a lillte help form the number crunchers - please

jazzrack · 01-12-2014, 11:49 PM

http://jazzrack.ohbah.com/corp/files...er%2Blist.xlsx

so we have a simple scoring formula to determine out CLB owner champion, this is designed reward those who best balance winning and making money, with money favored a bit, but I have a fear our formula is a bit... simplistic

Quote:

The owner score is designed to give making a profit some value in an on-line league without handcuffing owners with restrictive regulations.

� Owner Score - teams are ranked in wins and give points 1-20 in inverse order. points are split between ties.

teams are ranked in profit & given points 1-20 in inverse order, points are split if a tie should happen.

cost per win (as determined by ootp) ranked and points given 1-20 in inverse order, points are split in case of a tie.

cost per loss (as determined by ootp) ranked and points given 1-20 in inverse order, points are split in case of a tie.

points are totaled for the final owner score, ties are decided by highest profit.

do any of you statheads or number crunchers have any ideas for a more complete formula?

koohead · 01-13-2014, 06:11 PM

I'm thinking you may want to consider weights. Afterall, winning is still the most important thing when determining the success of a team/owner. Maybe the points for wins are given a 3x or 4x weight, while the other catagories are just 1x?
You could try to factor in market size too if that is a variable in your league.
How about a +/- score when compared to wins and/or profit from the previous year?

le receveur · 01-13-2014, 06:32 PM

i think the weighting is a good idea. be careful with the 1 to 20 points also. is finishing 6th as much better from 11th place finish, than worse from 1st?... some grouping may be good... (lets be honest 12th is not that much better than 14th, but would get as much bonus as 2nd over 4th)

The one thing i would recommend is once you have come up with some formula, test it on past years to see if it accomplishes what you want..

The Wolf · 01-13-2014, 07:13 PM

Quote:

Originally Posted by le receveur

i think the weighting is a good idea. be careful with the 1 to 20 points also. is finishing 6th as much better from 11th place finish, than worse from 1st?... some grouping may be good... (lets be honest 12th is not that much better than 14th, but would get as much bonus as 2nd over 4th)

The one thing i would recommend is once you have come up with some formula, test it on past years to see if it accomplishes what you want..

Always god advice.

jazzrack · 01-13-2014, 10:04 PM

winning is already naturally weighted, with winning come revenue, if you can't make a profit while winning this should actually punish that. the first version of my league had an awesome formula, but it was designed by a member & him and the formula are lost to history. and I am not a numbers guy, by any stretch of anybody's imagination.

the compensation for doing well is bragging rights and the 3rd round draft order, so we want teams that are not winning at the moment but managing finances very well to be able to do well, maybe not be able to win...often, but crack the top 5 on a regular basis. the current formula seems to be ok at it, be we have a sense that it's missing something.

jazzrack · 01-13-2014, 10:11 PM

past results

1951

1950

1949

1948

1947

le receveur · 01-13-2014, 10:25 PM

do you only care for the championship (ie who is number 1) or for a a ranking of best teams?

Le Grande Orange · 01-13-2014, 10:29 PM

I'm not sure the numerical dollar profit is the value to use. Which is more impressive: $1.5 million profit on $50 million in revenue, or $1.5 million in profit on $150 million in revenue? I'd consider using the profit margin (i.e. profit divided by revenue) as one of the measures instead of the numerical profit. The profit margin is an indicator of the efficiency at which a club turned revenue into profit.

le receveur · 01-13-2014, 10:30 PM

problem on profit, is you are penalising does who drafted well, but now these guys are getting expensive...

to value management, how about adding ranking of (Actual Wins- Pre Seasons projected wins) in your value? not as the only value because of injuries and bad luck-maybe (Actual-Expected-Pythagorian) if you want to get fancy...

le receveur · 01-13-2014, 10:31 PM

Quote:

Originally Posted by Le Grande Orange

I'm not sure the numerical dollar profit is the value to use. Which is more impressive: $1.5 million profit on $50 million in revenue, or $1.5 million in profit on $150 million in revenue? I'd consider using the profit margin (i.e. profit divided by revenue) as one of the measures instead of the numerical profit. The profit margin is an indicator of the efficiency at which a club turned revenue into profit.

agree on this

Le Grande Orange · 01-13-2014, 10:33 PM

I recall in an issue of one of SABR's publication there was a proposed system for ranking GM performance. It may be applicable or adaptable to what jazzrack is attempting to do. I can try to dig up the issue if requested.

jazzrack · 01-13-2014, 10:58 PM

Quote:

Originally Posted by le receveur

do you only care for the championship (ie who is number 1) or for a a ranking of best teams?

it's reward is bragging rights and 3rd round draft order, if that's what you mean, otherwise I am not understanding what you are asking.

Quote:

Originally Posted by Le Grande Orange

I'm not sure the numerical dollar profit is the value to use. Which is more impressive: $1.5 million profit on $50 million in revenue, or $1.5 million in profit on $150 million in revenue? I'd consider using the profit margin (i.e. profit divided by revenue) as one of the measures instead of the numerical profit. The profit margin is an indicator of the efficiency at which a club turned revenue into profit.

you are probably right, it wouldn't even be much extra work once the spreadsheet is set-up.

