What is going on with my research about betting on under or over in Basketball? Nothing so special, I would say, and the reason for that is the fact that the ROI has dropped from 17% to just four percents. However, there is good news and this is Gauss method for least square approximation. This is the most profitable method. Yes, with just four percents ROI, but it is a start. May be the way that the data is being collected is wrong, or maybe 4% is not so small profit. We will see.
Here are the results:
Under-over percentages from previous matches give 95.5 ROI
Under-over percentages from previous special matches give 101.3 ROI
Gauss method with data collected from last 10 matches for every team give 104 ROI
Gauss method with data collected from last 10 mutual games give 98.39 ROI
Average score made by teams in last 10 games for every team give 101.29 ROI
Average score made by teams in last 10 mutual games give 89.71 ROI
The last row is really a surprise for me. How is it possible the average for mutual games to be the worst result? Really strange, but a fact.
I am asking myself this question right now and the reason is simple. If you take a look at the table below you will see that the gap between matches finished over and under has never been so huge like right now. As far as I remember some previous seasons in NBA and WNBA the bookies are trying to hold the percentages 50-50 between unders and overs.
Would it be the same way this year? I do not know, but I will try it.
Here is the table:
|
Teams
|
U
|
O
|
|
Atlanta
|
35
|
65
|
|
Chicago
|
30
|
70
|
|
Connecticut
|
58
|
42
|
|
Detroit
|
47
|
53
|
|
Indiana
|
37
|
63
|
|
Los Angeles
|
69
|
31
|
|
Minnesota
|
40
|
60
|
|
New York
|
44
|
56
|
|
Phoenix
|
32
|
68
|
|
Sacramento
|
45
|
55
|
|
San Antonio
|
37
|
63
|
|
Seattle
|
59
|
41
|
|
Washington
|
44
|
56
|
If we speak in general all under finished games are 43.6% and over finished games are, well here the math is easy
So, if those numbers above must go to 50-50 percent there is clear profit. I will try it. Crossing fingers.
I cannot say I am satisfied with the results from this research. I definitely expected more from Gauss less square approximate method. I see that there is profit, but I expected more. Even so, 17% pure profit from 31 matches with betting on under over in basketball is not bad.
Here are the stats.
| % L + |
%i L + |
G L + |
Gm L + |
Avg. L + |
Avg.m L + |
| 61.3 |
54.8 |
41.9 |
58 |
45.1 |
38.7 |
If we start talking in profit percentage, best results shows my old method of betting, where you betting under or over depends from all previous matches for both teams. The profit there is 17%. However, as I learned from previous NBA regular season this method is going to be less profitable at the end of the season, because bookies start to level under – over percentage to 50-50 for almost all teams. Of course this can be used too, but you have to be good enough to use it.
The other method which is giving profit is Gauss less square approximate with using last 10 mutual games as data. The profit there is 10%. I hope this method to be more stable and to keep those 10% to the end.
I have profit in other section too. This is a method very similar to the first one I wrote above, but with dividing of data.
The other three methods are below the line, but I will wait until the end of regular basketball season in WNBA for final conclusions.
Last week my betting study was occupied only with the research in basketball under or over ending. I wrote a few times about it and I will do it now, again. It is time for evaluation after the first 18 matches already played.
I won’t be explaing again which methods I use. You can read more about them here.
There is 16.1% profit in two of the profits. The other four are below the line after those 18 games. However, as we all know we have to wait for the final matches for making analysis.
Here is the table with the result after 18 games in percentages:
|
% L +
|
%i L +
|
G L +
|
Gm L +
|
Avg. L +
|
Avg.m L +
|
|
61.1
|
50
|
33.3
|
61.1
|
38.8
|
44.4
|
One mark here. All odds here are 1.91
It is a bit strange for me but my basketball under over research shows similar results. I use 6 methods for picking matches and all of them give results around 50%. This of course is 100% unsatisfying, but my research will continue.
Here are the results after 9 matches on WNBA:
When I use my old method with all previous matches – 4 from 9;
When I use my new method with all previous matches from the sector where bookie line is – 5 from 9;
When I use Gauss theory for less square approximate with the last 10 matches for each team – 4 from 9;
The same method used with the last 10 mutual games – 5 from 9;
When I use average scores from data above both are 4 from 9.
As I posted earlier I was going to bet on Under for Phoenix with Martinagale. Finally it gave profit and I am satisfied with full pockets. 
Crossing fingers this to continue.
I am starting a quick research on a few different strategies about betting on under or over at basketball games. I will use WNBA because it is only one major championship where matches are on right now. Few words about those different strategies: the first one I have used before in NBA last regular season. I follow the under – over stats for every team and when the match is going to be played I just take a look to numbers. For example Washington and Atlanta will play tonight – Washington has 50-50 until now and Atlanta has 30-70 under – over percentage. So, I will bet over here.
