The Half-Point Calculator and the value of the 1

[the_fredrik]

I have been looking at the half-point calculator with interest:
It seems that most pushes in basketball occurs between 3.5%-5% for every x-point margin except for the 1-point push in NBA that only occurs 2.27%! (In NCAA the 1-point push occurs 3.47%)

Can anyone come up with some reasonable explanation for this? Can it have something to do with the way teams play in the last minutes of a tight game??

[Munson15]

Not sure, but I know some smart guys feel that 2 and 3 are key numbers in baskets, because they fall within the margin of a one-possession game.

[MrX]

I would think that the high ratio of 3-point attempts vs 2-point attempts by the team trailing by 3 in the late stages of the game would significantly lower the likelihood of the game ending in a 1-point advantage.

[Data]

If a team is down by 3 with seconds to play they are not going to try for 2 but obviously will go for 3.

[donjuan]

Not sure, but I know some smart guys feel that 2 and 3 are key numbers in baskets, because they fall within the margin of a one-possession game.

They aren't very smart.

[raiders72002]

7 is the key number in the NBA. NBA teams give up when they are down around 7 very late in the game and the winning team just holds the ball and they won't be fouled.

Because the total in the NBA is so high it decreases the value of 1 point.

Say an unrealistic total of a game is 5000, you can see how the value of one point doesn't mean much. It's proportional to the total.

[raiders72002]

but I know some smart guys feel that 2 and 3 are key numbers in baskets,

NCAA is 2. Buy off/on 2 in the tourney.

[Munson15]

NCAA is 2. Buy off/on 2 in the tourney.

Tell DonJuan.

[donjuan]

NCAA is 2. Buy off/on 2 in the tourney.

You can say that until you are blue in the face but it is quite simply not true.

http://www.sbrforum.com/Betting+Tools/Half+Point+Calculator.aspx

[Data]

If a team is down by 3 with seconds to play they are not going to try for 2 but obviously will go for 3.

Another scenario. An underdog is down by 2 with seconds to play. With rare exeptions, they will go for 3. About 35% of these games will end with 1 point difference and 65% will end with 2 point difference.

[raiders72002]

You can say that until you are blue in the face but it is quite simply not true.

read it again. I said NCAA tourney time.

[donjuan]

read it again. I said NCAA tourney time.

Because college basketball games magically change their distributions come tourney time? Hilarious.

[donjuan]

As for the actual reason for 1 point wins being more common in basketball I would guess it has to do with poorer free throw shooting in college basketball than the NBA and a closer 3 point line. Just a guess though, and it really doesn't matter for handicapping purposes.

[raiders72002]

Because college basketball games magically change their distributions come tourney time? Hilarious

I don't like to bash people in the forums that we are suppose to be respectful in but you're an idiot.

Check out the corresponding ML to spread in the NFL during the SB. The correlation is much different then it is during the regular season.

There was a time when it was also this way during the NFL playoffs. Books use to disallow teasers during the playoffs but this has changed.

Watch some big name college football games where public money has much more effect on a line then sharp money. The correlation is different.

Getting back to the Super Bowl most of the time you'll see money on the favorites for the spread and money on the dogs for the ML.

For NCAA basketball games, public money has a greater effect then sharp money.

Quite honestly I have not run the stats on NCAA tourney spreads but I rely on people that are a hell of a lot smarter than I am and that's what they told me.

Some people share info with each other and each has their own area of expertise. I see that you lack in all of these areas if you think that you can equate all regular season games with playoff type games.

[raiders72002]

SBR - Forgive this rant but it really angers me when a piss-ant like donjuan comes into a serious thread with a mocking or bashing type of tone.

Everyone else in this thread was respectful and added an opinion. I may be dead wrong on the 2 but it was my opinion.

Others may be right or wrong but it's nice to see their opinion. I don't care if you are a beginner or a pro you have the right to post in these types of threads in a respectful manner.

Ganch is the best math guy that I've seen on any of the forums. There are also some great math guys here that don't post much.

Even though Ganch is great by his own admission he would say that these charts are somewhat generic to professional gamblers in extreme circumstances such as very high or low total along with other things posted previously.

Like I said, and I could be wrong a pro told me to buy 2. Pros work a little different then forums.

