Here is a 20 shot group

[Dan]

From a purely stochastical point of view i wouldnt agree with you Nicole. If you drop 10%-20% extrema which would be the best 2 and worst 2 you get a set of 16 shots which are close enough to give meaningful data in the given target since they seem to easily fit within the 9 ring. You do not need 200 shots in that area to predict a tendency. Of course the more data you have the more precise your prediction will be, but i think its pretty clear that there is a readable tendency in those shots we see here.

[David Levene]

You do not need 200 shots in that area to predict a tendency.

The problem is that you will never have a valid sample that big under match conditions, it's 3 match-fulls.

If you cannot make a decision to alter your sights within 5-10 shots, or at the top level within 1-3 shots, then you have lost the match (or at least several points).

[Fred]

If the objective is to have all "technically perfect shots" in the 10-ring even if only the occasional shot is technically perfect and all the rest display have the same flaw, e.g., a consistent error in the sight picture, then sure, be my guest, adjust your sights for those occasional shots.

But if the objective is to maximize score, then it's not about where the technically perfect shots end up on the paper, it's about what happens to the overall distribution of shots, regardless of what skill or lack thereof is displayed in any given shot.

I can't think of a better argument for adjusting your sights in a match, based on what you see on the paper, than what is written above. The only remaining question would be, "after how many shots?" The answer to that question is far more dependent on how many shots the match consists of, than it is on whether 100 or 200 or 300 shots would provide a statistically optimal sample. Every skill and confidence level would imply its own practical tolerance-of-error point, but the operative word would have to be "practical." It is highly unlikely that waiting for more than 20 shots in a 60 shot match could be called a practical solution.

FredB

[CR10XGuest]

Ok, here's my 2 cents (both short and long). The following is my personal opinions and process. YMMV, etc.

Yes, from that group I would move the sights down several clicks to gain the points. Why? Because for those conditions under which this series was shot, the group it centered vertically above the 10 ring. It doesn't matter why, because a series of shots were involved wherein the shooter (or I) was shooting for a score.

(If the shots are falling in a consistent patten or group of x size, why fight about the change? If that's how it winds up after x number of shots, you might want to take advantage of it, rather than being stubborn and losing points. Your body and performance may be telling you something you might want to take a moment and listen.)

I am the master of my sights, I can't necessairly control all the other parts such as fatigue, lighting, what I ate yesterday, eyes focusing differently, getting tired, change in blood pressure, etc. It is a dynamic system, the shooter needs to observe and make notes to that effect. Sometimes its even the range and differences in lights (natural or artificial) or even target height. Move the sighs, record it, sketch or keep the group and compare it to other days, times, etc.

Now if this was just training, that's different. I would be working on only 1 or at most 2 parts of the shot process (alignment, sight picture, position, press, etc). In that event, I would not be concerned with the position of the group, but with the size and consistency (assuming it was even training where an actual shot was involved).

From my perspective, I'm either training or shooting for score. The traning is obvious. Shooting is practice or whatever you want to call it but it is seeing how all the parts come together.

By the way, this leads me to another differing opnion. I am not searcing for the "tecnhically perfiect shot", not by a long shot (pardon the pun). I'm searching for a process that will allow me to shoot a group approximately 3 pellets across, consistently and throughout a 60 or more shot match.

That "perfect" goal would result in me getting positive feedback only about 1 time out of about 1000 (leaving that whole inner 10 ring intact). My process gets me positive feedback about 70% or so of the time. So I don't go in for "technically perfect", I just want a bunch of really acceptable parts to the process that I can consistently repeat.

Then I'll move the sights to wherever I need to in order to get the most points from my "non-perfect, dynamic, old, failing eyesight, dark range, temperature variable, mentally unbalanced, etc. system).

Just my opinion.

Cecil Rhodes

[Benjamin]

"Without more data, yes, I would leave the sights alone."

Yeah Nicole. Us engineers should stick together.

The writer who said it is incorrect to assume a static condition, is correct on that point. Which is why adjusting the sights after that group would be a bad idea!

The shooter already said this group was unusually good, which would imply the next one is likely to be different. So which way should we adjust the sights, to correct a next group that we know nothing about?

