Value of 1/2 pt in basketball

David · 03-13-06, 04:54 PM

Depends on the line. Pinnacle's drop downs give a good idea of what a half point is worth.

pags11 · 03-13-06, 05:28 PM

I lost by a half of a point 15 times this year...that was using the free half point...probably won or pushed with the free half point 20 times...out of about 440 plays this year...

Ganchrow · 03-13-06, 10:56 PM

Originally posted by Dark Horse

If you know, how many basketball games have you lost this season by 1/2 point (divided into sides and totals)? And out of how many total basketball wagers?

Trying to get a percentage of what a half pt is really worth. I know someone posted a link before with numbers that showed the percentage between 3 and 4%, but I have a feeling it may be higher.

There have been at least a couple of threads related to this topic over the past month:

-and-

But anyway, using Covers data and offering absolutely no guarantees of any sort regarding accuracy of either my data culling process or of my calculations, here are sides for the 2005-06 season to date:

Real pushes ATS:

Code:

SPREAD	PUSHES	TOTAL	%
1	1	47	2.13%
2	0	51	0.00%
3	1	51	1.96%
4	4	73	5.48%
5	7	59	11.86%
6	2	53	3.77%
7	0	34	0.00%
8	2	43	4.65%
9	1	30	3.33%
10	0	22	0.00%
11	0	16	0.00%
12	0	12	0.00%
13	0	2	0.00%
14	0	4	0.00%
15	0	6	0.00%
16	0	2	0.00%
TOTAL	18	505	3.56%

Half-point dog losses ATS:

Code:

SPREAD	LOSSES	TOTAL	%
1.5	0	24	0.00%
2.5	1	52	1.92%
3.5	1	64	1.56%
4.5	2	58	3.45%
5.5	3	43	6.98%
6.5	0	46	0.00%
7.5	1	43	2.33%
8.5	0	24	0.00%
9.5	0	26	0.00%
10.5	1	15	6.67%
11.5	2	16	12.50%
12.5	0	5	0.00%
13.5	0	3	0.00%
14.5	1	3	33.33%
15.5	0	2	0.00%
TOTAL	12	424	2.83%

Half-point fave losses ATS:

Code:

SPREAD	LOSSES	TOTAL	%
1.5	0	24	0.00%
2.5	3	52	5.77%
3.5	1	64	1.56%
4.5	3	58	5.17%
5.5	0	43	0.00%
6.5	1	46	2.17%
7.5	1	43	2.33%
8.5	1	24	4.17%
9.5	5	26	19.23%
10.5	1	15	6.67%
11.5	0	16	0.00%
12.5	0	5	0.00%
13.5	0	3	0.00%
14.5	0	3	0.00%
15.5	0	2	0.00%
TOTAL	16	424	3.77%

I'll see if I can't coax the totals data out of my database later.

Jay Edgar · 03-13-06, 11:03 PM

Maybe I'm seeing in this data what I want to see, but for me it confirms a mantra:

if you like a favorite and are able to give no more than 5, feel better,

if you like a dog and can get at least 5, feel better

Ganchrow · 03-14-06, 12:22 AM

Originally posted by Jay Edgar

Maybe I'm seeing in this data what I want to see, but for me it confirms a mantra:

if you like a favorite and are able to give no more than 5, feel better,

if you like a dog and can get at least 5, feel better

Sorry Counselor, but even if I could figure out where within the data you found confirmation of this, I'd still caution you not to jump to too many conclusions on the basis of so small a sample size.

pags11 · 03-14-06, 12:25 AM

god damn ganchrow, you are good...if you ever come to Reno I'm buying you dinner...you the man...

