[Ganchrow]

For updated data please see my Half-Point Calculator.

Spread	N	Freq.	Std. Err.
1	6,510	2.29%	0.19%
2	8,059	3.92%	0.22%
3	9,476	3.84%	0.20%
4	10,330	3.50%	0.18%
5	10,286	4.30%	0.20%
6	9,626	4.11%	0.20%
7	8,679	4.09%	0.21%
8	7,563	4.17%	0.23%
9	6,273	4.69%	0.27%
10	5,119	4.12%	0.28%
11	4,098	4.05%	0.31%
12	3,163	3.48%	0.33%
13	2,397	3.84%	0.39%
14	1,811	3.37%	0.42%
15	1,277	3.13%	0.49%
16	850	5.18%	0.76%
17	509	3.14%	0.77%
18	286	3.15%	1.03%
19	162	3.09%	1.36%

---

Methodology:

• All NBA final scores and closing point spreads (from http://www.covers.com/) from the 1990/1 season through November 2^th of the 2008/9 season were analyzed (21,847 games in total) for various favorite margins of victory.
• The push frequency for a given point spread was determined by the percentage of games with a closing spread within 2 points of the spread in question ending with a favorite margin of victory equal to that spread. For example, the push frequency of a spread of 3 was determined from all games with a closing spread between 1 and 5.
• This is based on Stanford Wong's methodology as described in Sharp Sports Betting.

[Ganchrow]

For updated data please see my Half-Point Calculator.

Total	N	Freq.
170	548	2.37%
171	713	3.51%
172	869	2.53%
173	1,031	2.52%
174	1,202	2.25%
175	1,375	2.69%
176	1,563	2.18%
177	1,760	2.44%
178	1,960	2.40%
179	2,187	2.70%
180	2,348	2.21%
181	2,586	2.09%
182	2,819	3.23%
183	3,086	2.79%
184	3,379	2.78%
185	3,626	2.48%
186	3,842	2.19%
187	4,005	2.22%
188	4,123	2.38%
189	4,322	2.31%
190	4,432	1.96%
191	4,570	2.23%
192	4,600	2.33%
193	4,589	2.31%
194	4,580	2.16%
195	4,519	2.43%
196	4,432	2.62%
197	4,293	2.63%
198	4,152	1.93%
199	4,127	2.42%
200	4,029	2.08%
201	3,917	1.79%
202	3,814	2.60%
203	3,715	2.45%
204	3,636	2.50%
205	3,514	2.19%
206	3,372	2.28%
207	3,168	2.37%
208	2,943	2.14%
209	2,723	2.13%
210	2,531	2.69%
211	2,292	2.14%
212	2,087	2.30%
213	1,926	2.02%
214	1,741	1.78%
215	1,598	2.00%
216	1,447	1.80%
217	1,261	2.14%
218	1,134	2.03%
219	986	1.52%
220	884	2.60%
221	779	2.05%
222	685	1.46%
223	598	2.34%
224	522	3.07%

---

Methodology:

• All NBA regular season final scores and closing over/unders (from http://www.covers.com/) from the 1990/1 season through November 17^th of the 2006/7 season were analyzed (18,368 games in total) for various total point scores.
• The push frequency for a given over/under was determined by the percentage of games with an over/under within 4 points of the over/under in question ending with a total point score equal to that over/under. For example, the push frequency of an over/under of 200 was determined from all games with a closing over/unders between 196 and 204.
• This is based on Stanford Wong's methodology as described in Sharp Sports Betting.

[Jay Edgar]

Great work as usual. Many thanks.

Is it a simple matter to report on margins of victory in those same games? I can imagine how it might not be -- but if it is, I for one would be happy to see it.

[Ganchrow]

Originally Posted by Jay Edgar

Is it a simple matter to report on margins of victory in those same games?

Not sure I understand what you mean by this, Jay.

Could you clarify?

[Jay Edgar]

I didn't phrase it very well. Should have asked about "margin of victory in all NBA games in that time period."

Meant to ask this: without regard to the spread, what % of all games "land on 1" (the winning team's straight-up margin of victory is 1 point). What % of games "land on" 2, and 3, etc?

I'm routinely living in terror of a 1 point win by the team I chose to back at -2 rather than with the ML, etc. So it'd be nice to see a full MOV picture to get a sense of how many games that are in doubt in the last minute (say, MOV of 7 or less) actually land on 1 or land on 2.

[Ganchrow]

MOV	Freq.
1	4.214%
2	6.131%
3	5.820%
4	5.788%
5	6.327%
6	6.212%
7	6.245%
8	6.043%
9	5.869%
10	5.031%
11	4.797%
12	4.236%
13	3.800%
14	3.267%
15	2.995%
16	2.700%
17	2.428%
18	2.161%
19	1.960%
20	1.846%
21	1.677%
22	1.421%
23	1.024%
24	1.187%
25	0.985%
26	1.007%
27	0.730%
28	0.626%
29	0.550%
30	0.479%
31	0.343%
32	0.414%
33	0.338%
34	0.207%
35	0.180%
36	0.136%
37	0.109%
38	0.103%
39	0.131%
40	0.098%
41	0.060%
42	0.049%
43	0.027%
44	0.049%
45	0.044%
46	0.038%
47	0.027%
48	0.022%
49	0.011%
50	0.005%
51	0.005%
52	0.011%
53	0.005%
56	0.011%
58	0.005%
62	0.005%
65	0.005%
68	0.005%

---

All NBA regular season scores from the 1990/1 season through November 17^th of the 2006/7 season (available from http://www.covers.com/) are included (18,368 games in total).

[Jay Edgar]

Jeez, Ganch, we could time you with a sundial. 38 minutes!!!