Quote:

Originally Posted by le receveur

problem on profit, is you are penalising does who drafted well, but now these guys are getting expensive...

to value management, how about adding ranking of (Actual Wins- Pre Seasons projected wins) in your value? not as the only value because of injuries and bad luck-maybe (Actual-Expected-Pythagorian) if you want to get fancy...

I am not one who like to use statistics to predict future human activity, far too many variables, but the general idea is interesting.

Quote:

Originally Posted by Le Grande Orange

I recall in an issue of one of SABR's publication there was a proposed system for ranking GM performance. It may be applicable or adaptable to what jazzrack is attempting to do. I can try to dig up the issue if requested.

if you wouldn't mind, I know where my strengths and weaknesses are, I am fine with the mechanics of math, but I lack numerical imagination and have no problem admitting it.

le receveur · 01-13-2014, 11:16 PM

Quote:

Originally Posted by jazzrack

it's reward is bragging rights and 3rd round draft order, if that's what you mean, otherwise I am not understanding what you are asking.

this is what I mean... so you don't care if the stat doesn't reflect no. 8 vs no.9 properly.. all you care is who gets on top.

jazzrack · 01-13-2014, 11:22 PM

Quote:

Originally Posted by le receveur

this is what I mean... so you don't care if the stat doesn't reflect no. 8 vs no.9 properly.. all you care is who gets on top.

oh, my bad. I see now, says the blind man.

yes the order matters. we are talking draft picks #41-60 so getting say the top 5 right is more important than the bottom 5 but it should work top to bottom...more or less...

01-13-2014, 06:11 PM	#2
koohead Hall Of Famer Join Date: Dec 2001 Location: Phoenix, AZ Posts: 3,211	I'm thinking you may want to consider weights. Afterall, winning is still the most important thing when determining the success of a team/owner. Maybe the points for wins are given a 3x or 4x weight, while the other catagories are just 1x? You could try to factor in market size too if that is a variable in your league. How about a +/- score when compared to wins and/or profit from the previous year? __________________ GM - New Jersey Bears of the NPBL;

01-13-2014, 10:04 PM	#5
jazzrack Hall Of Famer Join Date: Oct 2003 Location: still kicking Posts: 3,229	winning is already naturally weighted, with winning come revenue, if you can't make a profit while winning this should actually punish that. the first version of my league had an awesome formula, but it was designed by a member & him and the formula are lost to history. and I am not a numbers guy, by any stretch of anybody's imagination. the compensation for doing well is bragging rights and the 3rd round draft order, so we want teams that are not winning at the moment but managing finances very well to be able to do well, maybe not be able to win...often, but crack the top 5 on a regular basis. the current formula seems to be ok at it, be we have a sense that it's missing something. __________________ . "Never confuse composure for ease" Was once Head Cheese of Corporate League Baseball Last edited by jazzrack; 01-13-2014 at 10:18 PM.

01-13-2014, 10:11 PM	#6
jazzrack Hall Of Famer Join Date: Oct 2003 Location: still kicking Posts: 3,229	past results 1951 1950 1949 1948 1947 __________________ . "Never confuse composure for ease" Was once Head Cheese of Corporate League Baseball

01-13-2014, 10:29 PM	#8
Le Grande Orange Hall Of Famer Join Date: Feb 2002 Location: Up There Posts: 15,644	I'm not sure the numerical dollar profit is the value to use. Which is more impressive: $1.5 million profit on $50 million in revenue, or $1.5 million in profit on $150 million in revenue? I'd consider using the profit margin (i.e. profit divided by revenue) as one of the measures instead of the numerical profit. The profit margin is an indicator of the efficiency at which a club turned revenue into profit. Last edited by Le Grande Orange; 01-13-2014 at 10:30 PM.

01-13-2014, 06:32 PM	#3
le receveur All Star Starter Join Date: Oct 2013 Location: canada Posts: 1,736	i think the weighting is a good idea. be careful with the 1 to 20 points also. is finishing 6th as much better from 11th place finish, than worse from 1st?... some grouping may be good... (lets be honest 12th is not that much better than 14th, but would get as much bonus as 2nd over 4th) The one thing i would recommend is once you have come up with some formula, test it on past years to see if it accomplishes what you want..

01-13-2014, 10:25 PM	#7
le receveur All Star Starter Join Date: Oct 2013 Location: canada Posts: 1,736	do you only care for the championship (ie who is number 1) or for a a ranking of best teams?

01-13-2014, 10:30 PM	#9
le receveur All Star Starter Join Date: Oct 2013 Location: canada Posts: 1,736	problem on profit, is you are penalising does who drafted well, but now these guys are getting expensive... to value management, how about adding ranking of (Actual Wins- Pre Seasons projected wins) in your value? not as the only value because of injuries and bad luck-maybe (Actual-Expected-Pythagorian) if you want to get fancy...

01-13-2014, 10:33 PM	#11
Le Grande Orange Hall Of Famer Join Date: Feb 2002 Location: Up There Posts: 15,644	I recall in an issue of one of SABR's publication there was a proposed system for ranking GM performance. It may be applicable or adaptable to what jazzrack is attempting to do. I can try to dig up the issue if requested.