Most of all tournaments are played in MP or IMP.
I know there are other methods for doing it.
Anyone can tell me where to see them?
Someone know any method combining MP and IMP?
I was trying to draw one, and maybe any matematician can say me if it has enough sense to try it.
1º Board scores
2º Points A=Points from sorting the list (schema delow)
3º IMP's from abs(result-average)
4º Points B=imp/10
5º Points for this board.
2500-----------3--------1940----------18----------1,8----------3----------1,8----------4,8
1430 3 870 13 1,3 3 1,3 4,3
1100 2 540 11 1,1 2 1,1 3,1
800 2 240 6 0,6 2 0,6 2,6
680 2 120 3 0,3 2 0,3 2,3
650 2 90 3 0,3 2 0,3 2,3
650 2 90 3 0,3 2 0,3 2,3
650 2 90 3 0,3 2 0,3 2,3
650 2 90 3 0,3 2 0,3 2,3
630 2 70 2 0,2 2 0,2 2,2
620 2 60 2 0,2 2 0,2 2,2
620 2 60 2 0,2 2 0,2 2,2
620 2 60 2 0,2 2 0,2 2,2
-50 1 610 -12 -1,2 1 -1,2 -0,2
-100 1 660 -12 -1,2 1 -1,2 -0,2
-800 0 1360 -16 -1,6 0 -1,6 -1,6
Average without the highest and lowest: 559,375
Typical deviation: 702,8984635
-99999-----------143,52----------0----------=>min----------av-td
-143,52----------559,375---------1----------=>av-td---------av
559,375--------1262,27-----------2----------=>av-----------av+td
1262,27---------99999------------3----------=>av+td--------max
Each result is one of this brackets, points are awarded as 0,1,2,3
<av-td---average-typical deviation-------average+typical deviation--->av+td
Parameters used
average wihout the highest and the lowest
sorting points 0,1,2,3
points B as IMP/10
Page 1 of 1
pairs scoring
#2
Posted 2007-June-01, 09:54
Combining MP and IMP Scoring
I am not sure there is an advantage to scoring this way. It causes mixed strategy based on almost random field results. In IMP scoring, your goal is clear: Make or set the contract. In MP scoring, your goal is to maximize your score against the field. Although the field may be difficult to gauge, once gauged, your goal is clear.
With mixed MP and IMP scoring, your goal becomes murkier. Not only do you have to determine the field result, you may have to determine how the field outcome will affect the complex scoring.
With that in mind, I came up with this methodology. My objective was to combine the MP into the IMP scoring to reward good MP scores or lessen the damage of a bad IMP score that would have been a not so bad MP score.
This method rewards good scores and does not further punish bad scores. It gives extra advantage to good MP scores that might not be reflected in the IMP score. I believe it satisfies my objective by retaining IMP goals.
IMP + MP scoring:
1. MP the result against all other scores into a percentage, Call it PctMP
2. IMP the result against all other scores. From this, produce three numbers: MaxImp, AvgImp, MinImp
3. Final score = MAX(AVERAGE ((PctMP * (MaxImp - MinImp)) + MinImp, AvgImp), AvgImp)
Therefore, you are always guaranteed to get the AvgImps on the board. However, if you do particularly well, you may benefit from a good MP score.
I was thinking that an alternate way to score is if the AvgImp is positive, then score as #3, but if the AvgImp is negative, then score MIN. However, this would punish a bad score worse than AvgImp, which I think is unfair.
Example:
Ten pairs get to a non-vul 4H, 2 go down, 7 make 4, and one pair makes 5
Using my formula:
Notice that the FinalScore rewarded pair #1 greatly for the top, pairs 2 and 3 were unpunished further and pairs 4 through 10 were rewarded a little extra. The defensive teams would score as follows:
I understand that in large games, it is common to eliminate the top and bottom one or two scores before cross IMPing the results. I assume you would want to do the same when determining the AvgImp, MaxImp and MinImp.
I am not sure there is an advantage to scoring this way. It causes mixed strategy based on almost random field results. In IMP scoring, your goal is clear: Make or set the contract. In MP scoring, your goal is to maximize your score against the field. Although the field may be difficult to gauge, once gauged, your goal is clear.
With mixed MP and IMP scoring, your goal becomes murkier. Not only do you have to determine the field result, you may have to determine how the field outcome will affect the complex scoring.
With that in mind, I came up with this methodology. My objective was to combine the MP into the IMP scoring to reward good MP scores or lessen the damage of a bad IMP score that would have been a not so bad MP score.
This method rewards good scores and does not further punish bad scores. It gives extra advantage to good MP scores that might not be reflected in the IMP score. I believe it satisfies my objective by retaining IMP goals.
IMP + MP scoring:
1. MP the result against all other scores into a percentage, Call it PctMP
2. IMP the result against all other scores. From this, produce three numbers: MaxImp, AvgImp, MinImp
3. Final score = MAX(AVERAGE ((PctMP * (MaxImp - MinImp)) + MinImp, AvgImp), AvgImp)
Therefore, you are always guaranteed to get the AvgImps on the board. However, if you do particularly well, you may benefit from a good MP score.
I was thinking that an alternate way to score is if the AvgImp is positive, then score as #3, but if the AvgImp is negative, then score MIN. However, this would punish a bad score worse than AvgImp, which I think is unfair.
Example:
Ten pairs get to a non-vul 4H, 2 go down, 7 make 4, and one pair makes 5
Using my formula:
# Score PctMP AvgImp MaxImp MinImp FinalScore 1 450 100 3.2222 11 1 7.1111 2-3 -50 5.556 -9 0 -11 -9 4-10 420 55.556 2.1111 10 -1 3.6111
Notice that the FinalScore rewarded pair #1 greatly for the top, pairs 2 and 3 were unpunished further and pairs 4 through 10 were rewarded a little extra. The defensive teams would score as follows:
# Score PctMP AvgImp MaxImp MinImp FinalScore 1 -450 0 -3.2222 -1 -11 -3.2222 2-3 +50 83.333 9 11 0 9.0833 4-10 -420 44.444 -2.1111 1 -10 -2.1111
I understand that in large games, it is common to eliminate the top and bottom one or two scores before cross IMPing the results. I assume you would want to do the same when determining the AvgImp, MaxImp and MinImp.
It costs nothing to be nice -- my better half
#3
Posted 2007-June-01, 10:02
In Patton scoring, IMPs and MPs are combined. For example, if you play four boards against each team, there are 8 matchpoints to distribute. So the IMP -> VP scale is made so that there are 8 VPs as well. You Patton score is the sum of your MPs and your VPs.
I suppose you can also combine MPs with IMPs in a butler or XIMP pairs tourney but I'm not sure about the aritmetics. I can imagine several ways of doing it. Probably someone else can explain.
I like Patton, but it's a matter of taste of course. If you strongly prefer, say, MP, of course you should choose that. If you think that IMPs and MPs both have their merrits and demerrits, Patton may be for you.
I suppose you can also combine MPs with IMPs in a butler or XIMP pairs tourney but I'm not sure about the aritmetics. I can imagine several ways of doing it. Probably someone else can explain.
I like Patton, but it's a matter of taste of course. If you strongly prefer, say, MP, of course you should choose that. If you think that IMPs and MPs both have their merrits and demerrits, Patton may be for you.
The world would be such a happy place, if only everyone played Acol :) --- TramTicket
Page 1 of 1

Help
