some algorithm. I was wondering if there are other sites that may have better models for future games. Any help is appreciated.
 

intlzncster

i fart in your general direction
Joined
Aug 24, 2011
Messages
27,716
Likes
33,419
I participate in a football pool every year where you have to pick two teams to win each week, no point spread, just pick the winner. Once you select a team, you can't pick them for the rest of the season. Week one correct picks are worth 1 point each. Week 2 are worth 2 points, etc through week 16 worth 16 points each and that exhausts all the potential teams. Obviously you have to get rid of bad teams early and count on big points for the final 8 or 10 weeks to have any chance of winning. This involves having to really plan for when your best matchups are.

I've been using the game predictor on ESPN for the last several years which gives win% for each team based on some algorithm. I was wondering if there are other sites that may have better models for future games. Any help is appreciated.
Don't know if this is the same one you are already using, but if memory serves, fivethirtyeight does something like this using their ELO ratings.

EDIT: Just went to Google for a memory refresh, and stumbled on this article. Might be useless for your purposes, but it has some interesting ideas.

 

jleves

Awesomeness
Joined
Aug 27, 2011
Messages
3,275
Likes
4,695
Don't know if this is the same one you are already using, but if memory serves, fivethirtyeight does something like this using their ELO ratings.

EDIT: Just went to Google for a memory refresh, and stumbled on this article. Might be useless for your purposes, but it has some interesting ideas.

I'm not quite sure that's what I'm looking for. It may be too early for them to have future predictions, but I tend to do an 8 week look ahead in order to try and maximize points.

As an example, I start the season by ranking all the teams according to their odds to win the superbowl from three different online betting sites. All models have the Dolphins with the worst odds, then Redskins, Bengals, Giants, and Bills and Cardinals tied for 6th worst in the league. Typically that means I want to dump the Bills by week three or maybe 4 even if I don't get any points for picking them. However, they play the Dolphins in Buffalo in week 7. I can look up that game currently on ESPN and they give the Bills a 74% chance to win which is very good. So it makes sense for me not to dump the Bills early and save them for week 7. A bit risky, but worth it considering the stats.

I can't make those decisions without some data and ESPN is pretty good - but I'm guessing there has to be better out there. I've just never found it.
 
Joined
Aug 24, 2011
Messages
8,353
Likes
5,804
There's a lot of free stuff on this site, but if you want to get additional info, you can subscribe.

 

whaler11

Head Happy Hour Coach
Joined
Aug 27, 2011
Messages
44,107
Likes
64,478
I participate in a football pool every year where you have to pick two teams to win each week, no point spread, just pick the winner. Once you select a team, you can't pick them for the rest of the season. Week one correct picks are worth 1 point each. Week 2 are worth 2 points, etc through week 16 worth 16 points each and that exhausts all the potential teams. Obviously you have to get rid of bad teams early and count on big points for the final 8 or 10 weeks to have any chance of winning. This involves having to really plan for when your best matchups are.

I've been using the game predictor on ESPN for the last several years which gives win% for each team based on some algorithm. I was wondering if there are other sites that may have better models for future games. Any help is appreciated.
Take the money lines from Pinnacle.

Remove the juice.

Convert to win %.

Easy to google formulas/calculators.

But pretty much use Pinny $ lines over anything.
 

whaler11

Head Happy Hour Coach
Joined
Aug 27, 2011
Messages
44,107
Likes
64,478
There's a lot of free stuff on this site, but if you want to get additional info, you can subscribe.

whatever you do dont use garbage like this
 

whaler11

Head Happy Hour Coach
Joined
Aug 27, 2011
Messages
44,107
Likes
64,478
The NFL is the most liquid gambling market on the planet. The lines at legitimate sports books lay out the strength of every team.
 

jleves

Awesomeness
Joined
Aug 27, 2011
Messages
3,275
Likes
4,695
Take the money lines from Pinnacle.

Remove the juice.

Convert to win %.

Easy to google formulas/calculators.

But pretty much use Pinny $ lines over anything.
Interesting - how far forward do they have games?
 

whaler11

Head Happy Hour Coach
Joined
Aug 27, 2011
Messages
44,107
Likes
64,478
Interesting - how far forward do they have games?
You have to adjust every week starting with teasing out the relative power rankings
 

jleves

Awesomeness
Joined
Aug 27, 2011
Messages
3,275
Likes
4,695
You have to adjust every week starting with teasing out the relative power rankings
I tried to send this via PM, but apparently I can't send to you... anyhow:

I do update my spreadsheet weekly.

