formula help

[markj11]

My league has the following catagories and I was wandering the best way to come up with players values.
Batter Cats: 1B, 2B, 3B, R, HR, RBI, SB, K, BB, AVG, OPS
Pitching Cats: W, SV, SHO, CG, K, BB, ERA, WHIP, K/BB, K/9

I was just using
(1B+2B+3B+HR+R+RBI+SB+AVG+OPS+BB-K)/AB
and
(W+SV+SHO+CG+K+ERA+WHIP+K/9+K/BB-BB)/IP

Pretty straight foward so I would like an experts advise.

[Amazinz]

This is not an easy question to answer. There have been quite a few threads on this topic. I believe Madison has a few of them bookmarked in his FAQ forum.

That formula you listed won't work too well because it assumes that each unit of each category is equally valuable which isn't true. AVG, for example, is going to be severely undervalued using that formula.

[Nerfherders]

Go here:

http://www.baseballmonster.com

It has all the stat categories that you describe, except for singles. You can get a pretty good idea of value from it.

[markj11]

Go here:

http://www.baseballmonster.com

It has all the stat categories that you describe, except for singles. You can get a pretty good idea of value from it.

Thanks.

[DK]

I think the easiest thing to do (and even this is difficult) is to make an excel sheet and for each separate player put their totals over the average starter or replacement level player at that position. One of those two, doesn't really matter which.

For instance, say the average 1B in your league (assuming your league is 5x5 in this case) hit .275, 25 HR, 80 R, 80 RBI, 5 SB. Now, you have 2 first basemen - One hits .290, 20 HR, 90 R, 75 RBI, 10 SB, and the other hits .250, 40 HR, 80 R, 100 RBI, 0 SB. You'd then divide by the amount of categories (in this case 5).

For these players, the formula would be:
Player 1 - ((.290/.275)+(20/25)+(90/80)+(75/80)+(10/5)), all divided by 5.
Player 2 - ((.250/.275)+(40/25)+(80/80)+(100/80)+(0/5)), all divided by 5.

Now this has obvious flaws, but it's a quick and dirty method to give you an idea where your players stand. You'll also have to take into account your team needs and such.

(Player 1 is more valuable according to that system, FWIW.)

[DK]

I would also like to note that since SB's are a huge variant, it may be easier to leave them out of statistical observation such as this, and add in their value after calculating the others.

[markj11]

How do I find out the avgerages for each position. I would really like to come up with something for next year because I went into this years draft basically thinking 5X5 and did good in those cats but not so good in the others.

[markj11]

I think the easiest thing to do (and even this is difficult) is to make an excel sheet and for each separate player put their totals over the average starter or replacement level player at that position. One of those two, doesn't really matter which.

For instance, say the average 1B in your league (assuming your league is 5x5 in this case) hit .275, 25 HR, 80 R, 80 RBI, 5 SB. Now, you have 2 first basemen - One hits .290, 20 HR, 90 R, 75 RBI, 10 SB, and the other hits .250, 40 HR, 80 R, 100 RBI, 0 SB. You'd then divide by the amount of categories (in this case 5).

For these players, the formula would be:
Player 1 - ((.290/.275)+(20/25)+(90/80)+(75/80)+(10/5)), all divided by 5.
Player 2 - ((.250/.275)+(40/25)+(80/80)+(100/80)+(0/5)), all divided by 5.

Now this has obvious flaws, but it's a quick and dirty method to give you an idea where your players stand. You'll also have to take into account your team needs and such.

(Player 1 is more valuable according to that system, FWIW.)

How do you take in to account the number of at bats?

[RugbyD]

I would also like to note that since SB's are a huge variant, it may be easier to leave them out of statistical observation such as this, and add in their value after calculating the others.

good point. what i generally do is severely discount the value fo SB for C and 1B and moderatley for 3B.

[United_I_Stand]

mathmatically speaking

calcuating value as a sum of the ratio's of cat/mean doesn't take into account the fact that different catergories are spread to varying degrees

for instance

compare

.290 avg with 5sb's (gives ratios of 1.074 and 1)
.250 avg with 10sb's (gives 0.93 and 2)

in fantasy is a drop of 40 points worth 5 sb's? nope, but infact the guy with sb's gets double the value almost according to this calc.

what you need to do is this sum:

sum of:

(score in a catergory - mean in the catergory) / standard deviation of that catergory

that'd give a fairer comparison of the stats. where you get the means and standard deviations from i don't know.

hope this makes sense!