SGP Rankings

[revo]

I'd like to come up with a truer value of players based on their actual 2K12 contributions and their effect on my league's roto standings. For instance, in my 16-team league, the average team had 139 SBs, meaning 1 SB is worth about .0072 fantasy points (1/139). So a guy like Rajai Davis, who had 46 SBs, is worth about 3.3 fantasy points in the SB category.

But I was reading about Standings Gap Percentage rankings, where the average of the league is not used, but the gap between ranks are used. I'm not sure how to go about coming up with a "point per stat" in using SGP Rankings.

Anyone have any insight on using SGP, a point per stat ranking system, or any other system you use on ranking players?

[rhd]

I'd like to come up with a truer value of players based on their actual 2K12 contributions and their effect on my league's roto standings. For instance, in my 16-team league, the average team had 139 SBs, meaning 1 SB is worth about .0072 fantasy points (1/139). So a guy like Rajai Davis, who had 46 SBs, is worth about 3.3 fantasy points in the SB category.

But I was reading about Standings Gap Percentage rankings, where the average of the league is not used, but the gap between ranks are used. I'm not sure how to go about coming up with a "point per stat" in using SGP Rankings.

Anyone have any insight on using SGP, a point per stat ranking system, or any other system you use on ranking players?

I've never heard of SGP but what you described basically is how I do my stat calculations for my leagues. I figure what the average gap between the rankings is for each category, figure how much each point in the standings is worth in the league and calculate my dollar values accordingly. The amount of a stat that will gain another point in the standings is part of what the formulas are based on. The cumulative dollar value for all the categories is what the player's stats are worth in a particular league. Each league will have its own formulas that are unique to it.

[Kevin Seitzer]

The SGP method you describe is what I used for vintage drafting.

[revo]

I've never heard of SGP but what you described basically is how I do my stat calculations for my leagues. I figure what the average gap between the rankings is for each category, figure how much each point in the standings is worth in the league and calculate my dollar values accordingly. The amount of a stat that will gain another point in the standings is part of what the formulas are based on. The cumulative dollar value for all the categories is what the player's stats are worth in a particular league. Each league will have its own formulas that are unique to it.

So how do you calculate the values? What's the formula you use?

[Todd Zola]

The main problem with SGP is it doesn't account for the baseline number of stats necessary to begin to accrue points. This is non-uniform (mainly because of stolen bases and saves).

[wannabegriffey]

The main problem with SGP is it doesn't account for the baseline number of stats necessary to begin to accrue points. This is non-uniform (mainly because of stolen bases and saves).

Hi Todd

Over at masterball, can you use league specific categories, # of teams, etc to determin a players worth in a particular season, and can I get values based on this information going back 3 or 4 years?

[rhd]

So how do you calculate the values? What's the formula you use?

There's a lot that goes into it, but say you have a league w a 5x5 format, there are 10 teams, you use a standard budget of $260 per team and you have the standard roster of 14 hitters and 9 pitchers. So there are 55 points for each category (10+9+8...+1=55), and 10 categories so 10 x 55 = 550 total points. The total $ to be spent are $260 x 10 teams = $2600. Each pt is worth $2600 / 550 pts = $4.73. That's how much you'd have to pay on average to gain a pt in a category. Now, for each category, you calculate both the average stat and the average value between each place. E.g., for HRs, add up the total HRs for the entire league and divide this by the # of hitters. So, say there were 1960 HRs in the league, and there are 14 x 10 = 140 hitters, so 1960 / 140 = 14 HR per hitter on avg. The avg value between each place is just the team w most HRs minus the team w least HRs divided by the # of place differences, which is 1 less than the total # of places. So, say the most team HRs was 285 and the least team HRs was 150, so (285 - 150) / (10 - 1) = 15. The formula is ((# of HR by player - 14)/15) * 4.73 = HR value for player. So if a player hit 20 HRs, the value of his HRs is ((20-14)/15)*4.73=$1.89. Do this for each hitting category and sum all the values.

Still not done. I use an average price per player. This is different for hitters and pitchers. Figure your average split between $s spent on hitters and pitchers in your league, say 68/32. Total league budget is $2600, so multiply this by .68 and divide by the # of hitters (140) and you get $12.63. This is the avg price of a hitter. So a hitter w avg stats in each category would have this value. For pitchers (total # of pitchers is 10 x 9 = 90), it would be ($2600 x .32) / 90 = $9.24 (so hitters are worth about 37% more than pitchers). Add this value to the cumulative values you figured above to get the total value for the player.

For the "non-counting" stats, the formulas are a little different. Also, much better to use at least 3 years of league stats to figure the averages.

Just my way of doing it; there are different methods. Some use the 2nd highest team total and the next to lowest to figure the average gap between places. Some, instead of using an average value for hitters/pitchers, figure that a zero value for a "counting" stat gives a dollar value of zero. My method I think works well for me.

Hope this made sense.

[Todd Zola]

Hi Todd

Over at masterball, can you use league specific categories, # of teams, etc to determin a players worth in a particular season, and can I get values based on this information going back 3 or 4 years?

Short answer is yes, I have an Excel tool that can do this.

Longer answer is the tool is designed for standard 5x5 (you can change the number of teams), but can be tweaked to handle league specific categories (though it can handle OBP for H and K/9 for P). I also have a 10x10 tool I keep for myself so I can provide values for leagues using more categories on a custom basis.