Determining a Field Army's Winning Percentage!

[Antony Barlow]

I agree with Francois, you must include all the results, including draws, to give an overall and accurate picture of combat performance.

Maj V : +2
min v : +1
draw : 0
min d : -1
Maj D : -2

I think Doug Shaw's formula has merit too.

[David Danner]

I see some value in all points of view and I draw your attention to baseball for the mathematicians nightmare. A pitcher goes deep into the 8th inning for his team with a tied score and receives no credit for his work except from the coach, because his teammates did not score enough runs. A hitter walks, but gets no help with batting average, however garners an increase with his on-base percentage. And there are numerous analogies that bring us to our Army's winning percentage and confusion.

I believe that whatever criteria we use should be calculated simply like the determination for the quarterly Sheridan award. It's up front and easy for all to understand and tally--no room for "moneyball."

[Joe Meyer]

Just a couple of points for clarification and perhaps additional expansion...

The current criteria for the Sheridan Combat Badge reads as follows:

Awarded quarterly by the Army Commander to the officer in each corps with the best combat record; such record determined by allowing 2 points for each MV, 1 point for each mV, 0 points for each DRAW, -1 point for each mD and -2 points for each MD as recorded from the DoR. The recipient's point total must equal 5 or more points. In the event of a tie, two or more awards may be simultaneously issued.

There's a minimum value that must be met in using that formula! If we apply it to the UA Combat Medallion, what minimum would be selected. Would we need a minimum? Remember that the Medallion is awarded on a quarterly schedule, based upon the entire field army's performance.

"...awarded by the CoA to the army with the best combat record for the previous quarter."

Also, I don't quite understand how you can say that DRAWS should count, if you give them a value of "0" in a point system! Isn't part of an overall performance based also upon the number of games that were originally undertook or attempted? How would you explain the merit of a 6-0-60-0-0 record against a 6-0-4-0-0? If you attempt 66 games and only win 6, is your performance better than winning 6 of 10 attempts?

Hey, I'm only asking some questions! I'll keep track of all of the suggested formulas and then identify them for your final vote.

[nelmsm]

Does it really matter, doesn't it go to the Army of the Cumberland each time??

[cameronm]

Nah, it usually goes to the Army of Alabama

Gen Cam McOmish
Georgia State Volunteers
1/1/1
A0A
CSA

[Antony Barlow]

Joe,

Draws would come into the calculation if you divide whatever scoring mechanism you decide on for MV, mv, D, ml, ML by the number of games played, if you decide to base the award on the ratio of wins to number of games played rather than just the absolute number of wins.

[Joe Meyer]

After studying all of these proposals and actually penciling them out for comparisons, I am of the view that the system advocated by Doug Shaw has the most inclusive merit, and I invite all of you to conduct the same experimentations! A field army's monthly combat record is first converted into a point total based upon the schedule given below and then divided by 4x the number of games played. This system accurately and inclusively portrays the values of each game result and provides a consistent, stable and accurate yardstick. As Dave says, it is very similar to a player's batting average calulation in the game of baseball.

Within his system the following points are awarded for each result:

MV = 4 pts.
mV = 3 pts.
DRAW = 2 pts.
mD = 1 pt.
MD = 0 pts.

EXAMPLE A - A field army with a monthly record of 7-1-5-0-7 would obtain 41 total points (28-3-10-0-0). Those 41 points would then be divided by 4x the number of total games played. In this case there were a total of 20 games played, which would equate to a dividing factor of 80! By dividing 41 by 80 a result of .512 is achieved.

EXAMPLE B - A field army with a monthly record of 22-1-4-4-4 would obtain 103 total points (88-3-8-4-0). Those 108 points would then be divided by 4x the number of games played. In this case there were a total of 35 games played, which would equate to a dividing factor of 140! By dividing 108 by 140 a result of .771 is achieved.

As Doug indicates, it is easy to see at a glance how successful a field army was that month, a perfect score being 1.000 and a total losing score being .000!

This same formula may be used to determine the Union Army Combat Medallion Award made each quarter. In that instance the formula is applied to the combined results of each field army's record for three months' of play. To be sure, a minimum number of games must have been played within the quarter by each field army to qualify. In reviewing the records on file, I think that 25 games would certainly be sufficient! Unless there be great objection to this system, I will revisit those UA Monthly Reports for which no previous Combat Medallion Awards have been made and re-calculate the percentages. I will also reconvene the UA Awards Committee to approve and authenticate the formula.

[ernie sands]

The above formula seems fair and covers many points discussed above.

[cameronm]

Gents I must disagree with Gen Meyer as follows;

EXAMPLE A - A field army with a monthly record at the bar of 100 bottles of beer, 52 bottles of rot gut and 32 bottles of Chateau De Goat 1861 (100+52+32) each, would equal a very drunk army

EXAMPLE B - A field army with a monthly record of only 10 bottles of beer, 3 bottles of rot gut and 6 bottles of Chateau De Goat 1861 (10+3+6) each, would equal a army full of girls

As Doug indicates, it is easy to see at a glance how successful a field army was that month, as the really drunk one is of course the winner.

However should example A come up against the Army of Alabama then they would find that they would come up short, by 1002 bottles of beer, 102 bottles of rot gut, 0 bottles of Chateau De Goat 1861 and a hogs head of rum (1002+102+0+ Lots).

So no matter which way you cut it the AoA is always going to be much drunker than any Union army.

As always keen to help.

Gen Cam McOmish
1/1/1
AoA
CSA

Tavern raiding because I care