Further, this group is high - but only by a small amount, substantially less than the horizontal spread of the group - it is really well centered already in my opinion. True, it would be a couple of points higher if the group were lower - but you can't adjust the group after it has been fired. Someone else already pointed out that he tends to shoot lower (and compensate by adjusting the sights higher) as he shoots the match. Assuming another shooter is the same, adjusting the sights downward now, would push the likely to be perfectly centered next group off the center downward.

My philosophy is, which I sometimes ignore but usually with bad results, that if in a match you shoot a better than normal group (higher than your average score), CHANGE NOTHING!! Think about what you did right, and shoot another one just like it! If you ever find you can shoot better groups consistently, but they are consistently off in the same direction, then by all means change the sights. Likewise if you shoot several shots with your normal group size during warmup, and most of them are off in the same direction with no other explanation, the sights should be adjusted. But one excellent group alone is something to be happy about, not a problem to fix.

- Benjamin

[Steve Swartz]

If your shooting process is broken, would you:

1) Change the non-broken part of the system (the gun) to improve the "holes in paper outcome?"

or

2) Fix the broken part of the system (why your group is off center in the first place) to improve the "behavioral/process?"


There *is* merit in focusing on "outcome" during a match for the purpose of temporarily improving "really sucky" to "sucky" performance levels. Yes, if you are doing something wrong, and are confident that you will consistently do the wrong thing throughout the match, by all means- grab that screwdriver and crank away.

There is also merit in focusing on "process" during a match to prevent the performance level from wallowing in "sucky" in the first place.

This is a "symptoms vs. causes" argument.

Just me, personally, no offense to more "hardware focused" practitioners- have achieved more performance improvement from focusing on process and causes; even during a match; than on alleviating symptoms and focusing on holes in paper.

Again- assuming nothing mechanical is at work (dropped the gun on the sights, radically different pellets, radically different lighting, etc.).

Steve Swartz

[aurorapolice02_11]

That's what they are made for...people are too afraid of moving their sights. Sure don't move your sights on every shot, otherwise you will be chasing shots. If you call a shot as a ten and it is a nine, you might want to start thinking about the sights. If you shoot a second shot into the same area, you better move them. This is where following through and analyzing your shot comes in. If you call the shot where it ended up, then a sight change is not necessary.

Waiting twenty shots to decide if sights need to be moved is unacceptable. Think about it, in matches where tenths of a point can determine a winner or loser, if a sight change is needed, then do it. You're looking at whole points in the match portion, those are hard to make up in a final, especially against really good shotters.

On the other hand, I do not recommend using the sights to correct for a shooting error, i.e., jerking, anticipation, etc. Ideally if you have a technnical error in a match you might be able to dry fire and take a break to work the problem out, but you are against the clock. With all this said, I would not rule out moving your sights in a match and continue using the technique that might be sub par. After all, a match is about score, your training should have been done a long time ago.

Mike Douglass

[Nicole Hamilton]

Your suggestions may be correct form a purely mathematical / statistical perspective but shooting is neither. You will not get a statistically meaningful dataset from a series of shots the population is just too small.

What possibly prompts you to say this? Of course it's a statistical distribution. And if you collect enough targets, you can get as large a dataset as you like.

The problem is that if you're a knob-twiddler and don't track your changes to the sights -- something knob-twiddlers rarely do -- you invalidate your history each time by making it impossible to compare the new series of targets to those generated prior to the change.

... the problem with Nicoles theory is that it assumes a stable system. Pistol shooting is not stable so you do not get a Guassian distribution. The process of a shot has many variables that can be slightly different on each shot, muscle fatigue, tension, stance, vision, balance, weight distribution to just name a few. So adjusting the sights is just a response to one of the other varibles changing to maintain an equalibrium.

It's as stable as any other stochastic process, which is to say you get a distribution of events that can be described probabilistically. You're right, it may not a perfectly round Gaussian distribution symmetric about a single midpoint due to specific systemmatic error modes. A finding of distinct clumps representing specific errors would not be surprising. For example, a shooter that makes a specific error, e.g., thumbing the shot some percentage of the time, might have two clumps, one when he makes the error, the other when he doesn't. But this doesn't change the the fact that it remains a probability distribution. All we may be arguing about is the shape. And I'm saying that if you collect enough targets and resist knob-twiddling, you get to find out what that is.

