[epikk]

How can I calculate points expectancy in a game ? (various sports)

Is there any golden rule/method to do it? I scanned through this forum and I couldn't find any good thread regarding this. A lot of guys were asking how to convert moneyline to spread but wrong thinking.
The question should be, how to convert data to totals as it is something completely separate from win probability.
And that's why I'm here - is there anyone with knowledge who could show us different methods of such calculations?



Thanks!

[Waterstpub87]

There is no simple formula. If there was, sportsbooks wouldn't exist, and certainly wouldn't give out bonuses.

[Bsims]

I think I understand your question and the answer is relatively easy for football and basketball. In these sports, generally the spread and totals are given lines of -110 each side. This says there is a 50-50 chance of each team covering and the totals going over or under. In this case, it's realtively simple. Divide both the spread and totals by 2 and add half the spread for the favorites and subtract half for the dogs.

So yesterday in the WNBA, LA was -12 over Dallas and the total was 172. Divide the 172 by 2 giving 86, then add 6 (12/2) to LA and subtract 6 from Dallas. The predicted scores from this line is LA 92, Dallas 80.

Now it gets a lot more complicated when the lines are not -110, like in baseball.

Is that the question you were asking?

[tsty]

stop wasting your time with totals

they are useless

only bums bet totals

[Mike1029]

I try and stay away from totals in general. I never seem to be on the the right side of the ov/un.

[dimaggio8]

I try and stay away from totals in general. I never seem to be on the the right side of the ov/un.

Me too Mike and I've been playing for years. I usually try to stay on teams. I am sure there are a few out there who know the good tricks of over/unders and which lines are good to take advantage of.

[Waterstpub87]

Probably 60%+ of my action across all sports are totals. Lines are much less efficient when it comes to totals, especially in college sports.

[KVB]

I've posted a few times about why Totals are inherently softer markets and have even posted evidence.

This post came to mind so I searched for it...

Here are the forecast results using only a half point discrepancy with the market…

Against the TOTALS:
Open: 59-36 (62.11%), Close 60-39 (60.61%)…


…You have heard that Totals can be softer markets, especially NCAAB Totals. There is evidence in the line movement as well as the results above....

As far as the original question, there is more than one way to handicap. Some combination of past and recent performance will get one so far and then factors can be used to sharpen the prediction.

This post offers a little something, even though the play lost, for a starting point..

https://www.sportsbookreview.com/for...l#post23787665

Understand that books often have lower max bets on Totals than they do on the side. Books will also limit you faster if you are even beating the closer on your Total bets.

Also, know that the lines move, even in the Pros, significantly more for Totals than for sides. While this may be less so of a late, I believe the difference is still prevalent.

Less max bets, more line movement, and a sharp long term forecast exposing Totals more than spreads are evidence that Totals are a softer market than sides.

[KVB]

Here are bookmaker limits. Notice the Football and Baseball limits.

https://www.bookmaker.eu/wager-limits

For the NFL, there's $10,000 limit on the opening sides, but it's $3,000 for opening Totals.

On game day, there's a $50,000 limit on sides but it's $20,000 on Totals.

For overnight NCAAB lines, big bad Bookmaker will only take $500 bets.

There's a reason Totals are so much less.

If you scroll down the link, you'll find that only the NHL has the same limits for sides, totals, and moneylines. They probably thirst for action.

The bottom line is, and has been since before there was an internet, that the books are a bit more scared of sharp Total bettors than anyone else in the book.

[epikk]

I'm not sure we understand each other here. Maybe it's because my English is a bit limited.

Each book for every sport has to ingredients: Moneyline (or match result) and point expectancy. From that you derive all other markets and that's including handicaps. That's often called Supremacy and Totals.

I thought that this forum holding a name "handicapper think tank", would have this thing sorted out long time ago, mathematically speaking.
After all what's the point of using Kelly Criterion if your estimation is biased (or taken from the sky) - instead of having an excel file with scraping statistics and math formula? To use Kelly (which appears to be popular here) you need to have your own estimation based on mathematical calculation...

So my question is - have any of you built mathematical model for estimating point expectancy in any sport?

[KVB]

...So my question is - have any of you built mathematical model for estimating point expectancy in any sport?

The answer is yes and a link to another post I put above explains one way to start. Use a weighted combination of recent and longer statistics to get a base start. You can come across market openers just doing that. But it's not enough. You need to enhance that prediction with other relevant factors.

The key here is "relevant factors." You need to make a line that is closer to reality than the book is able to or is willing to post. You then exploit the difference. I gave one suggestion in that link regarding the level of play of the teams facing each other.

For the Think Tank to have it "sorted out" makes little sense. The line you make must be dynamic. You have to be able adjust as the market and sport progresses. To publicly post one successful formula would jeopardize the very success of that formula.

Like WaterSt. said in post #2...

There is no simple formula. If there was, sportsbooks wouldn't exist, and certainly wouldn't give out bonuses.

