Poll

2 votes (50%)
No votes (0%)
1 vote (25%)
1 vote (25%)
2 votes (50%)
1 vote (25%)
No votes (0%)
1 vote (25%)
2 votes (50%)
No votes (0%)

4 members have voted

Wizard
Administrator
Wizard
  • Threads: 1491
  • Posts: 26436
Joined: Oct 14, 2009
Thanked by
Rigondeauxsmoothgrh
October 13th, 2019 at 6:52:22 AM permalink
I noticed a prop bet at the Rampart yesterday on the first team to score a touchdown on Sunday, Oct 13, 2019. It was, of course, limited to the 10:00 games (PST). The following table shows what each game pays and my analysis of it. The winner is determined by the first touchdown on the game clock, not actual time. Lines posted here I assume were opening lines and may have changed.

The following table shows the spread, total, estimated points scored by that team, estimated touchdowns (subtracting six points for field goals for each team and then dividing by 7), share of total touchdown scored by that team, fair line, and expected value.

The fair line assumes that the probability of having the first touchdown is proportional to total expected touchdowns.

As you can see, this model shows a 22% advantage on the Bengals.

My question is what are your thoughts on this approach?

Team Pays Spread Total Exp Points Exp TD Prob First TD Fair line EV
Bengals 20 12 48 18 1.71 5.80% 16.25 21.74%
Ravens 6 -12 48 30 3.43 11.59% 7.63 -18.84%
Seahawks 11 2 47.5 22.75 2.39 8.09% 11.36 -2.90%
Browns 10 -2 47.5 24.75 2.68 9.06% 10.04 -0.36%
Texans 8 5.5 55.5 25 2.71 9.18% 9.89 -17.39%
Chiefs 5 -5.5 55.5 30.5 3.50 11.84% 7.45 -28.99%
Saints 10 -1 44 22.5 2.36 7.97% 11.55 -12.32%
Jags 10 1 44 21.5 2.21 7.49% 12.35 -17.63%
Eagles 11 3 43.5 20.25 2.04 6.88% 13.53 -17.39%
Vikings 8 -3 43.5 23.25 2.46 8.33% 11.00 -25.00%
Redskins 12 -3.5 40.5 22 2.29 7.73% 11.94 0.48%
Dolphins 12 3.5 40.5 18.5 1.79 6.04% 15.56 -21.50%
"For with much wisdom comes much sorrow." -- Ecclesiastes 1:18 (NIV)
SOOPOO
SOOPOO
  • Threads: 121
  • Posts: 10942
Joined: Aug 8, 2010
October 13th, 2019 at 7:14:48 AM permalink
Quote: Wizard

I noticed a prop bet at the Rampart yesterday on the first team to score a touchdown on Sunday, Oct 13, 2019. It was, of course, limited to the 10:00 games (PST). The following table shows what each game pays and my analysis of it. The winner is determined by the first touchdown on the game clock, not actual time. Lines posted here I assume were opening lines and may have changed.

The following table shows the spread, total, estimated points scored by that team, estimated touchdowns (subtracting six points for field goals for each team and then dividing by 7), share of total touchdown scored by that team, fair line, and expected value.

The fair line assumes that the probability of having the first touchdown is proportional to total expected touchdowns.

As you can see, this model shows a 22% advantage on the Bengals.

My question is what are your thoughts on this approach?

Team Pays Spread Total Exp Points Exp TD Prob First TD Fair line EV
Bengals 20 12 48 18 1.71 5.80% 16.25 21.74%
Ravens 6 -12 48 30 3.43 11.59% 7.63 -18.84%
Seahawks 11 2 47.5 22.75 2.39 8.09% 11.36 -2.90%
Browns 10 -2 47.5 24.75 2.68 9.06% 10.04 -0.36%
Texans 8 5.5 55.5 25 2.71 9.18% 9.89 -17.39%
Chiefs 5 -5.5 55.5 30.5 3.50 11.84% 7.45 -28.99%
Saints 10 -1 44 22.5 2.36 7.97% 11.55 -12.32%
Jags 10 1 44 21.5 2.21 7.49% 12.35 -17.63%
Eagles 11 3 43.5 20.25 2.04 6.88% 13.53 -17.39%
Vikings 8 -3 43.5 23.25 2.46 8.33% 11.00 -25.00%
Redskins 12 -3.5 40.5 22 2.29 7.73% 11.94 0.48%
Dolphins 12 3.5 40.5 18.5 1.79 6.04% 15.56 -21.50%



Are you NOT factoring in the defense of the opponent? Edit... I see you use the pointspread which includes both the offense and defense of the opponents. I surmise that you would need defense specific stats.