Ganchrow · 03-14-06, 12:38 AM

For 2005-06 totals to date:

11 real pushes out of 497 integer totals (2.21%)
10 half point over losses out of 438 non-integer totals (2.28%)
12 half point under losses out of 438 non-integer totals (2.74%)

Dark Horse · 03-14-06, 03:28 AM

Originally posted by ganchrow

Real pushes ATS:

Code:

SPREAD	PUSHES	TOTAL	%
1	1	47	2.13%
2	0	51	0.00%
3	1	51	1.96%
4	4	73	5.48%
5	7	59	11.86%
6	2	53	3.77%
7	0	34	0.00%
8	2	43	4.65%
9	1	30	3.33%
10	0	22	0.00%
11	0	16	0.00%
12	0	12	0.00%
13	0	2	0.00%
14	0	4	0.00%
15	0	6	0.00%
16	0	2	0.00%
TOTAL	18	505	3.56%

Thanks ganchrow. Good stuff. Basketball has such a small maring for error. Small sample size notwithstanding, it's probably no coincidence that a huge percentage of the real pushes (13 of 18) hit at medium size spreads 4, 5, and 6. If that would hold up with greater sample size those would be the type of wagers to play at skybook...

Dark Horse · 03-14-06, 03:32 AM

[QUOTE=ganchrow] Half-point dog losses ATS:

Code:

SPREAD	LOSSES	TOTAL	%
1.5	0	24	0.00%
2.5	1	52	1.92%
3.5	1	64	1.56%
4.5	2	58	3.45%
5.5	3	43	6.98%
6.5	0	46	0.00%
7.5	1	43	2.33%
8.5	0	24	0.00%
9.5	0	26	0.00%
10.5	1	15	6.67%
11.5	2	16	12.50%
12.5	0	5	0.00%
13.5	0	3	0.00%
14.5	1	3	33.33%
15.5	0	2	0.00%
TOTAL	12	424	2.83%

Not sure if 424 games is a small sample size. Betting dogs, 6 of 12 losses by 1/2 point came for mov's 4, 5, and 6... (3.5, 4.5, and 5.5 lost). The same three numbers. Hmm.

imgv94 · 03-14-06, 03:35 AM

Ganchrow you are amazing! You are awesome. Thanks for this info bro.

Dark Horse · 03-14-06, 03:37 AM

Originally posted by ganchrow

Half-point fave losses ATS:

Code:

SPREAD	LOSSES	TOTAL	%
1.5	0	24	0.00%
2.5	3	52	5.77%
3.5	1	64	1.56%
4.5	3	58	5.17%
5.5	0	43	0.00%
6.5	1	46	2.17%
7.5	1	43	2.33%
8.5	1	24	4.17%
9.5	5	26	19.23%
10.5	1	15	6.67%
11.5	0	16	0.00%
12.5	0	5	0.00%
13.5	0	3	0.00%
14.5	0	3	0.00%
15.5	0	2	0.00%
TOTAL	16	424	3.77%

I'll see if I can't coax the totals data out of my database later.

Two areas where favorites lose by 1/2 pt. Spreads under 5 for close games (7 of 16), and spreads around the number 10 -mental boundary that many teams keep as a guideline to protect their lead?- for 6 of 16.

More numbers needed, but some very interesting early patterns.

Dark Horse · 03-14-06, 03:41 AM

To summarize (based on medium sized sample size of 400-500+ games):
-It pays to get a free half point for whole spreads 4, 5, and 6.
- It also pays to get a free half point for dogs +3.5, 4.5, and 5.5,
- For favorites, the free half point is helpful for spreads under 5, and for spreads around 10.

May get a small skybook account just for these numbers. Thanks ganchrow.

Ganchrow · 03-14-06, 05:43 AM

Originally posted by Dark Horse

To summarize (based on medium sized sample size of 400-500+ games):
-It pays to get a free half point for whole spreads 4, 5, and 6.
- It also pays to get a free half point for dogs +3.5, 4.5, and 5.5,
- For favorites, the free half point is helpful for spreads under 5, and for spreads around 10.