The other way of betting is similar, but a bit more difficult. The principle is the same, but the way of collecting the data is different. For every team I make different sectors divided by the border given by the bookie for under or over. Then I check the previous matches for this particular team for that particular border and compare the percentage between both teams on that border and bet for under or over with greater percentage.
The third way I will check to the rest of the season is far more difficult. I read in the forum about using Gauss’ least squares approximate principle for betting. It was about football, but I think it’s best to put it into practice with under – over betting. So I chose to use it here. For data I will use the last ten matches for every team – so 20 games.
The same method will be used for my fourth hypothesis, but the difference again will be of the way I collected data. Here I will use the last 10 mutual games between the teams.
I will check other two ways of making tips, but they are far less effective I think. I will check them only because it is easy to collect data. I will check the last 10 matches for every team and I will calculate the average scores so it will be up to me to guess if the match will be under or over.
Finally, the same method will be made, but with data collected from the last 10 mutual games between the teams.
Crossing fingers.
I am going to bet under on next five Phoenix games with martingale. I know it is a bit dangerous and I can lose all my bank, but I’ve got some reasons. From the beginning of WNBA Phoenix played 8 matches and all of them are over.
The longest strike of over I know about was New York’s last season with 12 overs in a row.
I think Phoenix will make at least one under in the next few games. So, crossing fingers.
nba chart for under and over
Do you see the pattern? I do and I will try it next season. I am really curious will the pattern repeat next season.
Something else I want to share is that fact that made impression on me. I wrote above that every period consist 60 matches and from the chart we can see that after 60 – 60 (or 30-30) under – over period we have a period with deviated results. Under or over does not matter. Is it possible from the first 20 games to guess where this period is going to and to bet for that guessing? What do you think?
Finally I found some profit in under over streaks in NBA. The trick is to bet on every 5th game in the row and to place only one bet. Not more, only one bet. However, the ROI is so small, that it is not worth to play it at all. If, I have to talk with numbers the ROI is exactly 105.82%.
Not so good for the entire regular season.
So, I have to say it. I am really disappointed. I have read so many times the streaks are the best way of betting. Things like a punter will always has profit if the bet is placed on the 4th match in the row etc, but all of this is simply not correct for under – over in NBA.
I checked also what will happen if the punter is placed only one bet on the third match in the row, but the result was disappointing – only 96% ROI.
Here comes the good part of this post. Using the streaks in under over in NBA gives 50 to 56% correct results. This can be used for good profit. I am quite sure about it and I will try it.
Down is the table, team by team with the results of straight betting with streaks.
|
|
On 3rd
|
On 5th
|
|
Team
|
+
|
-
|
+
|
-
|
|
Boston
|
12
|
11
|
2
|
3
|
|
New Jersey
|
12
|
5
|
3
|
3
|
|
New York
|
9
|
6
|
5
|
1
|
|
Philadelphia
|
9
|
9
|
2
|
2
|
|
Toronto
|
6
|
14
|
2
|
1
|
|
Dallas
|
10
|
11
|
0
|
2
|
|
Houston
|
9
|
7
|
3
|
0
|
|
Memphis
|
15
|
7
|
1
|
6
|
|
New Orleans
|
11
|
8
|
5
|
1
|
|
San Antonio
|
9
|
9
|
4
|
2
|
|
Chicago
|
7
|
15
|
2
|
1
|
|
Cleveland
|
6
|
16
|
2
|
2
|
|
Detroit
|
8
|
14
|
4
|
1
|
|
Indiana
|
10
|
7
|
5
|
1
|
|
Milwaukee
|
15
|
6
|
1
|
6
|
|
Denver
|
7
|
11
|
2
|
3
|
|
Minnesota
|
10
|
11
|
4
|
2
|
|
Portland
|
5
|
10
|
2
|
1
|
|
Oklahoma
|
8
|
9
|
2
|
2
|
|
Utah
|
11
|
8
|
6
|
2
|
|
Atlanta
|
8
|
8
|
3
|
3
|
|
Charlotte
|
11
|
9
|
2
|
3
|
|
Miami
|
8
|
9
|
1
|
4
|
|
Orlando
|
11
|
10
|
2
|
1
|
|
Washington
|
14
|
5
|
3
|
3
|
|
Golden State
|
11
|
10
|
5
|
1
|
|
LA Clippers
|
12
|
6
|
3
|
4
|
|
LA Lakers
|
10
|
9
|
3
|
2
|
|
Phoenix
|
7
|
13
|
1
|
1
|
|
Sacramento
|
8
|
13
|
2
|
2
|
|
|
289
|
286
|
82
|
66
|
|
|
551.99
|
575
|
156.62
|
148
|
|
|
96.00
|
105.82
|