On a forum you break down and dissect games. It's fun to do. When a pro that you respect says take Pho -3 tonight. You say thanks.

There is a very small pool of tourney games lines up at 1.5, 2 and 2.5 that it would be very difficult for someone like Ganch to write a program determining the value of a 1/2 pt at 2.

I went with what a guy told me, right or wrong. The problem I have is some jackass bashing on something he knows nothing about.

[HedgeHog]

Personally I don't buy points in Basketball, but I keep an open mind. If college tourney action shows that 2 comes up frequently enough to warrant a 10 cent buy on/off then I would adjust my betting.

I agree that we that we can disagree in a more tolerant way.

[donjuan]

Check out the corresponding ML to spread in the NFL during the SB. The correlation is much different then it is during the regular season.

That's a well-known inefficiency in the market. The ML is not significantly lower because the distribution changes, but because the public loves to bet the dog ML in the Super Bowl.

Also, you claimed 7 was a key number in the NBA, which is amusing as out of numbers 6-10, 7 is the LEAST key.

You may get pissed off at what you claim is bashing, but I get annoyed when people make idiotic comments backed up not by anything besides thin air and am not afraid to point that out.

[LT Profits]

Because college basketball games magically change their distributions come tourney time?

Well...maybe!

Fact: Post-season games have a higher percentage of close games than regular season games. Now does that mean the distribution should be adjusted for post-season, or that key numbers exist in post-season? I will leave that up to you math gurus.

Since 2000, there have been 39,032 regular season games through last night (2/28/08), and the average winning margin regardless of pointspread has been +12.8 points. During the same time, there have been 712 post-season games (NCAA + NIT), and the AMOV is two full points less, at +10.8 points.

Now down below, I will list side-by-side the percentage of games there were decided by specific margins from 1 to 10 for regular season on the left and post season on the right. Take note of the increased percenatge of one-point games specifically in post-season and of games decided by five points or less in general:

Reg Season	39032	Post-Season	712
1            	4.38%	1            	6.32%
2            	5.45%	2            	5.90%
3            	6.28%	3            	6.46%
4            	5.18%	4            	5.20%
5            	5.32%	5            	6.04%
6            	4.83%	6            	4.35%
7            	4.89%	7            	4.21%
8            	4.81%	8            	6.32%
9            	4.53%	9            	5.62%
10          	4.49%	10          	6.04%

[Quebb Diesel]

As for the actual reason for 1 point wins being more common in basketball I would guess it has to do with poorer free throw shooting in college basketball than the NBA and a closer 3 point line. Just a guess though, and it really doesn't matter for handicapping purposes.

this happens with good ft shooting games too...

[donjuan]

Fact: Post-season games have a higher percentage of close games than regular season games. Now does that mean the distribution should be adjusted for post-season, or that key numbers exist in post-season? I will leave that up to you math gurus.

Another fact: teams are closer in ability during the post-season and the point spreads are, surprise, smaller. The actual percentage of the time that a game lands on a certain number is meaningless for handicapping; what's important is how often it pushes on numbers around the spread. What Raiders is suggesting is basically that games with point spreads between 1 and 3 land on 2 more frequently in the NCAA Tournament than during the regular season.

[donjuan]

this happens with good ft shooting games too...

Do you have data to show this?

[Quebb Diesel]

not off the top of the head...but ive seen plenty of games in my time where the ft's go down and the game is decided by a missed fg in the last seconds causing a 1 point win/loss

[donjuan]

Sure, but we are talking about the difference between 2.27% and 3.47%, which really can't just be eyeballed.

[Quebb Diesel]

i dont think you could really apply poor free throw shooting to nba games anyway...the lowest ft shooting % is above 70% and very rarely do the worst ft shooters get to the line in the last minute of the game. i would assume the reason 1 point games are so infrequent in the nba is the fg % of the trailing team in the final minute of the game.

[donjuan]

i dont think you could really apply poor free throw shooting to nba games anyway...the lowest ft shooting % is above 70% and very rarely do the worst ft shooters get to the line in the last minute of the game. i would assume the reason 1 point games are so infrequent in the nba is the fg % of the trailing team in the final minute of the game.