A note of caution: I presented this data as I did specifically to answer your original question as to how frequently NBA bets lost by ½ point this season. This was NOT meant to be used as posted as a predictive tool. If you really wanted to use it in such a manner it would make considerably more sense to consider the three tables I posted as a whole and not individually.

Even better than that would be to base your decisions on considerably more than just one partial season's worth of data. If you didn't want to compile and sort though it all yourself (and don't look to me on that -- I do still have to keep at least some of work product proprietary) you could do a lot worse than using turbobet's NBA Half Point Chart.

imgv94 · 03-14-06, 05:46 AM

Mr:Ganchrow? Who is that man on your avatar? You had Don Rickels
earlier.

Ganchrow · 03-14-06, 05:59 AM

Originally posted by imgv94

Mr:Ganchrow? Who is that man on your avatar? You had Don Rickels
earlier.

The great magician, skeptic, and debunker of psuedoscience James Randi. He's currently recovering from bypass surgery. We nonbelievers wish him well.

Dark Horse · 03-14-06, 01:22 PM

Originally posted by ganchrow

A note of caution: I presented this data as I did specifically to answer your original question as to how frequently NBA bets lost by ½ point this season. This was NOT meant to be used as posted as a predictive tool. If you really wanted to use it in such a manner it would make considerably more sense to consider the three tables I posted as a whole and not individually.

Even better than that would be to base your decisions on considerably more than just one partial season's worth of data. If you didn't want to compile and sort though it all yourself (and don't look to me on that -- I do still have to keep at least some of work product proprietary) you could do a lot worse than using turbobet's NBA Half Point Chart.

The parameters I identified made sense to me, but I never said I'd chisel them into something for Moses to carry down the mountain.

Didn't trust the other chart, based on the iffy football numbers.

Chi_archie · 03-16-10, 11:10 AM

how would I figure out the potential long term (2-3 season) rate of middling by trying to beat the closer at one book with a free half point, and taking the opposite side at another book with a free half point, hopefully at an optimal line.

for example get Kansas at -5 during the overnight with a free half point bringing it to -4.5 and getting Kentucky +6 with the free half point to +6.5 what are my chances of middling with a 5 or 6? i'm thinking 4.5% or so as a guestimate. but i'd like to maybe figure it out for real.

so where can I get the #'s and results from past seasons with opener/closer lines also listed

Fishhead · 03-16-10, 11:23 AM

Originally posted by Chi_archie

how would I figure out the potential long term (2-3 season) rate of middling by trying to beat the closer at one book with a free half point, and taking the opposite side at another book with a free half point, hopefully at an optimal line.

for example get Kansas at -5 during the overnight with a free half point bringing it to -4.5 and getting Kentucky +6 with the free half point to +6.5 what are my chances of middling with a 5 or 6? i'm thinking 4.5% or so as a guestimate. but i'd like to maybe figure it out for real.

so where can I get the #'s and results from past seasons with opener/closer lines also listed

A 1.5 pt middle is basically breakeven if your laying -110 on both sides..............in the NBA with spreads under -10 it is a very, very, very slight overlay.

This would also be the case with BIG DANCE games on spreads under -10.

Fishhead · 03-16-10, 11:25 AM

Originally posted by Fishhead

A 1.5 pt middle is basically breakeven if your laying -110 on both sides..............in the NBA with spreads under -10 it is a very, very, very slight overlay.

This would also be the case with BIG DANCE games on spreads under -10.

If one wants to middle/scalp on a regular basis, having a MATCHBOOK account is an absolute must have.

maersksealand · 03-16-10, 11:28 AM

Originally posted by Dark Horse

If you know, how many basketball games have you lost this season by 1/2 point (divided into sides and totals)? And out of how many total basketball wagers?

Trying to get a percentage of what a half pt is really worth. I know someone posted a link before with numbers that showed the percentage between 3 and 4%, but I have a feeling it may be higher.

Reason for asking is that Skybook offers free half point on sides.

rarely. I think the 3-4% is about right in my case.