I agree that the specific outcome for a given shot and for the distribution as a whole is likely determined by a large number of variables. And if enough data is collected, some of the relationships may be discovered, e.g., by linear regression, gradient descent or other techniques. For example, it's entirely possible that with suitable instrumentation to log individual shots and computer analysis (since it's probably beyond what could be done by hand), that we might discover that a given shooter's shots tend to start high (or low, or left, or right, or whatever) then drift low (or whatever) through a match. If so, we might generate some rules for the shooter, instructing him to adjust his sights in a specific manner (e.g., reset at the beginning of the match, then go 1 click down after each 20 shots throughout the match.)

Our imagination is our only limit on the number of such possible relationships we might investigate, if only we're willing to do the instrumentation and analysis. Perhaps the shooter's expected distribution of shots could be related to his heartrate or temperature in the room or the number of hours of sleep he got the night before.

The problem is that while it may be entirely possible to discover any number of such relationships, doing this requires collecting the data and doing the analysis. It's not valid to shoot a couple targets and assert, as you put it, that "one of the variables" has changed and that you're now merely correcting for it. How do you know it's one of the variables? Could it be more than one? Maybe it's none of them. Maybe nothing's changed and all you're looking at is a random selection. Even if it is a single variable that's changed, it might itself be a random, independent selection from a distribution. For example, perhaps the variation is due to slight variation in grip. But if you adjust your sights based on one target, what happens when you set down the gun to change targets and then pick it up again? You'll have made a new, possibly Markov (sequentially-related) but probably independent selection. The only way to know if you've found a relationship is by collecting the data and doing a proper analysis.

For most of us, including me, the instrumentation and the work required to discover these other possible relationships is simply beyond anything we can afford. So the best solution is to resist knob-twiddling, making changes to your sights only very occasionally as the evidence mounts from a lot of targets to tell you what your overall distribution is and whether it should be moved on the paper.

If you drop 10%-20% extrema which would be the best 2 and worst 2 you get a set of 16 shots which are close enough to give meaningful data in the given target since they seem to easily fit within the 9 ring. You do not need 200 shots in that area to predict a tendency.

Elimination of extrema is a valid technique when you don't have enough datapoints and you're concerned that a few outlyers may skew your results. Often, the rationale is that outlyers may represent experimental error, e.g., human error in measurement or recording. But given a choice between casting extrema and simply collecting more data, most researchers would prefer to collect more data, in part because in trying to identify extrema, you have a new opportunity to make an error, which happens to be just what you've done here.

In the example at hand, while it may be reasonable to cast extrema, they are not the two best and two worst, they are the ones furthest from the center of the distribution of shots!

If the shots are falling in a consistent patten or group of x size, why fight about the change? If that's how it winds up after x number of shots, you might want to take advantage of it, rather than being stubborn and losing points.

We're back to the same question: How do you know you have a consistent (new) pattern versus a simple random selection from a distribution? The only way to know is by collecting the data. You might, e.g., collect and analyze the data from a large number of matches and discover a Markov (again, a serial) relationship that if you start a match with all your shots grouping high at 2 o'clock that that's likely to continue. But to assert this without doing that collection and analysis is just guessing. Absent the data, the better hypothesis is that with more shots, you'll see the same overall distribution develop that you have in the past.

The shooter already said this group was unusually good, which would imply the next one is likely to be different. So which way should we adjust the sights, to correct a next group that we know nothing about?

Precisely. The condition that should be presumed -- and will exist if the shooter avoids knob-twiddling -- is that the current settings reflect all that is known historically, though it's certainly fair to weight recent outcomes more heavily than those far in the past, e.g., with a moving average technique, assuming you're up for the analysis involved. So day-by-day, as you collect new datapoints from new targets, you should ask if these new data, not by themselves, but in combination with the accumulated past data, suggest that it is now more than 50% probable that if you made a change to your sights, that your future scores will go up.