Also, a running team is way less likely to score first, as running plays gain fewer yards per minute of clock running.
Wizard
Administrator
Wizard
  • Threads: 1491
  • Posts: 26436
Joined: Oct 14, 2009
October 13th, 2019 at 7:26:34 AM permalink
Quote: SOOPOO

Also, a running team is way less likely to score first, as running plays gain fewer yards per minute of clock running.



Good point. It looks like the Bengals have 7 passing touchdowns and only 1 rushing. They are listed as 7th passing yards and 31st in rushing yards. Both would make me even more inclined to bet them. I still have 2.5 hours before kickoff.
"For with much wisdom comes much sorrow." -- Ecclesiastes 1:18 (NIV)
SOOPOO
SOOPOO
  • Threads: 121
  • Posts: 10942
Joined: Aug 8, 2010
October 13th, 2019 at 7:34:33 AM permalink
Quote: Wizard

Good point. It looks like the Bengals have 7 passing touchdowns and only 1 rushing. They are listed as 7th passing yards and 31st in rushing yards. Both would make me even more inclined to bet them. I still have 2.5 hours before kickoff.



This is what you need to do..... Check the last 100 weeks of NFL games. What is the average time to first TD if there are 6 games only included? I'll guess 2 minutes. Then see how many times each team this season has scored a TD in 2 minutes or less from the time they get offensive possession.. You might find the Chiefs have 9 and the Bengals 1.
This is a very interesting prop bet with I think way more complicated analysis than you are giving it.
Rigondeaux
Rigondeaux
  • Threads: 30
  • Posts: 2549
Joined: Aug 18, 2014
October 13th, 2019 at 8:15:44 AM permalink
It seems to me that your approach is good enough that if it suggests a 22% advantage, it's very likely that there is some advantage.
AZDuffman
AZDuffman
  • Threads: 240
  • Posts: 13886
Joined: Nov 2, 2009
October 13th, 2019 at 8:21:35 AM permalink
Quote: Wizard



The following table shows the spread, total, estimated points scored by that team, estimated touchdowns (subtracting six points for field goals for each team and then dividing by 7), share of total touchdown scored by that team, fair line, and expected value.



I am assuming "total" is the over/under total?

Well, not the approach I would use.

I would try to find out how the teams score in the first quarter only. Reason being that your approach does not take garbage points into account. Example there is Jets. They covered against the Pats, when it did not matter to the game results. Throws it off by 2 TDs. On a pathetic offense team like the Jets just 5 games in, that is a bunch. All bad teams are like this.

Quarter by quarter scoring is easy to get though time consuming. Then I would figure "possession efficiency." IOW, points scored per minute. IOW, if a team has the ball 30 minutes and gets 28 points, then they would be expected to get 15 points for 15 minutes of possession.

Then figure punts vs. score possession. Then figure TD to FG. This should scrub things enough to get an idea how long it will take to score a TD.

Then we need to make that a 50/50 thing to account for coin toss win/loss/decision.

In simper terms, this is not a bet I would spend time on doing a ton of math work. I would just take the 2-3 best offensive teams available. I would NOT take the Pats since hey like to choose kick not receive. Then play by "feel" taking into account things like weather.
All animals are equal, but some are more equal than others
TomG
TomG
  • Threads: 16
  • Posts: 2426
Joined: Sep 26, 2010
October 13th, 2019 at 8:54:36 AM permalink
Great start.

First impression: when only one sportsbook in the entire world puts up a line, it's very likely to be off and there should be an edge somewhere.
Second impression: 15% house edge is going to be very hard to beat

I really like the Bengals to score first in their game at +220. That means they are only about a 30-33% chance, to score before the Ravens. Let's set all TDs at random in all the games and make every team's true odds 12-1. Then make the Ravens a 67% favorite to score the first TD in their game. Doing that pushes Cincinnati to 18-1, which is still a nice 11% edge. But that doesn't account for eight of the other 10 teams having stronger offenses and that even if the Bengals do score first, it is more likely to be a field goal than if the Ravens score first.

Definitely think the 20-1 is a lot closer to true odds than -15% that this prop implies. If it was horse racing where they give huge comps, I would have no problem betting it. As it is, can't get too excited about it.
AZDuffman
AZDuffman
  • Threads: 240
  • Posts: 13886
Joined: Nov 2, 2009
October 13th, 2019 at 10:04:48 AM permalink
Quote: Wizard



As you can see, this model shows a 22% advantage on the Bengals.



