### 4 Responses

1. Alex at |

What if my rotisserie league has different point values for each category? For example, I’m trying to apply your wonderful method to my rotisserie hockey pool using the same principles you’ve outlined in all the great guides. My Rotisserie league has 12 team,s but most categories are worth 0.75pts which means 1st place gets 9 points and 12th place gets 0.75. Other categories get up to 1.75pts (1st place = 21 and 12th place = 1.75). How do I account for this using the slope formula? Do I still let the x-values be equal to 1-12 or should I “weight” the x-values based on their roto point contribution?

Thanks for posting all this great information. Very helpful.

2. Tanner Bell at |

Hi Alex. You’re welcome, thanks for hanging around the site. That sounds like an interesting rotisserie wrinkle. I think I played fantasy hockey once years ago and it was a H2H deal, not rotisserie. Is that .75/1.75 point system something that’s unique to your league?

I played around with the data from the first table above. The slope formula I got above when I plotted things using 1, 2, 3, 4, etc was 7.1958x+211.89.

When I plot the same number of home runs against a scale of 0.75, 1.5, 2.25, etc. I get 9.5944x+211.89.

And 9.5944 / 7.1958 = 1.33333333. Which I think would be your factor if you were going to weight things afterwards (1/0.75 = 1.333333).

So it looks like you would get to the same result if you plot against the 0.75/1.75 or if you plot against the 1.00 scale and weight it afterwards.

In the end I think you just want to double check yourself and make sure you’re answering the question, “How many HR (or goals or assists) does it take to get one more point in the standings?”. In a world where 12 roto points go to first place it would have taken 7.1958 HR to get one more point. In a world where only 9 roto points go to first place, it would have taken 9.59 HR to get one more point (it’s harder to get 1 point in the 0.75 scale).

3. Alex at |

Thanks for the reply. I think I’m going to use places 1-12 with NO weighting to determine the slope since I think this better estimates how many goals (for example) you need to score to move up a place in the Goals standings.
Then when I use this slope in the SGP calc, I will then multiply the SGP for each category by their roto points weighting (1.75, 1.25, 0.75 etc) and then sum them up to get the total SGP. I think this makes the most sense.

Two questions:
1)Does that make sense to you?
2) Do you often see that the highest ranked players are usually well over-valued relative to your SGP calculations?

Thanks again for the help.

4. Tanner Bell at |

I don’t think it matters if you figure out the slope with the 1-12 or the 0.75-9 as long as you multiply by the weights at the end (when using 1-12).

I’m not familiar with hockey’s player pool to know if that makes sense, but for baseball going into this 2014 season I had Mike Trout and Miguel Cabrera valued at \$46-\$47 and the next closest player valued at \$36, so it’s entirely possible (or about 8 SGP over replacement level vs. about 6 SGP for the next closest players).