Seriously, many thanks.

[Ganchrow]

Originally Posted by Jay Edgar

Jeez, Ganch, we could time you with a sundial. 38 minutes!!!

Seriously, many thanks.

Well to be fair, I didn't actually see your last post until 11:40 ...

[Jay Edgar]

Was there really an NBA game decided by 92 points in the last 15 years?

(I'm sure that if someone identifies it the rest of us will say, "Oh yeah, I remember that.")

[Ganchrow]

Originally Posted by Jay Edgar

Was there really an NBA game decided by 92 points in the last 15 years?

(I'm sure that if someone identifies it the rest of us will say, "Oh yeah, I remember that.")

Unless Detroit really beat Seattle by a score of 92-0 on March 10^th, 1992, then this was just bad data.

(The final score according to Database Basketball was DET 92 SEA 98. I've modified the chart to reflect the correction.)

[zxcjason]

Ganchrow, where do you find the point spreads in Covers. I can't find the data on the web site and I'd like to do some analysis to test out my systems.

Jason

[Ganchrow]

Originally Posted by zxcjason

Ganchrow, where do you find the point spreads in Covers

Go the Covers NBA Team Page. From there you click on a team link and then click "Past Results" at the top of the page beneath the team's name.

[Newguyintown]

Can someone explain what the difference is between the first chart with the push frequencies and then what looks like the same chart that says MOV on it?

N G

[Ganchrow]

The first chart lists push frequencies ATS. In other words, for a spread of, say, ±3, the first chart indicates that there is a roughly 3.91% frequency that bet pushing. This can be extrapolated upon, so that one might say that for a spread of about ±3, there is a roughly 3.91% frequency of the favorite winning by exactly 3 points.

The second chart is simply a margin of victory chart. It indicates the historical frequency of the winning team's margin of victory being as specificed.

[raiders72002]

Ganch- That's very good information on push frequency. If you include 1/2, that is 6.5 and 7.5, and look for value instead of just push info I think that you'll find that 6 and 7 will be the key numbers in the NBA and 2 in the NCAA tourney. I haven't personally run the numbers but looked at charts from others.

Some say that there is value to buying 1/2 pt off/on 7 in the NBA and there's definitely value buying off/on 2 during the NCAA tourney.

[Ganchrow]

Originally Posted by raiders72002

Ganch- That's very good information on push frequency. If you include 1/2, that is 6.5 and 7.5, and look for value instead of just push info I think that you'll find that 6 and 7 will be the key numbers in the NBA and 2 in the NCAA tourney. I haven't personally run the numbers but looked at charts from others.

Some say that there is value to buying 1/2 pt off/on 7 in the NBA and there's definitely value buying off/on 2 during the NCAA tourney.

Hi Raiders, welcome back.

I gotta admit I'm that I don't really understand what you mean about "include ½ ... and look for value instead of just push info". This is, after all a push probability chart from which one would have to derive the value of moving on to a given spread.

I've certainly heard that about 6s and 7s and NCAA BB 2s as well, but I've not seen any evidence of this in the historical data.

[gridironguy]

Great work here. Thx

[the_fredrik]

So, looking at the spreads 2-10, the push frequency varies between 3.5%(4 points) and 5.2% (9 points)

The actual MOV frequencies for 4 and 9 points are about the same though.

Do any of you see any reason as to why a 9 point push ATS should be more common than a 4 point push ATS?

If we don't, does it not make more sense to estimate all the push frequencies for spreads of 2-10 to the same percentage, maybe 4% or so? (Or rather 4.x%, where you can estimate x better than I can)

[newb411breaker19]

thanks a lot for posting the push frequencies for all the sports ganch, we all appreciate it very much

[THE HITMAN]

Interesting stuff...thanks

[Goldfisch]

I dont understand the importance of these numbers?

[St.Anger]

I dont understand than number.
hay i'm new member!!!
thank you!!!!!!

[flyingillini]

Originally Posted by St.Anger

I dont understand than number.
hay i'm new member!!!
thank you!!!!!!

All new members are encouraged to please pm Kellen and ask about a therapy session. Thanks and welcome to SBR. Nicky, Mathdotcom, JJ and Charlotte run things.

[bconngemini]

thanks

[spot]

I am trying to write a simulation program for prediction of betting NBA, but I am lack with some info of the following. Please illustrate me step-by-step how to calcuate the following things:

1. How to calculate NBA Power Rating and Line spread, O/U ATS?

If you could, please reply to me so I can finish my program. This is part of my research in statistic.

[pats3peat]

mistake post

[danger]

Originally Posted by Ganchrow

Methodology:

• <span style="font-size: 11px;">The push frequency for a given point spread was determined by the percentage of games with a closing spread within 2 points of the spread in question ending with a favorite margin of victory equal to that spread. For example, the push frequency of a spread of 3 was determined from all games with a closing spread between 1 and 5.</span>
• <span style="font-size: 11px;">This is based on Stanford Wong's methodology as described in Sharp Sports Betting.</span>

Ganchrow, thanks for posting this and your half point calculator, excellent data. How do you handle 1/2 point lines, for example if the line is 2.5, do you count a 2 or 3 point margin as 1/2 a push each?

[Ganchrow]

Originally Posted by danger

How do you handle 1/2 point lines, for example if the line is 2.5, do you count a 2 or 3 point margin as 1/2 a push each?

If a game with a spread of 2.5 finished with a favorite margin of victory within 2 points of 2.5 (i.e., fave by 1, 2, 3, or 4) it would count as a (full) push for that M.o.V. (and only that M.o.V.)

[cito89]

bulls or celtics?