BOOM!
All animals are equal, but some are more equal than others
SOOPOO
SOOPOO
  • Threads: 121
  • Posts: 10942
Joined: Aug 8, 2010
October 13th, 2019 at 10:06:28 AM permalink
So did you get that $5,000 bet down like you were planning to?
michael99000
michael99000
  • Threads: 9
  • Posts: 2113
Joined: Jul 10, 2010
October 13th, 2019 at 10:20:50 AM permalink
Not too many sports bets cash in 12 seconds
7craps
7craps
  • Threads: 18
  • Posts: 1977
Joined: Jan 23, 2010
October 13th, 2019 at 11:07:39 AM permalink
Quote: Wizard

They are listed as 7th passing yards and 31st in rushing yards.

the reason for that should be obvious. They are almost always behind in the games and losing and are 0-5 and NEED to pass the ball to try and catch up.

Quote: Wizard

Both would make me even more inclined to bet them.

This makes NO sense at all.

Is this from the same person that always bets AGAINST a safety in the super bowl?
2 years in a row I bet the 1st score would be a safety and won both. just lucky
winsome johnny (not Win some johnny)
sodawater
sodawater
  • Threads: 64
  • Posts: 3321
Joined: May 14, 2012
October 13th, 2019 at 11:11:28 AM permalink
Nice hit at 20-1, Wizard. Hope you got some money down.

One correction to your post, though. The games are 10:00 PDT. Vegas will not be in PST until next month.
Wizard
Administrator
Wizard
  • Threads: 1491
  • Posts: 26436
Joined: Oct 14, 2009
October 13th, 2019 at 12:17:31 PM permalink
Quote: SOOPOO

This is what you need to do..... Check the last 100 weeks of NFL games. What is the average time to first TD if there are 6 games only included? I'll guess 2 minutes. Then see how many times each team this season has scored a TD in 2 minutes or less from the time they get offensive possession.. You might find the Chiefs have 9 and the Bengals 1.
This is a very interesting prop bet with I think way more complicated analysis than you are giving it.



Those are good suggestions, but I'm not going to analyze it to death for one prop bet.
"For with much wisdom comes much sorrow." -- Ecclesiastes 1:18 (NIV)
Wizard
Administrator
Wizard
  • Threads: 1491
  • Posts: 26436
Joined: Oct 14, 2009
October 13th, 2019 at 12:19:18 PM permalink
Quote: sodawater

Nice hit at 20-1, Wizard. Hope you got some money down.



Unfortunately, I was lazy and never went.

Quote:

One correction to your post, though. The games are 10:00 PDT. Vegas will not be in PST until next month.



Good catch.
"For with much wisdom comes much sorrow." -- Ecclesiastes 1:18 (NIV)
SOOPOO
SOOPOO
  • Threads: 121
  • Posts: 10942
Joined: Aug 8, 2010
Thanked by
smoothgrhAyecarumba
October 13th, 2019 at 12:28:11 PM permalink
Quote: Wizard

Those are good suggestions, but I'm not going to analyze it to death for one prop bet.



I didn't actually expect you to. As we saw today, there are special team touchdowns. And effectively half the teams are eliminated by coin toss. There are really to many variables to do a 'perfect' analysis of this type of bet.

I truly mean this. The fact that you spent a few hours doing what you did for this obscure prop bet is what makes you who you are. And this a great site.
Johnzimbo
Johnzimbo
  • Threads: 4
  • Posts: 1140
Joined: Sep 29, 2010
October 13th, 2019 at 12:40:56 PM permalink
I think that South Point has been doing this prop bet but I might be wrong. I recall reading about it but I might have the wrong casino in mind
Rigondeaux
Rigondeaux
  • Threads: 30
  • Posts: 2549
Joined: Aug 18, 2014
October 13th, 2019 at 12:41:39 PM permalink
I think most of these suggestions are splitting hairs/false precision. I wouldn't worry too much about predicting exactly when teams are going to score based on what happened in 6 prior games. Especially in football, where so much random crap happens.

Probably the only thing that I'd really consider of the above is if the team always defers on the flip and their opponent always receives.

There is a lot of intelligence baked into the team totals. Also, the wiz model looked about what you'd expect for all the other teams. It wasn't spitting out all kinds of weird numbers.

As I said, if the totals suggest a 22% edge it is very likely to be good. It could even be better.