The only time you should adjust your sights based on only one or two targets is when that's all you have. For example, if you get a new gun, you don't have any history so of course, you adjust on every target initially. It's also possible you might invalidate your history or reduce its presumptive value due to specific events, e.g., changing brand of ammo or changing the trigger pull. Knob-twiddlers who make changes to their sights but without recording them also invalidate their history, making it impossible to relate old targets (and old distribution) to new, and this is a mistake which unfortunately self-perpetuates: Since they never have a history, each target is always significant and always a reason to change. But don't do that! That's a mistaken procedure.

[CraigE]

After reading all that.....and appreciating the careful thought put into it......seems to me to come down to observing and controlling the fundementals over which we do have the power. Practice/train to assure a repeatable process and analyze execution of fundementals rather than twiddle. If we shoot good groups ....in the right place......at least once or twice....then it falls to the shooter to replicate the process that produces those results, not to the adjustment of or purchase of new equipment. Perhaps said more simply, the fewer things we change, the more observable the impact of individual alterations. IMHO. YMMV. CraigE

[aurorapolice02_11]

It absolutely floors me when shooters try to use statistics/data analysis, etc., etc. when trying to talk about shooting. Let's keep the shooting simple guys. Train (really hard), shoot the match and then work on what did not go well in the match at the next session. DO NOT cloud your minds with senseless information. It all comes down to releasing the trigger while the sights are aligned. Call the shot. If said shot did not imapct where you called it determine if it was your poor technique or the sights. Correct this for the VERY NEXT shot!! DO NOT wait for halfway through the match.

All this post match statistical analysis is unnecessary. It takes away from training time.

Mike Douglass

[RobStubbs]

Nicole,
You seem to keep ignoring the fact that there are multiple reasons why shot point of impact may change and that will change over the course of fire. So your statictical approach is meaningless because you never have a flat reference line to refer back to. As the previous poster says, call your shots, see where thet went and if they didn't go where you called them then change something. I really fail to understand how you can ignore those simple facts - unless the shooter is hopeless at shot calling. Which is a more serious problem and not for now.

Steve,
I don't dissagree with your approach however for a small series of shots where you called them and they didn't end up where called then I'd change the sights. A single click or two is nothing much in the great scheme of things and if it doesn't work then click em back again.

I think overall we're all gonna disagree on this so I'll leave it there (for now ;-) )

Rob.

[David Levene]

It all comes down to releasing the trigger while the sights are aligned. Call the shot. If said shot did not imapct where you called it determine if it was your poor technique or the sights. Correct this for the VERY NEXT shot!! DO NOT wait for halfway through the match.

I couldn't agree more Mike. The problem is, how many shooters bother to learn how to call their shots.

I consider it a skill to be practiced every time you fire a shot. To do that however there can be no such thing as training ammunition and match ammunition. Train with the equipment you will use in the match. To call the shot you must have total confidence in the accuracy of your gun/ammunition combination. If you are using an inferior "training" ammunition then you cannot have that confidence and are missing the opportunity to learn a valuable skill.

It is perfectly possible to learn to call shots within half a ring (or better). It is also vital however to learn to recognise when you are calling accurately. There will be times when you will not have the confidence in a particular call to enable you to make a sight adjustment. Thats fine, accept that it can happen and move on to the next shot. If you are confident in your call and know how accurately you can call, and the hole is not where it should be, then make the adjustment.

[aurorapolice02_11]

Calling shots is not the most exciting part of shooting, but it is necessary to becoming a better shooter. There is no way to learn from your mistakes without calling your shots. This is where follow through becomes very important. It is very hard to call a shot if there is no follow through.

Without the follow through to call the shot, you cannot make a correction. It's a big cycle. This is where I add my dislike for those shooting wheel of errors cards. Sure you have a shot that goes to the right. Because you did not follow through and did not really feel what error you made, you consult a pre-made piece of paper with a general/possible solution to the placement of the shot. Keep in mind several factors in the shooitng process can make a shot do the same thing, in this case go to the same location on the target.

I am disappointed in some of the coaching shooters receive sometimes. I am even more disappointed in the lack of interest in a coach by some shooters. I recently called several (5) clubs in the Denver area to offer coaching assistance. I got one call back and still have not been able to contact that person. But, I digress...

Mike Douglass

[Nicole Hamilton]

If your shooting process is broken, would you:

1) Change the non-broken part of the system (the gun) to improve the "holes in paper outcome?"

or

2) Fix the broken part of the system (why your group is off center in the first place) to improve the "behavioral/process?"

... seems to me to come down to observing and controlling the fundementals over which we do have the power. Practice/train to assure a repeatable process and analyze execution of fundementals rather than twiddle.

I agree with these comments. Leaving out the long explanation of how I got there, this is pretty much all I was trying to say. Strive for repeatability and small groups, not some imagined better setting du jour for your sights.

It absolutely floors me when shooters try to use statistics/data analysis, etc., etc. when trying to talk about shooting. Let's keep the shooting simple guys. Train (really hard), shoot the match and then work on what did not go well in the match at the next session. DO NOT cloud your minds with senseless information. It all comes down to releasing the trigger while the sights are aligned. Call the shot.

Let me guess: This isn't about anything I said, it's about you not liking math and pointy-headed people who think too much, it sounds like.

I didn't argue in favor of analysis, I argued in favor of not doing it because I think a proper analysis is probably beyond what most of us can afford. Instead, I argued for leaving the sights alone until you'd collected enough targets to make it obvious you could justify changing them. If it wasn't clear, let me make it so: I don't analyse this stuff on a computer, I simply eyeball my targets -- I just do it using a lot of them -- and that's all I'm recommending. If you leave the sights alone, what is there left to do except perfect your technique?

I didn't realize this thread was about calling the shot or I'd have come out in favor of it much sooner. Allow me to state for the record that I'm a great fan of calling the shot. I'm also a fan of motherhood and apple pie. And my experience is that I can call my shots just fine, within limits. (I can call the direction and whether I'm off by a little or a lot; I can't consistently call the difference between a 9.5 and 10.) But then again, my sights are set for me. If you try calling your shots and it doesn't work for you and you feel you need to change your sights, maybe that's true; maybe you've screwed them up, fiddling with them too much.

You seem to keep ignoring the fact that there are multiple reasons why shot point of impact may change and that will change over the course of fire. So your statictical approach is meaningless because you never have a flat reference line to refer back to. As the previous poster says, call your shots, see where thet went and if they didn't go where you called them then change something. I really fail to understand how you can ignore those simple facts...

Well, it's certainly possible one of us is ignoring the facts and failing to understand. Let's see if we can guess who that is.

There are indeed reasons why POI might change, even over the course of a match. In outdoor high-power matches, changes in crosswinds and lighting are well-understood to require compensation. That clearly is not the shooter failing to do his part at a repeatable performance, that's the environment changing. But this thread began with the discussion of an airgun match, where conditions don't change, not even very much from one match to the next. They're pretty much invariably indoors and even the lighting is usually very similar. If the conditions don't change -- and heck, we know the guns aren't changing, they're all good enough to put pellets through the same ragged hole all day, every day -- but you get different results on different shots, the only thing left to explain this is that you've changed.

I also agreed that some of that change might be systemmatic and could be compensated for by changing sights even during a match. I offered examples, e.g., a hypothetical shooter that always starts a match shooting high but whose shots drift lower through the match in a predictable fashion, or another hypothetical shooter whose performance exhibits a serial characteristic that if he starts high on a given day, he will likely continue. I also tried to be clear that the better you are, the smaller the discrepancy required to justify changing your sights.

The only place where we seem to be in disagreement is on the matter of whether a couple of targets at the beginning of match would be sufficient to justify changing your sights. So allow me to concede that maybe for you this is a good thing to do. I've certainly agreed that if you're good enough and did the data collection and analysis that you might discover clear evidence to support this practice. Maybe you've done the analysis (even just by eyeballing it, which I regard as perfectly adequate, just so long as you look at enough targets) or maybe you've just made a good, lucky guess about your own performance. Even if you're 100% wrong, it's still no skin off my nose. It's your gun, your target, your match, your score, your choice. You get to do whatever silly or not so silly thing you like!

My experience is just different. I find -- and I've a good 10" stack of targets on which to base this -- that my own ordinary shot-to-shot random variability is more than adequate to explain the occasional target with an otherwise beautiful cloverleaf a little high in the 9-ring without having to conclude that my sights are off. I know from examining my targets that my ordinary variability could mean the next one is low -- or maybe just not a very good group at all. Maybe you're just plain better than me.

What works for me is to try to "get in the groove" where I'm simply doing the exact same thing -- same sight picture, same stance, same grip, same trigger pull, same follow-though -- every time. When I succeed in doing that, I get a 10 and no adjustments of my sights are needed. My problem is that I just don't do that exact same thing often enough. To get better scores, I need to shrink my group -- and I don't mean just the group on one target, I mean the overall distribution of my shots across lots of targets. If I can do that, I already know I can adjust my sights to put it anywhere on the paper I like. My sights are fine; what's not as good is my own performance.

Your experience may be different. Maybe you're already perfect. Or maybe you're not.

[David Levene]

What works for me is to try to "get in the groove" where I'm simply doing the exact same thing -- same sight picture, same stance, same grip, same trigger pull, same follow-though -- every time.

That's fine Nicole, as far as it goes. What about the myriad of other factors that can affect shot placement (especially in a match). These could include, in no particular order:-

Different range lighting to what you are used to.
Different target height (still within the rules).
Match tension.
Electronic targets (they can all have different zero).

There are so many reasons why your sights could be off in a match. You just do not have enough time to gather the amount of the data you are talking about. You simply have to get on with it and maximize you score.

Adjust the sights as soon as you note that centalizing the group you are shooting would give you a higher score. If you are good at shot calling then you can adjust based on 1 or 2 shots.

[aurorapolice02_11]

You mentioned shooting with your sights as they are for numerous shots to "collect data" regardless of where the shots were landing. Unfortunately a match only lasts for 60 shots, you do not have time to shoot several hundred shots to "collect data". Two shots in the same area after calling them elsewhere is enough data in the shooting world to change your sights.

This has nothing to do with math or pointy headed people. I am trying to provide quality advise for new shooters. Unfortunately, your experience in the science world had led to you providing an opinion that a new shooter may think is correct. I hate to say it, but your reasoning for leaving the sights "as is" is incorrect.

Mike Douglass

[CR10XGuest]

Yep, I have to agree with Mike on this one.

My comments were based on the major assumption that the shooter was calling the shots as "good". Not "looked OK but can't remember", not "don't know", not "maybe it's a..."; but making good, solid calls.

If the shooter can't call the shot consistently, then the group position is the wrong thing to be thinking about anyway.

To put it another way. On the line for the Presidents 100. If the first shot is not where I called it... out comes the screwdriver. You know, the light is really different at 7:00 am depending on the rain, sun, fog, hangover, etc. And if I'm not sure about the call, then away with Number 2.

But with only a limited number of shots and no stinking sighters...one thing you really train for is calling the shot.

Cecil

[ColinC]

I have been away for a few weeks and was pleased to return today and catch up on this fascinating topic. The number of replies and hearty debate makes good reading.
Only question I have is why people are talking about 20 shots on a target. In normal club competition I shoot 10 x 6 shot cards and often have trouble scoring them. In open competitions it is normally two shots a target and even then both shots go through exactly the same hole.
Surely targets aren't that costly in the US.

[Steve Swartz]

David:

Aside from your point about the calibration of electronic targets (never more than one click out of calibration- at the absolute worst; assuming competent range personell), the rest of your points about lighting etc are, as the engineers would say, "insignificant."

Well inside of one click.

Steve Swartz

(p.s. "match nerves?" c'mon. that is a human [process] problem; not a mechanical problem to be solved with a mechanical adjustment)

Steve Swartz

[aurorapolice02_11]

with David. Numerous times I experienced having to move sights due to different ranges, electronic targets, after travel, etc. It is more than just one click in most instances. Then again, sometimes none at all. Regardless, move the sights if necessary, no matter how they were set the last match you shot.

Mike Douglass