I think you have what I was saying backwards. NBA teams are better free throw shooters than college teams and the 3 point line is further from the hoop in the NBA. It's just a guess though, and pretty irrelevant to sports betting unless the reason behind it is the mentality of the coaches and that mentality changes in the future for some reason.

[raiders72002]

Don- You used Ganch's calculator as fact in determining percentages in CBB. You were dead wrong about them being the same come tourney time. No one was an ass to you because your statement was so wrong. That's the point I was trying to make.

[raiders72002]

I was wrong on some statements here because I was shooting from the top of my head. I'll come back later with what I have.

[raiders72002]

That's a well-known inefficiency in the market. The ML is not significantly lower because the distribution changes, but because the public loves to bet the dog ML in the Super Bowl.

The distribution does change to a high degree because of the human emotional element in a Super Bowl. Also look at the totals. The same game played during the regular season would have a lower total.

[LT Profits]

What Raiders is suggesting is basically that games with point spreads between 1 and 3 land on 2 more frequently in the NCAA Tournament than during the regular season.

Well, it is hard to get a meaningful sample size for post-season then. Since 2000, their have been 188 NCAA/NIT Tournament games with a favorite of between -1 and -3, and that team won by exactly two points just four times (2.13%).

FYI, here is a breakdown using games between PICK and -2 that land on 1, between -1 and -3 that land on 2, and between -2 and -4 that land on 3.

Regular	Games	Pct	Post	Games	Pct
1	3871	3.23%	1	116	2.59%
2	5415	3.99%	2	188	2.13%
3	5302	4.94%	3	179	2.23%

[raiders72002]

I do NFL and bases and a few games in other sports. The rest I just piggyback. I don't even have data for the other sports so a lot of my numbers may be wrong but I'm sticking with the 2.

I should say on to 2 but not off of 2.

[donjuan]

LT Profits,

It is, indeed, tough to get a meaningful sample size. Nonetheless, my point stands.

Raiders,

Don- You used Ganch's calculator as fact in determining percentages in CBB. You were dead wrong about them being the same come tourney time. No one was an ass to you because your statement was so wrong. That's the point I was trying to make.

I'm dead wrong about them being the same? What is it that magically changes during the NCAA Tournament?

The distribution does change to a high degree because of the human emotional element in a Super Bowl. Also look at the totals. The same game played during the regular season would have a lower total.

All you've done is convince me that certain Super Bowl lines are inefficent, not that the distribution of outcomes changes.

I do NFL and bases and a few games in other sports. The rest I just piggyback. I don't even have data for the other sports so a lot of my numbers may be wrong but I'm sticking with the 2.

I should say on to 2 but not off of 2.

And the pertinent question here is why?

[raiders72002]

don- If two teams are playing in the SB would the line be the same in a regular season game? No, you do the math on the rest.

Billy Walters buys on to 2. That's good enough for me. Maybe you're more profitable in sports betting then he is.

[donjuan]

don- If two teams are playing in the SB would the line be the same in a regular season game? No, you do the math on the rest.

As already noted, Super Bowl MLs are a well known inefficiency.

Billy Walters buys on to 2. That's good enough for me. Maybe you're more profitable in sports betting then he is.

Buying and selling points on and off certain points can always be correct, but it depends on the price you get as with anything in sports betting. If you let me buy onto 2 from 1.5 for 1 cent, of course I'll do it. If it costs 20 cents, I won't.

[raiders72002]

Buying and selling points on and off certain points can always be correct, but it depends on the price you get as with anything in sports betting. If you let me buy onto 2 from 1.5 for 1 cent, of course I'll do it. If it costs 20 cents, I won't.

do you even bet? The industry standard for buying 1/2 points is almost universal with the exception of 3 and 7 in football. Now I understand why you don't get it.

[donjuan]

do you even bet? The industry standard for buying 1/2 points is almost universal with the exception of 3 and 7 in football. Now I understand why you don't get it.

No, I don't bet. Gambling ruins your life and is very bad. Now get back to telling me how the 2 is a key number in the NCAA Tournament despite having zero evidence to back it up. I'd also love to hear again how the 7 is key in the NBA despite it pushing less than any number from 6 to 10.