Tom makes an important point about the fact that this is a unique line at a relatively small book.
Johnzimbo
Johnzimbo
  • Threads: 4
  • Posts: 1140
Joined: Sep 29, 2010
October 13th, 2019 at 12:50:21 PM permalink
I verified, South Point has been doing this for a few weeks. They also have a prop on which team will be the highest scoring team on Sunday
RS
RS
  • Threads: 62
  • Posts: 8626
Joined: Feb 11, 2014
October 13th, 2019 at 12:55:00 PM permalink
I don't think there's anything particularly wrong about assuming the first TD is proportional to total expected TDs. I'm not saying that's necessarily 'right', but that it does logically follow. There are obviously other factors that I'm not super smart on -- but for example if the bet was who is to score last in the 2nd quarter, Patriots would probably be lock of the century millennium....that sorta thing. Some teams may prefer to receive in the first half while others may prefer to receive in the second half.

Looking at it more pragmatically, I think that's the right word, what does 20-to-1 odds translate to in a probability if that's a fair line? It'd be 1/21 or 4.76%. Your model puts it at 5.8%. Unless my model has shown success already, my assumption is the casino's model is more accurate than mine. That being said, do I want to bet a perhaps large amount of money on a 1% difference in "opinion"?

I'm sure you could also do this analysis 10 different ways and have 10 different results all wildly different from each others. Your model could be based on first half spread and total or first quarter spread and totals -- those are (almost assuredly) going to have different outputs on the model. Someone else earlier in the thread wrote about analyzing about TDs within the first 2 minutes. And probably several others I can't think of right now.


I think it'd be interesting to see a side by side by side of what the casino projected vs what you projected vs what actually happened. EG:


Casino | Model | Reality
Chiefs | Chiefs | Bengals
Ravens | Ravens | whoever got TD 2nd
Vikings/Texans | Texans | 3rd team
Vikings/Texans | Browns | 4th team
etc.


If it isn't overly time consuming and if you have access to it, perhaps try doing the same analysis with 1Q and/or 1H bets, although that's definitely going to be more skewed since the lines (likely?) aren't going to be as accurate from the casino as well as just the complexity in dealing with those numbers -- since first half generally aren't all going to be -110 lines.
Wizard
Administrator
Wizard
  • Threads: 1491
  • Posts: 26436
Joined: Oct 14, 2009
October 13th, 2019 at 2:22:07 PM permalink
Quote: SOOPOO

I truly mean this. The fact that you spent a few hours doing what you did for this obscure prop bet is what makes you who you are. And this a great site.



Thanks SOOPOO! I actually spent only about five minutes on the math of that.
"For with much wisdom comes much sorrow." -- Ecclesiastes 1:18 (NIV)
SOOPOO
SOOPOO
  • Threads: 121
  • Posts: 10942
Joined: Aug 8, 2010
October 13th, 2019 at 2:32:10 PM permalink
Quote: Wizard

Thanks SOOPOO! I actually spent only about five minutes on the math of that.



Mike, this is why you don't attract women! Brutal honesty does not always work. The correct response was.....

Yes, the three hours I spent were well worth it! Anything to help the loyal members of the forum.
gordonm888
Administrator
gordonm888
  • Threads: 60
  • Posts: 5005
Joined: Feb 18, 2015
October 13th, 2019 at 3:32:55 PM permalink
I think you've identified a reasonably valid approach to analyzing this prop bet. There are objections I could make based on 3-pt field goals, etc. but I think you main idea is fine for identifying large edges.

My only caveat: -if a team is making a quarterback switch to start the game then I would tend to not pick them -for example, Sam Darnold was starting for the Jets for the 1st time in a month, thus I would not pick the Jets because teams tend to be conservative (calling running plays, etc.) for the first few plays/series with a "new" quarterback. The same applies to Hodges starting a game at QB for the first time for the Steelers -that implies conservative play calling on offense for the first few series of the game.
Last edited by: gordonm888 on Oct 13, 2019
So many better men, a few of them friends, are dead. And a thousand thousand slimy things live on, and so do I.
smoothgrh
smoothgrh
  • Threads: 86
  • Posts: 1251
Joined: Oct 26, 2011
October 14th, 2019 at 10:20:24 AM permalink
Quote: SOOPOO

The fact that you spent a few hours doing what you did for this obscure prop bet is what makes you who you are. And this a great site.



I second this!

If a random buddy decided to take on this prop bet and won it, I'd be amazed! I'm still amazed that you predicted it, but it's more like being not surprised when a pro golfer sinks a 10-foot putt.
  • Jump to: