Tuesday, January 4, 2022

Points and Speed Bias

Club points ought to reflect how well a shooter has shot at an event.  On the national level, the top gun system works pretty well for those shooters in contention, say the top 30 or so shooters in the sport.  However at the club level the system does not always work that well because of the speed bias.

The speed bias is necessary to make the computer work.  Placings are based on x count and speed.  This results in quick shooters generally getting more points than they deserve.  For example, at an event at Rio this year, Half Cock and I had the same x count. Half Cock had beaten me in the main match and a reasonable conclusion would be that he had shot better than me.  However, I got 6 more points than Half Cock because of the speed bias.  Rio, of course, really emphasizes the speed bias because they do not consider the shoot off.   

On national level speed bias is not a problem because of the rule that recognized place must be shot off.  Since to be competitive, one must have 6 or more good events and these points will have come from recognized places.  Therefore, for the shooters in contention, points do not result from speed bias. 

We do a fair job of mitigating the speed bias by awarding points after the shoot off, but we can do better.  I ran a trial system based on only x count this year but I have concluded that it was not worth the effort it took to administer.

However, a small change in our shoot off format would have a substantial benefit in mitigating the speed bias.  I suggest the three shooter 2x magnificent shoot off.

We currently divide the field into three brackets based on x count and speed.  We then shoot a 1x magnificent shoot off in each bracket.  This can result in the speed bias affecting the ability of a shooter to mitigate the speed bias if he does not make it into Bracket A because of the speed bias. The split between Bracket A and B can really be unfair.  What if we divided the field into 2 brackets with about 1/2 the field in each bracket.  Looking at 2021, 100% of the points that were counted for all of the award winners were earned by being in the top half of the field.  Of all the points earned by the award winners, being in the top half of the field amounted to 93% for Master Gunfighter, 82% for Gunfighter, 71% for Sheriff and 93% for Deputy.  To do well, you have to be in the top half of the field.  Sort like at the national level, you have to be shooting for recognized places to get enough points to compete.

The division of the field would be at the break between 2 and 3 x or between 3 and 4 x.  You need to divide at a clear break to avoid the speed bias.  The wording for the rule would be "The field will be divided into two brackets at the nearest x count break to the center of the field"  For 2021 the break would have been at 2-3 for January, February, March, April, August, September, October, November  and at 3-4 for May, June and July.  I guess we don't shoot as well in the heat of summer.

Three shooter 2x Magnificent Shoot Off:  Three shooters shoot on three lanes (two or three) winning hits.  When a shooter gets (2 or 3) winning hits, he/she steps back and allows the other two shooter to finish the match (2 or 3) winning hits.  Loser  is eliminated having been beaten by two shooters.  In essence the loser has lost to two shooter and therefore has 2 xs. Next seeded shooter shoots against the two winners.

Big advantage of this is:

1.  All places are determined on the line and not by speed bias.     

2.  It is a two x elimination event.  

We need to test for time on the mountain. 

Time Test: 

We tested this on the mountain with 21 shooters shooting.  For the first 45 minutes we shot 3 winning shots.  We got 19 losses per hour.  We then tried 2 winning shots.  We got 22 losses per hour. A average turnout shoot at Pioneer has 38 shooters and it would take 2 hours to do a 3 winning shot shoot-off.  It would take 1 hour 45 minutes to do a 2 winning shot shoot off. Three winning shots may not be practical considering time but two winning shots may be.  For example, 8:30 to 11:30 for seeding no x matches, then 11:30 to 1:15 for 2x three shooter two winning shots shoot off.

Scoring: Winning shots were carried over from the three shooter portion of the match to the two shooter portion of the match.  Prior shots by contestants were not re-scored between the two shot contestants. (You could restart at 0-0, but that would lengthen the matches)

Match Length:  With three winning shots the majority of matches were resolved in less than 8 shots for both winners to be determined. However there were two matches that went to 14 rounds. (14 is the maximum on the score sheet so that was the dreaded two loss maximum.)  With two winning hits the majority of the matches were determined in less than 6 rounds with only one match going to 8 rounds. (14 was maximum in this event, but we could reduce it to 8 for 2 winning shots, since shooter are familiar with that. I would prefer keeping it at 14.)

Best Length:  Two winning hits was clearly better than three winning hits because it resolved the issue quicker while still doing a good job sorting. See next paragraph below.

Speed Bias vs. Gunfighter Bias: The three shooter match has an inherent speed bias built into it but that did not appear to be a problem.  I think the speed bias was overcome by the 2x factor which of course favors the better gunfighter. A quicker shooter may eliminate a better gunfighter in a single elimination type event but the better gunfighter is normally going to prevail in a multiple x fight. Examples of this were No Name surviving 4 rounds, Wyoming Ranger surviving 8 rounds, Desert Gator surviving 9 rounds and Half Cock Willie surviving 11 rounds. (For those not familiar with our shooters these shooters tend to be our better gunfighters while maybe not our quickest)

Conclusion:  The 2x three shooter shoot off two winning hits would be a good format to try at Pioneer.  It does a good job of insuring that top gun points are determined on the line and not just by the speed bias inherent in the computer program.

We would still have the speed bias in seeding, but I think that is actually a good sorting tool.  We have a fair number of shooters that succumb to the pressure of trying to go fast in the last round and end up being penalized by the dreaded 22.