[gojetsgomoxies]

grunching...

but i thought OP was asking if there is some sort of mathematical formula - or heuristic sp? - for coming up with an approx total for two teams given their points for and against AND assuming equivalent schedule so far.

we had a good thread where OP asked if there is formula for win % of 70% win team vs. 50% win team - again same schedule - and there was.. now that doesn't mean the formula is correct IRL. think of the formula as a die where the rolls are based on those winning %'s.

[Waterstpub87]

No, apparently not. He apparently wants a ready made formula to project the totals in the game. Based on statistics and data, not through pricing it via moneyline and spread.

[KVB]

...but i thought OP was asking if there is some sort of mathematical formula - or heuristic sp? - for coming up with an approx total for two teams given their points for and against AND assuming equivalent schedule so far...

Here...lol...

...Anyone can take the season average offense and defense scoring for these two teams, add them all up, divide by two, and viola you have a crude, basic total...

That's as basic as it gets and believe it or not, sometimes the market does little more than that when lines are created and maintained. Sure times have changed and a lot of subjective input has been hardened into what will adjust the line, so the line will move more often than it used to.

Every sport is different, and one can get very detailed in trying to predict the number of possessions and potential scores per possession or the number of hits or shots on goal.

But all these factors have base lines that change. Even the basic one above doesn't offer much unless you start comparing the teams to the league.

I'm saying use weighted averages so that recent performance is calculated into the prediction. Those who stick just to season averages will fall behind quickly.

It may or may not provide value in getting into micro stats and estimated players time on the court, etc. but I wouldn't go around the block to get next door...that too can be a mistake.

[BigdaddyQH]

In College Football, generally speaking, the Total posted on Sunday in Vegas drops during the week, and then climbs back to the original number or very close to it by game time. Of course there are exceptions to this rule, as there are exceptions to any rule in gaming, but generally speaking, this is the way it goes. So if you determine that a total is a point or so too high, just give itsome time and it will probably drop.

[ChuckyTheGoat]

Here are bookmaker limits. Notice the Football and Baseball limits.

https://www.bookmaker.eu/wager-limits

For the NFL, there's $10,000 limit on the opening sides, but it's $3,000 for opening Totals.

On game day, there's a $50,000 limit on sides but it's $20,000 on Totals.

For overnight NCAAB lines, big bad Bookmaker will only take $500 bets.

There's a reason Totals are so much less.

If you scroll down the link, you'll find that only the NHL has the same limits for sides, totals, and moneylines. They probably thirst for action.

The bottom line is, and has been since before there was an internet, that the books are a bit more scared of sharp Total bettors than anyone else in the book.

Agree w/ this sentiment. Posted Totals TEND to be not as sharp.

But Totals also (arguably) harder to handicap. And Totals seem to have higher variance. So, u have to shoot for a bigger edge and deals w/ the fluctuations. Not ez.

[DegenKing]

Lots of good reading in this thread - thanks fellas

[RudyRuetigger]

what a dumb question

given the spread and total obviously you can predict the totals for each team based on the lines

anything else you want to know is like asking what will be printed in tomorrows paper

[epikk]

In College Football, generally speaking, the Total posted on Sunday in Vegas drops during the week, and then climbs back to the original number or very close to it by game time. Of course there are exceptions to this rule, as there are exceptions to any rule in gaming, but generally speaking, this is the way it goes. So if you determine that a total is a point or so too high, just give itsome time and it will probably drop.

That's because bookies react to initial turnover on the game. If they get hammered on moneyline everything else will move as well. The most accurate spread line is the one just before kick-off.



Rudy: It's not a dumb question mate. Unless you are not interested in professional betting and you base your bets on a random chat with the lads.

- Oh Dave I think it should be -4.5 on that game
- rly why?
- my daughter turns 4.5 today.

Oh and Rudy - Spreads come from Totals - not the other way around.

Of course nobody want's give away winning formula, but I had in mind that kind of model which would point out obvious mistake in spreads - not the most accurate calculation in the world but one that would give decent results based on recent stats. And there is no such website which would provide it on a daily basis. Interesting because there are countless websites with value calculators, blogs and stuff - nothing about suggested odds with self-check.

And remember that most of the bookies probably use the same model. Especially when they have the same odds provider.

[tsty]

spreads come from totals? lol heard it here first

[tsty]

Agree w/ this sentiment. Posted Totals TEND to be not as sharp.

But Totals also (arguably) harder to handicap. And Totals seem to have higher variance. So, u have to shoot for a bigger edge and deals w/ the fluctuations. Not ez.

Higher variance? that's just stupid

If it's easier and you can beat it then the variance will be less...

[Bsims]

given the spread and total obviously you can predict the totals for each team based on the lines

Granted this is generally true in basketball and football. But what about baseball. Can you calculate the expected total for each team given the money line and total (plus it's lines)?

[evo34]

stop wasting your time with totals

they are useless

only bums bet totals

Bums who want to win...

[evo34]

That's because bookies react to initial turnover on the game. If they get hammered on moneyline everything else will move as well. The most accurate spread line is the one just before kick-off.



Rudy: It's not a dumb question mate. Unless you are not interested in professional betting and you base your bets on a random chat with the lads.

