Ok so using these definitions yield should always be higher than ROI if on a comparable time frame. Example: My starting bank is 100 units. I wager 50 and make a profit of 5. Yield: 5/50 = 10% ROI: 5/100 = 5%
According to his example. If you have a winning system the yield will get shorter the more selections you have throughout the season and the ROI will grow as the season goes on as you continue to add points to the total. For example 2017/18 I have a points win of +33.5 I layed (shorted the draw) a total of 494.35 units. So 33.5/ 494.35 is 6.77% yield But an ROI for the season of 33.5/100 = 33.5%. The reason I read the article was as he suggested there is a lot of confusion regards yield and ROI. I had always heard the ROI was actually calculated like the yield is but that's wrong according to the article.
ok I get it now. These ratios are calculated completely differently to the accounting ratios used in business. As long as you know what's going on. Looks like a good system. If you paper trade it back ten years and it holds up I'd recommend starting with small $ and then gradually scaling up if it continues to hold up.