I'll ask you in a week or so how to 'remove the juice and change to winning %.'

Thanks for the input.
 

whaler11

Head Happy Hour Coach
Joined
Aug 27, 2011
Messages
44,107
Likes
64,478
Taking Pinny money lines and getting to a win percentage is easy. For your contest you want to think about it the day before your first pick is due.
 
Joined
Nov 18, 2012
Messages
2,199
Likes
2,594
I tried to send this via PM, but apparently I can't send to you... anyhow:

I do update my spreadsheet weekly.

I'll ask you in a week or so how to 'remove the juice and change to winning %.'

Thanks for the input.
Example: Patriots -750/ Broncos +604

The first step to remove vig from Patriots -750 / Broncos +604 is to calculate the implied probability for each team. We do this by using the formula risk/return=implied probability (were return is stake + win). So risk for -750: $750 to win $100 returns $850 ($750 stake + $100 win) and the math is 750/850=0.8824 (88.24%). Risk $100 on +604 and the return is $704 ($100 Stake + $604 win) and the math is 100/704=0.1420 (14.20%). Add our two percentages together 88.24%+14.2% and you’ll see these equal 102.44%. The extra 2.44% is the result of bookmaker advantage (vig), to remove it we just need to divide each implied probability by the overall percents market (in this case 102.44%). So 88.24/102.44=86.14% and 14.2/102.44=13.86%, as you can see these now total 100%, so the vig is removed.
 
Joined
Apr 24, 2018
Messages
495
Likes
483
So This guy uses the Yahoo public projection matchup percentages for wins and the ESPN calculation for his information ? That’s about as basic as 2+2 - what is he a high school gambler?
 

whaler11

Head Happy Hour Coach
Joined
Aug 27, 2011
Messages
44,107
Likes
64,478
Aye, oh, just remove da juice eh, forgetaboutit

Tough guy
lol - i know this topic inside and out and you are working off one of the dumbest sites on the internet. maybe shut your trap and learn something.
 
Joined
Aug 24, 2011
Messages
8,353
Likes
5,804
lol - i know this topic inside and out and you are working off one of the dumbest sites on the internet. maybe shut your trap and learn something.
Maybe you should be less of a douche
 

jleves

Awesomeness
Joined
Aug 27, 2011
Messages
3,275
Likes
4,695
Take the money lines from Pinnacle.

Remove the juice.

Convert to win %.

Easy to google formulas/calculators.

But pretty much use Pinny $ lines over anything.
So looking at the Pinnacle site, they only have a few days forward for games. Do they allow you to bet weeks out once the season starts? I really need to look ahead about 8 weeks at a time.
 

whaler11

Head Happy Hour Coach
Joined
Aug 27, 2011
Messages
44,107
Likes
64,478
So looking at the Pinnacle site, they only have a few days forward for games. Do they allow you to bet weeks out once the season starts? I really need to look ahead about 8 weeks at a time.
we just need to help you tease out power rankings from week 1 lines. a solid set of power rankings + the home field advantage values and youve got the lines for the whole season
 

jleves

Awesomeness
Joined
Aug 27, 2011
Messages
3,275
Likes
4,695
we just need to help you tease out power rankings from week 1 lines. a solid set of power rankings + the home field advantage values and youve got the lines for the whole season
Once again, I'm way less versed than you. How do you tease out the power rankings and how do you get home field advantage values?
 

whaler11

Head Happy Hour Coach
Joined
Aug 27, 2011
Messages
44,107
Likes
64,478
Once again, I'm way less versed than you. How do you tease out the power rankings and how do you get home field advantage values?
we’ll help you out. the home field advantages are pretty simple. goes from about Jacksonville 2 pts to Green Bay 3 pts.
 

jleves

Awesomeness
Joined
Aug 27, 2011
Messages
3,275
Likes
4,695
@methodology @whaler11
OK - I got my spreadsheet setup for figuring the Pinnacle site. If you're interest the formula looks like this:
=((100/(G19*100)))/((100/(H19*100))+(100/(G19*100)))
where G19 is 3.01 for Dolphins and H19 is 1.444 for the Ravens. According to Vegas, the Fins have a 32.42% chance of winning the game.

Now how do I go about predicting further into the season? Where do I find these home field advantages and 'solid' power rankings and exactly how do I apply them once I find them?
 
Joined
Aug 24, 2011
Messages
5,391
Likes
7,297
lol - i know this topic inside and out and you are working off one of the dumbest sites on the internet. maybe shut your trap and learn something.
lol. Yet another topic you know more than anyone on. You're awesome!
 
Top