- Oh Dave I think it should be -4.5 on that game
- rly why?
- my daughter turns 4.5 today.

Oh and Rudy - Spreads come from Totals - not the other way around.

Of course nobody want's give away winning formula, but I had in mind that kind of model which would point out obvious mistake in spreads - not the most accurate calculation in the world but one that would give decent results based on recent stats. And there is no such website which would provide it on a daily basis. Interesting because there are countless websites with value calculators, blogs and stuff - nothing about suggested odds with self-check.

And remember that most of the bookies probably use the same model. Especially when they have the same odds provider.

Odds originators are not going to have their models publicly available, so I have no idea what you are looking for. The only way to check your own model by testing it -- first on historical data and then live.

[ericc]

I think he is asking, can you come up with a number by adding points scored (team 1 + team 2), points allowed (team 1 + team 2) and then divide by 2, or some formula like that.

[Auto Donk]

[jtoler]

I think I understand your question and the answer is relatively easy for football and basketball. In these sports, generally the spread and totals are given lines of -110 each side. This says there is a 50-50 chance of each team covering and the totals going over or under. In this case, it's realtively simple. Divide both the spread and totals by 2 and add half the spread for the favorites and subtract half for the dogs.

So yesterday in the WNBA, LA was -12 over Dallas and the total was 172. Divide the 172 by 2 giving 86, then add 6 (12/2) to LA and subtract 6 from Dallas. The predicted scores from this line is LA 92, Dallas 80.

Now it gets a lot more complicated when the lines are not -110, like in baseball.

Is that the question you were asking?

The bold is a mistake gamblers assess to over/unders, spreads etc. It is not a 50-50 chance of each team covering and/or total going over or under.

[maggiethebestdog]

You may be able to get a general idea of points scored in any game because the lines are so accurate, but to think you can develop some system to help you win games is insane. If it was that easy....

[Bsims]

The bold is a mistake gamblers assess to over/unders, spreads etc. It is not a 50-50 chance of each team covering and/or total going over or under.

Just looked at some past NFL games, 3,560 finished over while 3,625 were under. That's 49.5% over and 50.5% under. You're right, nowhere near 50-50. What data do you have?

[tsty]

The bold is a mistake gamblers assess to over/unders, spreads etc. It is not a 50-50 chance of each team covering and/or total going over or under.

You're wrong unless you are trying to say that odds are not always correct in some stupid way.

[jtoler]

Just looked at some past NFL games, 3,560 finished over while 3,625 were under. That's 49.5% over and 50.5% under. You're right, nowhere near 50-50. What data do you have?

I was responding only to exactly what you said since this isnt coin flipping.

[jtoler]

You're wrong unless you are trying to say that odds are not always correct in some stupid way.

Im paid to be a jerk.

[evo34]

The bold is a mistake gamblers assess to over/unders, spreads etc. It is not a 50-50 chance of each team covering and/or total going over or under.

100% pure trolling (not from concentrate).

[gojetsgomoxies]

Here...lol...



That's as basic as it gets and believe it or not, sometimes the market does little more than that when lines are created and maintained. Sure times have changed and a lot of subjective input has been hardened into what will adjust the line, so the line will move more often than it used to.

Every sport is different, and one can get very detailed in trying to predict the number of possessions and potential scores per possession or the number of hits or shots on goal.

But all these factors have base lines that change. Even the basic one above doesn't offer much unless you start comparing the teams to the league.

I'm saying use weighted averages so that recent performance is calculated into the prediction. Those who stick just to season averages will fall behind quickly.

It may or may not provide value in getting into micro stats and estimated players time on the court, etc. but I wouldn't go around the block to get next door...that too can be a mistake.

this is completely wrong... a 70% win team does not have a 70% chance of beating a 30% win team. the % is higher. your math on points assumes that would be analogously correct. it is not.

[KVB]

this is completely wrong... a 70% win team does not have a 70% chance of beating a 30% win team. the % is higher. your math on points assumes that would be analogously correct. it is not.

I have no idea what this post means.

I don't know what it means to be a 70% win team, unless you mean they've won 7 out 10 games they've played. I don't use win-loss records when making my Totals forecast.

This thread is about Totals, and the OP was looking for a basic formula. There are as many ways to handicap as there are handicappers. I gave the most basic starting point and suggestions to build upon it.

Adding together and dividing by two is hardly much math. I don't recall posting any other math. The link I posted offers another suggestion for a direction to look.

Then again, I don't recall anyone else posting any sort of way to arrive at a Total forecast in this thread, even if to give the OP a very basic start.

You can use the offered lines to make a judgement on win percentage and Totals, but we already know the market isn't really as efficient as it is often claimed.

This is why making your own forecast that is closer to reality than the book wants to or can offer is the best way to judge efficiency and win long term. That is, making your own line before ever seeing the opening line.

Also, in a sport like the NBA, don't neglect the possibility of making a spread forecast and Totals forecast independent of each other, making them completely separate. This can help the bettor avoid the influence of the originators, who understand how one can be derived from the other; and thus influence the apparent efficiency of the market.