Kelly simulator 1st results

I’ve completed the first runs of my Kelly Criterion simulator. Following is an Excel spreadsheet to simulate wagers made against a spread at -110 odds. It uses two random numbers per game, the first to generate an edge running from 3% to 7%. This edge is added to 0.50 to get a probability of winning the wager. The second random number is compared to this probability to determine the results of the wager. This guarantees that the expected return is precisely accurate in making the wager thus removing this uncertainty from the resulting Kelly wager.

Edge (3-7)	Prob Win	Random	Win/Lose	Odds	Kelly %	Balance	Bet	Net
4	0.54	0.999	Lose	1.909	3.4%	$1,000.00	$34.00	-$34.00
7	0.57	0.189	Win	1.909	9.7%	$966.00	$93.70	$93.70
4	0.54	0.216	Win	1.909	3.4%	$1,059.70	$36.03	$36.03

Here is a description of the cells so that you can build your own model.

Column	Heading	Contents
A	Edge (3-7)	=(INT(5*RAND())+3)
B	Prob Win	=0.5+(L2/100)
C	Random	=RAND()
D	Win/Lose	=IF(L4<L3,"Win","Lose")
E	Odds	=210/110
F	Kelly %	=(L6*L3-1)/(L6-1)
G	Balance	1000
H	Bet	=+L8*L7
I	Net	=IF(L5="Win",+L9,-L9)

I then translated this code into a program, so I could simulate a large number of wagers multiple times. I elected to simulate 100 wagers (a reasonable number for someone wagering on an NFL season). I added code to read the results of each 100 wagers collecting some counts. I then ran these 1,000 times and saved the results in another spreadsheet. In effect, I simulated 100,000 wagers spread over 1,000 seasons.

The results have something for everybody, pro and anti-Kelly enthusiasts. Here is a link to the 1000 case log, http://bit.ly/2MhtXbh . You can download it and analyze it anyway you choose to draw any conclusion you’re hoping to prove. In my next post, I’ll use this file to answer some of the questions that I have. As usual, I’ll use Twitter, @ole44bill, to tweet when this is posted.